Motor transistors keep dying on me

[JohnRoberts]

@CJ here's the part:



And @JohnRoberts I just measured the transistor and it shows 100 ohm in both directions base to emitter and a normal continuity between base and collector.

Back in the 60s (as a junior tech) I used to grind the tops off of TO-3 transistors like that to analyze the failure modes. I am not familiar (or don't remember) base-emitter shorts like that, but I expect with the top ground off you may see visual evidence of the base-emitter fault. The base collector junction appears intact (?). The 100 ohm resistor in series with the base makes that an unlikely path for over current failure, but perhaps a reverse voltage breakdown(?). Base-emitter junction breaks down with about 7V reverse.

JR

 

[EmilFrid]

@JohnRoberts interesting. I might grind down the top and have a look. And possibly the fault is because of reverse voltage, as I'm having problems seeing any other possibility. The question then is why this started now and what I can do to fix it.

 

[CJ]

Wow, those transistors are rated at 1000 volts, so I would rule out motor back emf,

If the E-B junction is going south I would check the driver transistors as mentioned upstream.

 

[JohnRoberts]

@JohnRoberts interesting. I might grind down the top and have a look. And possibly the fault is because of reverse voltage, as I'm having problems seeing any other possibility. The question then is why this started now and what I can do to fix it.

The logical troubleshooting path is look for what changed. Its always the last thing you fix...

JR

 

[EmilFrid]

@CJ I will have a look at those. The strange thing is, though, that they seem to be ok as they put out ~1-4 volts during operation and 0 volts in stop mode. I'm thinking that they should behave differently if broken. Still, I should unsolder them and measure.

@JohnRoberts I will have to think about that. Haven't really done much else than changing two caps (one blew up)

 

[EmilFrid]

I'm really grateful for all the advice. I really need some outside perspective or else I'll go insane

 

[Cqwet Dbdfte]

The motors are split phase synchronous, powered by 100VAC.
The cap between windings provide needed phase shift to start the motor
A catch diode will not work here, as the HV transistor i buried behind a bridge rectifier, and no diode across the motor will work. The rectified motor current thru the motor is passed thru the HV transistor.
It operates as a current controlled device. I can't find the Hfe for BUX48A.
It is listed as a switching device. Too low gain may strain the driivers.

 

[Hubbub]

I find that a somewhat scary approach to controlling an AC motor. Yes, it works fine if the AC source and load are fully floating, but are they? I'd also expect the motor 100V supplies to be isolated from each other, are they? If there is ANY connection between control circuit 0V (i.e. "ground") and the motor or its 100V supply then there is the risk of unintended current flow through the transistor to the 0V node. I'd check for any leakage to 0V from the disconnected motor and from the 100VAC source. Consider checking with a 200V DC supply using a resistor to limit current to 1-2mA or so, and look for any significant voltage across the resistor as you probe around.

 

[Hubbub]

C111 and C112 are 47uF capacitors and critical to circuit operation. Make sure they are in good condition, within 20% of 47uF or better, and decent ESR. With a 50 Ohm series resistor they don't need to be low ESR parts, although that may help. I'd make sure any replacement cap was rated 200V DC (for margin) and good for at least 3A ripple current. Also proper insulation and heatsink compound on Q102 and Q103 is mandatory.

 

[EmilFrid]

@Cqwet Dbdfte yes, perhaps that transistor isn't the best pick. I might try the original 2sc1114 again.

@Hubbub yes, I actually changed those as one blew up. It had been there since the machine was new so first thing I did was replacing those along with all caps in the PSU. And you're correct – there are three separate supplies for the motors. A really good idea to check for leakage. Haven't even thought of that. Do I understand correctly that I should measure across the added series resistor?

 

[Hubbub]

Using a 100K resistor makes for a good current limiter (2mA given 200V, less than 1/2 watt dissipation max.) Measuring the voltage (with a 10 MOhm DVM) across the resistor scales up any leakage current to a significant voltage such that 10uA leakage gives 1V across the resistor.

 

[EmilFrid]

@Hubbub thank you, I will try this out!

 

[sahib]

Hubub and Cqwet Dbdfte have covered quite a lot. So no need to repeat.

...... And why does my meter read 300VDC that gradually climbs down to 150VDC on the collector cable when it isn't connected to the transistor? Confuses me a lot.

For this have a look at the motor wiring schematic. Have a look at the take up motor for example. On start up C110 is charged through the one winding, and C104 in series with C109 is charged through another winding. They sum They then discharge and settle at (roughly) 150V.

What I find strange is that the short is between base and emitter. That must have to do with it getting reverse biased, because the maximum voltage that the base reaches is 3.8VDC and the transistor should be able to handle this

Now that's indeed interesting. I have not checked the data sheets for both transistors. May be you have to limit the base current further with the replacement transistor.

 

[EmilFrid]

@sahib oh I see now. Thanks for clarifying this. I totally "ignored" the motor caps.

And yes, it is very strange. Never happened to me before nad I've had quite a lot of tape machines with these heavy duty transistors. Would adding a diode going from emitter to base prevent the reverse bias voltage? Someone mentioned a catch diode might not work

 

[sahib]

It won't work as now the current will split. You will still have the reverse bias condition However, you said that the transistor is failing at turn on, hence the base/emitter is failing at turn on. We have to think why as this was a working circuit. I would normally consider things like the current across the inductor flowing through emitter-base path but that would imply path being open on the motor wiring end. In which case the motor would not run. But you said it runs for a few seconds and then the transistor fails.

I am dashing out to collect my wife from the airport. I'll look into this when I am back. It started to bug m now.

 

[EmilFrid]

Ok I see. One weird thing: my motor control card doesn't have the inductors - their connections are just bridged. But it has worked since 1978 so shouldn't be a problem.

Thanks for the input. It bugs me too (a lot)

 

[CJ]

Maybe there is a lot of friction in the drive train which is causing excessive current draw, check to see that the motors spin freely,

 

[Evertide]

Whatever it is the two tabs are shown shorted which makes it and the capacitor above useless.

This is wired as an 3-phase AC motor , the top capacitor makes a phase-shift for the 3e leg...

 

[sahib]

This is wired as an 3-phase AC motor , the top capacitor makes a phase-shift for the 3e leg...

I understand that but I would imagine 3 phase meaning 120 degrees apart.

Ok I see. One weird thing: my motor control card doesn't have the inductors - their connections are just bridged. But it has worked since 1978 so shouldn't be a problem.

Thanks for the input. It bugs me too (a lot)

Thank you too.

That's irrelevant within the context of DC.

However, I have just checked the photographs you posted and the transistor seems like TFK (Telefunken). Have you bought these on e-bay by any chance? If so, are you sure they are genuine?

 

[EmilFrid]

@sahib no I bought them from a reputable dealer here in Sweden. Would be surprised if they sold counterfeit transistors, but who knows