Background:
Let me begin by saying I'm a tinkerer with a basic, limited understanding of electronics.
Long story short, my treadmill got some liquid spilled on it and probably shorted something on the motor control board. Immediately after the spill, it tripped the GFI on the outlet. I let it dry, and can now plug it into the wall without it tripping the GFI.

Symptoms:
The treadmill runs normally, except no power goes to the motor.
I've tested the motor with a cordless drill battery, it works fine.
The power supply seems to be working, outputting 12V, and the PWM light flashes.
Very little voltage (1V?) is output at the motor spades on the control board.
After disassembly, the resistors at RPS 1 and RPS 2 were broken. I tried re-soldering them to no effect.
Now, the "CUR LIM" LED is lit, which it is never supposed to be:
"D16- Labeled CUR LIM- This light should never come on. If this light does come on, it means that the treadmill is exceeding the current limit of the controller. Continued use in this condition typically will cause repeat failure of the controller."


I've followed all the steps here:
http://www.treadmilldoctor.com/Symp...6pcw6xwhWXJgBFq9yQKJtX21pT3625rkmHy!307018887


I've found this old thread:
http://forum.allaboutcircuits.com/showthread.php?t=18895
with a nice hand-drawn schematic from trolley1 in the second post.


Now I'm at a loss. As soon as power is presented to the treadmill, all four LEDs on the controller board light up. How can I determine which component is causing the CUR LIM to light? Is this due to a short or a faulty component? How can I fix this to regain a working treadmill?

I blew the fuse on my multimeter today by being an idiot, so I need to get a replacement before I can use it. I can take more pictures if needed.


Here are some pictures (Click for huge size):
Overview of the board



RPS 1 (out of focus, dur....)


RPS 2


CUR LIM LED



Thanks!

 

It's possible the output FETs/BJTs (the three large ICs on the heatsink) are failed, which usually leads to them becoming shorted. Can you read the part numbers?

If the fuse on your multimeter is blown, you can still measure voltage, only current is unavailable.

 

If the fuse on your multimeter is blown, you can still measure voltage, only current is unavailable.

That makes sense .... I must've fried something else then, because it no longer reads anything? ...... damn

It's possible the output FETs/BJTs (the three large ICs on the heatsink) are failed, which usually leads to them becoming shorted. Can you read the part numbers?

These guys?

The outer ones read S4020L with 677 vertical down the side.
The middle one reads D4020L with 113 vertical down the side.


There are an additional 2x D4020L on the opposite side of the board.

Thanks for your reply!

 

Tough luck on the meter. However you can pick up a basic multimeter for less than $10 which will do everything you need and more for this project. Cheap meters do tend to fail suddenly, but I still like them because they do the job fine for their price.

After closer inspection that is intriguing.

It turns out those S4020L are actually SCRs (silicon controlled rectifiers). And the D4020 are diodes, which matches up with their part number. Obviously this board is designed for AC motors, or an AC power source. It is not outputting a DC voltage normally, right? SCRs aren't designed for DC; not usually, at least. I suspect it is just fancy PWM control for an AC motor. What voltage is the motor rated at?

Once you get a meter, you can test the diodes. Set to diode test, place the terminals on any combination of pins both backwards and forwards. If you always get "1 ." or overload, then the diodes are fried, and if you get "0" (or a close value, less than 100) the diodes are fried. Let us know.

 

Thanks so much for your help.

I should have another multimeter by Thursday. I'll try testing the diodes then.

It appears to be AC power supply -> controller board -> DC motor.
I don't know how I measured DC 12V over the blue and white wires coming from the power board, which appear to be carrying AC.... lol (maybe this kind of stuff explains my dead multimeter )

Maybe this sheet I found will give you a better picture of what's going on :

*EDIT: fixed the image

 

Well, if there's a DC motor it probably is DC. I don't know though. My suspicions point to one of the diodes failing open. This leads to no current and hence your symptom. It could also explain the symptom when the motor was locked at full speed: the SCR had failed, and this was pumping all the current through the motor, which overheated the diode.

 

Fixed the image in my previous post.....

Would one of these diodes failing cause the CUR LIM light to come on, or is that a different problem?

Thanks!

 

If the diode shorted out, it could. However, the CUR LIM probably comes from measuring the motor or diode drop voltage; if it is too low, the light will illuminate, because the motor is dropping too much voltage. There is no current shunt, so I cannot see any other way it could work.

 

Still white stuff on one of the ICs in the pictures. Also, sadly, sometimes water damage can be worse than "just drying it out" will fix.

As to the guy with the multimeter - some do just put the fuse in series with one of the leads so it can kill the entire meter, try replacing it first.

 

White stuff on ICs is just leftover residue from a label.

Bad (good?) news:
All 3 of the D4020 show .47 in the diode test mode. Reversing the leads gives 0.
Both the S4020L show .3

Does this mean the diodes are still good? What other components should I test?

 

That sounds about right for Schottky diodes. So the diodes are good. The SCRs usually can't be tested with a meter, you're probably measuring gate to cathode, which is not particularly useful (though it does tell us the SCR is probably OK.)

I am curious about the zero when the leads are reversed. Does it do the same when the leads are disconnected? It should go "1 ." or "OL" or something like that.

 

Yep, my mistake. OL when the leads are reversed.

Any other ideas for components?

 

Do you have a 'scope handy? If not, can you borrow one?

 

Unfortunately, no.

So I'm out of luck, then?

 

No, all is not lost, we can try other avenues.

Can you get a few pictures of the ICs on the board. I'm thinking that maybe they are damaged or some component around them is damaged. Also, I'm interested in the bottom side of the board (with solder) to see if there is damage/oxidization.

You can also try and see if you can get a sound card oscilloscope working. Be careful though, you'll need to limit the 150VDC which may be present in parts of the circuit board. It looks like a low frequency design so a sound card oscilloscope will be fine.

 

A sound card oscilloscope, huh? Never heard of it, sounds cool.
A bit of googling around leaves me wanting something like this, but something that is cheaper while still being safe.....

This is a stretch.... and I could probably make this, but it doesn't look very safe.

What would you recommend in terms of probes?
Keeping in mind that I don't have a dedicated soundcard anywhere, so if I fry my onboard one, my motherboard goes with it!

Not to mention I'd need to learn how to really use an oscilloscope, my only experience with a real one being 30 min in a lab... and that was a while ago.
If I ended up making my own probes, how many would I need, 1 or 2?

I'll grab lots more pictures tomorrow.

Thanks for your continuing help with my silly little project!

 

You can test the SCR with your meter in the diode function. Disconnect the component from the circuit if possible.

1. Place the meter in the diode check position.
2. place the negative lead on the cathode and the positive lead on the anode. Reading should be as if the leads were open.
3. Reverse the leads and you should get the same reading.
4. Place the negative lead on the cathode and the positive lead on the anode. Jumper the positive lead to the gate and the voltage should read approx .7 V for silicon devices (some meters might read approximately 700 - mine read 632 or 0.632V). Maintain the Anode and Cathode connection and disconnect the gate lead. The reading should remain close to the previously observed number.

I've used those cheap transistor checkers (audio tone type) to check SCRs also. It was for a basic go-no go status and nothing too complicated.

What you are looking for is the obvious shorts or opens. Opens when it should be low and shorts when it should be high.

The SCR I tested was a lower power device, the 2N5060 SCR.

 

Here are some pictures I grabbed today. It's really hard for me to get decent macros with my point-and-shoot... so I'm sorry for a bit of poor focusing on some.

ICs:

 

Underside of board:

I'll try testing the SCRs the way you described, joe.

Thanks!

 

Well, the design looks to be fully isolated from the mains...

However, I wouldn't want to try probing it without proper protection.

I cannot see any oxidization problems.

Those resistors you mentioned on the first post, can you post some pictures of them?