Strange CA problem

Hey guys,
I've got a bit of a strange CA problem.
I build a kit for a guy a while ago and all was well until he cooked the HT3525 going up a hill at 5km/hr with a heavy load on board.
The motor was replaced with another but ever since the CA has been doing strange things - the voltage overreads by 0.8v which isn't alot but it's noticabley too high when you see the full charge on a lipo pack. I tested it back to back with 2 other new CAs and and a multimeter and they all read the same (correct) voltage.
I could live with that (the guy didn't even notice) - the main issue is the speed is out of whack, it jumps around all over the place and is reading way too high, causing the bike to promptly cut out as soon as you start accelerating. All the settings were as they should be, and the same as another CA I tested which worked fine.
The problem is he's in another state so I can't test it all on the bike.
So I got him to send the CA and his controller to me, both of which functioned fine when I bench tested them (aside from the voltage over-read).
I tested his CA on 2 different controllers (his and one of mine) and it worked fine.
I tested his controller on a new motor with my CA and his and both read the same.

So the only other thing I haven't tested is his motor (which I can't due to being interstate). The voltage over-reading aside, the speedo being whack has to be a hall signal issue right ? All I can think is it's not getting the hall signal correctly from this new motor, but the motor itself functions fine (with the CA disconnected) - so the controller is seeing a proper hall signal.

Any thoughts ? I'm stumped!

Not sure. Is it a direct plugin model? Is it the same type of motor he has now? Different motors I assume may have different hall signals, enough so the controller works but the CA had an issue. Maybe he just had a dodgy hall signal on his setup, that could explain why it works for you.

I have spare CA shunts, and wheel some wheel sensors if you want to try converting it to a stand alone.

Voltage over reading may indicate he CA board is damaged. Perhaps if a ground reference is lost this could kill the voltage and speed sensor inputs.

I suspect there could be excessive voltage fluctuation somewhere between the battery, which would have the potential to cause the step down voltage circuitry in the Cycle Analyst get confused. The other possibility could be the voltage fluctuation from the battery to the controller rendering hall sensors input voltage and output signal unstable. Hard acceleration may drive weak cell of the battery string happened like that. You may want to ask the guy what type of battery and the age of his battery pack(s) and go from there.

If the motor cooked, could the current flow also have overheated the controller? A bad solder joint for the CA wires to the controller's pads at shunt and hall could cause the kind of thing you're seeing.

Also a wire broken inside the insulation or bad crimp (or corrosion between contacts) on the wires between CA and controller could cause the speed problem. When the connection is good it will read fine, but as the wire vibrates for any reason it could partly break the connection and get wierd readings. (this happened to me on my wheel speedo sensor when one of the dogs bit into the wire).

adrian_sm said:

Not sure. Is it a direct plugin model? Is it the same type of motor he has now? Different motors I assume may have different hall signals, enough so the controller works but the CA had an issue. Maybe he just had a dodgy hall signal on his setup, that could explain why it works for you.

Click to expand...

Yep DP model, same type of crystalyte motor (slightly different wind though) and with a crystalyte controller. So it should be all plug and play.

I have spare CA shunts, and wheel some wheel sensors if you want to try converting it to a stand alone.

Click to expand...

Thanks but I'll try and sort it as is if possible - with this guy being interstate it's a bit hard to sort out over email.

Voltage over reading may indicate he CA board is damaged. Perhaps if a ground reference is lost this could kill the voltage and speed sensor inputs.

Click to expand...

The voltage seems consistent, just a little over reading - the speed worked ok for me on on my controller and his on 2 different motors though, which is weird. I checked for corrosion and bad contacts and all looked OK. This guy soaked it in WD40 prior to sending it back to me (I could still smell it and feel it on the wire) but everything was working as it should on the bench. And its not particularly vibration / bump related because I talked the guy through some tests on the phone and it does it even when freewheeling.

Lyen said:

I suspect there could be excessive voltage fluctuation somewhere between the battery, ... hard acceleration may drive weak cell of the battery string happened like that

Click to expand...

No the battery is fine, it did actually have a weak cell which I replaced at the same time as trouble shooting the CA. When I tested it (and sent it back) all was working fine and the battery was sag free (especially freewheeling - which as above causes the same issue)

amberwolf said:

If the motor cooked, could the current flow also have overheated the controller? A bad solder joint for the CA wires to the controller's pads at shunt and hall could cause the kind of thing you're seeing.

Click to expand...

Yeah I thought it could be that, but his controller worked fine with both his CA and one of mine when I tested them - if it was his controller causing the issue I should have had the same problems when I bench tested it.

Also a wire broken inside the insulation or bad crimp (or corrosion between contacts) on the wires between CA and controller could cause the speed problem.
When the connection is good it will read fine, but as the wire vibrates for any reason it could partly break the connection and get wierd readings.

Click to expand...

Thats one of the few only other things it could be, which is hard to test. Maybe with the wire stretched out along the bike a bad joint develops, but when laying in a pile on the bench it conducts ? Unlikely, but stranger things have happened to me before.

I had a piece of medical equipment at work that was giving funny readings when testing patients (who sat in a seat opposite me)
The usual thing I did is swing the arm the equipment is mounted on around to my own seat so I can test myself while operating it to see if it works for me or not (to narrow down the problem). Every time I tested the patient the results were whack but when I tested myself it was fine. I scratched my head for ages trying all different things and in the end I found it was a broken wire in the moving arm. When stretched out to the patient the wire went open circuit but when I folded the arm back to test myself the kink in the wire conducted again and worked as usual.

My CA-LRC recently started reading two thousand + mile per hour max speeds intermittently too which is probably the cause of the cutouts I've been randomly experiencing and haven't found the cause of yet. No idea what's causing it. I'm using the magnetic trigger on the spoke setup. It's strange. It only shows up on the max speed screen but I noticed that when I ride and get a cutout here and there that the speed readout on the ca blinks for a split second.

Jay, my money is on a dodgy hall connection.

mdd0127 said:

My CA-LRC recently started reading two thousand + mile per hour max speeds intermittently too which is probably the cause of the cutouts I've been randomly experiencing and haven't found the cause of yet. No idea what's causing it. I'm using the magnetic trigger on the spoke setup. It's strange. It only shows up on the max speed screen but I noticed that when I ride and get a cutout here and there that the speed readout on the ca blinks for a split second.

Click to expand...

I had similar issues (not with the CA but my home brew ebike computer), caused by the sensor not getting a clean signal from the magnet, which caused enough weird spikes for the debounce filter to see more than one event. Check the sensor/magnet spacing and tweak it, worked for me.

I had some weird readings with a new large screen DP CA a few months back. I sent them an email and Justin suggested a bypass cap on the +5. Some of the regulators were oscillating under some voltage conditions. On mine it mostly affected the current and power readings so I couldn't get it to stay calibrated over the discharge voltage range. In any case drop them a line and see if they can help.

I installed the cap and readings did change a bit (which a bypass cap on +5 would not do unless there was an oscillation), but I have not really tested it out fully yet to verify it is fully fixed.

I'm having similar problem starting last week and am grounded until I fix. Lyen controller, CA direct plugin, 9c hub. Intermittent power loss coupled with abnormal CA readout (either mph wildly fluctuating or showing 0.0 mph readouts, and total mileage also way off). I don't know much about electronics but am trying to isolate where the fault is. Last night I:

checked battery packs and all wiring, all seems ok, CA voltage and AH readouts do not appear to be affected and are not dropping when power loss occurs.
test run on new wiring just be sure: problem still occurred.
checked halls and all other connections, inc controller connections - looks fine. ( Everything is inside the evg casing so no physical damage or exposure). 9c hub seems ok, def not a case of cooking the motor.

I managed a short test run with CA disconnected and no problem, but run was too short to be sure as sometimes problem doesn't surface for a few miles, and then it can get increasingly frequent - far from home.
Tonight, I will do a longer test run. Also I will try with CA connected, but with speed sensor taken off wheel. Is it possible the problem is a faulty connection between the speed sensor on the front fork and the CA? Or perhaps an internal problem with the CA, perhaps moisture or dust has gotten in there? The only other possiblity I can think of is a faulty CA dp interface inside the Lyen controller. Anything I missed?

I had a problem with intermitent cutting out on my Stealth Bomber about this time last year, John at the factory was brilliant, not only did he put up with all my moaning, he replaced my BMS and battery, then sent me a new controller, i rewired the whole bike, ran it with the CA disconected, replaced the halls, what ever we tried it still ran the same, sometimes fine and then sometimes like a large twin cylinder motorcycle cutting in and out on one cylinder, in the end it turned out to be a dodgy pot in the Magura throttle, what exactly was happening in the pot i dont know (i still have it) but changing the throttle cured it. It would seem you have tried everything else, dont rule it out for the sake a of a simple swap.

Simon.

I had/have an issue with my CA giving me funny readings also. And like has been mentioned, I found the connector between my CA and the controller to be iffy. If I jiggle it the readings will jump from normal operation to weird things like a negative Amp readings to normal operation again..

I've replaced and thoroughly checked throttles and all wiring connections are tight and have good connections. The magnet is riding 1.5 mm from the pickup sensor and it's still doing weird things like intermittently cutting out, reading insane max speeds and really high total mile readings. I'll try adding a cap next time I'm in there nerding though. I never did get an oscilloscope but I sure need one!

regarding my issue - went for a 20 mile ride this evening without the CA plugged into the Lyen controller. Everything works perfect! Next up, I think I should go for a ride with the CA connected but I will take off the magnet on the spoke so there will be no mph readings fed into the CA. My suspicion now is a faulty speedo cable or damaged magnet so I will inspect that with a fine toothcomb. Good to know everything else is ok though!

Have heard of a situation where the CA gets speed pulses both from the controller and from the sensor. That causes confusion, certainly.

mdd0127 said:

I never did get an oscilloscope but I sure need one!

Click to expand...

You can certainly bring it down here and use one of my scopes to try to check it out. Dunno if it'll replicate while on the bench, but worth a shot.

It definitely has to be loaded to have issues. It could be emi related. I'll figure it out eventually!

The regulator oscillation problem I was (supposedly) having changed with supply voltage. Things seemed to work okay at 75V under light load, but went whacky later in the trip at lower voltage under heavier climbing load.