Amberwolf's DayGlo Avenger, MkII

[docnjoj]

Hey Amberwolf! My wife is a treker and has a lot of "stuff" from collecting Star Trek memo. She asks "Isn't that stuff valuable"?
otherDoc

[amberwolf]

It might be...but I was in an auction with a room full of Star Trek fans and I got it for a dollar. So, valuable is relative. If no one else wants it, it's hard to get much for it.

If you know someone that does want it, as it is, I'd certainly sell it to make money for other projects and/or household budget. I'm not sure how else to go about finding a buyer (ebay is such a crapshoot). I did about 20 minutes of searching on the web and never even found another one out there to compare it with (and I'm typically pretty good at searching).

I also have my homebuilt ST:TNG transporter console, if you know anyone that would like an upright-piano-sized prop for their living room. I was offered from $1500-$4000 at various conventions for it, but the potential buyers were always from out of town and after they tried to figure out shipping arrangements to take it back with them they always cancelled out.

FWIW, the vacuum cleaner pods I have somewhere around here would probably look better and are certainly more durable; this just looked like an interesting shape, plus the trekkiness of it. Figured I might stick in on the Trek frame, for giggles.

[amberwolf]

Yesterday I was also donated this roll of doublesided velcro; it'll probably become tie-downs for cables and stuff as I have to replace sunrotted zip ties (only takes a few dozen hours' exposure, at most, to make them disintegrate). Should last a lot longer, and is easier to undo when I need to modify wiring.

DSC02911.JPG (38.78 KiB) Viewed 723 times

On my dogfood run today (see here:
viewtopic.php?p=281511#p281511 ),
I broke another rear spoke, and one more feels loose and is probably cracked and ready to break off (not visible in pic):

DSC02908.JPG (62.96 KiB) Viewed 723 times

Since there was no real load on the rear wheel that isn't normally already there, and actually perhaps a little *less* load as the trailer sometimes would lift the rear wheel a bit as the caster on the loaded trailer went over speed bumps and the like, I'm not sure why it picked right now to break as opposed to some other time. But I think I need to swap out this whole wheel, and rebuild it for later use. I have a few 26" rears.

I am thinking of trying out the airless tire I got on the freecycled bike a while back; the only issue is it's a knobby tire so I'd like to first get the thing off the rim and swap to a more road-ish tire. I have a damaged Kenda Kross(roads?) tire almost identical to the ones I used for the old friction drive, with the slick center and knobby edges. The damage is on the sidewall, so an inflated tube can herniate thru it, possibly enough to get cut by the brakes, but an airless tube would have no problem with it. Might use that one just to see.

But I couldn't get to any of that today: that stupid washer died again over a week ago, and I *had* to fix it today--put it off as long as I could.

This time it didn't melt/strip out the plastic power transfer part, it *shattered* it, into three nearly-equal pieces.

DSC02912.JPG (28.95 KiB) Viewed 723 times

I still had one unbroken one left, so after 30 minutes of work and cleanup, it's installed, and I guess I'll just run small loads, to reduce the stress on that part. I think I may try to make a pair of them out of steel or aluminum. Steel I can weld together from pieces, aluminum I'd have to find a single block large enough to do it.

Or, alternately, if I can find two skinny pulleys (or one I can cut in half), I can just put a bolt with thin nut thru it in the three places the transfer tabs would be. If they are already double-d pulleys, I can probably just stick them right on the shafts, but if they're just set screw types I can put them on that way, too--it's just harder since there's very little room to work right there, and the original transfer parts were designed to slip on the double-d shafts.

[amberwolf]

BrandonB donated these to the projects; they arrived today:

DSC02913.JPG (53.36 KiB) Viewed 823 times

I'd wanted one to experiment with when I first saw them, and I'd tried to catch them on sale a couple of times, but both times the sellers ran out before I could decide to get one; I really have a hard time making decisions on where to spend money when it's not on stuff I *have* to have, like food, utilities, rent, etc.

I still have an idea in the back of my mind to adapt this to help control my pedal-chain-tension throttle, but I cannot find where I wrote/drew the original idea, and can't recall the details of it now. I'll just have to open one up and see if it reminds me.

DSC02914.JPG (50.37 KiB) Viewed 820 times

For now I have just stripped the wire ends on the short-cable unit, so I can hook it up to DGA's Lyen controller to test out; then we'll see how well it works for my purposes.

DSC02915.JPG (38.37 KiB) Viewed 823 times

I don't usually have really long stretches of road that I need a constant speed for, but there are times I ride a part long enough between lights or stop signs to make it worth having around, to save my thumb a bit of work, or free it to do finger exercises to help the numbness that can happen (one reason I like CB2 much more than DGA; there is no good way to angle anything on DGA to not make my hands go numb after a while, but on CB2 it VERY rarely happens).

I'll probably put it on the left bar-end, so that I can use it with my left hand to give the right a break, and on the bar-end so I can use it to control the throttle even when I have my hands on the bar-ends instead of the bars (to switch off for anti-numbness; I have not yet ridden extensively with hands only on bar-ends to see if that helps but I think it would).

[amberwolf]

(crossposted from another thread as it also belongs here)

Now that I have it, regen is wonderful for braking, with a Lyen 6-fet and 9C 2807 front.

But I don't get a lot of recaptured power out of it; the most so far is about 1Wh. Since it takes me about 39Wh to ride that route (measured on a different day), I suppose that makes about 2.5% recaptured power. I have a lot of stops (a dozen or so) that must be made, depending on traffic, but I usually coast down to a stop rather than maintaining speed right up to hard-braking distance (partly to save energy in motor and legs, partly so I have more time to observe and deal with whatever happens at the intersection/etc., since other traffic may do unexpected and unpredictable things at the last moment).

Once I finalize my brake-switch I will be doing testing of the regen by hard-braking with only regen as much as possible for at least one full ride, and see how well it works that way. Then compare my power usage during the ride with the recaptured power. If there is little extra power usage vs the recaptured power increase, taking into consideration the efficiency losses (1.5x) of recharging the NiMH, I might change the riding style where it's safe to do so.

[amberwolf]

ianmcnally donated a spare NiMH pack to the project; 36V f-cell, so 13Ah in theory. It arrived just as I was leaving for work today. Got home and checked, it wasn't quite fully charged. First thing I did was just pull the shells off the little powerpoles so I could insert them under the tips of the much larger multiple contacts on my charger, then I stripped the insulation off the thermistor wire (it didn't have a connector on it) and stuck that bare wire end into the thermistor powerpole on the charger.

DSC02918.JPG (66.68 KiB) Viewed 694 times

Started charging just like it should, at 4A. Gets warm a LOT quicker than my other pack (9Ah D-cell) does, though. Not sure if that is because of the cell resistance (higher?) or if it just conducts the heat out of the cell better.

Last working picture of the Turnigy Watt Meter:

DSC02919.JPG (22.7 KiB) Viewed 694 times

before the unfortunate event later in this post.

Once charged, which verified operation of the pack, I took the wrappings off to verify it's structure/integrity, and found it's actually a 24V pack with half of another identical 24V pack, in series:

DSC02921.JPG (28.4 KiB) Viewed 694 times

First I carefully wiggled the powerpole contacts until the wires pulled out of the crimps (which was a shame to do because they looked PERFECT, but i saved them for later use on whatever I next need them for), then I crimped on and soldered some ring lugs to bolt to my anderson multipole cables (which all have ring lugs on them from their original sources).

Since I will need 36V for the Fusin headlight (and it's power meter), I decided that since this pack is 13Ah and my other is 9Ah, nominally, I could just run it off of this larger pack, even if I wire the packs in series as I intend to do at least for testing. So I bolted two multpoles with the low-voltage keying on the ring lugs just added above. One is for the light, and the other is for the series harness. Added a single powerpole connector for the thermistor to match what I used on the other pack, for charging.

Then I proceeded to bundle it up tightly, by white-gluing the cardboard rings around the sets of cells to each other, then strapping down with packing tape, and a little duct tape when I ran out of packing tape (which doesn't stretch as much as duct tape, so I'd rather use the PT). I also used some old nylon "clipstrips", which we regularly toss out at work--lots of products come prepackaged on these, ready for hanging up on store aisles for easy display--just add pricetag. Often we need them on the shelf peghook instead, so we end up tossing the strips. Sometimes I save a few, and have a fair collection of various thicknesses and types of plastic.

DSC02920.JPG (23.28 KiB) Viewed 694 times

I used nylon ones here, snapped at their scoring lines (so you can break off the empty sections as product is purchased). I only need them to cover the ends of the cells, so that I can be sure they will be spaced away from anything that could short them even if the tape comes off in the heat, since my cargo pod is metal. Also to help absorb a little vibration.

DSC02922.JPG (14.73 KiB) Viewed 694 times

Then taping down and adding a second, wider strip, of clear plastic clipstrip just a hair wider than the pack, wrapped from the bottom up both sides and just about an inch over each end at the top.

DSC02923.JPG (20.27 KiB) Viewed 694 times

I dozed off in the chair with the pack in my lap somewhere around then, can't remember exactly. Woke up probably a few minutes later and finished up the taping/etc.

DSC02924.JPG (22.2 KiB) Viewed 694 times

Then I started working on the wiring harness. You might have noticed the red, black, and gray multipole shells. Black and gray will fit each other, but none of the other colors will. THe blue on on my charger shouldn't, but I dremeled and filed it's key off so that it fits black/gray too, as at the time I didnt' have any more of the others not in use. The red ones will be used to prevent plugging the 72V pack output into the 36V charger, or the 36V light, and blowing stuff up.

DSC02927.JPG (16.3 KiB) Viewed 694 times

Red on controller, red on series harness output. Black for one pack and gray for the other, just so I can always match them up and remember which one is plugged in, if I use only one. To use one, I made a "jumper" multipole, which just has a wire from + to - on it, which shorts the empty battery pack connector, to allow me to use just 36V. It's original intent was so that if the 9Ah pack ran out but the 13Ah pack still had lots of power (should), then I could disconnect the 9Ah pack so I don't overdischarge it, and still have assist.

DSC02925.JPG (31.44 KiB) Viewed 694 times

The new pack now goes up front, scooting the other pack back a little. Tehy're snug in there now. Controller will bolt to right side of pod, just inside and forward of the lock that's on the lid.


Then I decided to start testing it. First I figured I'd see what the voltage was, and as I already had the Turnigy Watt Meter out, I temporarily swapped it's blue housing for a red one, plugged it in, and got zilch.

At first I thought I wired the pack series jumper cable wrong, but a meter showed correct voltage and polarity out of it. Then I realized I'd just put 86V on a 60V device, and I got really sad for a minute, until I remembered Doc Bass's thread about fixing that, which I am now going to need.

For some reason I'd gotten confused and thought it was 150V not 150A, then hooked it up before my brain could realize the mistake and stop myself.

Then I also decided before I hooked it up I'd better recheck the Lyen controller thread, to make sure they were already setup for "72V" packs (FETs and phase caps are 100V, but coudln't remember the rest), and found I'd need to mod it for above 60V for the LM317 input divider. That'll have to wait for a bit, so I guess no 72V tests yet. But at least I didn't blow it up first and THEN check.

Guess the Turnigy's death was more useful than I thought, or I would've blown up the controller instead!

Since I am way too tired to be modding electronics right now, and I have a Freecycle trip tomorrow afternoon to pickup a couple kids' bikes, I left well enough alone and will just use the system as 36V, swapping batteries when I need more power on the way back home (since either one of the packs will probably run dry before I can get all the way back, as I won't be going as slow as I was on the Hachi/park trip, or pedalling very much, as it will be really really really hot outside).


And now it is sleepy time.

[amberwolf]

Lyen advised me that the only mod I need to make (other than LVC programming/etc) for 72V is to add another resistor in series with the existing one for the low voltage regulator's input. No time for that before the trip, though, so I went with just 36V and figured I'd swap packs when the first runs dry.

[amberwolf]

After leaving on the trip, riding for a few miles just fine, then something in the controller began to fail. It still operates but there is a sizzling sound (probably a capacitor) so I cut my ride short and disconnected it from power. I was only a mile from a friend's house that has a truck, so I detoured over and got a ride back from him.

I suspect it is a combination of the Phoenix heat, as it said 105F when I started out around 2pm, and having to run at partial throttle to prevent damage to the NiMH from overcurrent. The controller and batteries are in that left side pod, and I had several icepacks in there below them to keep the pod around 15F cooler than ambient, but the cotnroller still got quite hot, uncomfortable to hold but not burning. Don't know it's exact temperature, but I'd guess at least 110F by the time I had the problem.

Problem started out as a wierdness in the power meter display on the Fusin headlight/meter I am still using. Even at barely any throttle, the display would flicker to "dead" and back very fast, as if it was in time with the wheel rotation (minus the response time of the LED bargraph meter).

Then I suddenly got no motor power, and figured it was a poor connection somewhere. The LED meter still showed full power but it has a separate connector to the battery. Battery was running a little low during accelerations, but still at full when no load on it. Measured 38.6V with a multimeter, no load. (this is the 13Ah pack that just arrived)

Connections were all ok, but now that I had the pod open, I could hear a sort of sizzling sound, like you get with bacon but as if it was tiny and very far away. Traced it to the little 6FET controller, opened the end and could clearly hear it. No unusual smells, just that of warm rubber from the seals on the ends like it has had from the start. Pulled the board out and could see no problems, but since I figure it's a cap failing for some reason that I had better not continue to use it, or when the part goes it'll probably take stuff with it. Much easier to repair one part than several! (I'll troubleshoot it tonite if I can, but I have to work at 6am so I may not have time).

I shifted down to first gear in front, and 2nd in back, and was able to go the mile or so to my friend's house with the truck, but with the wind now against me waffling from about 5-10MPH I was pretty wiped out by the time I got there, with the trailer and all the extra stuff I was carrying "just in case":

--8lbs of Tools
--1lb of Fasteners/clamps (for trailer mods if needed)
--6-foot extension cord
--NiMH charger
--Both NiMH packs, 9Ah and 13Ah
--2x 2-liter bottles of water, frozen
--about 5 pounds of frozen water-gel bags (for cotnroler/battery cooling)
--trailer minus top half of kennel
--bungees/paracord
--12v 20Ah SLA (15lbs, for lighting/signals since the other little SLA probably won't run it long enough for this trip)

So all that plus my 160lbs plus the bike which isn't light, plus the 9C motor (20lbs?), mostly against the wind for about a mile, with no motor assist.

I think I am going to redo the connectors on both controllers (Lyen and Fusin) to be the same, so I can just carry them both, and swap them out if one fails.

I'm also going to change back to the lower-current Fusin gearmotor for longer trips, since holding partial throttle is heating the controller a lot. Not a big problem on my short work commute with no trailer/extra stuff, but apparently a huge problem with this kind of trip.

I really like the quietness and pulling power of the 9C 2807, and it's ability to get me to speed quickly without killing my knees, but it takes so much current to do it that the NiMH would be damaged (at least 3C to 4C on the 9Ah) by that if I rode it full throttle for more than a few seconds at a time. Plus it'll use up all my range.

The Fusin does it with 2C or less *max*, and IIRC about 1C or less cruising WOT. Takes a lot longer to accelerate, but I can live with that to get longer range for this kind of trip.


A couple of alternate possibilities to switching motors out for each trip type:

--go to 72V for those trips instead, carrying both NiMH packs, which will let me cut down the battery current in trade for voltage. But this may make the controller heating problem even worse, as I'll be at even lower throttle settings to keep the speed down.

--leave out the NiMH entirely, and use the 3x 12V 31Ah SLA U1s, on the back of the trailer at the wheel axle line. Those can handle whatever current I need, and still give me about an hour's riding time if I don't do WOT all the time. (but not going WOT could cause the same kind of controller heating problem again).


A third possibility is to rehouse the controller to something with active cooling for the entire controller.


If I ever get the LiCo pack(s) made, they may solve the problem of current draw and range for just the bike itself. Using the trailer, I could use the 33V TS pack I will shortly have (at 50lbs and 60Ah I think I could ride for nearly an hour and a half, WOT with a heavy load).

[amberwolf]

I was so worn out Saturday night that I kept dozing off trying to test and fix the controller, and for it's safety I just stopped and went to bed, though I could not stay asleep until it started getting light outside (typical problem I have); had to be at work at 6am so I rode without batteries or controller on DGA. Even without the motor hooked up to anything, the 9C's weight is a lot to pedal without help, and I'm sure it's magnetic drag doesn't help. I had to stay in 1st gear in the front and mostly the first four gears in back.

If I had not been so tired I would've changed it back to the Fusin before going to bed; wished I had by the time I got home, since I also had to run by the grocery store on the way home from work.

Anyhow, before I fell asleep Saturday night, I actually had been starting to pull caps to test for which one is dying, and the first one pulled was the 470uF 100V main cap. It tested good for capacitance (430uF) but took longer than it should have to come up with a reading, so it is probably leaky, so I replaced it with a 1000uF 100V that is a bit longer and fatter but still fits inside (barely).

I did not get others pulled to test yet, since I kept dozing off at the table so I had to stop before I broke something.

I also made the resistor mod for 72V, but since I'll be using high and low voltage packs for experiments and whatnot, I used a Methods' switch to short across the extra resistor when I don't need it:

DSC02933.JPG (55.75 KiB) Viewed 678 times

The original resistor is left in place, on the right, with just it's long lead unsoldered from the PCB. The added resistor (pulled out of some old power supply) is the same lead configuration, short on PCB end and long on top end. I mounted it mirroring the original, in the PCB hole for the long lead of the original. Wrapped and trimmed the lead of new to top of original, and folded the original's long lead to tape down in case I need it to go back to the PCB hole again later. Switch is not yet connected.

Circled in magenta is the replaced cap; I used one off a scrapped Methods controller he'd sent me for parts. Somewhere around here I have some Panasonic or Rubycon caps to replace these with; I trust those brands better than any of the others I have had experience with so far, especially as regards to heat and leaking.

Here's the switch, connected, and set to high voltage mode (open, clicked in):

DSC02934.JPG (50.83 KiB) Viewed 678 times

One wire goes to top of original resistor, and the other to the empty V+ pad next to the other end of that resistor, on the same PCB trace the resistor is soldered into.

When clicked out, the switch shorts across the original resistor, setting it to low-voltage mode (below 60V).

DSC02935.JPG (51.83 KiB) Viewed 678 times

The bottom of the PCB, with the pad circled where the switch solders in:

DSC02936.JPG (36.88 KiB) Viewed 678 times

I'm off work tomorrow (today, Monday), so hopefully I'll get the other caps checked out and replaced as needed. Then I will make a little bracket to hold a small fan to be powered from the LM317, so the fan will suck air thru the controller, helping to prevent future capacitor failures from heat.

[amberwolf]

crosspost from battery thread:

The TS 60Ah cells arrived today; a bit lighter than I posted, at 42lbs for 10, rather than 50lbs. All at 3.31-3.32V.

DSC02943.JPG (44.85 KiB) Viewed 772 times

Gotta dig out some bolts, make connecting cables, and then perhaps test them out on DGA with the 9C after I finish fixing the caps in the controller.

They'll fill up the bottom half or so of my cargo pod, almost exactly a perfect fit side to side; I think I can almost fit a 6th pair of them front to back but not quite.

DSC02944.JPG (54 KiB) Viewed 772 times

Won't have a BMS yet but they should be ok for a quick "what can they do" test.

Even at just 33V, with no significant sag, they should be pretty good for acceleration even with the extra weight (about 3x the 36V 9Ah NiMH pack's weight, IIRC). The NiMH can sag to 30V during acceleration, after a couple of miles of typical riding. The TS should not sag at all for some time.

I might build up a couple of "cells" to series with these, made of 30p LiCo, to get the same Ah rating, and give me 36-40V to play with instead (closer to what the NiMH has typically).


Before I charge them up again or use them deeply I'll want to finish the v2.6 BMS LiFePO4 versions, at least for the LVC/HVC part if not the shunting. Gotta finish making the list of what I still need for them, and post that back in this thread:
viewtopic.php?f=8&t=18333&start=0

[amberwolf]

Fixed controller. Main cap was probably the issue, but I won't find out until I get it all hot again, which I am going to try not to do. All caps passed the tester, within a few uF of their marked values; the only ones that took (a lot) longer than they should have to come up with a reading were the main 470uF, C2 the 100uF 63V on the LM317, and two caps on teh 7805 100uF 25V. So I replaced those, just in case.

All the caps on there are "ChengX" brand, which I have seen on a handful of controllers so far, but I do not trust (haven't seen enough of the brand yet). The ones I replaced with were varied, from Jamicon 1000uF 100V to Nichicon for the small 100uf 25V ones; I forgot what the middle sized one was. If it weren't for shipping costs, I'd just order a bunch of Panasonic low-ESR caps for all the caps in this (and the Fusin), plus some extras for later. But for now, the ones I used are tested pulls from various salvaged electronics PCBs.

Afterwards, I tried just putting it back into the case and leaving it in the sun to heat it up, then powering it on to see if any caps sizzled, but nothing happened. So I guess the remaining caps are ok for now (they may get a lot hotter in use than just from direct late-afternoon sun, despite the 100F+ temp at the time). I still don't know for sure if any of the replaced caps are actually the failing ones, though.


I then put it back on the bike, hooked it up to the motor and one battery, verified it was set to low-voltage switch setting (out), and...nothing. Disconnected it all, opened it, and saw no obvious issues, so hooked it to just the battery and started probing, and found no voltage input at all.

Tracing back found the battery cable I had used on the old 9Ah NiMH had a wire pulled entirely out of the crimp, on the SLA-type slip-on connector that had come already crimped onto the other end of the negative wire of the Anderson I had used. I removed the entire connector on that end, interwove the wire strands, and lap-soldered them together. Did the same on the positive side. Now there is a soldered connection from the Anderson's contacts all the way back to the battery's cells.

However, in the process of finding no voltage into the controller, I also had found that the resistor I had used in series with the existing voltage divider resistor for the low-voltage section was wrong. I knew it was supposed to be 330 ohm, but had used a 33 ohm. I scratched my head and tried to figure out why I had done such a dumb thing, and then realized that the color coding on the original resistor appears to be wrong--it is marked Orange Orange Black! That's a 33 ohm, and I probably just kept the colors in my head when I looked for the one I used, rather than looking for Orange Orange Brown like I should have.

So....I pulled the 33 ohm out, and moved the 330 ohm over to it's spot. Then as I couldn't quickly find a 330 ohm in that size, although I'm sure I have at least one somewhere, I used a 470 ohm instead, in the spot that gets shorted by the switch for low voltage. That way I still have the right one for 36V operation, and actually have higher resistance total for higher voltage operation, which should still be safe enough.

Retested the controller on the bike, and vroooom, it works now.

Connected both packs in series, and tested ok holding wheel off the ground--dang that is FAST!

Rode it around the block that way, and it has harder acceleration with the seriesed packs, and of course goes much faster. I didn't have the PDA/speedo on there so I don't know how fast it was but it was at least 20MPH before I realized it.

That was when I noticed that regen braking wasn't working, and had to use mechanical brakes to slow down...they dont' work very well from those speeds--either grabbing too hard or not hard enough, doesn't seem to be a middle ground. I prefer the regen brake; it's much smoother.

I think that there is a regen voltage limit on this; I recall reading about this and a mod to fix it but that might be on a different controller. If there is no specific reason controller-wise to limit it, I will see about defeating it as I really would want the regen *more* at these faster speeds.

I could probably get scary fast with this setup, without pedalling (though the battery wouldn't last long--it was sagging significantly even with both on there). If I pedalled too, I'd have to regear to keep up!

The motor is significantly louder growling with the higher voltage, especially at lower throttle settings. Those lower settings also heat up the controller faster, of course, and they are awfully hard to hold--the throttle itself feels like it goes from zero to 30% or more right thru the bottom couple of degrees of movement. Really touchy.

I can definitely see the allure of higher voltage setups using this motor. I think I would want to actively cool the controller before I did so, though.

[amberwolf]

I ran with just the 13Ah NiMH pack today, for 36V, on my work commute. Without my watt meter I don't know for sure what the Wh used was, but it should be about the same as previously with the 9Ah pack, as I rode about the same way; might've been a little higher since there was a lot of 10-15MPH steady wind against me most of the way there; not much on the way home.

Plugged in when I got home, and the charger only took about 7 minutes to turn green. That's just wrong; has to be; I know there is more than 18Wh used, which is about what that comes to (~40v * 4A = 160W, 7min / 60min = 0.116h, 160W * 0.116h = 18Wh). I powered off the charger and turned it back on, and it ran for about a minute before turning green (maybe less). So it thinks it's full for sure; a meter shows ~43V, which is typical.

I know that this pack is changing as I use it; seems to be gaining capacity (similar to what appeared to happen to the 9Ah); so if I can fix my watt meter or put wires/connectors on the WattsUp then I'll be able to verify what capacity it actually gets and what it takes to charge it back each time. I've also been leaving it on the charger after the charger goes green, for all the time I am not actually riding it. This keeps a 7-9mA current thru it, which should compensate for the self-discharge and continue to condition the pack, I hope.


An interesting side note: Since the cap repair, regen feels smoother. There used to be a cogging sort of feel to it when engaged, so a slight vibration of the forks and bike would occur, making it easy to tell when it was working. Now, however, there is very little if any of that. Another thing I need to fix the watt meter for, to test regen current and see if it is indeed still working as it was before. I'm pretty sure it *is* working, and providing braking; I just am not certain if it is as strong as before or if I"m simply thinking it's not because I'm not feeling the same vibration it had originally. A regen current measurement would tell me if it was the same or not.

If it is the same, then it means (most likely) that the extra capacitance, more than double the original main cap, is smoothing out the regen pulses more and that smooths out the current draw from the motor, which smooths out it's braking. Just a guess.

[BikeFanatic]

Amberwolf,
That charger has to be malfunctioning, unless your commute to work was 7 minutes at 4 amps? Know what I'm saying.
I use 17-23 watts Per MILE, and my bike does not weigh in at 150 - me and the bike =200 lbs

unless your commute is 1/4 mile I do not see how this 18 watts is possible ...

Now one of the Highpower chargers used to cut off early some times I think I had to unplug it then redo it again. Also there is some kind of safety feature if the battery is too hot it will not charge until battery cools down.

Do you have a working watt meter? I havent been riding lately as I hurt my neck, maybe I cold loan you another charger?
( but I would need it back).

[amberwolf]

That charger has to be malfunctioning, unless your commute to work was 7 minutes at 4 amps? Know what I'm saying.
I use 17-23 watts Per MILE, and my bike does not weigh in at 150 - me and the bike =200 lbs

unless your commute is 1/4 mile I do not see how this 18 watts is possible ...

It's not possible; that's the thing. But it's still ok, since I have two working battery packs for now, and can use the other one even if the first runs low unexpectedly due to not really charging all the way. I am going to be using the 9Ah pack today, then recharging it when I get home, and if it also doesn't charge long enough I'll know it's the charger. If it does charge properly then it is probably the 13Ah pack. Either way I'll know which one needs fixing, and can start troubleshooting something.

Now one of the Highpower chargers used to cut off early some times I think I had to unplug it then redo it again.

That didn't help in this case, either from AC or from the battery. AFAIK it's never done it before (for me), as I am typically still at the bike sorting stuff for next day's trip for at least 10 minutes, with the dogs outside doing their thing, so I'd hear the fan cut off if it stopped that early. Could still cut off early but just not that early and I'd likely not know, since I do not always use the watt meter to verify charge, simply because I have to disconnect and reconnect it in reverse to read charge current.

I did used to check it every time when using the lab supplies, but got lazy when this charger was working consistently well.

Also there is some kind of safety feature if the battery is too hot it will not charge until battery cools down.

That was my first thought, but it was barely even warm. Even undoing the plastic I put on the end and touching each cell cap they were only a few degrees above room temperature.

If it was a thermistor open failure, the light would keep blinking yellow, and not try to charge, so that's not it, either.

Do you have a working watt meter?

Sort of, but the wires are too short to hook up to again and I need to solder new ones on there. Been lazy about it because I had the Turnigy (that I killed a couple days ago). Also have a CA now, but not usable till I fix it.

I havent been riding lately as I hurt my neck, maybe I cold loan you another charger?
( but I would need it back).

No, I have the other 36V charger that I will be fixing at some point, and I also have current- and voltage- limited supplies I can charge with, just limiting them to 41 or 42V so they never overcharge it will work (I did this before I got the ones from you ).

[amberwolf]

Been using the 13Ah pack with no problems last couple of days, no further instances of charger wierdness on it or the 9Ah.

I am now sure that it wasn't a pack thermistor or thermal issue, as when I got home from a grocery run several hours ago, the pack was still quite warm (pretty hot weather out there today plus extra load on the system), and when i plugged it in it started to charge, about a minute later it started flashing the yellow light at me and I found the pack kinda hot (not too hot, but uncomfortably warm). I put a fan on it and it cooled in about 15 minutes to the point where charging resumed and the light went steady yellow again. After a couple of hours, it went green, for a total charge of about 8Ah (or 6Ah or so actually used, since it takes about 1.5x the power in that I get out).


On the way home with the 45lbs or so of groceries (taking advantage of all the holiday sales I can afford to), there was a pothole I couldn't avoid; didnt' see it in time to do so without getting too close to a passing car. Missed with the front wheel but got it square on with the very heavy rear. Broke two spokes on the non-drive side and one on the drive side, all in the same quarter of the wheel, so it was severely warped from true. Even with the brake noodle pulled off the holder so the arms spread apart, it still rubbed for more than 1/8 of a rotation on the right pad. I could not loosen or tighten any spokes enough to make any difference to that, so I and the motor had a lot of extra work to do to keep speed up.

I decided that I need to totally break down the wheel and redo it, and I don't have time for that right now, so I just replaced the whole steel wheel with a narrower aluminum rim and steel hub off one of the various parts bikes I have around. It's not a great wheel but it was already nearly true when I started, so very quick to finish (about 15 minutes). The hard part was redoing the brake pads and arms to line up with the narrower rim, and sit flush with the flat aluminum brake surface as opposed to the angled steel one.

Because of the side panels for my rack and cargo pod, I cannot get a wrench in there to adjust the pads with the wheel on. I have to line them up with the rim, hold them while taking the wheel out, then tighten them. Put wheel back in, test contact angles/etc., and redo until they're right. That took well over an hour. I have contemplated several times drilling access holes in there, but it'd have to be very oversized, as the rack panels slide down a bit over time with vibration and heavy loads, so that by the time I need to adjust things again I end up taking the rack panels loose to get the holes lined up. (I already did this with the adjustment screws for arm return springiness).


Besides, I am thinking of rebuilding DGA as a whole other bike with a rear suspension built longtail style, using the Trek930 frame and a rear triangle from another bike. Then I would use disc brakes in the rear made by cutting a pair of nearly-identical threaded hubs I now have in half and welding the threaded ends together, plus a disc rotor to thread adapter. I'd have to weld on a mounting tab for the caliper to the seat stays. I would also keep the rim brakes, so that I can have both brakes available to slow down with when I have cargo onboard or the trailer loaded up. If the rim gets bent enough to require unhooking the rim brakes, then my braking in the rear will just be reduced, not eliminated. That plus regen braking in the front hub plus front rim brakes ought to be enough stopping power, I hope, since I can't put discs on the front hubmotor.


I would just add the suspension to DGA itself, but the way I want to do it will require a bit of welding to the frame, and DGA is aluminum.

[amberwolf]

"New" wheel is working ok so far, but I haven't hit anything with it yet.

I now have three days off in a row to try to accomplish something. The first thing will be to see if I can come up with a fully suspended bike out of the frames, forks, and springs I have here, without making it a monster like CrazyBike2. I'll have a new thread for that once I get started, possibly later tonight if things go right. (yeah, sure....)

It is highly likely that the Fusin will go back on DGA, and the 9C on the new design, along with the TS pack after I finish that, but using the NiMH at 72V for now (just cuz I can). Or possibly using the NiMH in parallel for discharge, via some really hefty diodes I've got out of some old linear power supplies. That should help on the voltage sag problem as well as help protect them from the heavy current pull of the 9C during accel.


The next thing will be to neaten up the handlebars of DGA or the newly-made FS bike, by making a dashboard of sorts, at least partly inspired by the idea of LI-ghtcycle's. I'll make it first from thin cardboard, since I have lots of that from food packages/etc., so I can figure out where things need to go before committing to it. Then I'll use some thin aluminum to make a dashboard face and box, to enclose the whole handlebar center area in it. Doing it that way will also let me "permanently" install things like the CA (once it's fixed, also hopefully in the next few days), without worrying that someone will rip it off the bars and take it just because it is a little box with a screen on it, even if they don't know what it is. (otherwise I wind up having to come up with a way to mount it so I can quickly take it off and take it in with me everywhere, meaning cutting the cables out the back and making connectors for them, another point of failure/etc.).

I can also make a visor-hood over the thing, which can be a partial fairing around the bars and potentially the whole front wheel, though the latter will probably come later. The visor will likely be immediate, to help sunshade things like the CA LCD, both so I can better read it and so the midday sun doesn't destroy it. I might even put a little sliding cover over the LCD.

Lots of ideas, too little time to try them all out.

[amberwolf]

I basically got nothing useful done in those three days; too many other things went wrong around the house and such, and I had to stop using the swamp cooler and setup the house with hanging blankets and such to divide off just the front three rooms for A/C using a pair of window units, one at each end. It's just too humid for the other to do anything to cool now, and the dogs'll get sick and potentially end up with heatstroke while I'm not there if I didn't take care of that.


I did get a little done besides that, just not much. Sunday I went with a friend to a few Goodwills and Savers thrift stores, and we found a number of useful things, but most were overpriced (for us), including a really really nice bike trailer meant to haul two fairly big kids, with quick-release wheels and all; it probably was a $500 trailer, but their price was $100 and neither of us could do that.

I found a smattering of nice helmets, most too small for me, or with damage from impacts (the stores won't pull them from the shelf even when I tell them about that, and the fact that they should not be reused).

Most only had tiny little kids' bikes, but one of them had a decent bike in a purple-front-fading-to-pink-rear frame, with a nice Shimano accessory set on it. No suspension, but otherwise decent. If I'd had the $25 to spend on it, I'd've picked it up just for the parts on it I could use. If it's still there in a week once I get paid for some side work, I probably will get it.

It would actually make a nice bike to motorize because I don't think anyone will really want to steal it, for the paint scheme if for no other reason. I like them bright and colorful because they're easier to see, and this bike could become DGA mark III, but most people would call it fugly.

I did find a couple of things that I did get (well, I got the first one and my friend bought the other for me, to my surprise). The first was a helmet that happens to fit me perfectly, and appears to be completely brand new with zero wear, other than scratches on it from being moved around with other stuff on the shelf. When I first saw it, I thought it was carbon fiber, but it's just a paintjob. It's still a good helmet, at least as good as the one I wear currently, with a thicker hard plastic shell instead of typical bike eggshell-thin.

DSC03004.JPG (44.5 KiB) Viewed 661 times

My current one has saved my noggin twice without damage to the shell or foam; although if the impacts had been severe enough to really need it the foam would be toast, at least. The only thing wrong with this one is that the plastic clip to hold the strap on one side to the one on the other side is missing. I can fix that easy enough, though I may have to drill out the rivet on one end of that strap to install a new clip, and re-rivet it on.

DSC03005.JPG (29.27 KiB) Viewed 661 times

For two bucks, I'm not complaining.

The other find was a pair of *brand new* never used, still has tags on it, Shimano MT20D bike shoes. No toe clips (cleats?) installed, but that doesn"t matter as I don't yet have a matching pair of that type of pedals anyway. Well, I don"t *think* I do, anyway; I have to go back and look at the three I have from that junkbox.

DSC03002.JPG (26.85 KiB) Viewed 661 times

But at $18 (paid by my friend, which I will pay back by helping him out with various things), in my size, fits really nice, can't complain.

DSC03003.JPG (17.06 KiB) Viewed 661 times

[docnjoj]

Those shoes are just like the ones I got for 50 bucks from the LBS on sale from 90! Good work! They really are comfortable and you can walk with the cleats. Oh (rub it in!) mine included cleats!
otherDoc

[amberwolf]

For $50 they better include cleats.

Even without them they are way better than regular shoes for pedalling even on just my regular pedals.

pedals1.JPG (45.77 KiB) Viewed 644 times

[amberwolf]

I'm still picking thru various boards for the Cycle Analyst parts, so for now I just moved the whole wire harness from the Turnigy Watt Meter to the Watts Up, by desoldering from one and soldering to the other. The wires are significantly thicker than the original WU wires, so I can't close the case up (unless I cut out the entrance holes some), so I have it just taped together for now.

Did some peak power tests with the 9C powered by just the 13Ah NiMH pack a few minutes ago, using the Lyen 6FET.

The peak current WU logged was 25.69A, but I saw 28.27A on the screen, during start from a stop by motor power only (no pedalling), while the voltage dipped down to 33.74V (from 41.81V) at that moment. It did correctly log that minimum voltage, but not the max amps, which seems odd. If it was too short to read, it wouldn't have shown up at all, so there must be some filtering taking place in WU data logging vs data collecting. Kind of annoying. I forget if the TWM did that too.

Then I flipped the WU connectors and did a regen test; motor and pedalled up to the fastest I could get on the short stretch of 29th Drive next to my house, probably about 17-18MPH (forgot to charge the Veloace PDA because I haven't been using it last few weeks). Then squeezed the brake lever just enough to engage regen but not rim brake, and let it go down to instability speed (about 2-3mph I guess) where I had to put a foot down.

Peak was 10.83A, into that poor little 13Ah NiMH pack. Nearly 1C; I guess intermittently (every few minutes) for very short periods (seconds at a time) that wouldn't do it too much harm. Recovered 1.1Wh in just that single braking, but at a guess it would take 5Wh to get up to that speed in that short distance at WOT.

I'd intended to also do some watts-at-speed tests, but since I didn't have the speedo and it's really hot out there (only 105F but it rained a lot last night so it is VERY humid, with just enough of a breeze to help pull more water off all the puddles), I decided to come in and recharge the pack first, then when it and the PDA are done charging to go back and try again. It'll be even hotter by then, but I won't be working *myself* very hard, as I intend to let the motor do all the work for these tests.

Then I can test again with the Fusin motor and controller, and compare readings. I expect the Fusin regen to be negligible based on prior tests, but will try it anyway just to be sure.

[amberwolf]

Took me a while to find the USB-serial sync cable for the PDA so I could put Veloace back on it (it forgets eveyrhting when the battery goes flat). Dropped to 95F in the meantime, out there.

I am not certain exactly what the circumference of my front tire is, as I can't find the info on that I was sure I had saved from before, and I also can't find anything longer than a 10cm ruler to remeasure it (my crazy sister has been back at the house for some days now and things are magically disappearing and moving around again). It comes out on a rollout close to 209cm, so I used that, but with 10cm measurements lined up end to end there is considerable margin for error.

There is some slightly gusty breeze that varies in direction, so I rode in all four cardinal directions with the streets around here and get these readings, using the 9C, 36V 13Ah NiMH, and Lyen 6FET. I forgot to look at the Watts reading but was instead looking at Amps, so that's what I've written down below:

5MPH = 3-4A
10MPH = 4-6A
15MPH = 5-7A
20MPH = 11-13A

To get to the desired speed for steady reading, I rode for at least 3 seconds at that speed with no spedometer variation before glancing at the WU readout. Would've been longer, but there is too much traffic this time of day in the neighborhood and the roads are too short in some cases, before coming to stop signs or other speed-altering situations.

Since the voltage for 5-15MPH is around the same, at 38-40V, and at 20MPH is around 33-35V, that puts calculated watts-at-speed at these numbers:

5MPH = 114 -160W
10MPH = 160 -240W
15MPH = 190 - 280W
20MPH = 363 - 455W

Accelerating up to those speeds takes a lot more power; the peak watt reading was 841W, and the peak amps were 24.63. Minimum voltage was 32.44, resting voltage after return was 38.5. Total of 1.544Ah used up.

Total mileage doing these tests was 3.396 miles, and total Wh was 56.3, giving a Wh/mile of 16.58. Surprisingly low given that I was not pedalling at all, and the heavy accelerations I did for extended periods for the 20MPH runs (which were interrrupted multiple times due to heavier traffic out on the faster longer roads).


(crossposted results to the Watts at speed survey thread).
viewtopic.php?p=292012#p292012

[amberwolf]

I didn't have time with traffic today on the way to/from work to try noting watts-at-speed during the trip, but here's the trip numbers with the 9C, Lyen 6FET, and 13Ah 36V NiMH:

I think my tire size *is* off, because my mileage is significantly different from what I had before, on either CB2 or DGA.

to work:
2.311 miles
24.77Ap
0.908Ah
33.15Vm
33.6Wh
863.8Wp
About 14.54Wh/mile.

to home:
2.9 miles
24.96Ap
1.178Ah
41.5Wh
837.4Wp
About 14.31Wh/mile.

Odd that Wh/mile is *lower* on the trip home, even though I deliberately rode faster (20mph for as many stretches as I could, vs the 13-15mph max on the way to work), and I was riding against a stiff breeze most of the way, and some significant gusts in certain places, unlike on the way to work which was almost windless.

Kinda confuses me; unless it is an efficiency thing with the 9c 2807 itself, taking significantly less power at higher speeds for whatever reason?

[amberwolf]

Did a trip to the store less than a mile a way, rode without pedalling except for startups just to try to keep currents down on the poor NiMH pack.

To store:
0.408Ah
14.9Wh
32.7Vm
830.4Wp
25.97Ap
0.967miles

From store:
0.553Ah
19.0Wh
31.47vm
797Wp
24.59Ap
1.151 miles (roundabout way back due to increasing evening traffic)

Trip back had 28lbs of groceries, so that helped increase the Wh/mile from 15.41 to 16.51, as did having to accelerate harder for longer to get into traffic exiting the store parking lot. Odd that the peak A and W are less, despite that, though.


I saw a storm coming in, with little gusty winds starting up, when I got home. I decided I wanted to test the bike in the rain, so I recharged the battery and found some clear plastic bags to put over the WU and the PDA, with rubber bands to hold them on.

The battery was already warm from the store ride, so it did not take a full charge before it shut off thermally. I forgot to turn the WU around so I don't know how many Ah went back in. but the voltage never got above 42.9V, as I kept watching it every minute or two to see when it was full (so I could go out and ride starting freshly full charge). Usually it goes up to almost or just beyond 45V and then drops to 42.something (nearly 43).

Results from my 34 minutes and 6.73 miles of wandering the neighborhood in the rain at various speeds up to 17MPH, but typically around 11MPH, as the rain made the streets slick with the oil on the surface.

Mostly it just drizzled, except for a burst here and there, and it was not enough to wash the oil off the streets, as evidenced by the number of screeches I could hear, and the several thuds from bumper-crunches a few streets away (probably at Northern, Dunlap, 27th Ave, and 31st Ave, as those are the higher-speed streets most likely to have traffic not paying attention).

*I* almost skidded out at Butler and 31st Ave, making a right turn, as I came to a stop for the 4-way stop, and then again as I started out just as I began to round the corner I lost all traction front (motor spinning out wheel and throwing water and gravel) and rear and started to skid sideways, but a foot pushed down kicked me upright as I cut throttle (and of course stopped pedalling) and I got traction back to get going. So no crash today.

Rode thru the deepest puddles I could find (8 or 9 inches at a couple of street corners, after the very short deluge part of the storm, including the leftover rain from a couple of nights ago, I'm sure). Wanted to test the water-proofness of the wiring, motor, cargopod, etc. No problems during or after the ride. Also was testing traction in water, and I didn't have any slip I could feel, even cutting throttle and then ramming it to WOT. Another reason I like these tires' tread pattern.

Ended up with these readings:
119Wh
3.369Ah
881Wp
25.12Ap
31.03Vm
37.16V end of trip resting

It wasn't a real drenching, but with the steady drizzle it was very wet and still the bike kept working for the whole 34 minutes, which is good. It's not that likely I'd be out in any rainstorm for that long, but good to know I shoudln't have a problem if I am.

Once I fix up the CA then I will go ahead and build that "dashboard" inspired by LI-ghtcycle's, though I probably won't have a fairing for it like his. Then I'll be able to seal up the whole wiring and electronics in the dash box so it'd be waterproof. Probably the controller will go in there, too, since it's closer to the motor and phase currents are going to be higher than battery currents so shorter phase wires is more critical than shorter battery wires, if either one matters at all in this setup.

I have been putting off the dash idea first because I did not really like any of the positions the PDA would end up in to keep it out of the way of the Fusin headlight/battery gauge unit and it's keyswitch, and later because I did not want to build one to hold the PDA and WU (or TWM) and then have to redo it to replace those with the CA.

Now I have to go change clothes, because I'm starting to get really cold all wet like this.

[amberwolf]

More data from work commute with the same motor/batt/cont combo. Again I only pedalled at startups, to cut down on the high-current draw on the NiMH. Based on the Wh numbers I must be riding a lot differently going home than coming to work, because Wh/mile is a lot less even though it's a significantly longer ride, and typically I ride faster for longer stretches going home, I thought. Apparently not, though.

To work:
Started at 40.9V, ended at 38.9V
1.207Ah
43.5Wh
24.66Ap
818.9Wp
32.42Vm
2.303 miles
18.89Wh/mile

From work:
Started at 39.1V, ended at 37.83V
1.415Ah
38.9Wh
24.86Ap
814.2Wp
31.58Vm
2.973 miles
13.08Wh/mile

[amberwolf]

Doing some experimenting with programming the Lyen 6FET, finally. Changed the current limit from default of 25A to 16A, which should basically match the Fusin controller. Doesn't affect the final top speed of the 9C but it definitely affects it's acceleration capability. Since my main reason for it is to help protect the NiMH without me having to think about it anymore, that's ok.

If I really want better acceleration, it'll give me more motivation to finish the stupid LiCo testing and/or the TS pack and BMSes for them.

I did discover that the preset for the speed percentage on the second (middle) "cruise control" settings directly affects the non-cruise max speed. In other words, if you have it setup so you can use a three-button speed switch instead of a throttle, this would be the middle setting.

But if you don't actually have such a switch hooked up, and instead have a throttle, it makes that your *max speed*. With it set to 100, my no-load wheelspeed is ~27mph, but with it set to 60, that drops to ~16mph (about 60% of 27).

So make sure if you're programming your controller, and you don't have one of those switches, leave that second one at 100%.

Interestingly, changing either the first or last presets to anything you want doesn't affect the actual throttle min or max. Just the middle one. Seems quite strange, as I would have expected that to be the last one to affect max speed.

I'll have to wire up one of those switches (as I have one from the Fusin) and see if perhaps it actually works by switching from 1 to 3 to 2, and thus if the program has these settings out of practical order (even though they're in numeric order). I do seem to recall that the PCB silkscreen on at least one controller (might not be this one) had some pads marked X1 X3 X2, in that order, which would then make sense of the way the settings actually work in the program.

I'm pretty sure this result will be applicable to all 116 controllers, and possibly to all the programmable Infineons in general.

I have not yet experimented with the regen settings, which are already set at the highest numbers listed (75V and strength of 2), but I would like to make stronger regen braking if it is possible.