Can't believe it still works: ancient Crystalyte controller edition

[Chalo]

So... the other day the funky old four-wire throttle that came with my old Crystalyte 5305 and 35A controller (bought direct from Crystalyte in China in 2003 or 2004) soiled its trousers, and the bike stopped working suddenly, without warning. I cut the controller loose from my cargo bike so I could take it to the pedicab shop to change out the throttle plug for a more usual three wire JST-SM plug. I bench tested all the parts together and everything worked just as before, minus the useless power-on indicator and red/green LED SOC indicator that were in the old throttle.

Later, when I was reassembling the bike and deleting all the old derelict wiring left over from earlier iterations, my buddy Dave picked up the controller and noticed it was rattling, like it had a foreign object closed up inside. So I opened it up for a look. To say I was surprised at what I found would be a strong understatement.

A big ole electrolytic capacitor (160V 220μF) fell out of the controller. Followed by another. And then another.



(I scratched the data from the controller's original paper labels into the case as the paper labels began to deteriorate.)

The caps had been floating around in there for some time, from the looks of them. Not only were the leads all smashed flat against the tops of the canisters, but the wrappers and unprotected free ends were textured from getting tumbled around in there for who knows how long.





Thing is, this controller never stopped working, never misbehaved at all. The only think I ever did to fix it over the almost 20 years I've had it was to replace crusty phase bullets, battery plugs, and JST plug for the Hall wires. Plus the throttle plug, just now. It still works just like before.

I love it! It just chugs along doing the thing it has always done, after jettisoning some seemingly important electronic components.

I intend to keep using it until it stops working right. I don't want to take it apart for a postmortem until it actually dies. Why mess with a good thing?

I guess I wonder this, though: if the big capacitors were put in there for a reason, why does the controller seem to function perfectly without them? What could be the drawback to rolling this thing without them?

Anyway, I reckoned I had to share this weird good luck with somebody else who could appreciate it, other than my buddy Dave.

 

[99t4]

I guess I wonder this, though: if the big capacitors were put in there for a reason, why does the controller seem to function perfectly without them? What could be the drawback to rolling this thing without them?

Since you can easily see the specs, why not replace them with new (good quality) while you have it opened up? (even though those three may still be functional)
And tie them down a little better.

Sorry I don't know the answers to all your questions.

 

[amberwolf]

Depending on where those caps are in the circuit (probably across the main battery bus to the FETs), then as long as the battery itself is low enough internal resistance, and the connection from controller to battery is low enough resistance and inductance, the caps would have less work to do in negating the current spikes (and voltage dips) as current flow within the controller/motor system changes.

Probalby also makes a difference what the motor resistance and inductance is, but I'm not certain of that. (if it is such that there are less current changes and thus less voltage changes across the phases, it probably also helps).

The main possible drawback to not having the caps on the battery bus is that if there are current spikes sufficient to create high enough voltage spikes to damage the FETs, they might fail catastrophically at some point. I'm not sure if there are any others.


It's been a while, but some years back when SB Cruiser was in it's infancy, a controller I used on it had it's caps fail in the typical way, and it's operation had changed over time as they got worse. It never stopped working or failed, but it didn't behave correctly. Replacing the caps fixed that. Same kind of problem with another controller back on DayGlo Avenger, didn't die but replacing caps fixed it's behavior.

I use the motors (and hence controllers) pretty hard, so I'd expect FET failures from this kind of thing...but I have rarely blown up a controller--only once in use that I can remember, and that was a Method's special that had already been repaired, and IIRC it blew under very high current testing (higher than it was originally designed to do, but had been modded for instead).

 

[Chalo]

Depending on where those caps are in the circuit (probably across the main battery bus to the FETs), then as long as the battery itself is low enough internal resistance, and the connection from controller to battery is low enough resistance and inductance, the caps would have less work to do in negating the current spikes (and voltage dips) as current flow within the controller/motor system changes.

That makes some sense considering that for the last half dozen years, to I use only EV cells that are rated for hugely more current than I ever ask of them. Despite this controller's advanced age and diverse life experience, most of its miles (and most of the shocks and vibrations that might amputate the caps) have accumulated during that time.

Given how satisfactorily the old thing works, and what size nuisance it would be to spudge it out of its housing and reinstall it, I think I won't be replacing the caps unless it tells me to.

 

[amberwolf]

If it ever does tell you to, it'll probably do it by the FETs failing shorted, since that's their most common failure mode from exceeding Vds.

But it might be more graceful about it, if you're not as lucky as you've been so far, but luckier than I was.