Fixing a dead Crystalyte controller

[patrick_mahoney]

Inspired by Knoxie's thread about fixing dead controllers, I decided to try my hand at fixing a dead controller that I have. I received it with my Crystalyte x5304, but it was dead when I went to hook it up. I spent a while messing with it, but I could never get it to work. I emailed the gentleman that I purchased the motor from and he sent me a replacement controller and told me to keep the dead one. The replacement worked great and I've been using it ever since. And the dead one has sat gathering dust since last summer.

It is a instant-start, 36-72V, 20A Crystalyte controller.

I took it all apart, nothing looks burnt or shorted. I individually examined all of the FETs and they all look good. I did the smell test - a highly technical procedure that involves sticking my olfactory sensor very close to all the chips and sensing their condition by taking a good sniff. They passed. I wired everything up and hooked up the battery. The motor hall sensors appeared to power up correctly when I checked with a DC voltmeter. The throttle wires didn't show any voltage at all on any of them. That seems wrong.

I hooked the motor up to a vice, powered it up and got nothing. Notably, the LED's on the 36V LED "fuel gauge" throttle didn't come on - while hooked up to the 36V battery pack. I didn't get anything there at all.

I checked the hall effect sensors by moving the motor and checking with a voltmeter and they look fine - and the sensors are receiving what appears to be the correct signal.

So I am currently suspecting an open in the wiring to the throttle - based on the lack of a power signal to the LED on the throttle. It's my guess for tonight anyway. The throttle is good - I tried it on my other controller and it's fine.

So, I'll wire up another set of wires for the throttle and will try this wire instead - which brings me to the question. How do you hook these things up?

Shown below should be a screenshot of the controller. What do the 4 wires that go to the throttle do? I presume, it's power, ground and hall-effect, but what's the fourth one for? And which is which? When I go to hook it up to the throttle, how do I know which of the 4 wires on that go to what?


Thanks for any advice.

 

[fechter]

The fourth wire is usually for a LED in the throttle. Not necessary to run.

You should definitely be seeing +5v on one of the throttle wires. This 5v usually comes from a 78?05 regulator on the board somewhere. This is usually a TO-92 transistor looking thing. Try to find it and measure the input and output legs.

What's the 3wire connector just left of the throttle connection?

 

[TylerDurden]

Me confuzed...

If the throttle checked OK on another rig, U are changin it?

Seems the prob followed the controller not the throttle.


BTW the IC indicated in the pic below looks like it got toasty, do the traces go to the throttle connectors?

 

[patrick_mahoney]

You should definitely be seeing +5v on one of the throttle wires. This 5v usually comes from a 78?05 regulator on the board somewhere. This is usually a TO-92 transistor looking thing. Try to find it and measure the input and output legs.

Ok. Thanks for the advice. I'll do that. I didn't read any voltages on any of the throttle wires.

What's the 3wire connector just left of the throttle connection?

I'm not sure. It goes to a male 3-wire header that doesn't look familiar. The only thing that makes for an obvious guess is the cut-off for brakes. The white wires go off to the FET (still attached to the heatsink, and not shown). The 5-wire set going off the top are the hall sensors. The only thing left is the brakes...

Me confuzed...
If the throttle checked OK on another rig, U are changin it?

The throttle is fine. I think it's the header or the wiring that goes to the throttle connector. I tried reseating the header. I should have tried removing it and checking voltages on the pins. I'll try that tonight - or maybe at lunch. I was thinking about just replacing all of these wires with a new set thinking there's an open or two inside... I don't get a continuity signal from the header wires to the pins of the throttle connector.

Seems the prob followed the controller not the throttle.

Yes, I see what you are saying. And you are right. It was just an idea. An easy idea. Before I move on to the hard ideas. The lack of continuity (only one of the four wires shows a resistance of "0" from the header to the connector. The other always show infinity) was the reason why I started down this.

BTW the IC indicated in the pic below looks like it got toasty, do the traces go to the throttle connectors?

Good eyes! Yeah, that does look suspicious. I'll look at it when I get home. Thanks.


Thanks for your help.

Patrick

 

[patrick_mahoney]

I checked out the chip that looks burned out in the photo - it looks fine under a magnifying glass. If looks mean anything...

I rewired the throttle - there were some loose wires, in fact when I took the connector apart, two wires just fell out on the table.

But this doesn't appear to be the problem - because I got the throttle hooked up correctly. The power lights come on. I see +36V on one wire, all others read 0 relative to each.

I see the TO-92 style regulator. I checked the input and output and ground and they are all "0" relative to ground.

With the throttle hooked up, and the motor hooked up and powered up, if I apply throttle, I get nothing.

If I hand-spin the motor while powered up, with throttle or without, it turns ok in one direction, but it resists the other direction strongly and moves with almost a thunk-thunk-thunk as it resists and then as I move past that it gets easy and then again meet resistance. Without power, the motor spins normally.

One final thing was that while I was messing with the throttle wires while it was powered up, I did actually manage to get the motor to spin - it spun at a constant speed - which was fairly slow - but it ran without problems. I haven't figured out what I did though - although whatever it was, it didn't have anything to do with actually moving the throttle. It's never happened again.

 

[knoxie]

Hi

I had the same problem was a break in the throttle wire, it was caused be a user fitting a 36V led throttle to his system then running it on 72V, the led type throttles are set to the system voltage, the throttle burnt out and in turn shorted the wires out.

Check the continuity from pin in the connector to the plug on the board, if there is a break or a short between wires change the cable, unplug the plug from the board and check between the leads to make sure there are no shorts, use the didoe check or beeper if it helps on your DVM.

You may of course still have a problem, the IC that looks warm is the driver IC chip for the blue phase high side fets, they can blow (it is unusual), the first check is to set your DVM to measure resistance and check between each phase and the battery + lead, if you have a low or short here you have a fet blown, do that first as its a good base test for a blown fet, the fets can blow with no burning or smell, so do that first.

If you have a blown fet its easy enough to change, you also want to go around all of the SMD components on the phase to make sure they are good, read off the resistor values and measure them with your DVM, if you are not sure of what the values should be measure them against another phase component, normally the gate resistors blow 22R and the 10R resistors on the low side, also the 2 little npn and pnp transistors go as well, check across them with a DVM for a short.

There wont be much wrong with it, I have fixed all but the one that caught fire which isnt worth fixing.

Cheers

knoxie

 

[fechter]

The fact that you feel resistance in one direction when the controller's powered up indicates the FET low side drivers, the commutator chip, and the hall sensors are working. This is a good sign.

With reference to Battery (-), the throttle wires should have one at 5v, one at zero and one that varies from about 1 to 4v with movement. The 5v should be coming from the regulator.

If you have zero on all three pins of the regulator IC, then you need to trace the input to see where it's supposed to be feeding from. I have no idea on that model, but try to follow the trace from the input pin to see where it goes to. On some controllers it feeds off the output of a 12v regulator. The one on my bench looks like it comes of the descrete regulator made from a transistor and a zener diode. This is a big transistor mounted on the heatsink. The input to that comes right from the power switch. There might be a resistor in series with the input.

 

[fechter]

If you get a chance, post pics of both sides of the controller so you can see the whole board.

 

[patrick_mahoney]

I have to say, at various times, I sit and think - this thing is way too complicated... I have no hope of fixing it. And then I log in here and you guys inspire me to keep going by pointing out the next step. I really, really appreciate your help.

Shown below are two photos - a better one of the topside and then one of the bottom.

I think I figured out what made the motor go - at least I thought of a good possibility - and that is that I must have accidently tied the +36V for the LED to where the +5V for the hall sensor must go on the throttle side. Yeah, I cringed when I thought of it too. Hopefully the throttle is still ok.

But if that is the case, then the repair might not be too bad - if nothing else, I can rig up a separate mini board that does the +5V regulation for the throttle.

Thanks again for the advice! I'll try to figure out the input to the 78L05 (TO-92 package) regulator tonight.

 

[Mathurin]

So you don't have power going to the throttle anymore, if you still have juice going to the motor's hall sensors, how about stealing some 5v off there?

 

[fechter]

Q1 appears to be the primary voltage regulator. The collector should go to the power switch, the emitter should be somewhere around 12v, which should be the feed for the input to the 78L05.

Check the voltages on all 3 legs with respect to battery neg.

Just for fun, you could also try Mathurin's idea. Jumper the +5v from the hall sensors to the +5 terminal on the throttle. I assume these would be the red wires on both connectors.

The hall sensor supply might sag under the load of adding the throttle.
It would be best to find the real problem and fix it.

Can you read the numbers on the FET's ? Just curious...