Fender Rhodes germanium to silicon transistor swap

[MarkusBass]

In the case of a power supply transistor, distortion is not an issue. We might not have to change anything.

This is exactly what I suspect. Most probably you don't need to change anything except the transistor. But to confirm it, we need to see the schematic. Michael said that the board is very similar to the one posted by me but there are slightly different components. Two good photos could help us to solve the problem.

PS: in other thread, JPB posted schematic of the circuit: http://www.fenderrhodes.com/org/ch11/fig11-10.jpg (available on fenderrhodes site). Isn't it the schematic that we are searching for?

Mark

[MichaelNuzum]

Hey guys, thanks for the replies. Maybe I should clarify a few things so we're all on the same page:

My qualifications to be working on this amp - none. Seriously, I am a musician who got handy with a soldering iron to repair his own cables. That sort of morphed into working on some effects and amps and whatnot. I have done literally hundreds of hours of self study in electronics, but only in the very limited and specific scope as to how they apply to the gear I normally work on, mostly Fender type tube amps and effect type units. I have developed a strong skill for having a unit and maybe the schematic and diagnosing a failed component, safely (I understand how to work around high voltages without killing myself and how to discharge capacitors, etc) and cleanly. Beyond that, as to the theory aspect of what it does or why it failed or how to modify it, I admit I am totally in the weeds. But I am always a sucker for a friend with a cool vintage unit, which brings me to the amp in question.

A friend got the Rhodes from an online site, sight unseen. Plugging it in, it had a pretty weak signal, and there were occasional low frequency oscillations and volume swells and pops, just a lot of instability that indicated the amp needed attention. Doing a bunch of research, I found out about the germanium output transistors being notoriously unreliable, and there was an easy swap, so I figured I get those changed and then evaluate the rest of the circuit to the best of my ability.

Now another point to clarify - I have not yet swapped out any components on the amp. I wanted to fire it up to get some before/after voltages to observe the changes, if any, on the circuit (trying to learn by observation). When I fired it up, it fired up, literally. One resistor emitted sparks. I had it on a variac, but I didn't have an ammeter or current limiter, my bad on that.

So, here I am looking for more info or documentation, in case anyone else might have run across this situation. The 80 watt Peterson power regulator schematic that keeps getting mentioned here is awfully close, but not quite the same thing. And, due to my previously mentioned limited understanding of how these things work, I am not able to see the similarities and differences enough to discern what I need to focus on. But even in my limited understanding, I can identify enough differences that I am not sure how to interpolate one set of info to the other.

I must say, I am very grateful for all the help so far, but I am not asking for anyone to do the work for me. I will try to draw out the circuit myself and post it, and will gratefully accept constructive criticism of what I will probably get incorrect, but I'd like to give it a stab if that's the next logical step.

I am sure that once you see the circuit you will find it laughable that this is stumping me, but again, I admit my skills are limited in breadth and depth.

I apologize again if I use the wrong terminology or don't know the proper ways to present information. I am trying to do the learning, but I am still on the front end of the curve, long way to go, so please bear with me.

I need to jet to my day gig, I will attempt the drawing later tonight or tomorrow. I will also post front and back shots of the board in question. Thanks again.

[Enzo]

It isn't a matter of being laughable, it is just we need to see it to do anything about it. No one criticizes a lack of experience.

From an experienced point of view, let me say something about "familiarity". When I open something, yes, it helps if I have seen such units befor, but what I rely on is not prior knowledge of the specific system, but really it is more about understanding what I see. I realize you lack that part of the experience, but we can feel really comfortable working on things we have never seen before. If you will allow the analogy, it is a bit like having studied LAtin makes it easier for me to study Russian.

[MarkusBass]

If you are afraid of making mistakes while drawing the schematic, you can start with posting photos of the board. As you can see there are some guys eager to help you (including me). If your photos would be of good quality, we could help you using the photos only.
One issue requires clarification: previously I thought that you replaced transistors is the power amp and you work on the power supply. But now it seems to me that the power amp failed but it's not fixed yet, right? In such a case you need to fix it first (by replacing germanium transistors with silicone transistors).

[Mick Bailey]

+1. A photo of the board can explain a lot.

My initial thought is that the transistor is a series regulator for the power supply. This is the most usual configuration and is perhaps easier to replace than an amplifier transistor.

There are a number of considerations when replacing Ge with Si. Ge's thermal characteristics are comparatively poor and there needs to be correction against thermal drift, otherwise the transistor can suffer from 'thermal runaway'. Basically, the hotter it gets, the hotter it gets. Until it (or something else) gets destroyed. So there are often additional components that are not strictly required for Si to correct for this tendency. In a power supply these can often be left as they are.
The circuit designs are different between the two types for the following reasons;

Si has a forward voltage drop of about 0.65v, whereas Ge has a drop of 0.2v. This is important when considering amplifiers, as the transistor needs to be biased to overcome this drop - it needs to be turned on sufficiently to make sure it conducts for the entire cycle of the waveform applied to the base. In a power supply this is usually not a consideration, as the regulation function often overcomes this requirement and in normal operating conditions the regulating transistor would be turned on sufficiently regardless. That is to say, it would not be biased to 'just on' as would an amplifier.

2. Ge has a higher leakage current than Si. This means that amplifier stages connected together (such as the stages in a preamp) can be adversely biased due to the leakage of DC from a previous stage. Sometimes there are coupling capacitors present to isolate the stages from such effects. These are not usually needed in Si amps, which can be directly coupled due to low leakage. Again, these capacitors can be left in place without adverse effects. For a given device number, Ge can have hugely variable leakage characteristics. So the biasing networks surrounding Ge devices are more complex than Si to accommodate this variation. This also provides greater stability from changes in supply voltage.

In an amplifier stage an Si device usually replaces Ge with just a minor component change to ensure the biasing is correct - a single resistor change (sometimes two). In power supplies I've just replaced the device and adjusted the output to provide the correct voltage. I can't recall making any other changes. Maybe if the regulator voltage is sensed off a fixed-resistor voltage divider then this would need correcting.

A start would be to cross-reference the device to the nearest silicon equivalent. Maybe input the values into the Alltransistor database and see what comes up.

When doing power supply work I always test the supply with the load disconnected (if at all possible) and bring it up on the variac, monitoring the output voltage and current draw. Then clip a load resistor across the supply outputs to ensure the voltage is stable.

[MichaelNuzum]

Thanks for the help everyone. Your encouragement is gratefully appreciated.

Mick- thank you for that explanation. It really took a few things I sort-of understood and helped clarify. Thanks very much.

So after attempting to draw out the circuit (and not doing a very good job), I realized that the circuit is indeed very close to the 80 watt version referenced above (I humbly eat crow for denying that they were alike). They do contain an awful lot of the same components, but there are a few differences.

http://www.fenderrhodes.com/org/ch11/fig11-10.jpg

Firstly- the one I have has two small transistors on the board and the third large one off the board. The schematic only shows two transistors.
Secondly- the schematic has a zener diode, I don't see one on the board I have or anywhere near it.

Here is the top of the board I have (there is a missing 820 ohm resistor that exploded and has been removed right under the 5 ohm resistor):



And the bottom:



The failed transistor is mounted outside the rear wall and fed from the E,C, and B terminals. The orange wire feeds (or is fed from?) one of the filter capacitors.


I have purchased MJ15016G transistors as replacements for the outputs. It seems the part number on the failed power transformer (Delco DTG-110B) is also a number that I have seen used for the output transformers. I am wondering if one of the MJ replacements might also work for the power transistor position?

Also, the failed transistor had two legs (E and B) with the case being the collector. It was mounted to the outside of the chassis, and one screw had a terminal lug for connection back to the circuit. It was mounted with a pretty old and squished isolation washer. If that thing dried out and the screw and the case developed continuity to chassis, would that create a problem or no? The actual case is isolated with a mica spacer, but the mounting screw has me concerned.

Again, thanks for all the assistance.

[MarkusBass]

Firstly- the one I have has two small transistors on the board and the third large one off the board. The schematic only shows two transistors.
Secondly- the schematic has a zener diode, I don't see one on the board I have or anywhere near it.

You can clearly see that the "second small transistor" is in parallel to 100uF capacitor and it has only two active leads - this is the Zener diode from the schematic. And there is just one small transistor on the board. You should have post these photos at the very beginning of this thread. This would save you a lot of time .

Also, I suggest that you continue just one of your two threads because sometimes answers to questions from the first thread are in the second thread.

Mark

[MarkusBass]

Looking at the photos you posted and at the fig11-10.jpg schematic, one can see that this is the same schematic with few minor changes:
470R resistor (to the trimmer) is replaced with 1k
680R resitor (from emitter of the small transistor to the ground) is replaced with 820R resistor,
620R resistor (from emitter of the external transistor to its collector) is missing (two holes over 5 Ohm resistor). Most probably it was not needed.
The "second small transistor" is in fact a Zener diode shown on the schematic.
So we may say that this is exactly the same schematic with minor adjustments. I hope this will help you.

Mark

[Mick Bailey]

If the collector shorts to chassis, then something will fail. You have the output of the rectifier being shorted to ground via the 5R 10W resistor. When the transistor is replaced, replace the pad at the same time and apply thermal compound, plus make sure the connection to the collector is suitably isolated.

For replacement, almost any Si transistor of the same or higher voltage and current rating will be superior. Gain in a PSU is usually not a consideration so long as the replacement is roughly similar - there can be considerable gain variation in transistors but power types are usually fairly low gain and the circuits tolerant of variation.

[Mick Bailey]

BTW, did you establish the reason the PSU failed? My thought is the +25v shorted out or was heavily loaded, burning up the 820R resistor and taking the transistor with it. Looking at the circuit it seems reasonable that your suggested replacement would work without any further changes. If I was doing this I would check the supply for any further damaged components, then install the replacement Si transistor with a new pad etc. Disconnect the plug (to isolate the 25v rail) and bring the unit up on the variac to monitor the voltage. If the voltage was creeping up over 25v, I would trim it to correct the increase as I went along.

At the same time, check any of the components for overheating.

[Jazz P Bass]

I believe it was mentioned that the offboard PNP transistor was a DTG-110B.
Here is the datasheet: DTG-110B.pdf

[MarkusBass]

BTW, did you establish the reason the PSU failed? My thought is the +25v shorted out or was heavily loaded, burning up the 820R resistor and taking the transistor with it.

I think that this was exactly the reason of the failure (with one difference - the resistor was 620 Ohms). Most probably someone has shorted the collector of the power transistor to the ground. So it was just a human mistake and not a mystery failure of the regulator.
You can either buy original transistor (and make sure that it is isolated from the heatsink this time), or rebuild the regulator with silicon components. To do it, you need a 12V/0.5W Zener diode, low power NPN transistor and high power PNP transistor. It is important that the missing resistor (620 Ohms) it put back into the circuit. Without it the regulator will not work.

Note: the circuit is quite unusual. Please note that the input voltage is provided to the emitter of the power transistor, the Zener diode is connected to the output voltage and there is additional resistor that bypasses the power transistor. This is not a typical circuit used nowadays. But I checked that it will work correctly with silicon components.

Mark

[Mick Bailey]

The schematic is slightly ambiguous - could be a 6 or an 8, but the OP mentioned the resistor was 820R so I just went with that.

[Enzo]

FWIW, 820 is a standard value resistor, 620 is not.

[MarkusBass]

Yes, it's hard to read it on the schematic. Maybe it was 820 Ohms. But this is not that much important. The regulator will work with any resistor with value between 560 Ohms and 1k. It is important that it is added back to the circuit.