fender rhodes power transistors

[pontiacpete]

Does anyone know if there is a substitute for the 120725 transistors in the fender rhodes?

[Enzo]

CAn you link or post the schematic? Fender made many Rhodeses. I hate to have to get out all my old Rhodes stuff from the warehouse to do enough research to find out what you have.

You got a good one? is it silicon or germanium?

[pontiacpete]

mouser and another source on the web say an NTE179 is a good sub, but I also heard that they need to be 'matched'. How important is that?
I have the all the rhodes schematics and am trying to determine which one is the correct one.
So far it looks like the one below
It's germanium. Both of them might be shorted. It looks like one side of the power amp is not working because the other module doesn't work on both sides.
The .5ohm/5w, the 820ohm and the 2.7ohm resistors blew.

[pontiacpete]

here's the power supply

[R.G.]

If the one pontiacpete posted is the schematic, it's almost identical to the Thomas Vox "Royal Guardsman" power amp, in that it uses a driver transformer to be a phase splitter driving a stacked pair of outputs. In that circuit, for that time, the fact that they're PNPs probably means that they're germanium.

In any case, replace them with PNP silicon, BVceo of 140Vdc or more, Ic of 10A or more, Pdiss of 150W or more. This will cost you about $3-$5 each, maybe. Theoretically, they should be matched, but the matching is solely to get the DC offset on the output to be low. I'd buy two from the same vendor, test the offset, and use it if it's under 0.1 - 0.2V.

If you go for silicon, you must rebias. The bias must be set for each transistor separately. You change bias by increasing the value of those 2.7R resistors. Kick that up to about 6.8 ohms each and look for crossover distortion, increasing both of them equally until crossover just stops. If you overbias with too high a value for these resistors, the output heat sink will get too hot. Back them down if that happens. When you get crossover right, check for DC offset on the output, and if necessary, you can diddle in the offset by paralling bigger (10K and up) resistors with one or the other 820 to change the DC current per device a bit. Don't go wild with this.

I have done this process to the Thomas Vox amps several times. It works fine.

If it's not transformer coupled as with the pontiacpete schemo, then other advice applies. But I'd go for modern silicon devices and rebias. Suitable power germaniums will be $30 or more each, if you can find them, and then you get to worry about matching.

[Enzo]

SOmewhere there is even a Fender bulletin on converting to silicon. Says basically what RG said.

[pontiacpete]

Thanks, it's a relief to hear there is a low cost sub.

So far this is what I've been able to find: MJ15025 -- will this work?

[Enzo]

That transistor will work just about anywhere you need a PNP power transistor.

That plus the conversion to silicon details.

[pontiacpete]

I finally got to work on this, installed the mj15023s and new resistors where needed. Now I'm only getting half the output, the wave on the output is clipped off on the positive peaks.
Could this mean the transformer is blown on one side?
I did a bunch of resistance tests in circuit and find that the 270ohm 2w resistor on the output to ground measures 8.2ohms, whereas on the good module it measures almost to spec in circuit. When comparing resistance tests to the good module on the transformer there are some inconsistencies, leading me to believe the tranny might be bad.
If it is, is there a replacement available?

[J M Fahey]

1) Check that the secondary windings have not blown open. Lift one end, and measure resistance, it should be alow value, say 2 to 12 ohms, and both practically the same.
2) You should have around 500mV bias ; from top to bottom your string should be something like RBias>820r>RBias>820r; RBias being around 15r in theory, but I should use a safer 12 ohms, or even 10r .
Crossover is ugly but a burnt or unstable amp is even worse, play it safe.
3) A Zobel network (10r1W in series with a .1 cap) from output to ground (or across that 270r 2W should help stability.
Scope it and drive a real world load (speaker).
Is the waveform clean or has "fuzzies" which indicate HF oscillation?
If so, try removing (lift one end) of those 1200 pF in the feedback loop.
With the old, *very* slow germaniums they might have helped; today with modern transistors it may contribute to instability.
This amp was obsolete already in 1968, go figure.

[R.G.]

I finally got to work on this, installed the mj15023s and new resistors where needed. Now I'm only getting half the output, the wave on the output is clipped off on the positive peaks.
Could this mean the transformer is blown on one side?

It could, but it's more likely that the resistor was bad. Pull up a leg on the resistors to measure them because the low impedances in the circuit will spoil the measurements.

Also, it is still a transformer. Pull the transformer leads loose, noting carefully where they went so you can put them back where they belong. Now feed the primary a test signal from a generator and look at the output voltages on the secondaries. If they're both correct, it's not the transformer.

I did a bunch of resistance tests in circuit and find that the 270ohm 2w resistor on the output to ground measures 8.2ohms, whereas on the good module it measures almost to spec in circuit. When comparing resistance tests to the good module on the transformer there are some inconsistencies, leading me to believe the tranny might be bad.

I'd eliminate the resistors being bad first, because...

If it is, is there a replacement available?

No, I doubt there's a chance on this, other than cannibalizing another similar amplifier.

I have reverse engineered the driver transformer for the Vox Beatle, which is quite similar in operation. It is possible to rewind this one if it's bad, but be really, really sure it's bad before committing to this.

To test the transformer,
(1) map out from the schematic what lead connects to what other lead through wire resistance, and which ones should show infinite resistance. Then get out your meter and find out if a winding is open. All of the windings on this tranny will be under 1K if they're any good, over 1M if they're open.
(2) then figure out whether there is an internal short. The transformer shorted turns tester at geofex.com will tell you that.
If the trannie passes both 1 and 2, and has ever worked, then it's not faulty for most cases. There are a few weird ways to fail that don't apply here.

The chances are that the resistors and/or transistor wiring are the issue, not the transformer.

[pontiacpete]

I pulled the other .5r/5w resistor that I hadn't changed and it measure quite bit higher than the new one, at .9ohms. I paralleled the 820s with 10k ohm resistors and I was able to dial out the flat top on the positive peaks. But now there is a massive hum. On the negative rail i measure -37 and the other is 34vdc. I changed both electrolytics on the board. Still hums.
What else could cause this hum?
There's quite a burn on this pcb and the traces are in bad shape.

[J M Fahey]

RBias>820r>RBias>820r

Can you please post the voltages you read across those resistors? as in, (say) .5V>32V>.52V>31V or whatever. No speaker attached.
Also DCV to ground from output terminal.
Did you measure the secondaries DC resistance?

[pontiacpete]

funny thing is that I messed up on my installation of the transistor, so re installed another with a new to3 holder and it still hums.
I'll get those voltages readings later today, got to run to a gig playing a bunch.
I thought the bad transistor install was it.

[R.G.]

I pulled the other .5r/5w resistor that I hadn't changed and it measure quite bit higher than the new one, at .9ohms. I paralleled the 820s with 10k ohm resistors and I was able to dial out the flat top on the positive peaks. But now there is a massive hum. On the negative rail i measure -37 and the other is 34vdc. I changed both electrolytics on the board. Still hums.

It is likely that the difference in the power supply voltages is from different loads on the + and - sides. That causes a much bigger ripple in the + side, and the difference in ripple can cause hum, too.

If it were me, I'd remanufacture the power amp - replace every single resistor and cap with new modern devices, possibly with bigger power ratings, and possibly all the semiconductors too. That sounds drastic, but it really costs very little more. It's much less costly than an hour of your bench time.

What else could cause this hum?

Many, many things, right down to where the wires run. I personally think it's from the difference in the plus side versus minus side power transistors idle currents, but it could be many things.

There's quite a burn on this pcb and the traces are in bad shape.

That's another vote for there being something in there that has been very damaged. If the PCB is charred, that charring is conductive. It has to be sanded away or it will permanently act like a resistance between any traces it touches.

I thought the bad transistor install was it.

It's important to remember that there may not be any "it". There is often "them". The single-fault assumption does not apply to very old, or very abused equipment. Those cases are much more like peeling an onion.