Maybe I am wrong here, but when we're talking three and four terminal devices, I don't see the problems some face. I have tried diligently to get my co-workers to get the hell out of the ECG book and compare things for themselves.
The appilacation should give you the needed specs. If this is an RF power transistor, which it was in the last post I read, you can go on a datasheet search and find what you need usually.
One of the things people in the field need to realize though, is the basic difference between a switching transistor vs. one that is meant to operate in it's linear range.
If we are taling an RF transistor here, they almost always operate in class C, although this is not true of small signal RF transistors.
As people ask for mundane things, like outputs for an audio amp and stuff like that, that is all easy. Personally the first thing I do is to find out if I can get the originals. THEN instead of looking in an ECG book, I look at the circuit.
For example, I got this Mac amplifier, I know it uses darlingtons, not from them telling me, just from the cicruit that drives it. It looked wimpy, and the thing had a decent damping factor, s those are darlingtons. Sansui used them as well, their SJ series of transistors were darlington.
But we weren't talking about that.
Let me roll out a newer experience. Reloze that I work for an ISC and info is sometimezs lacking.
A newer Sony came in, I looked around, had never seen this model before. Had a shorted HVOT.
Now I am figuring the 2SC5150 I was going to put in it would run long enough for me to cop a waveform. And if it didn't stay bad I would reconnect the flyback output.
Well, when I saw the zener diode across the source and gate I knew. I never thought MOSFETS would get up to the voltages required for HOTs and HVOTs. but the proof is in the lack of a driver transformer, but that is not conclusive. What is conclusive is the totem pole SMD transiastors that run the gate drive, so I knew I could not throw my favorite 2SC5150 in there, and we had nothing like an FET that would handle it.
We found out subsequently that the FET cost as much as the flyback, and this was the HVOT, not the HOT.
One acceptable way to go. Get two FETs and one flyback, if the FET deos not blow with the new fly, it was the fly. If the new FET blows again disconnect everything from the fly except the feedback loop. If the waveform stays bad, disconnect the feedback loop and watch for the good wavweform without the feedback (whatever it is) disconnected.
So, back to the OP, you have to know the frequency range, in what band does this operate ? Oops, that is another thead.
You need to know a few things, what voltage is it fed and meant to control. how much current and at what frequency ? Also you must consider the drive circuit of course.
When I have all that info, it is not hard to select a replacement.
Now when you get to obsolete types that need to replced with modern equivalents, you will most likely have to deal with lower intetrelectrode capacitance and higher speed devices. If tolerances were built into the circuit design you might be off scot free.
All you have to do then is choose the best device with the same case style and of course the type.
In the specsheets you will find the description of your basic bipolar, if so, when it says the original is like "meta diffused" you should use a replacement that is "meta diffused". Some processes are designed to make sweitching transistors and some are used to make linear transistors.
Same with MOSFETs, GEMFETS, all of it. It needs the base or gate drive that it needs, it is specialised.
Talking ICs ? yes, you need a datasheet, but three terminal device I really don't have a problem.
My latest problem is finding those 1200-1500 volt HOTs for direct views in the TO220 package. I want to know what the best one is, and buy a bunch of them. I'll just put up a sign "here they are".
I don't do alot of direct view sets anymore, but the company has my
100+% when I am there. I am used when I can get like 10 parts that can fix alot of sets.There is one thing I haven't been able to figure out yet, what is so special about the old Toshibe outputs, the 2SD2553 ? I have never found an adequate sub for them, even for temporary testing procedures.
And another thing, watch those lot munbers. I have replaced 2SC5148s with 5150s, and had them fail in warranty.The 5150 is the same family of transitors as the 5148, and has the exact same specs except for more current. Cost us alot of money, and you couldn't even trust the ones from B&D. Changed again they last, so that is proven.
I actually put 2SD1887s in PTK195s and they have not failed. They run even cooler as well.
So what I am saying in the end here, you do whatever you need to do.
Post the usage of this device you wish to replace, is it a final out on an amp of any kind ? Well what are the main filters rated at ? If RF. what frequency are we talking here ? Megs, Gigs what ?
You zero in on these things, and who knows where you start, but you need to zero in on the right solution. That is why some of us get the big bucks.
JURB