The SB Cruiser : Amberwolf's 2WD Heavy Cargo Trike & Dog Carrier

simple in concept...convoluted beyond belief in the little details. would take literlaly pages to write all that down, so we'll just say...it works, finally, after many hours. performance we'll have to see aobut after some tiem ina operation, but right now, it's bedtime.

the pic of the fatbike fork shows how badly bent it is, it looks like the front is facing my feet in the pic, but it is not--that's the back. :/
 

Attachments

  • dsc08270.jpg
    dsc08270.jpg
    46.1 KB · Views: 4,331
  • dsc08271.jpg
    dsc08271.jpg
    40.4 KB · Views: 4,331
  • dsc08272.jpg
    dsc08272.jpg
    32.4 KB · Views: 4,331
  • dsc08273.jpg
    dsc08273.jpg
    50.3 KB · Views: 4,330
  • dsc08274.jpg
    dsc08274.jpg
    38.3 KB · Views: 4,330
  • dsc08276.jpg
    dsc08276.jpg
    41.8 KB · Views: 4,330
  • dsc08280.jpg
    dsc08280.jpg
    54.4 KB · Views: 4,330
  • dsc08279.jpg
    dsc08279.jpg
    55 KB · Views: 4,331
  • dsc08278.jpg
    dsc08278.jpg
    48.7 KB · Views: 4,331
  • dsc08277.jpg
    dsc08277.jpg
    48.9 KB · Views: 4,331
Sorry your body is hurting. Glad you are back on the road.

:D :bolt:
 
So far it's a success. Much much stiffer than the fatbike fork, or the crappy suspension fork. About the same as the suspension fork I broke in the crash. I'm not totally sure but I think the headset is not perfectly tight either; have to check & fix that during the next step.

But the ride height is a little tall so I'll adjust the legs upward at least an inch, and retry. Not tonight, maybe tomorrow after work if I'm not wiped out again like tonight (I think I'm over the cold, but now I have to regain the energy I lost to it...and my voice, which is still gone). May end up shortening it as much as 2 inches, probably not more.

Full hard braking from 20MPH skids the wheel with just a little judder (of the tire on pavement, not disc/caliper), not quite as stiff as the rim brake was on the broken suspension fork; we'll see if the adjustments help with that. Gradual braking is normal.



There's an odd sound during turns on the disc brake; I think the wheel itself is "wiggling" on it's QR axle just a bit or something from the sideloads. I already checked spokes and they feel tensioned correctly. The bearings feel ok, and it doesn't wiggle like loose cone-nuts would.

But I did have to remove a couple of mm of metal from the inboard side of the dropouts, to get the axle between them (something wierd about the lowers of this fork, may have been messed up before I started working with them; I never checked to see if an axle fit in them right beofre using them). I first tried spreading them with the jack, but it wouldn't "stay", and I didn't want to force it and damage the fork legs. So I ended up removing some metal instead to give the space needed. And doing that required also removing the same metal from the disc caliper mount tabs on the lowers, but they don't appear to flex under braking abnormally.


I'll probably use some of my cargo ratchet straps around the crowns to force-compress the headset together while retightening the clamp bolts (since with the top clamp it doesn't really need a star nut and top cap to do that job; it didnt' come wiht one installed, and while i do have a spare I can use, I'd rather leave the one more component out if I don't need it.
 
amberwolf said:
There's an odd sound during turns on the disc brake; I think the wheel itself is "wiggling" on it's QR axle just a bit or something from the sideloads. I already checked spokes and they feel tensioned correctly. The bearings feel ok, and it doesn't wiggle like loose cone-nuts would.

Or the fork itself could be bending under load causing the caliper to go out of alignment with the rotor.
 
RunForTheHills said:

Or the fork itself could be bending under load causing the caliper to go out of alignment with the rotor.

I suppose, but that kind of bending would seem to require that the dropouts move with the fork legs, and the axle/hub/rotor would move with the dropouts, which should keep them in alignment?

It seems to my presently fuzzy mind that axle movement (which would move the hub and thus the rotor) is more likely?

I'll have to investigate it to find the root cause, regardless, and fix it.
 
When you put it that way, I agree that it is unlikely to be the dropouts bending. Does the hub use cartridge or loose bearings? If they are cartridge bearings, you can't really adjust them like loose bearings with cones. There will be some play in them.
 
plain old cup and cone, so I can adjsut them. it's a crappy old wheel off a "mongoose basher", which was a very heavy "bike". I used it before on the trike when I tried some crappy disc brakes (bso type) that basically didn't do a thing to slow me down even at walking speed. :lol: But otehr than this noise/etc issue during turns (sideloading) it works great with the Avid 200mm rotor and BB7 caliper.
 
I guess that is the downside of a larger rotor. It gives you less tolerance for lateral movement of the disc as the movement gets magnified more due to the larger diameter.
 
haven't gotten to deal with anything yet, still been sick and wiped out after work, but better today, hopefully get this stuff worked out tomorrow on my day off.

another thing to deal with is the mirrors. they've had issues staying in place for a while, but with no shock on the front at all now it's much worse, i cant' go 1/8 mile witout having to readjust them, pullling the left one back towards me cuz it just rotates forward like 40-60 degrees; worse and faster when i decelerate at intersections cuz of those forces plus many more and larger bumps (waves in the asphalt). the right one rotates up and back....

i can fix the left one relatively easily by adding friction washers with ribs between the mirror stem and ti's bar mount, then retighten the pivot bolt.

the right one...the plastic on the ball joint case is cracked, so there really isnt' a way to tighten it up. i'd have to rebuild the joint somehow; it's part of the mirror casing.


another issue is that with the tiller / stem clamp raised up by the amoutn it is, teh right mirror hits the canopy on any hard left turn of the tiller (which only happens at very low speeds, like parking lot stuff) but it messes up ti's aim when it does. this will b fixed by movign the tiller back down where it was, which i'll do in the other fixes for the fork/etc.


but i'm considering moving the mirrors off the bars completely, and making a fixture to hang them from the corners of the canopy instead. then they won't be affected by the road vibration directly from the fork, and also they wont' pivot with the steering which with the long tiller means i cant' even see the mirrors when i'm sitting waiting at a turn that makes me pivot the front wheel a lot, or when i'm in a parkng lot situation where i hae to make sharp low speed turns, etc.

anyway, it's a thought, dunno if i can do it this "weekend" but i can start thiking about it.
 
Is the right mirror necessary? I only have one on the left side, but I don't have a vehicle like yours. I ride in the bike lane on the right side of the road and use the mirror to check for car traffic on the left. With the canopy and cargo area blocking your view, I guess you can't just turn your head and look.

Motorcycle mirror mounts are only a few dollars on eBay and then maybe you can find some old motorcycle mirrors. Or you could get a cheap set of mirrors like these for $10: https://www.ebay.com/itm/Motorcycle...388488?hash=item25ee55c588:g:YkQAAOSw6IdcKHo2
 
These *are* motorcycle mirrors. ;) The mounts are fine, it's the (parts of the) mirrors themselves that just need fixing, because I've had them a while, and theyve been shifted/moved a lot in that time, and worn (or in the case of the right one, cracked at the ball pivot housing).

I find mirrors like these at Goodwill every so often, but haven't for over a year now. I might end up buying new ones, but will avoid spending money as long as possible (because I have too little of it, and spent more than I should on the brakes, though it was somewhat necessary). The hard part is finding ones that are large enough surface area for viewing--if it's not at least as big as my hand and fingers, it's not big enough, and I end up having to move my head and hunt around in it to see enough area behind me. My left mirror is the "domed" type which helps with that, but it's difficult to judge actual distances with it.

Bicycle mirrors are all so light that they shake like crazy and are unusable--you have to stare into them a long time to get used to teh shaking in order to tell what the hell is being reflected, so instead of being hit from behind you crash into whatever was in front of you.


The right mirror is very necessary, since I am not always in the farthest right, because of the way the roads are laid out (right turn only lanes, etc), and because I have to ride in other lanes to get where I"m going, often enough.

I *can* turn my head and look, but there is no time for that in most situations that I actually need to see behind me--by the time I've turned my head and back, the situation has changed and the info is invalid, because the cars are generally going significantly faster than I am.

I also need it for monitoring pedestrians and other sidewalk traffic, and in parking lots, though a headcheck might suffice for most of those.
 
Haven't tried epoxy, but I had already previously tried some of the super-sticky gooey glue used in mousetraps, so I could still move it by hand if I had to...it didn't really cahnge the problem much. :/

I didn't get teh chance to clean the gunk out and epoxy it in place, but I did think of two other things (taht I didn't get to try yet): I could use a hose clamp around the cracked portion, or I could put a pointy screw thru that plastic and into the plastic of teh ball joint itself.


I did fix the part about it hitting the canopy by angling it's support downward, movng the mirror itself down and out.

I corrected the left side by tightening the bolt and nut sufficiently, didn't end up using any washer/etc in there. Seems to be holding fine so far.

I corrected the ride geometry back to what it would've been with the good suspension fork.

The SF would, of course, compress with me sitting on the trike, and even more during braking / in turns (which did not occur to me until yesterday), so the effective fork length was shorter than I had accounted for both when working with the fatbike fork (which *was* the right length to match the *uncompressed* SF) and when setting up the new triple-clamp fork.

So I loosened the clamps and moved the fork legs up a little more than a couple of inches, a small amount at a time, then test rode it, utnil I reached the same "feel" it had before, when making turns. I am not sure it's exatly the same, but it's much better than any solution since the SF broke in the crash.

I might lower it even further just to experiment, but right now it rides well enough.

That's as far as I got, though.
 
i forgot--i also shimmed the element fork lowers where they expand in diameter near their tops, so the wiggle under braiing is gone
now.



today on teh way home from work, the right motor began to shudder as if it were shorting intermittently. wasn't an overheat, as it was only 39C, which was only about a dozen degrees above ambient. Same even without applyin gpower to it.

sstopped in a parking lot and disconnected the controller, and same, so wasn't in the controller.

had to just take it slow, less than 5mph, the rest of the mile and a half or so home from there. During the ride i discovered that whenever sidelaoding forced the wheel toward the right (pulling it rightward against the axle), the problem went away. this meant the problem was almost certainly a phase shorting against the motor cover. was only about a half mile from home by the time i figured that out, so ust kept going.

got home, suffering thru the thunderearthquake level bass from the party more than a block north of my house, which i could start feeling about the point i figured out the problem, half a mile north of there, and just got more intense the closer i got to home. all the dogs were hiding in the house, which is very unusual, cuz jelly always is waiting by the gate, and often the rest of them are. yogi was shaking laying on the bed, cuz he was so afraid of the thundering bass shaking the ground and house and him. had to feed him dinner in there cuz he wouldnt' leave. he didnt' wanna eat which is pretty bad for him cuz he's always super hungry and would eat any amount of anything he could get hold of. had to hand feed him a bit at a time. puppies ate fine, but didnt' wanna go outside and stuck with me wherever i went in the house. kirin didnt' wanna eat either and wuouldn't leave my side even when i went outside, but she was freaking out over the earthquake bass.

anyway, after all that i eventually had to go out to get the wheel off to fix it cuz i have to work tomorrow and have no alternatives right now. so fumbling for half an hour tyring to get the wheel off while being assaulted by the bass, which was rattling my tools, bones, and teeth, after having called the police and been told they'd already gotten so many complaints from the area they wouldnt' even take mine, i took the wheel in the house and opened it up.


found that indeed some of the insulation coating on three strands of a winding, really teh wires between windings, were rubbed off on the outboard side, and scratches / zaps on the inside f the cover. and the smell of ozone from arcing.

oddly there were other things sticking out further than those strands, but they didn't appear to be rubbign on the cover...whatever.

i recoated them with insulation and added thin spacer washers at the axle shoulder the inbaord side of the bearing id rests against, to push the cover farther away from the insides, taking them from the other side since that one appears to have plenty of clearance, so all the washers oare now on the outboard side.

reassemnled and spins easy no short, put back on trike, thankfully bass assault had finally stopped, there were flashing blue/red lights on the tall palm trees north of the area so i guess the police came from all the other people tired of this crap. i wonder if they had to shoot someboyd to make them stop.


anyway trike works agian but coudlnt' raod test cuz i can't see straigth and my head is poudning.

needed to charge trike so put in shed and plugge din but the satiator isn't working anymore. it powers on but no relay click when plugging into the lighting pack. cant' ell wahts hapening cuz of coure the screen is still not working cuz the cable to it is damaged from when i had to open it opu to fix the output ground that broke off at the pcb.

so had to dig out the rc charger to do the lighting pack charging. ists the only ohter thing i have laying around that can work at 16v as a charger besides the big sorenson, except for one smaller sorenson that only does like half an amp.


sorry for ht ranting have a whole body migraine from the fucktards with the earthquake machines that don'g give a flying frock aboutanyone buat themselves. had to turn the screen brightness all the way down just to be able to type this up after what feesl like half a bottle of tylenol and ibruprofen.
 
they're not neighbors; they're multiple streets north of here.

but there are neighbors that do this too, every day for just long enough to hurt poeple and make them mad, but not long enough for police to catch them and so they wont' do anything about it if it isn't hpaneing right at that instant

mostly it's one family right t a 29th dr and echo lane in the green house, mostly the guy in the copper truck and hte guy in the white truck, the whtite truck roars around hte neighborhaodo at varoiuas times day and ndight, jstu their ehaxst is loud enogjuh to set off car alarsm over a street away. then thye sometmis also turn up their bass while doing that and itll rattle your bonse and teeth and knock stuff off ashelvse and set fof car alarms and scare all the anmials, you can see birsds fly fof in clouds when they approach an area. sometism they have firends or whatever over with other fucktard cars that shake the world. sometims they leave their stuff runing long enoughf or policse to get there but nothing ever happens to stop it evne thne. the ymust be buddies with the police. tried talkign to them politlye aobut the bass befor involeinv police but they dont' give a frock and just threatn

been ogoing on for years nwo

nuff ranting at least its over tonite unlike otehr nights where i tgoes on practalyt till doawn.
 
motor survived the commute, so good news for today at least.


bad news for tonite is more assholery from the fucktards at 29th drive and echo lane shaking the neighborhood for blocks with their thunderquake bass trucks
 
Ok....so....I"m kinda tired of the problems with the MXUS motor(s). One thing or another keeps going mechanically wrong with them..breaking axles, mostly, then other problems either directly from the axle failure damaging things during the break, motor windings shorted to the stator laminations, etc.

Few days ago I when fixing the last problem posted above, I was going to unlace the 4503 from the Zero rim and lace the Stromer Mountain 33 Ultramotor into it instead, but the M33UM is smaller diameter than the 4503 by a lot, so ther'es no way to use the spokes I have.

So I ordered new spokes from Grin Tech, probably be a couple weeks before they get here (I picked the cheap slow shipping option). Using some of the donations from various ES members, I also ordered some of their motor cabling (3x12g, 7x data/etc wires) to replace some of the messed up motor wires in various motors I have around here that I could use for wheels if I only had good / long enough wires for them. (some have cracked insulation on phases, or damaged wires in various other ways, some are just so short they don't have enough length to even splice at the axle exit). Two of the clearance controllers that support sinewave and proportional regen, to replace the generic controllers that have fixed regen and sometimes unpredictable current limiting results). Some other bits and bobs that wouldn't be worth shipping on their own but that I can use, or need, for one thing or another (thermistors, speedo sensors, etc). If I had more money I'd've ordered a number of other things (like the 24pole PAS unit, and the slim digiaux button control unit, etc). Didn't really wanna spend the money I did, cuz this takes away from the dogfood fund, but I got a tiny raise at work that should just about make up for it over the year.

The reason I went ahead and bought the stuff is because I had to order the Satiator screen replacement anyway, cuz mine went out a few months back when I fixed the output ground that had broken, and I somehow damaged the cable from the board to the screen. (I could probably fix that, too, but I don't have any conductive paints, and it would take some fine work to solder wires across it, which I don't think I can do anymore). The screen doesn't really matter for my usage of it (though the data was interesting, it wasn't necessary, and I can change settings via computer usb anyway), but it recently stopped charging the lighting pack, and I can't see the error to know what's happening. :/

It clicks at AC power connection, so it's functioning. It could just be a connection between the charger and the port on the trike, but it tests ok for continuity. Didn't do any other tests besides a visual inside teh Satiator, whih looks ok. Guess I'll find out when the new screen arrives.




Regarding the M33UM motor, I was going to use it as the front motor wheel of the trike in the 26" rim it came in, (had used it on the now-defunct fatbike fork), but I think I need reliable rear wheels first. So it'll go in the right rear wheel for now, and I'll save the spokes and rim it came in if I manage to get a better solution for the rear (or if it doesn't work out back there). Maybe I'll figure out a way to add a disc brake rotor mount to the old front 9C wheel off CrazyBike2....
 
Finally had a chance to get back to testing the SFOC5, after a month of evyerhting else taking up my spare time.

But didn't get to do any "real" testing, cuz it has a problem with rapidly heating, even when disconnected from everything but battery power and it's little status led / switch panel. :( Still have a bit more investigating to do, but something is wrong.

https://endless-sphere.com/forums/viewtopic.php?f=2&t=30680&start=200#p1472052
https://endless-sphere.com/forums/viewtopic.php?f=2&t=30680&p=1472072#p1472072


I did the test on the 4504 rather than the 4503 which I did a month ago, when I first experienced this rapid heating problem.



In other news, I've been slowly drawing up the "schematics" for the dashboard switches that will control which throttle controlls which motor, and if the CA does any limiting/etc., and if the PAS controls one or both or no motors, and bypass switches to disable either throttle or either controller, and fwd/rev switches for each one or both. And switches to disable regen on either controller if needed. I should have somethign clean enough to post "soon", right now it's all unreadable sketches drawn during lunches at work. :oops:

The places where multiple throttle signals will come together will be buffered with 1:1 opamps. Partly this is just ot buffer the signal, but it's also to somewhat isolate the electronics of each section from the others in case of shorts, etc. I'd rather replace opamp chips than throttles or the CA, etc.

I still have to figure out the mechanical part of the steering proportionality control. This will feed more power to the motore that's outboard in a turn, and less to the inboard motor, so that the power can help me turn. Am thinking a potentiometer with a gear on the shaft, mounted on the trike frame, and a toothed partial "wheel" on the steerer that meshes with that to go from centered-balanced power to more left power to more right power depending on which way the steerer is turned. Then use that resistance to change the gain on the opamps that buffer each throttle's output.

A magnet with hall sensor would be easier to weatherproof, but I don't know how I can make it happen any simpler than the pot; I'd still have to proportionally gear things because of the small phsyical range of movement I can do for magnets vs sensor sensitivity.

Any other ideas out there?
 
More SFOC testing today, to narrow down the rapid heating issue.

details here:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=30680&p=1472322#p1472322

But basically, with two other firmwares besides the Oct18 fw (which worked fine back when i was using it before the long break due to axle problems/etc, AFAICR) while it does heat more than I would expect for a cotnroller doing "nothing", and not connected to a motor, it doesn't heat up rapidly and unusably like the Oct18 fw did.

I have not yet tested on-ground, under load, with any of these, so it's possible that it would still heat up like that when actually using it. But at least with offground tests with the 4504 motor, it appears to operate relatively normally.


Hooked back up to the generic 15fet for my commute tomorrow (and the rest of the week thru tuesday). Waiting to hear back from Incememed before attempting normal usage of the controller, about the results of the resistance tests done on the disconnected SFOC phases (which look normal to me).


The package from Grin Tech is on it's way here, theoretically arriving tomorrow (probably while I'm at work). Once it does arrive, I"m going to build a completley new wiring harness for the throttle, ebrakes, etc, to route to the new Grinfineons, rather than ripping out the old controllers and using the existing wiring from those. That way I can switch back to the old stuff easily simply by switching over the throttles/ebrakes and phase wires between them, via switches on the T/EB, and the PP75s on the phases. (which means changing out the PP45s on the Grinfineon phases to those). I'll add a pair of PP45 sets to the load side of the shunt for the Grinfineons to plug into for power.

Thus I can make the wiring up as I have time, instead of being rushed over a day off to get it all done (especially since it's over 100F nowadays, past around noon thru post-sunset), so I can use the next day off to test it all out thoroughly and fix anything that doesnt' work...and then being screwed if I can't get it to work, unable to quickly switch back to the old working stuff.

Then, once the Grinfineons and new wiring is all proven to function fine, I can dismantle the old stuff to keep as spares or use on other projects (like powering the big duallie trailer, the MkIVa, with a couple of front hubmotor wheels).


But the first thing likely to get done is lacing up the Stromer33 Ultramotor into one of the 20" rims, to replace the 4503 motor that keeps having various issues (axle primarily, and related).

I would prefer to lace it into the intact rim on the 4503 itself, but that means the risk of having to use the loose-axled HSR3548 on the trike if I can't finish the lacing...and since I would also like to still use the 4503 as a backup (it is better than the HSR3548, at least) wheel, I'd then have to delace the HSR3548 and use it's damaged but still usable rim on the 4503.

Probably not something I can do in one or two days if things dont' all work out right, so I may end up just putting the Stromer33UM into the HSR3548's damaged rim. (wish I hadn't cracked the bead seat area of the first of these four rims...it's essentially unusable until I can get someone that can TIG aluminum to fix that, assuming I can do that for a lot cheaper than a new rim).
 
I've started the process of putting the STromer Mountain33 motor into the damaged rim out of the HSR3548 wheel. So far, that's just the stages of unlacing both, and testing with 8 spokes (two on each flange in each of four quadrants) to make sure the 88mm spokes are the right length--they seem to be.

Next, I probably have to open the motor back up, and install the new temperature sensor from Grin, as well as probably changing out the old extension cord wire I used for phases, and old serial cable I used for hall sensors, with some of the new motor cable from Grin. I may not end up doing this, but since I ahve to unlace the wheel to do anything inside it, it'd be a lot easier to just do it now rather than wait till after it's laced. The idea is to make as reliable a motor wheel as I possibly can.


After that, I'll test it out in place of the MXUS 4503 on the right side, with the generic sensorless 12fet presently on there. I'll need to use a 10mm wrench as a torque arm on the outboard side, though, becuase that dropout was widened to 12mm for the wider axle end I welded on the 4503's outboard side, and I don't want to modify the dropout itself until I know this motor works as it needs to.



I also tested out the Satiator with the replacement screen, just plugged in, not installed, and at least it powers on and displays the normal stuff. Haven't plugged in a battery yet to charge, that will wait until I actually install the scrren so I don't destroy this one's cable, too.

Maybe if I"m really lucky I'll be able to fix the other one as a spare, eventually...but I'll probably have to have someone else with steadier hands and better eyes do that.


Didn't get anything actually *done* yet, but at least started.

Also still drawing up the dashboard switch panel box (that will also hold the CA, and a lighting-power-to-5v-USB DC-DC, turn signal blnker, etc), to make sure it will all fit in the space I have available on the handlebars around the stem.
 
Got the Satiator fixed, using the display kit from Grin.
https://www.ebikes.ca/shop/electric-bicycle-parts/charger-accessories/satiator-oled-repair-kit.html

Easy enough to install, but since it's for the newer versions that have the power cord coming out the end instead of the front (like my old probably original version), I had to cut a hole in the cover sticker (taht makes the waterproofing seal over the display/face), to accomodate the IEC plug mount on the face of teh Satiator.

It's supposed to come with both stickers according to the picture on the page linked above, but it only comes with the one without the hole.
oled_repair.jpg

oled_repair.jpg


Once that was done, the dipslay worked again, and came up wiht no errors. Hooked up to the trike's lighting pack, and didn't detect it was plugged in.

Re-continuity-checked the cable from Satiatoor to XLR, was fine. Same fro the plug inside the satiator, on i'ts pins to the wires on the back...and i could visually see the wires internally were ok, too.

To test those wires from the output connector board to the main board, I had to take the whole main board out. In hte process I almost destroyed the new screen's cable, too, but managed to only break off the connector's brown retention clip on the board it plugs into instead. :/ So to make the screen work again after reassembly I had to put a thin sliver of dense foam between the cable and hte PCB to force it to the right angle to hold it against the contacts, as I can't fix the retention clip. :/

Anyway, while taking the main board out, the signal and positive wires from the output board to the main board came right off the output board, so likely just one strand of wire was holding it in; maybe even just touching and not even soldered. Guesing this meant a continuity test worked but actual operation could not.

Resoldered the wires to the connector cups, and reassembled, and now it works fine.

Still cant' find a nut to hold the connector itself to the casing, so it will still wiggle around and break again...so I used a block of dense foam inside behind it to keep it from moving as much, and put a ziptie tightly around the threaded part of the connector to help hold it against the case. I"m sure this greatly reduces it's waterproofing, without anything holding the o-ring against the case internals, but best I could do without gluing it in. Might end up doing that, wiht silicone, if I ever mount it on the trike.

Forgot to take any pics while it was opne, though. :/
 
I really need to start over with the Cycle Analyst 3.13(b2, IIRC) settings; I've gotten nowhere over the last few months with trying to troubleshoot the dangerous problems with it's speed limiting.


Basically, I have three presets, with the only difference between them being the speed limit itself. I attached what I think is the settings file I last used, probably should just download the present settings in teh CA itself.

One is 5mph, one is 10mph, and one is 20mph.

The CA only controls the rightside motor, via either torquePAS or hand throttle. THe left side is a separate hand throttle.

The controllers are just generic trapezoidal types; the one the CA has conrol over is just a 12FET with about a 30A current limit, powering an MXUS 4503 in a 20" wheel. The other is a 15FET powering an MXUS 4504 in a 20" wheel.

There are no power limits (whatever the max in the CA is). It's in high range mode, not low range, so maybe it's a bug related to that (there have been others).

The problems:

--if i switch from any preset to any other, and was pedalling at the time (usually am), or the CA-controlled motor was running, the CA may or may not "decide" to obey the now-current preset's limit, usually until it has reached the max for the original preset and then droped to the minimum for the new preset (sometimes the minimum for the original preset) and then accelerated to the max of the new preset. Sometimes it never obeys the new preset, and I have to switch to a different one and back again--but no matter which behavior, it does always display the correct preset-changed message, with the correct preset number / name.

For instance, if I was on preset 3, the 20MPH limit, and stop pedalling as I coast into a driveway for a parking lot, then start pedalling to go up the driveway, and switch down to the 5mph preset1 as I enter the parking lot at anywhere from 3-4mph up to 7-8mph, continuing to pedal, the CA keeps full throttle on the motor. If I let it, it'll continue up to around 22-23MPH (when it would begin limiting down for the 20MPH preset3, another problem entirely), and then it won't reapply throttle until it gets down to 2-3mph. ONce it does that, it'll apply full throttle and accelerate back up to 7-8mph, then cut throttle, wait till down to 2-3mph, then reapply full throttle, and repeat that cycle until I stop pedalling or change the preset. (using the hand throttle does the same thing, except I have control of the actual throttle output to some degree, as long as I am not also pedalling).



--if I am in preset1 20mph, the minimum is extremly variable, though the max is much more predictable.

What I *expect* is that it will simply always limit to 20mph max, and never allow me to exceed that speed via PAS or throttle (for on the flats riding). And that if I am applying enough throttle (via PAS or hand throttle) that it will never drop below 20mph, either (as long as power output from the motor is sufficient to overcome wind resistance/etc.)

What actually happens is:
--accelerating from zero, it keeps accelerating until speed is 22-23MPH (most of the time, occasionally it keeps going up to 25MPH), and then it cuts throttle entirely, until speed decreases to 18.2-18.8mph, most of the time.

--sometimes, under the same conditions and the same place on the trip / road that previously it did the above, it will decrease speed down to 15mph, or even 10, in some cases almost to zero, before it will reapply throttle. (watching the screen showing throttle output, it's kept at minimum until that moment by the CA). When it reapplies it, it does so at maximum, as it should, but of course it takes significant time to reaccelerate back to full speed.

When it *does* actually do so, it *then* finally limits to the correct speed, 20MPH, within a couple of tenths, until I either switch presets, let off hand throttle or stop pedalling (whichever I was doing, usually pedalling). So the speed limiting does actually "know" how to do what it's supposed to, but only after what I consider wild excursions above and below the limit.

But if I let off throttle, or stop pedalling, even for an instant, then there's two possible behaviors when I resume:
--it may reaccelerate to 2mph and hold that as it should
--it may restart the whole extreme excursion cycle, with any of the possible extreme listed above.


--- On preset 2, the 10mph, it never settles at 10mph. Ever. It always goes way up to at least 11-12mph before cutting throttle, then goes down to 7-8mph (sometimes 5-6mph) before it reaccelerates back up past 11-12mph, and cycling like that continuously.

Preset1, 5mph, does the same thing, except the high is 7-8mph, and the low is 2-3mph.



The dangerous part is the 20mph behavior, because this is in traffic, where everything the trike does in response to my input needs to be completely predictable.

The surge in speed is easy enough to fix, just let off pedalling.

But then the decrease in speed requires me to use the other non-CA-cotnrolled motor to maintain speed, until the CA "decides" to get around to following the limit it is set to. This can take anywhere from a few to several seconds at minimum, to never ever reaccelerating at all (presumably because speed hasn't dropped below the absurdly low excursion value it seems to expect). This is true whether I keep speed just under 20MPH with the left motor and handthrottle, or only maintian just at 18MPH (which is below the "maximum minimum excursion value" of around 18.6MPH).

If I let the CA just do it's thing, I'd be hit from behind by the cars behind me as it slowed more and more, especially when it expects me to stop completely before reactivating throttle.


The problems are exactly the same whether using a 3-pole speed sensor (three magnets on the front wheel) or a single pole/magnet speed sensor on the front wheel, with the appropriate setting for taht in the CA.

I've forgotten at this point all the settings I've varied in attempts to fix this, but nothing so far has had any effect. I'm sure I'm just screwing something up...but havne't figured out what it is.
 

Attachments

  • SB Cruiser 030519 CA3-13b2 thermal sensor test 1 000001.zip
    663 bytes · Views: 67
Back
Top