Disc type Thyristors 1

[Drivesrock]

Hi all!

I probably know the answer but have been unable to convince my customer as to why it is necessary to clamp a disc type thyristor when wanting to test it's properties with a multimeter. Without contacting a thyristor manufacturer can you guys tell me me the reasons? Any suggestions as to how much to clamp a device (the force and area) and what sort of readings to expect? I would put them into an insulated vice and tighten and compare measurements with other thyristors to identify 'suspect' devices. Any other tips?

Thanks in advance.

Drivesrock

 

[ScottyUK]

The hockey puck type package only makes contact with the silicon die when it is under compression. Test on an uncompressed die are meaningless. I have tested properly clamped devices successfully using a 110V DC source with a couple of big halogen lamps or a resistor load bank to draw about 20A. Low current tests such as you might have built in to your multimeter don't bring the die fully in to conduction and can give weird and / or non-repeatable results. The gate drive requirement could well be quite high, perhaps as much as an amp.

Don't put the thyristor in a vice!! There are purpose-made heatsink / clamp arrangements which exert a uniform and carefully calibrated clamp force on the device. Too little pressure and you get inadequate contact to the die; too much and you distort the package and fracture the die. A fractured die will probably work for a while, then fail without warning. Usually the failure will take out a load of other expensive components (or is it just my world that happens in? )


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[Skogsgurra]

Yes, you know that already. I can see that from what you are doing.

Some (not all) puck thyristors rely on outside pressure for contact between disc and semiconductor. It is soldered (always?) cathode side, but pressure contacted anode side.

I wouldn't use a multimeter. OK, you can do that to test gate-cathode (anything from a few ohms up to 100, polarity independent), but for reverse and blocking voltage, I would use a megger.

Remember, these are sturdy things so a correctly used megger is not a problem. But, connect before starting megger. Some devices have a capacitor inside and if you connect after starting the megger, this capacitor is charged and will cause a current to flow (charge carrier separation) and this might trigger the thyristor "over head" and cause a very intense discharge current with a lot more A/us than the device is built for.

Same thing can happen if you try to trig the thyristor with voltage applied - that's why we have commutation reactors in the working circuit.

I made that mistake once. After havning destroyed two thyristors, I realised that something was wrong. Expensive lesson.

Meggers without that internal capacitor should be OK, though. Most are.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[LionelHutz]

The puck is basically built with 2 round copper pads having pressure contact on the round die in the middle. The ceramic insulates the 2 sides. The flat ring at the cathode side is the flex seal. The manufacturer puts the 2 sides together and then pressure welds the case together at that ring. Since it is just a touch contact inside you need pressure to ensure proper contact. You can usually get enough pressure with your hand for a resistance only check.

Most of the puck thickness is in the copper pads. The die is usually no more than about 1/8" thick so, for example, in a 1" thick puck the copper pads on each side are close to 1/2" thick.

Most of the pucks I've seen start at about 2000lb of clamping pressure. So putting some pressure on them with a vise is not likely going to hurt them unless you are dealing with the smallest of puck sizes. The largest puck I've personally seen required about 15000lb of clamping pressure.

If you really want to see one opened I can probably dig up a picture or 2 and post them.

 

[itsmoked]

Yes! yes! Dig away! Need picture.
I have never heard this "clamp to work" stuff even after years in the semi business. How bizarre! 15klb!!!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[LionelHutz]

OK, here's a couple of pics of a Eupec T1189 SCR. This is a 1200A SCR that failed and was taken apart. This device has a 48mm face (die diameter) and is 26mm thick.

This pic shows the device with the seal at the cathode cut open. You're looking into it from the cathode side. You can see the gate lead connection - the black piece coming in and going to the white part in the middle. The white part actually has a spring loaded pin that hits on the gate of the device. The red is just a rubber insulator around the outside. On the left is the cathode block that was removed.

Here's a different device with the parts separated. The disc on the top right is the actual SCR wafer. You can see the gate which is the small dot in the middle. This device has an auxiliary gate which is the second ring you can see in the middle. I believe you basically trigger a small device that in turn triggers the main device. The disk on the top left is just a plate that just turns over and lays on top of the wafer.

This device looks like it was an over-voltage failure. Also could have been temperature I guess.

As you can see, it's all held together by pressure. You can also see how thick the contact faces are. They're around 1/2" thick copper each.

 

[itsmoked]

WOW Net stuff! Thanks Mr.Hutz!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com