1. You can now help support WorldwideDX when you shop on Amazon at no additional cost to you! Simply follow this Shop on Amazon link first and a portion of any purchase is sent to WorldwideDX to help with site costs.

PAL 200MDX P/S transistor replacements

Discussion in 'Amplifiers' started by Aztec, Mar 20, 2017 at 9:59 PM.

  1. Aztec

    Aztec
    Expand Collapse
    New Member

    Joined:
    Dec 2, 2016
    Messages:
    17
    Likes Received:
    1
    Hey guys,

    I just aquired this old mobile tube amp. It's a PAL200MDX.

    Unfortunately I think in my excitement I ended up killing it. I seem to have killed the two transistors that drive the transformer. They are unmarked but the schematic says they are PE250 AKA EF250. They are stud style transistors with a bolt on the collector. I can't find any info on these. They are PNP types. I'm wondering if anyone here has any info about these or can suggest a replacement.



    Here is the schematic.

    http://www.firestik.com/Instl-Art/PAL200MDX-dc.pdf

    When I was setting it up for testing I placed the amp on top of something metal. This shorted the collectors of the two drive transistors on the bottom against the case. Apparently negative from the battery and the amps chassis ground are supposed to be isolated as the transistors are isolated from the case with mica insulators. Schematic doesnt show that but on the amp those two grounds are isolated.

    When I keyed it up it made a squealing sound but after that it kinda stopped. When I key it up now it stills squeals a tiny bit and puts about 40 Volts on the plates of the tubes but that's about it. Transistors are testing a little bit weird. Not totally shorted, still conduct in one direction on certain terminals but voltage drop reads around 100mv. Weirder still is that even though they seem to be shorted, when I key the amp up they don't get hot at all which I would expect if they where totally dead. I read somewhere that early power transistors of this type where germanium, maybe that's why I have such a low voltage drop reading? Who knows.

    I would love to get it to work again. Stupid mistake I made. Though in my defense this thing seems extremely vulnerable with those transistors being so exposed like that. With this being a mobile amp I imagine it was often at risk of being put against the metal floor or dashboard of a car.
     

  2. Robb

    Robb
    Expand Collapse
    Yup

    Joined:
    Dec 18, 2008
    Messages:
    9,827
    Likes Received:
    1,821
  3. Aztec

    Aztec
    Expand Collapse
    New Member

    Joined:
    Dec 2, 2016
    Messages:
    17
    Likes Received:
    1
    Don't think that's them. These come in an old TO-36 package they look late 60s early 70s. They are germanium and PNP. That's all I know.

    Anyone know what the plate voltage on these tubes should be?
     
    #3 Aztec, Mar 20, 2017 at 10:37 PM
    Last edited: Mar 20, 2017 at 11:11 PM
  4. nomadradio

    nomadradio
    Expand Collapse
    Analog Retentive

    Joined:
    Apr 3, 2005
    Messages:
    569
    Likes Received:
    118
    The transistor numbers are "house" numbers, custom-printed on them for a customer buying them in bulk.

    Pretty sure they should be type 2N4048, rated at 60 Amps, 30 Volts and 170 Watts, so long as the heat sink is big enough.

    Those were prone to fail for no good reason, or from overheating. A shorted tube could overload them all by itself. I would suggest you simply assume that the 450-Volt filter capacitors have shorted and replace them before installing expensive transistors. If the HV filter caps are original, they have no excuse to still work this many years down the road. That kind of capacitor can short, overload the transistors and then chemically "heal" the short when you go to check them with a meter. The next time high voltage is applied, the cycle repeats.

    Normal plate voltage is around 800 Volts DC in this amp, pretty sure.

    Those transistors are a bit like the tubes that amplifier uses. When they stopped making them the prices shot upwards. Would be cool if some design genius figured out a way to use five-dollar MOSFETs in that circuit.

    73
     
  5. Aztec

    Aztec
    Expand Collapse
    New Member

    Joined:
    Dec 2, 2016
    Messages:
    17
    Likes Received:
    1
    Well I fixed it kinda. And yes I figured out how to make it work with mosfets. I originally intended to rebuild the circuit as is. However even with modern day transistors it still wouldn't be efficient or stable. What I did was build a CD4047 oscillator circuit that drives two IRF3205 mosfets. The mosfets are rated higher than the original transistors and are much smaller. I kinda didnt think it would work unless I paired up more mosfets but it seems to be doing just fine.

    I can't get a reading of the plate voltage while transmitting as an RF loaded plate connection isnt safe grounds for my digital meter. Right now the oscillator has variable frequency as I don't know what frequency the inverter used to use. It's a laminated type ransformer not ferrite. .

    At around 500hz or above I get roughly 330-350 volts on the filter cap which is around 700V B+ past the doubler. At below 450hz I get 450 and up to 500 volts. Which is around 1000V. I havent gone below say 300hz. However at this lower frequency the current consumption goes way up, the pilot light dims.

    In summary below 450Hz more voltage but higher current draw. Above 500hz a little less voltage but also less noticeably less current draw. Not sure which is the right one.

    It would be nice for anyone to give me some input on the frequency theyve heard these inverter things work on. Again this one has a laminated transformer not a ferrite one so I would expect a lower frequency. I had a ferrite transformer amp once and it seemed to work at around 1khz or so.

    I need a jumper to hook up my wattmeter, when I do I'll try and report my findings about inverter frequency.

    Ill share the mod once everything is finished.
     
    #5 Aztec, Mar 22, 2017 at 8:54 AM
    Last edited: Mar 22, 2017 at 9:46 AM
  6. nomadradio

    nomadradio
    Expand Collapse
    Analog Retentive

    Joined:
    Apr 3, 2005
    Messages:
    569
    Likes Received:
    118
    The original frequency was loosely determined by the resonance of the transformer, more or less. The frequency would be highest with no load but the bleeder resistors and drop with increasing current load. The old-time practice of "dipping" a plate-current meter used the fact that highest RF output usually coincides with minimum plate current when tuning the output tank circuit for resonance. You could do that with one of these by tuning the whine for highest pitch. Does this one have a ferrite core transformer? Or one with an laminated iron core?

    The ferrites would whine at a higher pitch, but I never remember measuring the frequency. From memory, I would guess around 2 kHz or more unloaded, down to 600 or maybe 800 Hz under full load. Of course it would "warble" in step with your voice modulation.

    Finding the frequency where it's most efficient would be the goal, I think. Regulating the output by controlling the duty cycle should work, but keeping rectified RF out of the feedback-control circuit would be a priority, too.

    Rock on! You've gotten farther with this than anybody else I know.

    73
     
  7. Aztec

    Aztec
    Expand Collapse
    New Member

    Joined:
    Dec 2, 2016
    Messages:
    17
    Likes Received:
    1
    This is a laminated type transformer. I read those are supposed to be lower but from my testing everything seems happiest at a higher frequency.

    Earlier I ended up killing the mosfets by taking the frequency below 400hz, this time I put thicker wires on the battery thus more current was available and it took too much. The voltage goes up and so does the current consumption at lower frequencies. I would have thought that would equate more power but instead it made amp angry it seems. When the voltage goes up it also seems excessive. Based on the fact that the voltage goes up around 500V which is the caps maximum voltage I would have to guess that the actual voltage its supposed to put out is lower than that, closer to what it gives me when at 500Hz-1Khz. When at 500hz or above the voltage is more like 350 or 380 which seems a lot more reasonable. From 500hz to around 1Khz there doesnt seem to be much of a difference in how it affects the amp. Considering that you say that these tended to shift frequency depending on load, I imagine the transformer has a flat frequency response throughout a range of requencies. Considering there's audio output transformers that go from 70 to about 5khz completely flat, I would have to say a transformer that goes from 500khz to 1khz without complaining isnt a crazy idea.

    I replaced the mosfets again and the amp is chugging along again. I made new jumper wires so this time i was able to clock the wattage at around 20 watts in low and around 90 to 100 watts in high, looks like my tubes are good! Supposed to be a 200 watt amp but we all know that's BS.

    I'm driving a stock Uniden PC68 into it. When I modulate the wattage swings down, this is so far the only problem I'm seeing. I see this same thing on radios that have the carrier turned up. Not sure if its the radio that might be doing that or if the amp has an issue or if i need to turn the PC68 down. I{m using a micronta 21 524 which a simple meter. Could also be the battery getting low. Need to fix my PS.

    Anyway yeah so far it seems the magic frequency is around 500hz to 1Khz. The Xformer is at its most efficient at that frequency range and doesnt have one magic frequency. Setting it anywhere between those frequencies should be okay though preferably in the middle. The other question might be whether it NEEDS the frequency to shift with load, I would say no. That was probably just unavoidable with the original oscillator.

    I'll await any other comments you guys might have would be nice to hear from another person who owns this amp. I will post a schematic of the control circuit later once all testing is done. RF doesnt seem to be getting into it either. A simple decoupling cap should be enough.
     
    #7 Aztec, Mar 23, 2017 at 5:03 AM
    Last edited: Mar 23, 2017 at 5:13 AM
  8. nomadradio

    nomadradio
    Expand Collapse
    Analog Retentive

    Joined:
    Apr 3, 2005
    Messages:
    569
    Likes Received:
    118
    Sounds like comparing power consumption to HV power output is the only way to establish the frequency where it's most efficient.

    Since the oscillator feedback came through the original transformer, the frequency change was a side effect, pretty sure.

    Finding the 'sweet spot' frequency, and then using PWM control circuit to regulate the HV is probably the way to go. Have never done that trick, so I'm guessing.

    73
     

Share This Page