Help! Controller debug sequence?

8 posts / 0 new
Last post
mf70
Offline
Last seen: 1 year 4 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Help! Controller debug sequence?

I've been riding my 48V 700W e-Vespa-clone scoot, "Hummer," for about a year. Since I re-glued the stator magnets in the brush-type motor, it has had a pronounced buzzing "cogging" when not under power: hence the name. You can see it at:

http://www.austinev.org/evalbum/822

Now a strange new problem has arisen.

The controller "stutters." At 80% to full throttle, the controller cuts out for 1/2 sec or so. Larger throttle settings produce longer cutout periods. (This is particularly obvious as the cogging returns when the power is cut.) It behaves very much like the low-voltage cutout, but the voltmeter is showing fully charged pack voltage at the time. The pack-controller connections seem to be solid, as are the electrical throttle connectors.

Any ideas on where to start to "debug" this? I assume this is a hall-effect throttle - there are three wires to it, and none of them acts like "ground," "source," or "wiper," as I expected. What voltages should I be seeing on the wires to a hall-effect throttle? Is it DC or some sort of PWM?

I just found:

http://www.electricscooterparts.com/troubleshooting.html

and will be using that. Any other suggestions?

Puzzled in DC
Mark

Gman
Gman's picture
Offline
Last seen: 11 years 4 months ago
Joined: Monday, November 20, 2006 - 23:24
Points: 1720
Re: Help! Controller debug sequence?

`
This is off topic, but any luck getting Buttercup back?

Peace Out,
Gman

Now that we have clarified our beliefs, your invited to join us as we begin building on them to define our Community Mission Statement

Peace Out, <img src="http://tinyurl.com/ysafbn">
Gman

Fechter
Fechter's picture
Offline
Last seen: 10 years 4 months ago
Joined: Friday, November 17, 2006 - 07:01
Points: 199
Re: Help! Controller debug sequence?

Most hall throttles have ground and a +5v supply. The signal wire usually goes from about 1v at zero throttle to about 4v at full throttle. Put the negative lead of your voltmeter on the battery negative, and use the positive probe to measure the three wires going to the throttle. I use a skinny wire and poke it into the backside of the connector to measure. The throttle must be connected and the controller powered on to measure.

Your problem sounds like a bad connection between the battery and the controller that's causing the low voltage cutout to kick in. Check all the connections, fuses, switches, etc. and look for signs of heat (discolored metal, melted plastic...).

If you have a way to put a voltmeter directly across the battery input to the controller while riding, you could see if the voltage is dropping out at that point. The idea is to measure downstream of all the connections and connect the meter directly to the wires coming out of the controller.

rgx
Offline
Last seen: 9 years 10 months ago
Joined: Tuesday, November 21, 2006 - 13:01
Points: 137
Re: Help! Controller debug sequence?

Since the throttle seems to be working and the behavior is repeatable and consistent, I'd say it is rather what you suspected at first - low voltage. Have you measured voltage DURING the stuttering? If not, do that, to verify the batteries are good. My own experience is that after around one year the voltage sag of the batteries increased, the no-load voltage was fine but under load the voltage would sink. If the batteries are OK, check all the power connections and thick cables again, there could be corrosion or a loose bolt somewhere. Does the scooter have a relay for the power as well? You might have to measure the voltage right where the cables enter the controller, to make sure. (An old trick is to stick a needle through the cable, if you can't find any other good spot to measure.)

mf70
Offline
Last seen: 1 year 4 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Re: Controller debug: Fun With Electricity

Ahem.

First, no, Buttercup is still missing. We got a note from the police asking if we'd gotten it back - a laugh, since they refused to help look.

Second, I dug into Hummer last weekend. As had been mentioned, it acted like a low-volt cutout. I monitored the pack supply to the controller, but it remained at full voltage, even while the controller was cutting out. THe pack wiring was OK.

Next I looked at the throttle. It was being fed 4.3V, and the "wiper" lead varied smoothly by throttle position. The throttle was OK.

I opened up the controller, and hit paydirt. While there were no smoking craters on the PC board, one of the SCR's quickly became too hot to touch when the board was energized, even before the throttle was opened.

On the repair front, the good news was that it was entirely discrete components, so it should be possible to replace parts. The bad news is that all part numbers have been scraped off. Cute.

Oh well, that's a project for the future. When I had first started putting Hummer in commission, I saw that the nominal 700W motor was controlled by a nominal 500W controller. I became really alarmed when I measured current draw and saw that the motor was pulling 1200W peak current. I immediately bought a replacement spare from TNC:
http://www.tncscooters.com/LB37.php

I pulled the spare out. The first question was whether it would take a hall effect throttle. With power applied, the hot throttle lead was at 4.3V, so that was a go. Next was the "lock" circuit. My original controller had only one sense lead connected to the ignition lock, which went to ~48V when activated. The LB37 "lock" connector had two leads, one of which went to ~48V. I put 48V to the other lead, and got output to the motor. Oops, the wheel was turning backward. Not a biggee, just reverse the motor leads.

Next was the brake connector. In my original controller, a 12V level on the brake light line shut off the controller. On the LB37, the "brake" connector was two conductor, similar to the "ignition" connector. I tried grounding the red lead, but that didn't seem to work. I put 12V to the black, and that didn't work either. On a whim I connected the red "brake" line to pack positive. SNAP!. The magic smoke was out, and there was no response to the throttle.

Oops.

Time for recovery mode. I put 4.3V to the throttle leads, and got the expected "wiper" voltage as the throttle was advanced; I hadn't fried my throttle. I opened up the LB37; nothing was obviously burned. I turned the PC board over; there WAS heat damage. One trace was lifted from the board and melted through. I replaced the trace, powered the board up, and.... it WORKED!

In fact, the scoot works better than it has in months, and I am able to breathe again. The only functional loss is that the brake cutout doesn't work; it is possible to run the motor even though the brakes are engaged. This is tolerable since my wife and I are pretty good about not doing that anyway.

Time to order another TNC controller.

MarkF

chas_stevenson
chas_stevenson's picture
Offline
Last seen: 7 years 8 months ago
Joined: Wednesday, December 6, 2006 - 17:14
Points: 1309
Re: Controller debug: Fun With Electricity

Mark,

The 2 wires (red and black) which are for the brake switch on the TNC controllers are to be shorted together when the brakes are applied. This will disable the controller output to the motor. It may still work I connected one of these controllers bacwards and even after it released it's smoke when connected correctly it still functioned.

Chas S.
My Bicycle Pages

mf70
Offline
Last seen: 1 year 4 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Re: Controller debug: Fun With Electricity

The 2 wires (red and black) which are for the brake switch on the TNC controllers are to be shorted together when the brakes are applied. This will disable the controller output to the motor.

The problem is that I have a single sense line in the scoot that goes high (12V) when the brakes are activated. This is the common brake light power line. I'll have another look, but it seemed that neither of the leads to the "brake" plug responded. If it de-activates the controller when shorted, it is either looking at a sink opening up under the supply line or a source appearing on top of the signal ground.

It may still work I connected one of these controllers bacwards and even after it released it's smoke when connected correctly it still functioned.

As I noted, it DID work, after I replaced the trace I'd blown off the board. That's a tough little box! Since it was physically 1/2 the size of the original, I had been suspicious.

MarkF

decibel1
Offline
Last seen: 8 years 2 months ago
Joined: Wednesday, December 13, 2006 - 09:49
Points: 170
Re: Controller debug: Fun With Electricity

The LB37 is a fine little controller, for sure. I burned out the original controller on my old Evader 1000w scooter (my fault, I'm afraid - I was trying to "improve it") and replaced it with an LB37. I did take Fechter's advice and I soldered over some of the current monitoring shunt on the board to fool it into providing more current, and I have now been riding with it without problems for a year and hundreds of miles. The only problem I have is with the brake switches. Before I disconnected them, both the motor and the lights would go off whenever I hit the brakes. Now trat the brake switches are disconnected using the brakes does not shut off the motor power, which is not a big deal to me, but I also have non-functioning brake lights, which I will have to rewire.

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • Autosellworld
  • icandyb
  • jasonk
  • ridethelightning
  • va7db

Customize This