Failed RF power transistors...DC testing (i.e. with multimeter)?

[eccerr0r]

When a RF power transistor fails, is it still possible to test it at DC? Do they tend to typically fail in a way such that you can't detect at DC (unlike the fact most power switching transistors fail as short and can easily be detected at DC)?

Granted transistors can fail in more ways than one, but curious if anyone had experience with RF power transistor failures...

[stj]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

the gain can slowly drop over time,
i have seen them fail in the frontend of some nokia cable boxes.

if you have a meter/tester that shows HFE/gain and you have the datasheet then you should be able to check the spec.

[eccerr0r]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

The hFE/beta tester is also done at DC which I wonder if it's sufficient...

Using diode check in-situ, it seems okay, but I dread taking this device out of circuit however... going to be painful to desolder.

[CapLeaker]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

What kind of Rf power Transistor are we talking here?

[stj]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

i assume it's one of the ceramic discs with gold plated wings!!

[redwire]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

In-circuit testing rarely works because there are shunt inductors or transformers in tuned circuits, so you get low ohms across the diode-junctions.

Newer RF PA parts are mosfets, BJT's with Si-Ge and older stuff was Si BJT.
I find RF PA transistors fail short or open with not much in between. As I recall, they are arrays (like mosfets) of tiny transistors in parallel, so one shorting usually takes out the entire die.
ESD or bad antenna SWR usually kills them because they are low voltage parts.

[eccerr0r]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

In-circuit testing rarely works because there are shunt inductors or transformers in tuned circuits, so you get low ohms across the diode-junctions.

Newer RF PA parts are mosfets, BJT's with Si-Ge and older stuff was Si BJT.
I find RF PA transistors fail short or open with not much in between. As I recall, they are arrays (like mosfets) of tiny transistors in parallel, so one shorting usually takes out the entire die.
ESD or bad antenna SWR usually kills them because they are low voltage parts.

Thanks, yeah I guess that's what I was wondering, "short or open with not much in between". Yes it's a goldwing ceramic. But weird I did seem to get the characteristic C-B and B-E 0.7V drops of BJTs, which after thinking about it, should be near 0 with the inductors! I need to look at it again.

Still kinda weird, there is protection circuitry on the board. Also I'm curious how much of a SWR hit if you put a multimeter probe on the circuit while running...isn't that going to mess with the impedance of the wire? :o

The other question is whether it's worth repairing these radios if it turns out to be these RF power transistors...

[llonen]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

I guess that rather depends on the value of the kit in question, I used to repair quite a lot of amateur radio equipment when it was a more popular interest and still get work this of nature in from time to time. Your going to need an appropriate 50 ohm load and rf power meter as a basic starting point, and while a dvm in volts mode should present a high impedance however they are not generally that useful for measuring RF. You can use a scope, mostly these are very high impedance devices, some have 50ohm inputs for which you will need to take care with and use an appropriate 50 ohm attenuator.

Mostly the lower power devices are relative inexpensive, they go up in costs by power and relative rarity of device(s) though.

[eccerr0r]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

Yeah, that's the problem. It's an old single band mobile radio. I don't think the radio goes much for ebay prices (probably 30 USD at most, and shipping kills the deal) hence the hesitance on buying an ebay transistor for it... those gold wing transistors aren't cheap.

I've hacked up a homemade RF probe frontend and the readings I get are very much in line with the transmitting experiments that I've done. I don't have perfect correlation but it's very reasonable. The homemade dummy load is not very good but also seems to be fairly consistent (it doesn't get warm when it should!)

[redwire]

Re: Failed RF power transistors...DC testing (i.e. with multimeter)?

Depending on the frequency, hams are converting some transmitters from BJT to mosfet. It's also a popular mod for CB radios.

For your TX, the problem might be upstream at the driver stage, or bias for the final PA stage that need to be checked. It's odd the transistor looks asleep but tests OK.
The make/model of radio might help, if a schematic exists. I think multimeter leads do upset the SWR seen at a point but at a few ft of wire it is usually nothing compared to the wavelengths. It's usually capacitance that loads or affects an RF circuit then.