Fender Rhodes Student Jetsons model piano question

[craig]

Hey all -

Helping a friend restore his Jetsons 69 piano.  He wants to replace the germanium output transistors with silicon on the amp module before they fail. It's a Peterson amp module (w/820r,270r,.47 5w resistors in the output section), but the output transistors are Delco GM 801552, not the usual ones I'm accustomed to seeing on 80w Peterson suitcases.  Has anyone had success swapping these out for Silicon ones?  Would the Vintage Vibe silicon conversion kit (matched transistors, bias resistors) that they offer for the suitcase amps work for this? Thanks!

[The Real MC]

Germanium and silicon transistors are not interchangeable without a major redesign.

[Tim W]

Redesign being needed is true for complex designs, but not for the simple Peterson output stage.

Increase the 2.7 ohm resistors to 12 ohms, 1W.
Change out the 820 ohm 2W to 820 ohm 3W (2W was marginal to begin with)
Replace the 0.5 ohm 5W sandblocks with new ones.  0.47 ohms are OK too.
Replace the Delcos with ON Semi MJ15016s.  These need to be somewhat matched, as long as the DC offset of the amp output without a load is 150mV or less you are OK.  Otherwise swap out one for another MJ15016 and recheck offset.  Usually within a batch of the same date code they are close enough- it is rare we need to swap them out.

With the 12 ohm bias resistors, you should not have any crossover distortion when cold.

So if you buy 4-5 MJ15016s to guarantee you get a close match and the resistors from Mouser you will be out about $30, and have a few spare transistors.

The MJ15016s are a 15A 120V part, so they have a much better SOA (safe operating area) compared to the original Delcos.  Simply put, they won't run away and catch fire like the Delcos often would.  Plus, the Delcos were only rated for 60V.  Part of the reason they eventually run away is because the Rhodes power supply runs at +/-40 to 43 VDC.  When the amp swings to drive the speakers when loud, the transistor not conducting can see more than 60V across it.  Each time this happens it like shooting the transistor with a BB.  Enough BBs, and the transistor dies.  In a Rhodes, this results in a catastrophic failure.

 

[craig]

Tim - thank you so much for the info, it's extremely helpful. I've swapped the GE for SI transistors before on Peterson suitcase modules with success (w/12 r bias resistors). I just wasn't sure, since the +- voltage on the student piano is 30v (not 35-40v like the suitcase) which transistors to use, and which ones would work/how much to bias. One question though, how do I check the offset?  Unplug the speaker and measure there?

I've had mine catch fire before...in the middle of a gig. Not fun.

[Tim W]

For DC offset:

Measure the DC output voltage between the amp output and ground without a load connected.  It is best to do this first anyway, to insure you don't fry your speakers after rebuilding the amp.  ALWAYS check the DC output voltage and make sure the amp passes a signal (preferably with a scope) properly first with no output load, then with a dummy load, before connecting to your speakers.  Much easier to do this than paying for replacement speakers or a recone (especially with the oddball 32 ohm suitcase speakers).

Since your piano may have a lower supply voltage than a suitcase, measure (don't trust the schematic) the power supply on your piano.  If it is +/- 30V rather than +/-40 you may need to listen (or scope) the output for crossover distortion while connected to a load when the amp is cold (right at power up).  Crossover distortion will decrease as the amp heats up, but you shouldn't have any when it starts out cold.

If you have crossover distortion, you may need to slightly increase both the 12 ohm resistors.  And I mean slightly- maybe 1 or 2 ohms AT MOST.  We have never needed to do this in practice, and we've rebuilt many Petersons, but we might in the future.  You never know.

Good luck with the project- but be careful- you can get hurt working on this stuff!

[craig]

Thanks again Tim.  I tried this morning with some MJ15025G's that I'd gotten from Mouser last year and the 12r bias resistors but it distorted so I'm going to order some MJ15016s as you suggested and follow your directions.  Hopefully that will do it.  Oh, and the amp does measure just under +-31v.  Thanks again for the help!

[craig]

On a hunch I threw in 15r for bias with the MJ15025G's and it works great.  Just letting it run for a bit.  Thanks again for the help.

[Tim W]

Good to hear.  Those transistors are even tougher than the MJ15016s.  Not surprised you needed a little more bias at the lower supply voltage.  Hook it up to a dummy load that equals the speaker load impedance.  Then run it so that the output voltage peaks at about 63% of the supply for a good long time and get it good and hot to make sure it doesn't run away.  If it makes it you're golden.

Best
Tim