Those of you that understand electronic circuits, step right in

[AaronB]

Thanks in advance for looking. Sorry, but it's kind of long.

I have a small metalworking lathe that has an electoronic speed control. The speed control board was damaged today after I replaced a blown fuse with aluminum foil and then proceeded to do something with the lathe that would have blown the fuse but instead blew something else.

The foil was only supposed to be a short term solution. It turned out to be a very short term solution. Yes, I realise how stupid that was so let's move on.

Here is a photo of the board with some descriptions. I have made it a point to draw a diagram of the mosfet circuits because I have been led to believe that this is a relatively common failure on these boards.

This is what I know: (please excuse incorrect terminology).

-The motor that this drives is rated for 110 VDC

-There is 120VAC going into the board.

-There is 110VDC coming out of the rectifier

-(here comes the incorrect terminology) There is 110vdc at the "input" terminal of the mosfets, which is the red line on my photo. There is 75VDC at the "output" terminal of the mosfets, green line. There is 75VDC at the "control" terminal of the mosfets, yellow line. All voltages are relative to the positive motor output terminal.

-The potentiometer itself is functional but has zero effect on the above voltages.

-The failure was proceeded by a "pop", some smoke, and the typical burned electronics smell. Visual inspection revealed no burned components or traces. This is a fairly transparent two layer board so the traces are easy to follow. There is a chip in the right hand mosfet but I don't know if that is related. (it's the whitish spot where the red line makes it's downward turn)

Now for my questions:

I have a multimeter and soldering iron. Is there any way to diagnose bad mosfets? Is there anything else that I could test?

 

[GalvanizedYankee]

I'm not going to even try. FOIL! Sheesh!

Register at http://www.badcaps.net/ Give Galvanized the referral, please. Post pics that you load to thier server in the appropriate forum. Other Devices With Bad Caps will work. Give it 2~3 days. If no one can offer help....Only then do you PM Per Hansson and invite him into your thread. He maintains the electronic controls for machinery at a company in Sweden. He is sharp.

 

[Cattlegod]

measure all of the caps and resistors to check their value. then measure each trace along the board to ensure 0 ohms. if you don't find anything wrong in that pass through, replace the transistors. if that doesn't work, replace the ICs. If that doesn't work, you are screwed.

 

[Colt45]

The pot should change the voltage on the gate of the fet ("control")
well probably not the voltage per se, but the duty, which will look like voltage on the multimeter..

The little daughter board on the side must be the PWM circuit, which then drives the fets. the fets are 2SK790 (an asian part number, so probably a PITA to find here)

 

[Mark R]

There isn't much in the main circuit that could blow:

There's the bridge rectifier and the MOSFETs. Everything else is off the MOSFET gates, which apart from a few small resistors, diodes and a cap, is optically isolated from the rest of the board. I'd be surprised if anything other than the MOSFETs (or possibly bridge) has sizzled out.

I'd replace the MOSFETs with a pair of STW20NK50Z from Digikey and the bridge rectifier with a MB104-BP (double check that this will actually fit - I think that's a BR6 shaped one on the circuit board, but without measuring the spacing of teh pins, I can't be sure).

 

[AaronB]

Originally posted by: Mark R
There isn't much in the main circuit that could blow:

There's the bridge rectifier and the MOSFETs. Everything else is off the MOSFET gates, which apart from a few small resistors, diodes and a cap, is optically isolated from the rest of the board. I'd be surprised if anything other than the MOSFETs (or possibly bridge) has sizzled out.

I'd replace the MOSFETs with a pair of STW20NK50Z from Digikey and the bridge rectifier with a MB104-BP (double check that this will actually fit - I think that's a BR6 shaped one on the circuit board, but without measuring the spacing of teh pins, I can't be sure).

*Cough Cough*


Thanks for your help. It's really important to me that I get this lathe running quickly. The speed controller is an available part but no one seems to keep it in stock so there will be a wait. I know there is no guarantee that this will work but it's worth a try.


Oh, and extra credit goes to whoever can explain in one paragraph how the mosfets control the speed of the motor. I really don't know how they work. Do they hold a constant 110VDC or do they vary the voltage?

 

[Mark R]

OK.

Here you go.

Bridge
MOSFET

It's easy enough to test the bridge with a DMM. Select diode mode.

Put the red lead on the - terminal, and probe the 2 AC terminals with the black lead. Both should give a non-zero reading (no reading, or a 0 reading indicate failure). Reverse red and black leads, you should now get no reading (a reading indicates failure).

Now put the black lead on the + terminal and probe the AC ones with the red lead. Same principles apply.

(Thinking about that, I might have got the red/black the wrong way round - but it should be fairly obvoius once you start testing).

These are the most likely components to be fried - it's relatively unlikely that anything else has gone, as they aren't in the direct current path. Unfortunately, debugging anything else is likely to be awkward. These parts aren't expensive, so it's probably worth a go at fixing it.

The MOSFETs aren't an identical replacement, but are a nice upgrade.

As to how it works:
The bridge converts the 110V AC into DC.
The DC then goes to the MOSFETs, which act as switches, before going to the motor.
The control electronics on the board rapidly turn the MOSFETs on and off - so the motor is rapidly turned on and off. The amount of time that motor spends switched on determines it's power.

The little white chips on the control board are opto-isolators. The left side contains a light-sensitive transistor that activates the MOSFETs. The right side contains an LED which is controlled by the control circuits. This arrangement ensures that the 110 V mains voltage can never reach the low-voltage control electronics - so you could never get a shock from what are supposed to be low-voltage circuits.

 

[AaronB]

Originally posted by: Mark R

The little white chips on the control board are opto-isolators. The left side contains a light-sensitive transistor that activates the MOSFETs. The right side contains an LED which is controlled by the control circuits. This arrangement ensures that the 110 V mains voltage can never reach the low-voltage control electronics - so you could never get a shock from what are supposed to be low-voltage circuits.

Thank you again.

I do have a question about the above paragraph though.

I'm still a bit hazy on what these isolators do but are they affected by ambient light or light being emmited by a standard light bulb?

If I were doing all of my previous voltage tests with the board under a bright lamp would it have affected my readings?

 

[Mark R]

Originally posted by: AaronB
I'm still a bit hazy on what these isolators do but are they affected by ambient light or light being emmited by a standard light bulb?

If I were doing all of my previous voltage tests with the board under a bright lamp would it have affected my readings?

The isolators are a way of making one circuit control another, but without an electrical connection between the two circuits. This improves safety, because a problem on the high voltage circuit can't cause dangerous voltages flow into the low voltage circuit.

A simple alternative is a relay - by switching an electromagnet on and off, you can control a mechanical switch on another circuit. No electrical connection between the two circuits - so it's safe. The opto-isolator is the electronic equivalent - smaller and faster than a mechanical relay.

The isolators are sealed - so they aren't affected by ambient light. They can only detect light produced by the internal LED.