Identification of transistor/semiconductor

Thread Starter

Denesius

Joined Feb 5, 2014
119
I'm trying to repair an inverter board for a mechanical gyroscope (the gyro is a J.E.T. AI-904, 501-1170-02, and the board converts 28VDC to 3-phase 400Hz AC). I do not have schematics or parts list. I've been able to figure out 2 defective parts in the leg that doesn't work: one is a UA748H opamp, for which I have found a replacements.
The other part appears to be a transistor in a TO-12 case with faded markings and obvious overheat discoloration. As best as I can tell under a microscope, it's labeled "7943" and "U21101". An extensive internet search has come up empty.
My transistor tester sees it as 2 diodes, with Uf=659 & 748mV.
Other similar case devices in the circuit, without labels, are identified as PNP with B=82 and Uf=670mV. I'm not sure all the parts are the same. The other legs are all sealed in a gelatinous substance that's impossible to remove without possible damage to the components, so that route is not open.
Any help from you smart guys about the mystery 7943 part would be greatly appreciated.
A schematic for the gyro would win you accolades and a small token of appreciation!!
Thank you one & all.
 

Ian0

Joined Aug 7, 2020
3,525
Would you really expect to find a μA748 and a MOSFET on the same board???
So, it's either a double diode or a bipolar transistor, and it's silicon.
TO-12 is a 4-pin component with an extra lead to earth the case. That suggests a high frequency component.
Is that likely to be true?
 

Thread Starter

Denesius

Joined Feb 5, 2014
119
Start by posting clear focused photos of the board.
IMG_5645.jpg
This is the board. The small blue dots are on the Opamps, the large red dot is on the heatsink covering the assumed transistor devices x 7.


IMG_5654.PNG

These are the TO-12 devices, with the one removed from its heatsink shroud. The markings are barely legible. The other one I removed with difficulty is unreadable

IMG_5655.jpg

This is how my transistor tester sees this device.....
 
Last edited:

Thread Starter

Denesius

Joined Feb 5, 2014
119
Would you really expect to find a μA748 and a MOSFET on the same board??? ..........
Thanks for the input.
I don't know the base frequency of the inverter. I know it converts 28vDC to 3-phase 400Hz AC. I assume that's the role of the Opamps. I would also assume if the heatsinked devices are all 7943 transistors, then they serve as the drivers. Why there are 7 I don't know, and the remainder all test different. 5 read as double diodes (burned out gates??), one reads as a PNP device.
 

srobb2001

Joined Dec 5, 2017
2
7943 is probably a date code, 1979 and workweek 43 (end of October) and indicates the date the device was assembled. You should check other distributors. I may be available in Europe.
 

Thread Starter

Denesius

Joined Feb 5, 2014
119
7943 is probably a date code, 1979 and workweek 43 (end of October) and indicates the date the device was assembled. You should check other distributors. I may be available in Europe.
Thanks srobb2001, but the gyro has a date code of 2001. I can't believe they'd use transistors from 1979! Perhaps the U21101 is the part designation?
 
I saw the picture you sent. Extremely old analog technology so I would be not be surprised at all if that is in fact the date code. Regardless I agree with you that the part number is more than likely U21101. If you google that number it will show up on European distributor websites. Can you send me the markings on the parts that are standing up (rectifiers?) on bottom left of the board?
 

Thread Starter

Denesius

Joined Feb 5, 2014
119
I saw the picture you sent. Extremely old analog technology so I would be not be surprised at all if that is in fact the date code. Regardless I agree with you that the part number is more than likely U21101. If you google that number it will show up on European distributor websites. Can you send me the markings on the parts that are standing up (rectifiers?) on bottom left of the board?
Thanks for getting back.
If you're referring to the black devices, they are ceramic caps with various markings of their ratings/capacity. They are okay.
I have since had 2 revelations:
5 of the transistors are part# U2T101, which are readily identifiable on the web. They are old PNP devices, so I still have 2 parts, one of which reads as NPN, the other as 2 diodes (obviously cooked). At this point I think I have figured out that the inverter uses the opamps to generate the 3 400hz phases, and the U2T101 devices as drivers. Now to hunt down the remainder! If I work out the circuit I'll post it for the benefit of others.
In the meantime, if there's any guidance as to how this circuit may work, I'm open to suggestions.
Thanks again.
 

Thread Starter

Denesius

Joined Feb 5, 2014
119
Well, thanks to all for trying. After spending better part of a day on a search, I managed to round up the schematic, as well as the mystery parts. This device was originally manufactured in the 1970's. For anyone historically interested in how the US military generated 3phase 400hz AC from 28VDC in an aircraft gyroscope, the schematic is attached.

Schematic.jpg
 
Top