Short/arc fault in cold weather on XT-90

Yer bike busted? tell us all about it here.
bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Short/arc fault in cold weather on XT-90

Post by bombadero » Jan 14 2022 8:39pm

Hey, I had a similar problem to one seen in a few other threads here on ES, but mine is slightly different, so wanted to get everyone's take on it. I have a dual-motor, dual-battery, wired in parallel setup with components purchased from Grin. The wiring setup is basically two Y-cables plugged into each to form an 'x' between the batteries and two controllers to achieve a parallel wiring harness; works great. Initially, everything was stock as ordered from Grin, with APP connectors. I had issues with connectors out of one of the batteries melting and fusing one day while hauling a heavy load on a warm day. I discovered this is sometimes an issue with APP's (from ES, general web search and confirmed by a Grin employee), so I resoldered all 10 connectors with XT-90's, and have been riding it that way for several months. A few days ago I was taking the dog to the vet and then to the dog park; on the way to the vet, I experienced a sudden loss of power heading up a hill. The bike is 1.1 kW nominal power from both motors combined, so it usually crushes this hill, even towing a 55 lb. dog. I pulled over and power cycled and it went away, so I thought nothing of it. Later in the evening as I was coming home from the dog park, after cresting another hill and resting at a light for a minute or two, when I proceeded through the intersection I heard a brief 'pzzt' noise and the bike died shortly after starting out across. I has some momentum, so I was able to coast to the other side and pull over. After another power cycle, the bike seemed fine again and I finished the trip home without incident, although the bike smelled strongly of an electrical fire. It wasn't actually on fire or smoking, just had that smell.

After some troubleshooting, I narrowed it down to one of the Y-cables, one leg of which only registered ~8.5V on a multimeter, verified with a LunaCycle Power Meter. All other connections read ~58V, which is normal for this bike. The conditions on that night were humid and cold, but it didn't rain that day. Moreover, I covered all of my connections with cable repair tape, which if you're not familiar, sticks to itself like glue once you wrap it on; it's pretty much impossible to unwrap it, you have to cut it off. I have that wrapped from one half of the XT-90 on either side until it meets shrink wrap that had been applied by Grin. I call it 'poor-man's heat shrink', because the seal is easily as tight as shrink wrap. It's also extremely thick. As expected, when I cut that off, there was no moisture inside. What I did not do is apply any dielectric grease or liquid tape. The cables were burnt for a couple inches on the short cable leading to the Baserunner in the rear of the bike; on the Y-cable end, it was much worse, with gold-colored flecks on the outside of the cable insulation from the electrodes in the connector getting so hot they melted, and the carbonization from what appears to be a very small, very brief fire, extended several inches. Fortunately the cable repair tape and the Y-cable contained it. Back on the controller side, the connector had been damaged a little and the little nylon apron that clicks in to protect the solder points was going in at an angle. I was reminded on inspection that I had stripped my cables too much and about 5mm of exposed cable was sticking out of the apron for both hot and neutral. I repaired the connector for now as it reads correctly, supergluing it back into the right orientation with Loctite gel superglue clamped together, and reapplied two layers of tape this time, both high temperature electrical tape and a fresh layer of repair tape on top of that to seal it. I also wedged a piece of rubber between the two wires and taped them separately with electrical tape to insulate them from each other.

After some research, I will probably open that controller connector again and coat that exposed wire with Plastidip Liquid Tape, and grease them liberally with Permatex dieclectric as well for good measure, and do the same with all other connections on both sides. I ordered some new Y-cables to convert to XT-90 (a replacement and some back-up inventory) to replace the fried one. Going forward, I will never strip my cables too long and plan to always make liberal use of liquid tape and dielectric grease.

So here's my question though: does it seem like the 5mm of exposed wire on both hot and neutral was definitely the culprit, or does it seem like moisture may have played a factor? As I said, everything I examined was dry inside, but given how hot that tiny fire burned from the arc/short, if there was any moisture it was probably vaporized instantly. I don't like to use shrink wrap for the end of these XT-90 cables because the connector is so much wider than the cable, and it's not convenient to take apart and inspect when things like this happen. If you do that, to re-apply heat shrink you now need to de-solder and re-solder the whole thing, and then re-shrink it. So that second question is, does Plastidip + Permatex dielectric grease + high temp electrical tape + cable insulation repair tape seem like sufficient insulation against moisture? It seems like overkill to me, but I'd like to hear other folks' opinion. I live in San Francisco, so extreme winter weather is mostly not a concern, apart from the record-breaking bomb cyclone we had recently, but obviously I wouldn't be riding in that! Just light amounts of rain and extremely high humidity on some days.

I've attached a photo of the connection in question. BTW, in that photo there is a layer of automotive cable loom on top of the cable repair tape. I'm going to stop using that on cables because it holds moisture, but in this case there was the thick repair tape underneath it. Unlike the repair tape, the automotive loom can unravel a little, although usually it also has very good adhesion to itself once applied, which is why I use it on bikes for various things.
20220110_161809.jpg
20220110_161809.jpg (2.9 MiB) Viewed 737 times
Last edited by bombadero on Jan 18 2022 10:22pm, edited 1 time in total.

john61ct   100 GW

100 GW
Posts: 8438
Joined: Dec 18 2018 2:06pm

Re: Short/arc fault in cold weather on XT-90

Post by john61ct » Jan 14 2022 9:05pm

Are you sure your "peak" power drawn did not exceed the connector rating for more than a few seconds?

Put an Anderson SB350 on there

or even a welding style 500A if you think needed

for testing with a logging shunt based wattmeter capable of say a 1000A scale, to check.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Jan 15 2022 1:03am

bombadero wrote:
Jan 14 2022 8:39pm
I'ver attached a photo of the connection in question.
20220110_161809.jpg
That image shows arc residue, commonly referred to as KFF (Kentucky Fried Finger) on the female connector face, as well as down inside the shell of the male conector.

That usually means the connector was not fully seated, or at the least that it was not making an actual connection at the barrel contacts, but I see no sign of it inside the barrels or on the pins, so that's probably not it.

If the cables, where stripped too far back, had been arcing together, you would see definite pitting and burning, unmistakable, at the arc point. If it is that close to solder joints, and went on for any length of time at all, those joints would almost certainly have melted, at least partially, if not completely, and reflowed, and probably look different than they did when originally soldered. If you happen to have pics of them from then, you can compare them. But the arc point between two wires will have obvious pitting, burning, etc., either way. That might be what you're referring to regarding the damaged insulation and flecks, but without pictures of this I couldn't tell you if it was or not.

If it's not that, then the next most likely thing is that there is something (solder balls, usually) down in the back of the shell at the base of the pins, allowing an intermittent short to occur between the pins whenever they vibrated to the right spot. This would also cause heating of the wiring up to that point, possibly enough to set insulation on fire, before the solder vaporizes. It could even desolder the wires at the back of the contacts in the solder cups, and cause even further arcing even after the solder itself has evaporated and coated the inside of the shell with soot.

The next most likely is that a wire has come off the back of a male connector shell contact (usually cold solder joints cause this, very common on large gauge wires and solder-cup style power connector pins), and it is arcing when it is pulled away enough by vibration, etc., and the plasma leaked thru a gap (potentially burned by the arc plasma) in the back of the shell to douse the shell interior in arc soot. It may be held on well with the heatshrink or tape thats' back there, so it doesn't appear to be loose from the outside of the assembly. Requires removal of that tape to find out. Heat from the arcing is pretty intense (arcs are used for welding steel...), and conducted thru the wires can easily melt wiring insulation.

Note that if it was a cold solder joint, it would have been completely remelted by the arc heating, so there may be no obvious evidence left of this. If there wasn't much solder present because a cold joint didn't let it flow into the wire and cup properly, then you might notice that the visible conductor of the wire and the cup appear to have very little solder present, as what was there may have wicked up into the wire itself, and down into the cup/wire interface.

Similarly, any moisture would have been vaporized and probably chemically bonded into the arc products.


Another (unlikely) possibility is if the barrel of one or both contacts of the "male" connector (pins that fit inside the hollow barrels of the other) are actually broken inside the shell, so that the pin pulls forward away from the back of the shell enough to allow arcing between the pin and hte wire solder cup at the back. It's not very likely since the pin would probably come out stuck in the female barrel if that were to happen. But Iv'e seen this with some connectors in the past (not XTs yet, though, but other machined-barrel types).


I haven't used plastidip, but as long as whatever you use completely bonds to all of the surfaces involved, so that water or moisture in any form cannot penetrate between the surfaces and the covering, even a tiny bit, then it would keep water out. If water gets in even a little bit, it can degrade the surface bonding further over time (especially if it ever freezes in there, or vaporizes as steam), and "pry" the coating up off the surface, eventually allowing actaul water penetration of connector and connections.

If you fill the connector shell itself with dielectric grease, and pack it around the connection points, it will help keep out any water that gets past the outer layer of protection, and the protection coating will help keep the grease in place. Note taht if things get hot enough the grease flows, and if there is anywhere for it to go, it'll go there instead of staying in the connector. ;) (happens in Phoenix in the hot summers)

Electrical tape doesn't make a waterproof connection; the adhesive doesn't remain sticky in the right way long in hot conditions, even the best 3M stuff I ever used degrades too quickly to rely on for this. Cheap stuff sometimes doesnt' even last a few weeks here.

If the self-stick "moldable" tape does correclty stick to itself and become one piece, and doesnt' separate under heat and moisture conditions, then you're better off using that than electrical tape.

Keep in mind that anything meant to be waterproofing that does not *completely* and *totally* keep all moisture, humidity, water in any form, out, is more likely to keep it in once it does get in there, especially if conditions tend to be humid or wet more often than not. So...if you're not sure it will keep it out, it can be better to leave a way for the water to get back out. This applies more to containers, boxes, covers, etc., but can apply to connections as well.

For power connections, I use Anderson SB50's, and PP75s (with crimped contacts), because they are large enough with sufficient contact spring force to handle the battery wire sizes I need, without the wire tension causing contact misalignment (which is what usually causes PP45 failures resulting in heating/melting/arcing), and even though they are completley open connectors I don't end up with connection problems on them even when they get wet, because the water can get right back out of them and dry right out. For ones on the bottom of the trike where they can get submerged in flashflood waters (rare, but has happened more than once riding home from work in summer storm weather), I can fill them with dielectric grease to keep water out...but I have to refill them periodically because when its' hot enough out there, the road surface a few inches from them is even hotter (I think you could cook things on it :shock:), and vibration can shake the now-much-less-viscous DE grease out of the shells. :( It's still coating the contacts, shells, and wires, but isn't filling the housings anymore.


FWIW, in general, poorly made housings or contacts (which have happened with various "clone" connectors; hard to know which are actually genuine these days) can simply fit so badly together that they arc under sufficient current draw at a high enough voltage. It's one of the major contributors to problems in all sorts of wiring in all sorts of electrical systems. Doesnt appear to be your problem, but something to keep in mind when buying connectors.




BTW, in DC there is no hot and neutral, that's only for AC wiring like in household/building/etc stuff. ;) In DC, it's positive and negative, for that specific set of wires. Or positive and ground (when not referring to something earth grounded, though, probably technically incorrect, it's still commonly used).
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Jan 18 2022 10:10pm

john61ct wrote:
Jan 14 2022 9:05pm
Are you sure your "peak" power drawn did not exceed the connector rating for more than a few seconds?

Put an Anderson SB350 on there

or even a welding style 500A if you think needed

for testing with a logging shunt based wattmeter capable of say a 1000A scale, to check.
In terms of wattage this bike can peak around 3.6 kW, but exceeding 90A for a 52V/1,1kW e-bike seems insane.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Jan 18 2022 10:21pm

amberwolf wrote:
Jan 15 2022 1:03am
bombadero wrote:
Jan 14 2022 8:39pm
I'ver attached a photo of the connection in question.
20220110_161809.jpg
That image shows arc residue, commonly referred to as KFF (Kentucky Fried Finger) on the female connector face, as well as down inside the shell of the male conector.

That usually means the connector was not fully seated, or at the least that it was not making an actual connection at the barrel contacts, but I see no sign of it inside the barrels or on the pins, so that's probably not it.

...

BTW, in DC there is no hot and neutral, that's only for AC wiring like in household/building/etc stuff. ;) In DC, it's positive and negative, for that specific set of wires. Or positive and ground (when not referring to something earth grounded, though, probably technically incorrect, it's still commonly used).
Thanks for all of your advice and knowhow; this bike is really my first major electrical project and it's been a hands on and sometimes hair-raising learning experience to be sure. I'm also excited that I earned the right to use this emoji: :kff:. I didn't know what is was referring to :lol: . Now I do.

It seems like a dry socket from the solder breaking in cold weather or damage to the 'snubber' as a Grin tech called it in response to an email (the little nylon sheath on the XT-90 that clicks into place behind it to provide some insulation) during the aforementioned accident. It isn't clear in the photo, but it was chewed up and at an angle and that's the side the hit the pavement when I tacoed my rear wheel.

I now have to think carefully about replacement. The XT-90's seem fine, but having so many connectors in my dual motor set-up means there are multiple points of failure if they get damaged in an accident. Then again, if I hard wire everything together with solder joints, that makes life harder for potential maintenance. For instance, if either the rear battery or rear controller, which is a BaseRunner in a Reention casing for the rear battery inside a rear rack and over my rear wheel, are soldered in, and then my front battery is also soldered in, then if I ever need to remove that rear Reention, which I've had to a couple times, I also need to remove the front Reention from the bike, which is kind of a big mess. I think it'll probably be fine with XT-90's, some shorter wires, another pass at cable management plus the liquid tape and grease, but Grin called that in question so I'm second-guessing it now. Another factor is that an accident was involved, which has to do with the fact that I live in a hilly city and was towing a heavy trailer, which jack-knifed when I had to brake very suddenly going downhill. I've ordered parts and am in the process of devising a way to modify my trailer to fit it with overrun brakes which will allow me to avoid that circumstance in the future.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Jan 19 2022 2:11am

bombadero wrote:
Jan 18 2022 10:21pm

Thanks for all of your advice and knowhow; this bike is really my first major electrical project and it's been a hands on and sometimes hair-raising learning experience to be sure. I'm also excited that I earned the right to use this emoji: :kff:. I didn't know what is was referring to :lol: . Now I do.
Many people learn what it means by connecting things wrong, and actually getting injured fingers (or worse), so consider yourself lucky. :)

Iv'e found that even if I draw up a diagram of how I intend to wire a bike/etc., I may find practical challenges while actually doing it that change my plans, and I may ahve to stop in the middle and rethink it. :( Sometimes I end up undoing a fair bit of work to redo it when I realize there's a better, more practical, safer, or whtever other way. Sometimes I just sigh and keep going because I'm too tired to care at that point. ;)


It seems like a dry socket from the solder breaking in cold weather or damage to the 'snubber' as a Grin tech called it in response to an email (the little nylon sheath on the XT-90 that clicks into place behind it to provide some insulation) during the aforementioned accident. It isn't clear in the photo, but it was chewed up and at an angle and that's the side the hit the pavement when I tacoed my rear wheel.
Hmm. I must have missed that in your original post...but I just re-read it and still don't see a reference to a crash or impact damage, etc; I'm tired so maybe I missed it? This type of information may not be relevant to this case, but sometimes it is, and is usually good to have in troubleshooting (because crash damage can be much more extensive than is realized at the time, even if it's a minor crash with little obvious damage...I've experienced this before).

Solder won't break in typical cold weather (not counting arctic/antarctic conditions :lol: I have no experience in those), unless it isn't soldered correctly to start with. The most common failure of solder I see is with cupped-contact connectors, especially for large-signal stuff, like the XTs; the cup doesn't fill / isn't heated sufficiently, then the heavy wire draws out heat as fast as it's pumped in by the iron, and the solder just wicks up the wire instead of filling the cup and voids in the wire/cup connection, so very little is actually holding the wire to the cup. Or the wire isn't inserted into the cup, just soldered to the rim of it. That last one is the one that most typically results in the wire breaking off the contact.

(FWIW, I used to do a whole lot of solder-cup style RS232 25-pin modem / terminal cabling for my dad, long ago...quickly discovered crimped connections were much better, more reliable, and easier to do right every time once I'd had enough practice with the correct tools, if not quite as fast to accomplish initially.)

WIres that are soldered do break from vibration, at the point where the solder has wicked up inside them and meets the bendy part of the wire, if it's all not supported by something else sufficiently to prevent that. But tha'ts not the problem you saw, AFAICT.

Wire and connector damage from impact, etc. can certainly happen; I try to make sure all my wiring is protected by the frame, etc., wherever possible.
I now have to think carefully about replacement. The XT-90's seem fine, but having so many connectors in my dual motor set-up means there are multiple points of failure if they get damaged in an accident. Then again, if I hard wire everything together with solder joints, that makes life harder for potential maintenance.
Yes, it is a dilemma. There have been a number of structural repairs I've had to make over the years that having connectors either did or would have saved a lot of rewiring trouble, depending on whether i had them or not in that setup.

My call on my own stuff is that if there is more risk of a connector causing me a problem than being a help, I don't use one. ;) SOmetimes it takes the experience of having one to decide not to use one, and vice-versa, in specific situations.

So for the most part I use them only for things that have to be disconnected to work on them (like hubmotors), or for safety reasons (like batteries, since mine are not in removable cases with their own connectors). And then I try to choose connectors that fit all the requirements for that signal type and it's location/environment, but that is also a compromise, because of budget, primarily.

Some of the best connectors for specific uses are expensive to use, because not only is the connector expensive, but it and/or it's contacts require special tools to correctly wire them and set them up (generic tools may not do the job exactly right, and that may matter a great deal).

So for batteries, I have anderson SB50s on them, as close as practical to them. Motors, it depends--PP75 or Anderson SB50 for the big stuff; PP45 on the small ones. Then just ensuring alignment of wiring doesnt' allow torsion or tension or bending of the cabling to the connectors, so that the contacts cant' be twisted out of contact (whcih is where most people have problems with andersons; the rest is using too large a gauge of wire for the housing so the contact can't float inside it, and using too small a contact/housing size for the current).

Small signal connectors I may use JST-SM for if I know there is no chance of environmental contamination...in a few cases I've used them anyway, and had no problems, but other times I do have problems.

My SB Cruiser has quite a few around the Cycle Analyst, because it came with them and so I used them on my wiring to match...but there are some that are beginning to cause me problems, and I will probably end up direct-wiring those, with soldered and heatshrinked connections, just because it's easier than finding and installing better connectors at this stage, and I'm unlikely to need to swap the CA out at any point. (I say that, and of course something will force my hand :lol: )

IF I were to use ebike-specific connectors on small signal stuff, I'd probably go with the Julet / Higo types, because they are common now, and seem pretty water resistant, but they have to be installed along with "pigtails" of wire already mounted to them, by opening up whatever devices they are connecting, so if you have longer distances to cover like I do on the trike, you may have to use very short pigtails (or even panel-mount versions on the device itself) and then use extension cables to make the actual interconnects. (which can also make the removal / installation of individual devices easier, because all the wiring harness stays in place, and only the devices come off).



I think it'll probably be fine with XT-90's, some shorter wires, another pass at cable management plus the liquid tape and grease, but Grin called that in question so I'm second-guessing it now.
The tape is a potential problem because it could seal in whatever moisture does get in there. It also prevents you from just quickly checking a connection for whatever reason.


Dieelectric grease is messy, but doesn't really hurt anything to at least put on the conductive surfaces, as long as the connector contacts are sufficiently "wiping" style with sufficient contact force to push any of it out of the way that could otherwise interfere with good contact. (the andersons do this) If they don't really do contact wiping well, like some bullet-style and some pin-and-barrel-style connectors, and various old blade-styles, then the DE grease could cause problems of another kind while preventing humidity/moisture-related ones.

There are a number of kinds of it, for different situations (like if you have different metals in contact, etc).

XTs are good enough if they don't have problems in your usage normally, but you have to consider the potential failure modes, such as severe overheating (like from arcing) potentially causing a desoldering event, which could cause further arcing etc. That sort of event probably wouldn't happen unless something serious was already wrong with the connection.... They work fine as high-current connections in most situations and usage scenarios that they are used in, and yours is probably one of those.

If whatever connector you use has a plastic casing that isn't tough against bending, impact, fracture, etc., and heat resistant, etc., then it may not be as suited for mobile environments like ours as others. Another reason I use the SB50s for power/current connections is their shells are really tough. (PP45s, not so much, even PP75s that use the same contacts as SB50s aren't quite as durable, at least not the ones I've had).

Some of the connectors I've used, especially solder-only ones, have had shells that damage ro crack easily, or deform too easily under sustained heat, especially if the wires are under torsion or bending near the shell. I haven't used XTs in this situation yet, so don't have personal experience with their ability to handle it. Plenty of people out there using them, so if it is an issue, it should be possible to find posts / webpages about problems with them.
Another factor is that an accident was involved, which has to do with the fact that I live in a hilly city and was towing a heavy trailer, which jack-knifed when I had to brake very suddenly going downhill. I've ordered parts and am in the process of devising a way to modify my trailer to fit it with overrun brakes which will allow me to avoid that circumstance in the future.
YEah, that's a good idea; there are a number of ways to implement something like that; a few are discussed in my various trailer threads. I never have put them on, but my biggest trailer needs that (could really ahve used them when towing the piano home, and some of the biggest loads of dog food, which were comparable in mass).

[/quote]
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Jan 19 2022 9:36pm

Hmm. I must have missed that in your original post...
Sorry, been talking to Grin and Surly about that incident, I guess I didn't mention that here. Yeah, had an accident that tacoed my rear wheel. I had the same rim lying around in an older version, so was able to rebuild it pretty easily just swapping the spokes over then truing/tuning/tensioning.

I also misspoke a little describing the accident as the trailer jackknifing; it was more of a reverse-jackknife, i.e. the trailer stayed straight but the bike jackknifed once I had dropped under 10 MPH or so. The brakes on my bike are very good, they stopped me, who weighs about 185, my 100 lb. e-bike, my 55 lb. dog, a 40 lb. Surly trailer with maybe 40 lbs. of stuff on it and a 30 lb. Burley trailer (I sometimes chain them together in a train), so about 450 lbs., in the width of 1 road lane from 20 MPH. The problem is that there was a downward grade, and no brakes on the trailer. So the bike stopped, but the momentum of the trailer wanted to keep moving forward and it just slammed my ass into the pavement. I've had a couple of similar accidents in slightly different circumstances that were not as severe. Hence the project to put overruns on my Surly Bill. It will be a bit of a project because I will need to modify the bash guards by cutting a gap into them to make space for the Magura HS33 hydraulic rim brakes I bought, and welding a custom steel bracket to attach to the braze-ons on the bash guards and the brakes to mount them on the tires. For some reason I hadn't put together that the two incidents were probably related until yesterday. They still might not be, it could be a coincidence, but Occam's razor dictates that the crash probably damaged the connector.

For the wire harness, I've decided on lengthening the battery and controller pigtails in the rear so they feed between the chainstays and their connectors rest behind the seat tube, and I'll tape them together to keep a tighter profile. That will allow me to unhook the rear for continued convenience for maintenance. I'm also keeping the connector for the front controller; but then I'll solder together a custom parallel 'X' harness, rather than two Y-splitter cables, that also hard wires into the front battery cable from the down tube. So I'll be down to 3 connections from 5, with two of them protected by the chainstays and the seat tube, and the least vulnerable one exposed a little bit on the side of the seat tube down near the BB. Being near the cranks and pedals though will offer a lot of protection in a fall.

I also ordered a sidecar frame kit from Scandanavian Side Bike via their only NA distributor in Big Sur, a fancypants e-bike shop called Mad Dogs & Englishmen. They sell actual sidecars in fiberglass and carbon fiber as well, but they are a bit pricey and don't really meet my needs, plus you need to physically lift your dog into it. I'm strong enough to lift my 55 lb. dog into a sidecar, but why? So I got just the frame kit; I'll assemble and install it and see how it feels, then bash something up out of marine board and extruded aluminum DIY stock.

The sidecar will allow me to do my smaller cargo runs or dog park runs without a trailer and sidestep all of the issues I have experienced over the years with both trailers and panniers. Then when I still need it for larger runs, I'll have the Surly Bill but with overrun brakes to supplement painful experience in assisting me in staying safe. I like bike racks, panniers and trailers, but I've paid enough dues with all of them when hauling large and/or bulky and/or heavy items, and the sidecar solves problems with all of them. It's pretty nicely engineered so that it moves independent of the bike so you can still do full leans for turns (although I suspect right turns are a little more restricted), and it removes very easily from the bike with two short quick-release skewers from a custom bracket that goes on your chainstay. It will also add stability to the bike and make falls pretty much impossible. The outboard wheel doesn't have a brake, but doesn't really need one since it's attached to the bike. The tendency of the bike is to fall to the left, but the sidecar is on the right. This bike also has kickstand issues, which the sidecar somewhat addresses as well. On the actual sidecar I will fab, I plan to add a locking door in some fashion, maybe a ramp for the dog (at the very least, it is low enough to the ground even without a ramp), and a locking/folding top hatch that will let me convert it from locking cargo storage to open-air sidecar for the dog. Of course I'll have to invest in some cool goggles for my dog :D.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Jan 20 2022 12:22am

bombadero wrote:
Jan 19 2022 9:36pm
I also misspoke a little describing the accident as the trailer jackknifing; it was more of a reverse-jackknife, i.e. the trailer stayed straight but the bike jackknifed once I had dropped under 10 MPH or so. The brakes on my bike are very good, they stopped me, who weighs about 185, my 100 lb. e-bike, my 55 lb. dog, a 40 lb. Surly trailer with maybe 40 lbs. of stuff on it and a 30 lb. Burley trailer (I sometimes chain them together in a train), so about 450 lbs., in the width of 1 road lane from 20 MPH. The problem is that there was a downward grade, and no brakes on the trailer. So the bike stopped, but the momentum of the trailer wanted to keep moving forward and it just slammed my ass into the pavement.
Yeah, I had a similar crash a decade or so ago, maybe more, with Nana (my first St Bernard) in my Mk I dog kennel trailer viewtopic.php?f=2&t=18671 pulled by the DayGlo Avenger in it's front hubmotor configuration https://www.endless-sphere.com/forums/v ... hp?t=15570 . I forget exactly what the sequence of events was, but the bike crashing pulled the trailer over too because of how high my hitch was...and things could've been much worse than they ended up. It was no fun at all, though. I learned quite a bit from that crash. :oops: Nowadays my trailer designs are quite different....

I've had a couple of similar accidents in slightly different circumstances that were not as severe. Hence the project to put overruns on my Surly Bill. It will be a bit of a project because I will need to modify the bash guards by cutting a gap into them to make space for the Magura HS33 hydraulic rim brakes I bought, and welding a custom steel bracket to attach to the braze-ons on the bash guards and the brakes to mount them on the tires.
You should post pics of your bike, system, etc.; they would be interesting and probably useful to others to see the setup, for those that use their bikes in similar ways.

For some reason I hadn't put together that the two incidents were probably related until yesterday. They still might not be, it could be a coincidence, but Occam's razor dictates that the crash probably damaged the connector.
If it didn't damage it, it probably exposed an existing problem, one way or another. Either way, you know about it and can fix it. :)

For the wire harness, I've decided on lengthening the battery and controller pigtails in the rear so they feed between the chainstays and their connectors rest behind the seat tube, and I'll tape them together to keep a tighter profile. That will allow me to unhook the rear for continued convenience for maintenance. I'm also keeping the connector for the front controller; but then I'll solder together a custom parallel 'X' harness, rather than two Y-splitter cables, that also hard wires into the front battery cable from the down tube. So I'll be down to 3 connections from 5, with two of them protected by the chainstays and the seat tube, and the least vulnerable one exposed a little bit on the side of the seat tube down near the BB. Being near the cranks and pedals though will offer a lot of protection in a fall.
Less connections is good for reliability. :) You can add a "bash guard" for the cabling itself, if there is anywhere to mount one to (or you can weld mount points to the frame for it).

One thing I learned about routing wires between frame/etc sections, from others and my own projects, is that if there is a chance that any common kind of crash could pinch the wiring enough to short current-carrying wires together, they ought to be fused at their sources before that point. (small signal wires can be fused, too, at their sources, but it would usually have to be a very very tiny fuse to let it blow before the supply for it is damaged in a shorting event, and a short of a low voltage signal to a high voltage doesn't usually blow the fuse before damage occurs (like hall signals and phase wires).

Also, that if it is warranted, protective brackets/etc can be created around the wiring to help prevent such an event (usually it's not warranted...but the fuses always are).

I also ordered a sidecar frame kit from Scandanavian Side Bike via their only NA distributor in Big Sur, a fancypants e-bike shop called Mad Dogs & Englishmen. They sell actual sidecars in fiberglass and carbon fiber as well, but they are a bit pricey and don't really meet my needs, plus you need to physically lift your dog into it. I'm strong enough to lift my 55 lb. dog into a sidecar, but why? So I got just the frame kit; I'll assemble and install it and see how it feels, then bash something up out of marine board and extruded aluminum DIY stock.
That sounds interesting; you should post that project, too. :)

On the actual sidecar I will fab, I plan to add a locking door in some fashion, maybe a ramp for the dog (at the very least, it is low enough to the ground even without a ramp), and a locking/folding top hatch that will let me convert it from locking cargo storage to open-air sidecar for the dog.
Those sound like good ideas, and I'd like to see how you go with it. If I were to do a sidecar (rather than the trike I have), I'd set it up with a fold-down ramp, possibly at the side so it would also be the sidewall of the car, and could be flat and taller if necessary, and still not interfere too much with aerodynamics (if I did that at the back or front it would be a wall against the wind). (how this would work around the wheel I would have to figure out). Then dog can walk up it even when they get old enough to have trouble with that sort of thing (like Yogi and others I've had), or if they develop some other problem that causes the same result (like Tiny and Kirin did).

I'd also have a wire-mesh (like a dog crate) style of cover I could fold over it and secure, for the few dogs that aren't completely capable of just enjoying the ride, and decide they have to investigate stuff we pass while we're still moving. ;) It could also be used to secure cargo inside, so I could make more than one stop on shopping trips (like I can do with the trike) and not have as much risk that things i got earlier won't be there when I come out of the next place.


The sidecar will allow me to do my smaller cargo runs or dog park runs without a trailer and sidestep all of the issues I have experienced over the years with both trailers and panniers. Then when I still need it for larger runs, I'll have the Surly Bill but with overrun brakes to supplement painful experience in assisting me in staying safe. I like bike racks, panniers and trailers, but I've paid enough dues with all of them when hauling large and/or bulky and/or heavy items, and the sidecar solves problems with all of them.
That's what drove me to a trike (along with my own problems as I age). (plus, with the canopy and rack, I can carry long cargo that I couldn't even with a sidecar or a more typical trailer). :) I've thought about sidecars for a long time, and I love the idea, and read up on each new sidecar project I come across...I've just never convinced myself to build a bike with one yet. :oops:
The outboard wheel doesn't have a brake, but doesn't really need one since it's attached to the bike.

Easy enough to fix, if you needed to (depends on the sidecar frame design, but a disc brake caliper should be easy to implement; I've found the Avid BB7 MTN with a 203mm rotor is more than sufficient even for my SB Cruiser trike. (I've used rim and disc, and really only went with disc at this point because it meant a bent rim wouldn't take my brakes out, and it was much easier to add a disc caliper mount to the fork I had to build than getting rim brake bosses installed correctly).


Of course I'll have to invest in some cool goggles for my dog :D.
I only got one, Teddy, to wear goggles and leave them on:
Image
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Jan 20 2022 3:45pm

I only got one, Teddy, to wear goggles and leave them on:
Lol, you made my wife laugh. It's hilarious that his head is so big he can just wear human safety goggles. Speaking of which, my wife is disabled, and we actually have a trike as well. It's an HP Velotechnik Scorpion FS that we bought unassisted back in 2007 so they could ride, because they can't balance on a regular bike. Around 2018 I had it retrofitted with a factory pedelec kit, and it's great. I tried it for hauling, but the Go Swissdrive kit that was stock at the time (HPV've had to switch to another system since as GSD went under), while capable, has a tendency to overheat on hill climbs, especially on warm days, if I'm hauling more than 100 lbs. or so. Too bad, because otherwise it's a great system. It's extremely torquey for a 250W nominal system and often beats class III bikes off the line even though it's only a class I, and it has regenerative braking as well. The class III riders, roadies and single speed hipsters get really upset when I overtake them, usually from behind because they crowd you at lights thinking you'll be super slow. The class III's overtake me again of course, but the roadies and singlespeed hipsters end up playing endless leapfrog, getting more and more frustrated at every light, sprinting to keep up the whole time (having been perfectly satisfied with going 15 MPH before I passed them). It's annoying but also really funny. Really sucks that some good companies like GSD and BionX didn't make it.

Some day I plan to convert the trike to another Grin kit. Still won't probably use it for hauling, but it will have a throttle which will help for starts and when my wife's left leg loses motor control, which sometimes happens if they get tired towards the end of a ride. The SPD pedals and recumbent position mean they can keep going anyway, but even with the 250W GSD kit, it's a struggle with only one leg, especially since less leg torque=less assist. We used to run our other dog, an ACD on a dog leash next to it--they're great for running dogs--but she's too old now. So in the meantime, the trike is sadly in a 2nd fallow period hanging motionless from a bike hoist in my garage. The first fallow period was when we moved to San Francisco and they were intimidated by the hills (fair, as even the unassisted version of the trike weighs 55 lbs.).

For now, they ride the "first" e-bike we bought (it was an trade-up for a Kalkhoff bike that was actually our first e-bike, but it couldn't hang in SF--motor burnt out several times), a 2017 R&M Homage GT Nuvinci. It also serves as back-up cargo bike for my current ride as I work out these kinks. I rode the R&M for three years, and it's an amazing, if underpowered for my purposes, little machine. Since my current bike weighs about 100 lbs., the 65 lb. R&M feels like riding a track bike now :lol:. I had to ride it for a UPS store run the other day (down a hill with a 29% grade and back up via a more circuitous route), and couldn't believe how light and maneuverable it is. It is of course much slower, as my DIY can hit 34 in a sprint, and my typical cruising speed is about 32. 20 feels like walking to me now. And of course it's harder on the hills; but still a great bike.

I'll try to post some pics of the repairs, trailer mod and sidecar project. It will take a bit for the latter two. For the trailer mod, I basically have to re-learn welding. I bought a little multi-process TIG/MIG/stick garage welder a while back for various projects, but I haven't done anything with it yet, as I haven't welded for several years. I have to get Argon, which I've never had to do myself (learned to weld in art school, and the studio tech took care of all of that stuff), grab some scrap stock and play around a while to get back into the TIG groove. For the sidecar, I'm taking a constructed approach, more like you're MK I trailer, using DIY extruded aluminum framing stock. There isn't much to that technically besides precise cuts on my metal saw, but there will be a lot of parts and planning.

In the meantime, here is the bike from October of last year before the accident. He goes by "Bruce the Moose", aka "The Bat Bike":
20211021_192026.jpg
20211021_192026.jpg (3.3 MiB) Viewed 589 times
Overall it's a solid bike, built on an old 1990's stump jumper frame. It's a gonzo build, so he has a few quirks, but amazing to ride. It looks even stranger now, because one of those quirks is loss of traction on the front wheel under certain conditions, so I added spoke weights that motorcycle riders use to balance spoked wheels to add some additional weight to the wheel. I plan to finally wire up his 12V horns as well, and rebuild the rear wheel yet again on a slightly better/wider DM24 rim from the current DM18.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 11 2022 2:43pm

Finally made some progress on the bike, having finished up a new wiring harness. I ended up only removing a single connector, where the two previous Y-splitters hooked up, due to convenience for maintainence. I ordered a few new splitters from Grin, removed their APP connectors, soldered them together, shortened each arm of the resulting 'X' to a custom length based on what it needs to connect to, painted the solders with liquid tape, liberally packed the connectors with dielectric grease, heat shrunk the exposed wire insulation and painted the heat shrink junctions with liquid tape as well. It is still not technically water tight, but the liquid tape and grease should be able to handle any small amount of moisture that gets in.

I had previously also torn down my rear battery cage assembly and re-soldered the battery and controller cables. The batteries and both battery outs on the harness are equipped with XT-90S plugs for spark protection and also to prevent partial voltage. I also shortened and re-soldered the long cable coming from the front controller. All XT-90S's read a steady and smooth 0.56 on the DC 10 mA setting on the multimeter.

Unfortunately after plugging everything back in the Cycle Analyst appears to still be dead. It came on with a bright screen but no display for a bit, then unexpectedly shut off. I turned it on again and the screen barely lit up, with the light level randomly fluctuating at a very low level of brightness. The controllers receive power directly from the batteries, and are upstream from the CA3, so technically the other culprit could be the main controller. However, the front controller, which is slaved to the rear controller, is a Phaserunner that sits outside on the frame, and its red status light was solid, indicating it is powered and running properly. I cannot get status on the rear controller, a Baserunner, because it is inside that rear battery interface housing. There is a data cable and app I can use to inspect it, but it won't work without the CA3. The CA3 is also a much cheaper part than the Baserunner, so I ordered a new CA3 to swap out first. If the problem still occurs, then the likely culprit is the Baserunner. It seems based on the symptoms like the display on the CA3 is damaged however, so it will need replacing either way. It just occurred to me I could try to run the data cable to the CA3 anyway and see if I can access it. The new CA3 should arrive in a couple of days. I know the batteries are good based on the multimeter readings and the fact that they charge and are recognized by the charger. The motors are unlikely to have sustained serious damage, and were working right before the bike died, and a even a little after the blank screen occurred, via the throttle.

Posting some pics of progress (in no particular order). I took the old cable apart and noticed that the positive connection was a little loose on the affected connector. Also included are some pics of what the new cables look like, and an initial assembly on the bike. The colors on my phone for some reason blow out the colors, which are actually darker than shown here, especially the greens, which are a much darker olive drab color. It looks pretty hairy, but believe it or not it's far cleaner than the previous iteration which had two fat, messy bundles of cables wrapped in neoprene sleeves fore and aft of the seat tube, in addition to neoprene sleeves on each of the cables coming out of the rear assembly. Now there is only one sleeve behind the seat tube. I also eliminated a large sleeve to the left of the down tube in favor of some plastic channel. So I've eliminated 4 bulky neoprene sleeves, and there are far fewer cables in front of the seat tube. The harness has to make some pretty acrobatic contortions to plug into everything in the right way, but the lengths turned out perfectly for that.

Making extensive use of Rovaflex softties helps a great deal, even if it looks a little messier. It would look less messy if they were all black, but the black ones are hard to find in the States and usually sold out in the more useful sizes in the EU and UK. However, I discovered at my local gardener's supply that they are re-marketed under a brand called Peacock in the States, colored green and sold as garden stake straps, so I snatched a bunch up. They don't look as nice, but they work fine. They are literally the same product in green instead of black and in a different package. They're much better than zip ties in many cases, as they are reusable, and unlike reusable zip ties, which are still plastic, they won't get brittle over time. I use especially large ones as cargo straps now. I also included an image looking at the cable harness from the perspective of my front derailleur, with which, remarkably, it does not interfere. The inside of the derailleur cage just barely kisses the cable insulation when shifted into the smallest chainring.
20220210_193729.jpg
20220210_193729.jpg (1.92 MiB) Viewed 481 times
20220210_192830.jpg
20220210_192830.jpg (3.04 MiB) Viewed 481 times
20220210_192815.jpg
20220210_192815.jpg (2.43 MiB) Viewed 481 times
20220210_182431.jpg
20220210_182431.jpg (3.11 MiB) Viewed 481 times
20220210_182421.jpg
20220210_182421.jpg (3.19 MiB) Viewed 481 times
20220201_101228.jpg
20220201_101228.jpg (3.03 MiB) Viewed 481 times
20220126_173423.jpg
20220126_173423.jpg (1.6 MiB) Viewed 481 times
20220126_173342.jpg
20220126_173342.jpg (1.6 MiB) Viewed 481 times
20220126_173318.jpg
20220126_173318.jpg (1.96 MiB) Viewed 481 times

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Feb 11 2022 6:54pm

bombadero wrote:
Feb 11 2022 2:43pm

Unfortunately after plugging everything back in the Cycle Analyst appears to still be dead. It came on with a bright screen but no display for a bit, then unexpectedly shut off. I turned it on again and the screen barely lit up, with the light level randomly fluctuating at a very low level of brightness.
That sounds like the main voltage regulator in teh CA is damaged. If so, it can still be powered by a lower voltage into the low-voltage regulator. I don't remember what the specific wiring is, but one of my Cycle Analyst repair threads should have a post by Justin_LE detailing that.
I cannot get status on the rear controller, a Baserunner, because it is inside that rear battery interface housing. There is a data cable and app I can use to inspect it, but it won't work without the CA3.
That's very strange, because the CA doesn't communicate with anything, so the data cable/app should go directly to the BR. It might go thru a cable harness that also attaches to the CA, but there is nothing in the CA that would do anything to facilitate a data connection to anything (except itself).

It just occurred to me I could try to run the data cable to the CA3 anyway and see if I can access it.
If the CA's regulator is damaged (likely) then it wont' be accessible via the serial port either.
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 16 2022 2:09pm

Bruce lives!
20220215_144120.jpg
20220215_144120.jpg (2.46 MiB) Viewed 427 times
So it seems my gamble to replace the cheapest of the two possible components that might have failed, the CA3 and the Baserunner, thankfully paid off. I took the old CA3 apart to inspect it. I didn't go very far, just unscrewed it and removed the display, leaving the daughter board soldered in place at the GPIO pins. Prying it apart, Q1 is the only component that appears it might be damaged, although I'm not sure because I don't have a lot of experience hacking around with PCBs directly, and the yellowish coloration on the traces around it are very subtle. But the solder/traces/pins and what have you around it don't look silvery and shiny like the other components. In this post you and Justin identified it simply as the 'main transistor' for the board, so I'm not sure if it is the equivalent of a VRM on a motherboard as you mentioned. I suppose it would be fairly cheap to try to obtain a replacement for that transistor and solder it on to see if that revives the CA.
20220215_144303.jpg
20220215_144303.jpg (1.87 MiB) Viewed 427 times
The new CA3 unfortunately did not come pre-programmed with the configuration required for my bike, so I need to do a dive into restoring the config today. I thought I had remembered the correct settings for the battery, but it's doing what it normally does when the settings are incorrect, namely showing 0% charge even though both batteries in this system are fully charged. I'm using 19 cells, 39.6 Ah and 'Battery A Only', which I had remembered as the correct settings for this 52V dual-battery bike, but it doesn't seem to be working correctly. I may or may not have downloaded a copy of the old settings, I have to hunt around on my 'daily driver' PC for that.

I also did a quick throttle test and the rear motor, a GMAC, did not seem to respond even though it is the motor belonging to the master controller. The Grin All-axle in the front seems fine. However, as I said, the config is wonky at the moment and the rear motor has always had a higher threshold to kick in, so I'll have to get it on a stand/back on the road to test it out. It's also the motor on the drivetrain, so it has to overcome more mechanical resistance, and being in the rear also has to overcome more mass, and I was holding the front wheel off the ground. It's probably fine, fingers crossed. Motors seem to me like the components that would be least likely effected by an issue like this.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 16 2022 4:01pm

For my battery config, I was simply using the wrong number of cells. I had entered 19, but consulting the product page I verified it's a 14S6P battery, so changing it to 14 cells fixed that issue.

My torque sensor is an E-rider, which should have 36 poles according to Grin's documentation, but when I attempt to set it to that, the CA3 sets it back to 32. I think it needs to go back to the default of 23. I set the R-shunt value to 0.5 mOhm as I have a Baserunner/Phaserunner combo, but the CA3 overwrites it to 0.767 for some reason. I'm going to take another pass via the desktop software to dial in my presets, and perhaps I will be able to overwrite R-shunt resistance there.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Feb 16 2022 6:12pm

are you in high range or low range mode? the low mode will prevent using a shunt below a certain resistance.

the ca only allowed 32 poles in earlier versions of the fw, but i think there is a new one that allows 36.

23 poles sounds like the typical number used for the wheel speed sensor (rather than the torque sensor) when using a hall signal from a typical dd hubmotor.
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 16 2022 9:53pm

amberwolf wrote:
Feb 16 2022 6:12pm
are you in high range or low range mode? the low mode will prevent using a shunt below a certain resistance.
:bigthumb:
That did the trick. I verified it as well from the CA 3.1 manual, page 29:
Most high current shunts will be less than 0.8 mOhm and the Cycle Analyst must
be set to High Range (0.1A) mode to accomodate them. The High Range mode
will show power in kilowatts instead of watts, and current to the nearest 0.1A
instead of 0.01A.
Thanks for that.
amberwolf wrote:
Feb 16 2022 6:12pm
the ca only allowed 32 poles in earlier versions of the fw, but i think there is a new one that allows 36.

23 poles sounds like the typical number used for the wheel speed sensor (rather than the torque sensor) when using a hall signal from a typical dd hubmotor.
Yeah I'm confused what to enter here. 23 is the default most people recommend, but the Grin/manufacturer docs both state the ERider T9N has 36 poles. My CA states it has v3.14, which I thought was the latest version. Is there a beta version with that feature?

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 16 2022 10:28pm

I downloaded and flashed firmware version CA3-2b1, which let me select 36 poles for the PAS (and still the default of 23 for the speedo, which tops out at 30--not sure what to set here, so left at default). It has a couple of neat new features as well, like backpedal regen. It also has a setting that pertains to my build, but I'm not sure what to select. You can select a setting to double amperage for two motors, otherwise scale as normal based on the Rshunt, which in my case is 0.5 mOhm. I assume if I select 'yes', I need to set Rhunt back to 1.00 mOhm, but since it's new I want to verify that before I change from 'no' and 0.5 mOhm to 'yes' and 1.00 mOhm.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Feb 17 2022 12:07am

bombadero wrote:
Feb 16 2022 10:28pm
(and still the default of 23 for the speedo, which tops out at 30--not sure what to set here, so left at default).
you use the number of poles actually on your speed sensor.

if you're using a separate wheel sensor with magnet on spokes, put in number of magnets you used. same for if you're using a speed sensor built into a geared hubmotor; however many magnets that sensor uses, which is not the nubmer of magnets in the motor; it's separate. like a common bafang is 6, a dapu on a pedego is 1, etc.

if you're using a hall sensor in a motor then you use the number of magnets in the mtoor divided by 2. thats ofetn 23 for dd hubs, but not always. may be very different for geared hubs. you can count them by watching teh speed sensor pulse indicator in ca setting screen if your version has one while manually rotating the wheel very slowly, or your baserunner setup sw might be able to tell you how many poles, directly or with an indicator on screen for each pulse a hall sensor gets. or with a multimeter on any of the halls sensors, watch for teh change in voltage; each change is a pole.


It has a couple of neat new features as well, like backpedal regen.
that sounds useful for me on my trike, except it would cause me problems when backing up if there is no external disabling switch available.

It also has a setting that pertains to my build, but I'm not sure what to select. You can select a setting to double amperage for two motors, otherwise scale as normal based on the Rshunt, which in my case is 0.5 mOhm. I assume if I select 'yes', I need to set Rhunt back to 1.00 mOhm, but since it's new I want to verify that before I change from 'no' and 0.5 mOhm to 'yes' and 1.00 mOhm.
that i don't know. i would personally just use teh actual rshunt, if you only have one shunt in the system that is being used for one ca but measures all current in whole system out of battery(ies).

i am not even sure what benefit such a setting would have. i'm too tired to think clearly enough to figure it out but there must be some reason or they wouldnt' have wasted time coding it, especially when there are other features people have asked for that still aren't in there but could be easily enough.
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 17 2022 4:38pm

Took a short test run today and mysteriously the torque sensor isn't picking up in any preset :?:. I verified the proper settings in the Grin documentation for the ERider T9N, which are set correctly. I fixed the speedo; the correct number of poles for the GMAC 10T is 6. Throttle seems to work fine.

Not sure what to do about the T9N. It could be the cable (I noticed one of the pins is a little bent and loose), the config or the sensor itself, possibly another casualty of the arc fault. The barrel connectors are solid and nothing is pinching the T9N wire, the extension wire or the wire coming out of the CA3. I will start by rolling back to CA3-14 firmware and looking into the config. Failing that, I have another ERider lying around because I originally ordered the wrong one (68mm and I needed 73mm I think), so I should have another extension cable hopefully unless the replacement didn't ship with one, which I don't recall. If I don't have one, I guess I'll order another and have to wait again for that part to arrive before being able to eliminate the cable as the cause. I hope it's not the sensor, it isn't hard to swap out but they tend to be low in inventory and are pricey. I had to order the last one from Aliexpress and it took several weeks to arrive.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 17 2022 5:20pm

Flash back to CA3-14 didn't work. I couldn't find the old T9N anywhere and then remembered I sold it or gave it away to someone either here or on electricbikereview.com so it wouldn't go to waste. :?

So dead in the water again, apart of course from a functioning throttle. It's fine for it to be a moped for a little while, but it's not nearly as useful for hauling up hills without the addition of human pedal power.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 17 2022 6:10pm

Looks like I have to buy a new sensor. I probed the JST-SM side of the torque cable from behind the connector while plugged into the CA3, and it reads 0.011 A on the DC 10 mA setting on my multimeter when the bike is on; I have no clue if that is a good or bad value. I verified that amps are zeroed; they appear to be almost perfectly zeroed, reading 2.49V and 2.50V. There isn't a great way to prove the HiGo end that I'm aware of, and no way to probe the HiGo on the torque sensor without stripping insulation. Moreover, I searched high and low and no one sells just the JST-SM to HiGo cable, which I'm assuming is manufacturing by Grin, but they don't sell it alone either. There also isn't any real way to differentiate if either the cable or sensor is at fault. The cable appears to be fine visually.

I have connectors and pins, I suppose I can try to re-crimp a new cable end and see if that works. The only other thing I can think to do is tinker with the minimum starting watts. I had set it really low, 25 hW, because the bike is heavy and it is sometimes difficult to start with torque, and I often have to start with throttle when hauling a lot of weight or on hills. I read a number of threads you have participated in, amberwolf, regarding troubleshooting of issues with the T9N specifically, and it seems like there is nothing in the config that is wrong to cause this issue, so it must lie with the sensor or cable. When I'm in the PAS setup screen, the 'P' toggle does not move.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Feb 17 2022 6:16pm

all the sensors i've worked with don't output current, they output voltage. so measuring with a meter set to amps isn't useful. you'd need to use dcvolts instead.

the amps you are zeroing have nothing to do with the torque sensor, those are for your shunt for measuring battery current.

i think it would help you to read the entire ca info page including each fo the menu explanation sections, so you get a better handle on what each screen and function in the ca does what, since you seem to be confusing some of them with each other. ;)
https://ebikes.ca/product-info/grin-pro ... yst-3.html


is yoru erider the one that gives a direct throttle voltage output? if so you can measure it to see if it gives the right range vs torque with a voltmeter; if it is working you could then verify the connections between it and the ca's internal board, perhaps it's a broken wire. if you have a variable voltage supply you could set it to the typical sensor output range and hook it up to the ca's sensor input plug while monitoring the ca setup menu field for it and see if the ca is reading it's plug correctly.

note that the toggle isn't the torque sensor output, it is the cadence sensor. so if you aren't getting anything from either ouptput on the sensor, check to see the ca is outputting the correct voltage on the pas connector's voltage supply pin.

to probe higo stuff you can get extension cables and put them between the source and destination and just cut into the extension to do the probing. can be useful for troubleshooting. that's what i will be doing if i ever get to changing out the cabling/connectors on my trike.

you can set one of the watts-per-something settings (can't recall which one :( ) to a negative value to get more power at startup than later on. tjere is a post or few in the cav3 beta thread discussing that; it might be one of my conversations with justin_le, or someone else's, or both.
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 17 2022 7:19pm

amberwolf wrote:
Feb 17 2022 6:16pm
all the sensors i've worked with don't output current, they output voltage. so measuring with a meter set to amps isn't useful. you'd need to use dcvolts instead.
Good point. Measuring that way, I get 11.4V, which I believe is a little high (supposed to be 10V), but I don't think that should cause much of an issue. I guess that indicates an issue with the sensor itself.
amberwolf wrote:
Feb 17 2022 6:16pm
i think it would help you to read the entire ca info page including each fo the menu explanation sections, so you get a better handle on what each screen and function in the ca does what, since you seem to be confusing some of them with each other. ;)
https://ebikes.ca/product-info/grin-pro ... yst-3.html
I think I actually have a pretty good handle on that, but I've noticed there are some inconsistencies between the manual and the Grin website. The PAS page on the website says starting watts refers to motor power; the starting watts in the manual refers to the minimum threshold of human power. The screen on the CA3 itself refers to HW, however, which backs up the manual.
amberwolf wrote:
Feb 17 2022 6:16pm
is yoru erider the one that gives a direct throttle voltage output? if so you can measure it to see if it gives the right range vs torque with a voltmeter; if it is working you could then verify the connections between it and the ca's internal board, perhaps it's a broken wire. if you have a variable voltage supply you could set it to the typical sensor output range and hook it up to the ca's sensor input plug while monitoring the ca setup menu field for it and see if the ca is reading it's plug correctly.
Probing the TRQ pin gives 0.532V; when I step on the pedal, nothing happens. I have the brakes locked up while I'm doing that, idk if that makes a difference because it's in regen mode; I would think it would still send the torque signal, but maybe it is programmed in firmware not to because you are braking. The other way to code it would be to send the signal anyway but ignore it. The range is supposed to be 1.5-3.5V, but idk if 1.5V is flowing continuously even when there is no torque, or if it's supposed to be down near zero at zero torque, then range from 1.5-3.5V when torque is applied.
amberwolf wrote:
Feb 17 2022 6:16pm
note that the toggle isn't the torque sensor output, it is the cadence sensor. so if you aren't getting anything from either ouptput on the sensor, check to see the ca is outputting the correct voltage on the pas connector's voltage supply pin.
Yes, I think so (see above).
amberwolf wrote:
Feb 17 2022 6:16pm
to probe higo stuff you can get extension cables and put them between the source and destination and just cut into the extension to do the probing. can be useful for troubleshooting. that's what i will be doing if i ever get to changing out the cabling/connectors on my trike.
That's a great idea. I don't have any spare cables atm though.

It seems like the sensor is probably at issue, unless the somewhat high voltage on the 10V pin is a problem--it does seem like it is closer to 12V. Nothing seems to be happening at the green TRQ wire, and that's an output from the sensor to the CA3, correct? It could be a damaged cable, I suppose, but again those are not sold standalone. If the green wire is broken outright, there would be no voltage on that line; a wire with some strands broken I guess would return a partial voltage? But as I said, nothing changes when I step on the pedal. So changing out the connector would be pointless, as that end is fine, and I don't have HiGO extensions to test the HiGo mains, which would take a week or more to arrive after ordering. I do have an old wire harness for our trike has HiGo's on it, I could take a look at that to see if that has the right kind of connector that I could cannibalize, as I kept it crufting around to possibly use for repairs on the replacement wire harness (the old one had extensive damage due to a poor cable managemet job on the part of the installer).

I would need to completely take apart the PAS cable or the sensor for further testing, and I'm not sure how easy it is to get into the T9N housing to probe inside and see what's happening in there. There is a a 50/50 chance atm that it is the cable or the sensor it seems like, as the CA3 seems mostly ruled out. I'm leaning towards the sensor given the solid power reading on the cable, but it's a dissatisfyingly fuzzy diagnosis.

I'm not sure if I have the time right now to fully diagnose both the cable and sensor. So it's a trade-off between the luxury of spending a day or two taking those components off the bike and taking them apart and trying to diagnose them fully, and potentially then needing to repair them on my own, and the luxury of buying a new sensor kit from Grin. Buying the sensor kit might actually be the more economical solution. There is a small chance the brand new CA3 could have a broken TRQ wire on its PAS cable, but it seems like the least probable and I'm hoping Grin would replace it for me if that were the case based on past experience with their customer service in the sales department, which tends to be excellent. I could test the new one before installing simply by plugging it in and turning one end with a wrench while holding the other end still. If it turns out to be the cable rather than the sensor at that point based on testing, I would have a spare sensor which doesn't necessarily seem like a bad thing to have around since they appear to have frequent issues with the wiring. In the long view I could order up the wires and connectors to maybe build up some spare cables to have on hand in the future. I'm hard on bikes as this thread proves, so I like having spares of certain things on hand within reason.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Feb 23 2022 7:05pm

It seems to have turned out to be the torque sensor, as the old sensor did not respond at all. However, I now have a new, mysterious and currently unsolvable problem. The PAS will only work if I pedal backwards. This is very strange behavior, and I haven't been able to find another example anywhere on the Internet. At first I only replaced the sensor on the non-drive side, because it was a replacement part. So I thought maybe the magnets in the spindle on the drive side might have different polarity in the two components, so I replaced that side with the other half of the new T9N, but the behavior is the same. I've been over all the settings for the T9N backwards, forwards and sideways several times, and they are correct. I reversed the sign of the torque scale, flipped the polarity of the 5V line, and reversed the sign of HW required for starting; none of these things had any effect whatsoever. I reset to factory settings; no change. The only things I can think to try now are upgrading to the beta firmware suggested by Grin and emailing Grin about it. I now have a brand new CA3, a brand new T9N, and the batteries and motors seem to be fine. The controllers seem to be functioning properly. The wiring is brand new. I've checked every connection several times. I'm stumped atm.

bombadero   10 mW

10 mW
Posts: 30
Joined: Jun 19 2019 11:59pm

Re: Short/arc fault in cold weather on XT-90

Post by bombadero » Mar 31 2022 10:12pm

Okay I finally have some progress on this. I seem to have somehow bent the spindle on the new torque sensor, not sure if I mentioned that before or not. It still didn't entirely explain the issue, as I had the old spindle on hand, which was not bent, and I used that in conjunction with the new sensor from Grin. Given that the spindle just has a bunch of magnets in it as far as I understand, presumably it would not be affected by an arc fault. However, even in this configuration, the behavior was the same: the torque sensor only worked if pedaling backwards, no matter what I tried. I think this may have been a perfect storm of both a faulty sensor and my own mistake in somehow bending the new spindle. Alternatively, I surmised that maybe the arc fault flipped the polarity of the magnets? I don't even know if that is feasible, just thinking out loud.

I contacted Grin attempting to get some technical assistance, but after a brief reach out from Justin they ghosted me, even after an additional email in which I came up on my own with about 10 different options about what the issue might be. No worries, I realize everyone there is very busy. Kind of frustrating when your car replacement is dead though. In any event, I ordered another sensor from Aliexpress, which is where the previous sensor had been purchased that had worked. I did this to disambiguate between the two versions of the sensor that I discovered exist, one manufactured by ERider and the other by a licensed or unlicensed clone (more on that later). This Aliexpress vendor is also the only person that sells a kit without a pair of cranks, so it's cheaper and doesn't have extra parts I don't need. I ran into supply chain and communication issues with them, so I also recently ordered another one from Grin, which arrives next week. The Aliexpress sensor ended up arriving on Tuesday, much to my delight, fresh from Shenzhen.

I installed the Shenzhen sensor (as I will heretofore refer to it) today and after a short test ride it is (mostly) working. The main thing is that it provides assistance whilst moving in a forward direction, which is my preferred method of biking :lol:. Strangely though, it is a bit 'hesitant' to do so. I had the start treshhold set quite low, at +020 HW. However, at that very low setting, it doesn't kick in until more like 500W or something. I actually couldn't start with it, having to use the throttle. The PAS only kicked in over 20 MPH or so. So I set it to -020 HW, and now it works more or less normally, although I still have to start with the throttle. It's possible I might just have to tweak it a bit lower, or that something else in my setup is wonky, as I have not touched it in a few weeks. Since the Vancouver sensor arrives on Wednesday, I'll be able to compare it to the Shenzhen sensor and the hybrid sensor (last new Vancouver sensor + old Shenzhen spindle). I'm still on v3-14 CA firmware, so unfortunately can only configure for 32 poles. I'm a bit trepid to update the firmware to v3-15b2 after my experience with the last firmware update when this problem first occurred, but Nicole at Grin vouches for that version.

So I've learned a couple interesting details with this most recent sensor. First, as this is now the third such sensor I've installed, I've noticed that the sensor output wire seems to end up at a randomly different location radially around the BB in each sensor. The very first one I installed, which was a 68mm sensor that was too small and I had to give away, pointed forward a little. The first correct sensor, a 73mm from Aliexpress, pointed straight up. The last one from Grin pointed about 45 degrees forward; and the new Aliexpress sensor points straight down. A bit inconvenient, if trivial. Straight up is probably the safest position for the cable, and I believe the intended standard position, but once you get the cupware to the right torque setting it won't budge any further.

The second is the aforementioned existence of two versions of the sensor. The one manufactured by ERider has a product ID of ERS-T9-73; the other variant is KYD-T9-73. I assume the first 3 characters in each are abbreviations for the companies; I have no idea who KYD is though. ERider is based in HK. It seems like there is another company making their sensors in China though, and they even have a notice on their sensors product page that the KYD-T9-100 sensor is not licensed. I don't know if that means all of the other KYD sensors are not licensed either or just that one. Interestingly, if you search for this sensor online most people have the KYD version pictured on their website, including Grin. In fact, until my last sensor purchase from them, the previous 68mm sensor I purchased from them was also a KYD sensor. They seem to have only recently begun selling ERS variant T9 sensors.

So that's interesting from a general manufacturing mystery perspective, but more importantly, perhaps there are some sort of slight variations, say in magnet placement, between the two, that just coincidentally made the hybrid sensor behave the same way as the old, burnt out KYD sensor. Again, I will be able to test a little more when the new Grin ERS variant arrives. Until then, I'll do a deep dive into the config again as well.

I also added a new metal fatbike fender in the front, which I've been meaning to do for a while. This is off topic, but the install was a bit difficult and had to be customized because I have an inverted fork and two racks on the front. The racks are fixed to the uppers, and don't move, and the fender is attached to the bottom of the forks and moves up and down inside the racks. It works great, but clanks around a little on bigger bumps, which is not ideal, being entirely metal. If anyone has any suggestions about that, they would be welcome. I might just have to live with it. I'll probably start a separate post about that topic. I may just put some sort of rubber edging on the fender.

User avatar
amberwolf   100 GW

100 GW
Posts: 33669
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Short/arc fault in cold weather on XT-90

Post by amberwolf » Apr 03 2022 4:09pm

bombadero wrote:
Mar 31 2022 10:12pm
I seem to have somehow bent the spindle on the new torque sensor, not sure if I mentioned that before or not.
That should take quite a lot of force. Do you think it was caused in the original event? Or is this the replacement unit? If it's the replacement, I can't imagine how you could bend it accidentally without knowing it (or at least, something serious) had happened right then, so it's more likely it came bent from the manufacturer or seller.

It still didn't entirely explain the issue, as I had the old spindle on hand, which was not bent, and I used that in conjunction with the new sensor from Grin. Given that the spindle just has a bunch of magnets in it as far as I understand, presumably it would not be affected by an arc fault. However, even in this configuration, the behavior was the same: the torque sensor only worked if pedaling backwards, no matter what I tried. I think this may have been a perfect storm of both a faulty sensor and my own mistake in somehow bending the new spindle.
An arc fault on a battery line can generate a lot of high-energy RF, which could directly affect and even damage things that are directly connected to the same battery line. But it shouldn't affect things that go thru a regulator or two to get their power, like the torque sensor (which is powered from the CA...but if the CA's regulator was damaged or destroyed and passed battery voltage to the sensor, it could destroy that..but the CA would also be destroyed, and very likely everything powered from it, like throttles, etc).

Alternatively, I surmised that maybe the arc fault flipped the polarity of the magnets?
I don't think that could happen even if the magnets were directly in the path of the arc. ;) They probably wouldn't even be demagnetized by it.
I installed the Shenzhen sensor (as I will heretofore refer to it) today and after a short test ride it is (mostly) working. The main thing is that it provides assistance whilst moving in a forward direction, which is my preferred method of biking :lol:. Strangely though, it is a bit 'hesitant' to do so. I had the start treshhold set quite low, at +020 HW. However, at that very low setting, it doesn't kick in until more like 500W or something. I actually couldn't start with it, having to use the throttle. The PAS only kicked in over 20 MPH or so. So I set it to -020 HW, and now it works more or less normally, although I still have to start with the throttle.
If the readings you are seeing are from the sensor itself, interpreted by the CA, and they are that much higher than before, it almost certainly means the CA itself is not set the same way it was before, or it's firmware is a different version and is interpreting settings and/or input differently. If it were the sensor causing the problem by having too low an output, you would see the readings themselves be much much lower than before, for the same physical input.


The second is the aforementioned existence of two versions of the sensor. The one manufactured by ERider has a product ID of ERS-T9-73; the other variant is KYD-T9-73. I assume the first 3 characters in each are abbreviations for the companies; I have no idea who KYD is though. ERider is based in HK. It seems like there is another company making their sensors in China though, and they even have a notice on their sensors product page that the KYD-T9-100 sensor is not licensed. I don't know if that means all of the other KYD sensors are not licensed either or just that one. Interestingly, if you search for this sensor online most people have the KYD version pictured on their website, including Grin. In fact, until my last sensor purchase from them, the previous 68mm sensor I purchased from them was also a KYD sensor. They seem to have only recently begun selling ERS variant T9 sensors.
I would guess that the last number is the length of the spindle (or BB shell), with 100 being a fatbike version. If it is not, and they all have the same spindle/BB length, then the number may designate the type / sensitivity of the sensor inside, so they may operate differently for the same inputs.

I also added a new metal fatbike fender in the front, which I've been meaning to do for a while. This is off topic, but the install was a bit difficult and had to be customized because I have an inverted fork and two racks on the front. The racks are fixed to the uppers, and don't move, and the fender is attached to the bottom of the forks and moves up and down inside the racks. It works great, but clanks around a little on bigger bumps, which is not ideal, being entirely metal. If anyone has any suggestions about that, they would be welcome. I might just have to live with it. I'll probably start a separate post about that topic. I may just put some sort of rubber edging on the fender.
If you can show a picture of the issue, it may help inspire ideas. My first suggestion is to mount the fender to the bottom of the rack itself, rather than the fork, if that's possible; then it won't move around. ;)
If you found this advice helpful, supporting contributions are accepted here.

Beginning of a New Life


Give a listen to my music!
(More here, including experimental).

Post Reply