Announcement

Collapse
No announcement yet.

Considerations For Replacing Power Transistors

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Considerations For Replacing Power Transistors

    Members -

    In the past, I have swapped out Power Transistors based on recommendations, suggestions here on MEF, and other sources.

    As an example the 2N3055, 2N3773, RCA 4815, etc.... It has been recommended to use the MJ150033 as a replacement. But I would like to know, what do you look for in a replacement?

    Breakdown Voltage?
    Max Collector, Base, or Emitter Current?
    Hfe?
    Power Dissipation?

    What about bias? Should I assume that the MJE will run at the same bias level as the predecessors? If I am not mistaken, I once saw something here that stated "if you use an MJ15024 in place of a transistor listed above, you might need to tweak the bias." If that is the case, what is the spec I would look for that would tell me a transistor would behave similarly to the original?

    Thanks,

    Tom

  • #2
    There are some "gotta have enough" specs, and then there are some subtle ones.

    The ones you mention are largely "gotta have enough" types. If a transistor doesn't have enough breakdown voltage, current rating and power dissipation capability, it dies more or less suddenly. Hfe is about half subtle. If you don't have enough hfe, your drivers can't drive the transistor hard enough (in many instances) so the driver devices get overloaded and may die. Too much hfe could leave you more prone to leakage and thermal runaway, although not so much today as in the bad old days. In general, you want to have at least as much hfe AT FULL CURRENT operation as the originals. For normal, not Darlington, devices, hfe being larger up to 100 or so will not be a problem.

    Here are some caveats on the "gotta have enough" specs.
    Breakdown voltage ought to be at least twice the sum of the power supplies for the output stage. If the output stage runs +/-50V, use at least 200V breakdown devices. This is largely because the processes that make for good high voltage transistors also increase safe operating area, and you need all the SOA you can get.
    Collector/emitter current ought to be (IMHO) at least three times the peak current the outputs will see into the lowest-ohm rated loads. You can get big ones today, so go ahead and use them. 200+V and 20A devices are commonly available for $5 each. Use them.

    The vast majority of amp output stages run the outputs as either emitter followers or complementary feedback followers. In either case, the output transistor is expected to produce no voltage gain, only current gain. This operating mode produces the highest frequency response for the devices, and so with today's much better devices, frequency response and exact current gain are not critical like they once were.

    You can probably quit there for guitar amps. The semiconductor industry has changed, and the number of new complementary audio output transistors has shrunk. For many amps, you can simply pick one of a few types and that's what you use, period. The MJ150xx series is good, as are the Toshiba power outputs if you can find them. On Semi has some of its own take on the Toshiba "extended beta" devices, and they're good. Sanken has a few. But there are few enough and they are good enough in terms of high frequency response and thermal dissipation that you generally don't have to sweat the selection that much.

    This advice is NOT for snooty gold-plated hifi tweako amps where the owners swear they can hear the difference between tin-lead solder and tin-lead solder with 2% silver.
    Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

    Oh, wait! That sounds familiar, somehow.

    Comment


    • #3
      Thank you RG.... that is a great analysis. I appreciate the info.

      Comment


      • #4
        You're welcome. I forgot a big one - counterfeiting. Toshiba output transistors are very good - and widely counterfeited. The On-Semi devices less so than Toshiba, but the chances of getting a fake from ebay are HIGH. Even secondary dealers or surplus dealers may be tricked into buying stock that is counterfeited. The worst problems will be with fakes that look OK to low power testing but fail under any heavy current or power.

        The bigger distributors have all been bitten by fakes getting into the distribution channel, so they have instituted authentication trails with the manufacturers to (help) prevent it. I would trust (... but verify!) devices from Digi-Key, Mouser and some others. Everything else, I'd power-test before using it in a critical application.
        Amazing!! Who would ever have guessed that someone who villified the evil rich people would begin happily accepting their millions in speaking fees!

        Oh, wait! That sounds familiar, somehow.

        Comment

        Working...
        X