fender rhodes power transistors

[pontiacpete]

I removed both transformers to compare out of circuit and they appeared to have equal resistances between windings. Cleaned the board top and bottom with sand paper and files. Raised the value of the two 2.7rs to 15ohms you, R.G. suggested in your first post, and that did it., the sound cleared up, the notch in the wave is gone. I didn't add the 10K in parallel w/ the 820s, tried it but didn't notice any changes so left it out. Hope that's doesn't come back at me.
Also found another bad connection on the other power amp board where a pigtail fuse was mounted which affected that channel.
Thanks for the help, much appreciated.
pete

[pontiacpete]

R.G., I have a quick question, I have another fender rhodes here with blown output transistors. I replaced them and tried to bias with larger bias resistors. The 6.8Rs are not getting rid of the crossover distortion. I found that I have to bring the value of these resistors up to 15R to eliminate the distortion. You mentioned in your post above that if these resistors are too high the heat sink will get too hot. My question is, since these MJ15023s are the last i have here, should I just order some more and try a different pair?
thanks
pete

[J M Fahey]

I would not go over 12 ohms and stand a little crossover distortion.
Remember you have absolutely no thermal compensation there.

[pontiacpete]

so, do you think a different set of transistors is the alternative?

[J M Fahey]

No, it's not a *transistor* problem.
The *design* is flawed.(way too old).
Silicon is silicon, and no matter which transistor model you choose, Vbe will lower/fall 2mV/ºC , no matter what.
Modern (I mean 1969 on , go figure) amps compensate that with thermal sensing diodes or transistors attached to the heat sink.
A late 60's "patch" solution was to add some NTC thermistors in parallel with those bias resistors, but they are obsolete and practically impossible to find now.
In fact, what I would do, is to pull that obsolete power board and replace it with an LM3886 board, which will *love* those +/- 35V you have there.
Will sound *great* too.
jm2c

[pontiacpete]

hmm,, it seems strange, the last repair went well with the 6.8Rs in there for the bias resistors and a pair of mj15023s from the same batch as these. This amp won't bias in with anything less that 15ohms. I ended up ordering a new set output transistors anyway.
Do you think that the transformer could have taken a hit?
The owners says it stopped working after being on over night and there was a burning smell. Sure enough one the 820's and one of the 2.7s and one of the .5/5w s were burned. Also, This module had been repaired once with the germanium 120725.

[R.G.]

I have another fender rhodes here with blown output transistors. I replaced them and tried to bias with larger bias resistors. The 6.8Rs are not getting rid of the crossover distortion. I found that I have to bring the value of these resistors up to 15R to eliminate the distortion. You mentioned in your post above that if these resistors are too high the heat sink will get too hot. My question is, since these MJ15023s are the last i have here, should I just order some more and try a different pair?

The really pertinent question is - does the heat sink get too hot?

The "theory" about the bias resistors being about 15 ohms is my own, based on all the analysis I could do (not terribly much compared to what modern amps have had) and doing ratio-and-proportioning with germanium bias strings. And I have fixed a number of amps this way. It's very much a cut-and-try operation. If it took 15 ohms to eliminate crossover, well, just barely eliminating crossover is what they're there for. But you do want to just barely eliminate the crossover. And test it for thermal runaway when it gets hot, and in long periods sitting still. If it works OK under test and doesn't go into thermal runaway, 15 ohms was OK for that pair. You're right, it's troublesome, but I have had to go to the equivalent of 18 ohms on a similar amp.

I recently found and bought 24- 75C thermal cutouts for use in protecting amps on my personal hacking. I like having something in there that is watching for overtemp.

hmm,, it seems strange, the last repair went well with the 6.8Rs in there for the bias resistors and a pair of mj15023s from the same batch as these. This amp won't bias in with anything less that 15ohms.

I've been trying to understand that output stage for a long time. It's really quirky compared to modern all-semiconductor ones.

Do you think that the transformer could have taken a hit?

Not if you're getting signal output swings in both directions that are substantially symmetrical before clipping.

The owners says it stopped working after being on over night and there was a burning smell. Sure enough one the 820's and one of the 2.7s and one of the .5/5w s were burned. Also, This module had been repaired once with the germanium 120725.

That's the end result of thermal runaway. It heats up a bit, and that lets it get hotter, and that makes it run away more. Germaniums are really, really, really bad about thermal runaway. Germanium has a leakage (which is at the root of thermal runaway) about 1000 times (not a joke, nor picked out of the air, by the way) worse than silicon. Just going to silicon makes it hugely better.

It is also possible that the last germaniums were not mounted properly. You have to pay attention to having the transistor, mica washer and heat sink clean, and putting a very, very thin but continuous film of heat sink grease on both sides of the mica when you mount them, and getting the mounting screws tight but not fracturing the mica. I have run into Thomas Vox amps where the mounting screws on the power transistors got loose over the years. That lessens the clamping pressure on the power transistors, thermal resistance goes, transistor gets hot, gets hotter, gets HOTTER!! then dies.

[Rhodesplyr]

Interesting... I've had my own 80W/Peterson design Rhodes Suitcase since 1989. It had bad transistors in one power module when I bought it, so my brother and I replaced them.

About ten years ago, the transistors died in the other power module, so I replaced them with another matched germanium pair. It's still working, and it gets played fairly frequently, as my screen name might suggest :-) In both cases, the transistors were carefully installed with thermal compound. I've left it on all night by accident a couple of times with no meltdowns.

I won't debate the question of whether or not it's a primitive design--which it obviously is--but either it's not that bad or I've been extraordinarily lucky.

[pontiacpete]

The amp does sound that much better with the 15 ohm resistors, no crossover distortion, but the heat sink does get a little hotter than the other one.
It's funny you mentioned thermal grease, I swear these were mounted without any at all, same as the other one module, only mica.
For now I will go with the 15rs and run it for a long time and see what happens, after all I have new transistors on the way.
Thanks to all for the help.

[R.G.]

About ten years ago, the transistors died in the other power module, so I replaced them with another matched germanium pair. It's still working, and it gets played fairly frequently, as my screen name might suggest :-) In both cases, the transistors were carefully installed with thermal compound. I've left it on all night by accident a couple of times with no meltdowns.

I won't debate the question of whether or not it's a primitive design--which it obviously is--but either it's not that bad or I've been extraordinarily lucky.

To put things in perspective, the vase majority of any commercial device pretty much works. It's the small percentage that fail that give it a bad name, and that percentage can be very small, because people only see what their own personal unit did. Most of them work and keep on working for a long, long time. I'd say your experience is typical. Most of them work and keep on working.

This is a primitive design - but a nice one. It has some quirks as I mentioned; one of those is that it sounds good. It's distorted, but in a nice way for making music. That's one reason I thrashed through the design for so long. I like the sound.

Your careful job on replacement is one of the things that goes into it working for a long time.

It's funny you mentioned thermal grease, I swear these were mounted without any at all, same as the other one module, only mica.

ACK!

For now I will go with the 15rs and run it for a long time and see what happens, after all I have new transistors on the way.

Good plan!

[Rhodesplyr]

Heat is a major enemy of transistors, power transistors included. The heat sinks have to be properly sized to keep them cool. If you're turning up the heat by increasing bias current, but not making the heat sink larger, you may be asking for trouble. I suppose you could mount cooling fans inside, but there's not a lot of cross-ventilation inside the Suitcase base. On the other hand, given equal bias, a warm heat sink can mean that there's a better thermal interface between the transistors and the heat sink.

My brother used to maintain sound systems for dance clubs, and, despite his repeated advice on how to make amps last longer, he'd find that some moron had put the power amp inside a small cabinet with no airflow to "get it out of the way" or that the entire heat sink was clogged with dust.

Also, when you replace heat-sinked output transistors, it's a good idea to clean the mating surfaces carefully, removing all the old thermal compound, which can dry out over time. You want a thin, but complete coat of thermal compound between the surfaces, and I've read something to the effect that it's a good idea to get the mounting screws snug, but not to over-tighten them. I think the idea is that you don't want to cause any warping or bowing of the mating surfaces.

All this has gotten much more sophisticated with the need to cool high-performance computer CPUs.

[R.G.]

100% correct, all points!!

[vintagekeyboards]

Hello!

I am new to this group, and had a question I wanted to post in regards to this thread.

First of all:
1. I am a complete beginner to electronics. I have only taken a few community college courses on DC Circuits, AC Circuits, some Digital Logic (logic gates), and have attempted some minor builds and repairs. I realize I am most likely speaking to some veteran electronics hobbyists, technicians, and engineers.... and I also realize that many of you may not want to take the time to "dumb down" any basic knowledge of the repair I am attempting.... that being said, I would be extremely grateful for any suggestions/direction along the lines of websites and/or textbooks to bring my knowledge base up-to-par, before I attempt this repair. I am fully prepared to learn everything properly, in order, before I attempt any repairs with potentially dangerous voltages, etc.
2. My repair: I am attempting to convert the germanium transistors in my Fender Rhodes 80 watt Peterson model suitcase piano, to silicon transistors, and am just beginning my research into the necessary modifications when doing this, specifically re-biasing the output power amplifier.
3. To further clarify my level of knowledge, I have no idea what "biasing" or "re-biasing" means... and am aggressively seeking resources to educate myself with this.
4. I very much appreciate any of you who might provide any support with this, and am more grateful for those who can help me by directing me toward resources for me to learn.... instead of dismissing my desire to repair my piano, by saying that I'm "not up for this project" etc.

I have purchased, what I believe to be, all of the necessary parts, transistors, resistors, thermal compound, etc. and have my own digital and analog scopes, as well as a DMM, and other tools required for the job. I do not have, however, any type of wave generator (please feel free to suggest a reliable/affordable model) or a distortion analyzer (again, feel free to suggest a reliable/affordable model).

I realize this post may be lacking in specifics, and may be too vague/general.... but again, I am learning.... and very much appreciate any help/support with my attempt at repairing my piano.

Thank you in advance!

[vintagekeyboards]

I have read all of the conversations of the threads on this site, in regards to:

http://music-electronics-forum.com/t22175/

http://music-electronics-forum.com/t27388/

these two (above) articles.... (very helpful! thank you!) but find myself still puzzled with more basic questions, that might better prepare me to fully understand the workings of this power amplifier (output section) before I attempt any repairs.

again, thank you for any responses/help with this.

[TimmyP1955]

For a wave generator, you can get an app for your smartphone for a few bucks (if not for free). It might be wise to have an audio isolation transformer to protect the phone from any DC that might be on the input of a faulty device.