failing transistors

[bfhamr]

Newbie here. I repair machines for a living, I'm currently working on an EDM. Normally I would contact the manufacturer but I'm told they are no longer in business.
The machine has a set of 6 transistors (Hitachi 2SC1343). These can be switched into and out of the circuit depending on how much power is wanted while machining. The problem is these transistors are failing (shorted) and I'm not sure why. I had replaced the board which contained these and all 6 have failed again. Another board which had been in the machine had 4 of the 6 shorted.
I'm not looking for specifics on this machine but in general what would cause a transistor to fail shorted.

Thanks
Dan

 

[bfhamr]

stupid question, how do I attach a picture, or do I have to supply a link to it.

dan

 

[bfhamr]

See here for a copy of the schematic.

http://i100.photobucket.com/albums/m18/bfhamr/DSCF0369-1.jpg

Also I was wrong, the transistors are switched in banks of 3 not 6.

dan

 

[itsmoked]

That's useful! Nice shot.

You're saying you have +/-8V but the schematic is calling for 10V that will reduce drive!

Any chance of you putting a scope on the board while it operates? That is where I'd be trying to go.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[bfhamr]

I've talked to another EDM supplier who sells a nearly identical machine (it has the same circuit board). One of their engineers told me that the control voltage, when loaded, will drop to +/- 8 or even +/- 7 volts.
I may be able to get a scope on the board. At this point though the customer may not want to spend any more money trying to fix this machine unless a solution appears forthcoming.
I wish this guy was next door, I'd spend my free time working on it. I love a good problem. Thats the reason I lurk on this and similar boards.

dan

 

[OperaHouse]

I would just go with higher voltage transistors. Transistors are often selected out of batches for higher voltage at the builder. I used to buy the lowest voltage of a series and then put each transistor on a curve tracer. 90% were many grades higher. They have the time to do this. I visited a couple EDM manufacturers in China and they even solder up little pieces of copper sheating to make float switches. Your stock may just have a number of lower spec transistors. Does this machine also have a seperate supply for a teaser voltage. Many machines have this at about 300V and that may be causing the problem. When one transistor fails it may be feeding enough voltage to the base of the other transistor to partially turn them on. A hot transistor has a lower voltage breakdown.

 

[Comcokid]

I go with itsmoked most recent post. You indicated +/-8 V for the two supplies for the drivers, but stated the negative one read (-)5.5 to (-)6 volts and the schematic indicates +/-10 volts. I would thoroughly check out those supplies, especially the negative one before proceeding back to an examination of the power transistors. The drive circuits consist of complementary transistor pairs and if the supplies aren't right, the drive to the power transistors may not be right.

 

[bogeyman]

All of the comments above make good sense to me, however may I be so bold as to introduce the effects of second breakdown on the longevity of BJTs used in real power applications.
Any given transistor has max permissible values of Ic and Vce but none of them can pass max current at max volts for long (micro seconds or less) without giving up on life.
The manufacturers data sheet should provide an indication of how cruel you can be to their transistors.
The instantaneous Vce x Ic product is usualy much lower than the maximum continuous power dissipation and can be a common cause of failiure when power amplifiers are connected to reactive loads.
When this breakdown mechanism takes hold there may often be no apparent overheating or easy to find cause.
The poor driver supply voltage is a very likly candidate for you EDM machines repeated consumption of power devices wether its due to second breakdown or plain old overload.

There are may big fat NPNs out there, For the audio brigade On semi make some low saturation voltage "beta sustain" devices with Vce >250v, Ic max >16A with Hfe >100, and a genrous second breakdown area.

M D lamb

 

[OperaHouse]

I doubt that the low voltage supply is very complex (no regulation). The problem may actually be external to the machine. The low voltage supply is generally single phase and the lines this is connected to may be low due to voltage imbalance in the ship. Maybe they installed a new single phase machine in the shop that is dragging down the line. This is very common in shops that are 208V. Without knowing the value of the four resistors that make up the voltage divider in the driver, the effects of low voltage can not be determoned.

 

[cbarn24050]

I would forget all these attempts to redesign your machine, just employ an engineer to track the fault down and repair it.

 

[Snoball]

On your original question, I had an instructor demonstrate transistor failure. He said they can only fail open or short. He took a little one and stuffed the legs in a wall socket, whack, gave it to me, it was shorted. He then whacked it hard with a hammer, it was open.
I'll look at the schematic but when was the last time it worked, does it current limit?

 

[Snoball]

"C" to "B" controls transistors which control transistors which switch the outputs. To prove the board make 1-6 Hi Impedance to 7 and you can watch with an Oscope current and voltage.

 

[ScottI2R]

Our older machines used a circuit that was called current cutoff. It was designed to increase the off time of the pulse effectively taking the duty cycle to 0, and saving the output section in the event of a short in the edm gap. A quick test of this circuit was to short the gap with EDM on but not machining a part, and watch for deflection of the ammeter. Assuming your machine has this type of protection circuitry, you would NOT see an increase in current if is functioning. Oh, FYI, as of late, I have been seeing a lot of failures in smaller sprague electrolytic caps. I actually had one that was shorted about a month ago.

Hope this helps,

Scott

In a hundred years, it isn't going to matter anyway.

 

[socrace]

If you look closely enough at the schematic, it's clear it can't work as drawn, I guess that's what some of the dotted lines are supposed to mean; missing parts. For example, the driver stage can never swing positive enough to turn on the outputs the way its drawn. Just wondering what purpose this kind of "pseudo-schematic" is supposed to have? Put service techs on a merry chase maybe...
Bob D

 

[OperaHouse]

It doesn't have to swing positive above ground. This is evidently a pump charge from the capacitor. The capacitor is charged when the driver is at the off condition. when the driver goes to ground, the stored charge in the cap drives the transistor. That is the reason there is a resistor from the cap to B. Not exactly great for long on cycles and short off periods. A failed driver can never turn an output on.

 

[socrace]

OK, thanks for that OperaHouse, makes much better sense now. Have to wonder about using charge caps though, if they drift over time to a low value, then the outputs might not fully saturate. Wonder if they're using electrolytics?
Bob D

 

[bfhamr]

Sorry I haven't posted for awhile, the project is currently on hold. The customer isn't sure he wants to sink any more money into the repair of this machine. He is considering replacing the machine.

To answer a couple of questions:

-The low voltage supply is unregulated. It consists of a transformer, bridge rectifier and a pair of caps.
-There are 2 electrolytic caps on the transistor board.
-I don't know the value of the four resistors in the circuit with each transistor. I did take a picture of the board the last time I was at the customer's shop, you can see the board here.

http://i100.photobucket.com/albums/m18/bfhamr/DSCF0351.jpg

dan

 

[LionelHutz]

Two things I've "flagged" when reading this.

The negative voltage is low. With this circuit it seems to rely on the negative voltage to charge the caps in the gate drive circuit to actually turn-on the transistors. So, low -ve supply means less base drive.

The circuit relies on the capacitors in the final part of the gate drive to turn-on the transistors. Failing caps would again mean less base drive.

If transistors fail at random then it points to problem 1 and if the same transistors keep failing it points to problem 2.

It appears the output of the final driving stage approximately swings from -10V to 0V as it is switched.

Overall, troubleshooting it further seems fairly pointless without using an oscilloscope to watch the gate drive circuitry. It's not a complex circuit so once you can find the problem the solution should be easy.

 

[analogkid2digitalman]

I noticed some debris sticking to the transistor heatsinks. Any chance this could be conductive- remnants of metal being etched away by the EDM process? Could this be a possible root cause of failures? Just going off of past experience with brushed motor life testing- carbon brushes wore which caused carbon residue to eventually be deposited on PCB's causing lots of field failures.



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"It's the questions that drive us"
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 

[Snoball]

The board in the picture doesn't represent the schematic.

 

[LionelHutz]

Agreed, there seems to be a bunch of diodes in the picture that are not on the schematic as well as some terminals and such. Different revision maybe?