Help recapping Telefunken Opus 2430

[marsupilami]

Hello,

I have a Telefunken Opus 2430 EL95 push-pull amp from the 60s, it works great with no hum but as a preventative measure I'd like to replace the input caps. These are the only electrolytics that I could spot besides the two other 100 uf caps by the output (coupling caps?), and those have been replaced already by 150 uf caps.
I'm fairly handy with a soldering iron but don't have much knowledge in electronics so before I get started I'd appreciate it if someone could tell me that I'm doing things right (or otherwise heading for disaster).

First here's a photo of the capacitors housing and connectors :

Next here's what's written on the side of the housing (part of it is hidden from the rest of the chassis) :

Elyt 4x100 uf
350/385 V - ...
Minuspol - Gehäu...
-20 + 70 C
638
[] Auss. C. ([] being a square symbol)

And finally here's the link to the schematics :

http://soundup.ru/images/stories/library/schem/telefunken/telefunken-18/opus-2430.jpg

Now if I understand correctly inside the housing are 4 identical 100 uf electrolytic caps rated for 350/385V, the + end of each being one of the terminals with a little symbol and the - end being connected directly to the housing, itself connected to the chassis. From there my plan is to buy 4 100 uf / 400V panasonic electrolytic caps (I know, not ideal, but any other type of cap is huge and very expensive for those values) and simply disconnect the wires going to each terminal of the housing and connect it to the + end of the corresponding cap the same way, with all the caps' - ends connected together to where the casing is connected to the chassis. Is that the right way to do it or did I miss something?

Thank you in advance for your help, I really appreciate it!

 

[RWood]

You are on the right track. The Panasonics will do well if you have the space.
If you need something smaller or want to see about restuffing the can, take a look at Nichicon Pencil Caps . I do notice that the final two 100uF caps (reading from right to left on your schematic) seem to be in parallel so a single 220uF could work there if that helps with your layout.

There are two 100uF caps on the cathodes of each pair of power tubes. Replace those, but the voltage requirement is much lower and 25V or 35V caps would work well (and save a lot of space)

The two resistors in your photo have seen better days; I'd be willing to bet they are the two dropping resistors in the power supply, 220Ω and 390Ω. You might need to do some wire-tracing to verify which two they are. Replace with metal oxide resistors, 2W minimum.

I would also get a supply of .022 630V film caps for the grids of the power tubes - replacing them is next in line of "must-do's". Those will keep DC voltage off the EL95 grids and enable them to be biased correctly.

The crossover caps that are connected to the audio transformer secondaries can be very confusing. They come into play when various tone-compensating keys are selected. I would wait to replace those, as they won't have had high voltage run through them and they could be just fine.

You might also consider replacing the selenium rectifier with a modern silicon full-wave bridge. They are small, easy to mount and may be bolted right to the chassis or a piece of phenolic near wherever you mount the filter caps. The negative from the rectifier should be grounded along with the negative of the first filter cap (if not all of the filters)

let us know what questions you have, as you go.

RWood

 

[marsupilami]

You are on the right track. The Panasonics will do well if you have the space.
If you need something smaller or want to see about restuffing the can, take a look at Nichicon Pencil Caps . I do notice that the final two 100uF caps (reading from right to left on your schematic) seem to be in parallel so a single 220uF could work there if that helps with your layout.

There are two 100uF caps on the cathodes of each pair of power tubes. Replace those, but the voltage requirement is much lower and 25V or 35V caps would work well (and save a lot of space)

The two resistors in your photo have seen better days; I'd be willing to bet they are the two dropping resistors in the power supply, 220Ω and 390Ω. You might need to do some wire-tracing to verify which two they are. Replace with metal oxide resistors, 2W minimum.

I would also get a supply of .022 630V film caps for the grids of the power tubes - replacing them is next in line of "must-do's". Those will keep DC voltage off the EL95 grids and enable them to be biased correctly.

The crossover caps that are connected to the audio transformer secondaries can be very confusing. They come into play when various tone-compensating keys are selected. I would wait to replace those, as they won't have had high voltage run through them and they could be just fine.

You might also consider replacing the selenium rectifier with a modern silicon full-wave bridge. They are small, easy to mount and may be bolted right to the chassis or a piece of phenolic near wherever you mount the filter caps. The negative from the rectifier should be grounded along with the negative of the first filter cap (if not all of the filters)

let us know what questions you have, as you go.

RWood

Thank you very much for your detailed answer!

These Nichicon caps look like they would be easier to integrate, I'll look if I can find them here in France. The 2 100 uf caps you mention have been changed already by a previous owner, by 150 uf electrolytic, is that alright or would it be better to fit 100 uf caps again instead?

Regarding the resistors, alright I'll go ahead and change them, it's true they look a bit shabby. About other caps I'm not sure though, first of all some of them are difficult to access due to the way the amp is built, and second of all I'm not sure I want to mess with things too much, to be honest I was hesitant to touch anything in the first place (as the saying goes, if it ain't broke...). If you think it's really necessary to future proof the amp then maybe I'll give it a go, otherwise I might leave it as it is for now.

I have read on several occasions about the selenium rectifier being a high failure risk, I guess I'll go ahead and change this as well, seems straightforward enough. I'm going to do some research to see what part(s) I need exactly but if you have suggestions to point me towards the right direction please feel free!

Thanks again!

 

[RWood]

150uF for the final filtering section is probably more than enough - no problem there.

Yes, lack of easy access to the coupling caps in German radios is part of the deal. If you want to do it in stages, do the filter caps first and when you get your radio running, measure for DC voltage on the grids of the EL95s (pin 7). If there is less than a half a volt or so on the grids, you'll be okay leaving those caps in there.

Here is a pic of a diode bridge I installed in the last Telefunken I worked on. Before and after shots. I think the part was something like this one. Very common part, and cheap. Some people will put a dropping resistor on them to match the original output voltage, but these radios run at such low voltages already (relatively speaking), I don't bother.

Btw, nice Tele EL95s are not that tough to find. I have a set of 4 waiting for a custom amp build.

Bon chance!

 

[marsupilami]

150uF for the final filtering section is probably more than enough - no problem there.

Yes, lack of easy access to the coupling caps in German radios is part of the deal. If you want to do it in stages, do the filter caps first and when you get your radio running, measure for DC voltage on the grids of the EL95s (pin 7). If there is less than a half a volt or so on the grids, you'll be okay leaving those caps in there.

Here is a pic of a diode bridge I installed in the last Telefunken I worked on. Before and after shots. I think the part was something like this one. Very common part, and cheap. Some people will put a dropping resistor on them to match the original output voltage, but these radios run at such low voltages already (relatively speaking), I don't bother.

Btw, nice Tele EL95s are not that tough to find. I have a set of 4 waiting for a custom amp build.

Bon chance!

Hello,

Thanks again for the information. So I went ahead and replaced the input caps and resistors (they were both way out of spec), as well as the selenium bridge with the part you suggested, and the coupling caps too (they weren't that hard to get to after all, and I thought I might as well go all the way if I did work on it).

The good news is that I fired it up and it all works, meaning (I guess) that I didn't mess up anywhere.

The bad news is that I have one tube that develops a red glowing plate after a few minutes now. I swapped the tubes and still get the red glow on the same location (that's the tube at the bottom in the diagram) so that's apparently not a faulty tube. Now this tube was already glowing a little bit brighter before so I don't think it's a result of my messing up somewhere, rather I think it might have to do with the slight bump in voltage going from the selenium bridge to the silicon one. Am I right for thinking that?

I did some research and with this amp being cathode biased a likely culprit is the cathode resistor, so I checked its value and got a reading of 236 Ohms instead of 470 (so basically exactly half). Now the weird thing is that I checked the other cathodes' resistors and also got a reading of 236 Ohms for all of them. The color coding indicates that it should be indeed 470 Ohms as it is in the schematic. It seems weird to me that all the resistors would have drifted equally to exactly half their rated value, and so the most likely explanation is that I must be doing something wrong. I checked my multimeter and it is spot-on on other resistors so it's not it. And so I need your (or somebody else's) help again...

Thank you once again in advance!

 

[ggizzy]

Replace those with 470 wirewounds. I've used 5W in my 25W valve amp, though I've heard suggested they can be replaced with a higher wattage than that. Someone else may be able to clarify that position.

The good thing is these can be had on the cheap. If they are the two burned out ones in that first picture, I would leave extra lead on the new ones, heatshrink those leads, and mount them so they have some space between the tagboard and eachother. This way they might more effectively dissipate any heat that might buildup when you enclose them in the chassis.

 

[marsupilami]

Replace those with 470 wirewounds. I've used 5W in my 25W valve amp, though I've heard suggested they can be replaced with a higher wattage than that. Someone else may be able to clarify that position.

The good thing is these can be had on the cheap. If they are the two burned out ones in that first picture, I would leave extra lead on the new ones, heatshrink those leads, and mount them so they have some space between the tagboard and eachother. This way they might more effectively dissipate any heat that might buildup when you enclose them in the chassis.

Unfortunately they're not the ones in the picture (those I've replaced already), instead they're on the PCB and replacing them would mean A LOT of disassembly, so I'd rather make sure it's absolutely necessary first. Isn't it strange that all of them are at exactly at half their rating? Seems like a strange coincidence to me... Could it be that the way they're located in the circuit would give me such a reading? Also if they were all the same wouldn't all the tubes get a red plate the same way? Could it be that with the new silicon bridge the voltage is too high and I need a drop down resistor after all? I haven't checked any voltage yet, I'd rather limit the time the amp is on to a minumum right now...

I would appreciate it if someone could give me a 2nd opinion... thanks!

 

[RWood]

It is not a coincidence that you are seeing half the value of each - you are actually measuring "through" both of them and they are in parallel thanks to the bypass cap. The 100uF bypass' cap is common to both cathodes on each channel. This cap will also affect the resistance reading, so you'll need to lift it from the circuit or lift the resistors at one end, to accurately measure the cathode resistors. But they look to be right on.

Since the bypass caps should be replaced anyway, an easier way to do this is to clip out the caps but leave the leads. Then use those leads to make it more convenient to tie in your new caps. Positive end of the cap goes towards the tube socket, negative to ground, and this way you will not have to solder to the chassis.

This does not explain, though, why only one of the tubes was redplating. For that, I would measure for voltage on that tube's grid, pins 1 and/or 7. You said you replaced the coupling cap to that grid but since something is amiss it needs to be re-checked. DC voltage there should be 0. And resistance (with unit turned off and caps drained) should be a little over 1M + 47K.

While you are in there taking live readings (with one hand in your pocket) look at the cathode voltage, which should be around 10 vdc with respect to ground. If it is more than that it is also evidence that the grid is probably too positive (when it should be close to 0). Check filament voltages on 3 & 4 (6.3 vac between them) just to be safe, and of course plate and screen dc voltages up in the 260 - 280 range or so.

I doubt the silicon bridge would be causing this issue. Yes, the new bridge should increase the B+ voltages from the spec'd 270 vdc but it is still on the medium/low side, as voltages go. I wouldn't worry about it but would focus on what's going on with that EL95 socket.

Nice work on the re-cap, btw and congrats on the initial fire-up. Feels pretty good when it works, doesn't it? Also you did well to notice the replating on that tube!

RWood

 

[marsupilami]

It is not a coincidence that you are seeing half the value of each - you are actually measuring "through" both of them and they are in parallel thanks to the bypass cap. The 100uF bypass' cap is common to both cathodes on each channel. This cap will also affect the resistance reading, so you'll need to lift it from the circuit or lift the resistors at one end, to accurately measure the cathode resistors. But they look to be right on.

Since the bypass caps should be replaced anyway, an easier way to do this is to clip out the caps but leave the leads. Then use those leads to make it more convenient to tie in your new caps. Positive end of the cap goes towards the tube socket, negative to ground, and this way you will not have to solder to the chassis.

This does not explain, though, why only one of the tubes was redplating. For that, I would measure for voltage on that tube's grid, pins 1 and/or 7. You said you replaced the coupling cap to that grid but since something is amiss it needs to be re-checked. DC voltage there should be 0. And resistance (with unit turned off and caps drained) should be a little over 1M + 47K.

While you are in there taking live readings (with one hand in your pocket) look at the cathode voltage, which should be around 10 vdc with respect to ground. If it is more than that it is also evidence that the grid is probably too positive (when it should be close to 0). Check filament voltages on 3 & 4 (6.3 vac between them) just to be safe, and of course plate and screen dc voltages up in the 260 - 280 range or so.

I doubt the silicon bridge would be causing this issue. Yes, the new bridge should increase the B+ voltages from the spec'd 270 vdc but it is still on the medium/low side, as voltages go. I wouldn't worry about it but would focus on what's going on with that EL95 socket.

Nice work on the re-cap, btw and congrats on the initial fire-up. Feels pretty good when it works, doesn't it? Also you did well to notice the replating on that tube!

RWood

Thanks for the explanation regarding the resistors, makes a lot of sense now.

It does feel good when you plug it in and everything works and no big cloud of smoke coming from the cabinet .

I checked the bias on the tubes and all of them are well below .5V so I guess all is good on the coupling caps side of things. Plate voltage is around 340V. Screen voltage isn't much lower at around 335V or so. Now regarding cathode voltage it's around 13V but on the problematic tube it starts to rise after a couple mins or so, along with the plate getting red. Other tubes' cathodes stay right there at 13V. Interesting thing too is that when I checked grid voltage on the tube the glowing disappeared right away, only to reappear once I took the probe off. I checked the resistors you mentioned and on the problem tube I can't get a reading on the 1 MOhms resistor whereas the other ones all register a little over 1 MOhms like you said. Could that be the culprit?

 

[RWood]

It certainly could be; I would treat it as such and focus there first. See if the 1M has a good ground connection and is not shorting out against something else. If in doubt about the value, pull one leg and measure it or, if you've gone that far, go ahead and replace it. This resistor is the ground reference for the power tube grid.

The other area of concern is the high plate and screen voltage. Were all the tubes in the radio conducting when that measurement was taken? That is, was B+ under its' full load? If so that number is high and I see your concern about what it is doing for the rest of the circuit. When I have added silicon bridge rectifiers in place of selenium, the increase in voltage has not been as great as what you are seeing.

I think you should still focus on the one bad EL95 and after you've sorted that out readdress the power supply. Look at things like line voltage, transformer voltage setting, filament voltage, DC voltage immediately after the rectifier, after the first dropping resistor. If everything seems to be as it should but the B+ is too high then, as you asked previously, you should look at a method for reining it in. There are a few ways to do this but you've got some work to do first.

 

[marsupilami]

It certainly could be; I would treat it as such and focus there first. See if the 1M has a good ground connection and is not shorting out against something else. If in doubt about the value, pull one leg and measure it or, if you've gone that far, go ahead and replace it. This resistor is the ground reference for the power tube grid.

The other area of concern is the high plate and screen voltage. Were all the tubes in the radio conducting when that measurement was taken? That is, was B+ under its' full load? If so that number is high and I see your concern about what it is doing for the rest of the circuit. When I have added silicon bridge rectifiers in place of selenium, the increase in voltage has not been as great as what you are seeing.

I think you should still focus on the one bad EL95 and after you've sorted that out readdress the power supply. Look at things like line voltage, transformer voltage setting, filament voltage, DC voltage immediately after the rectifier, after the first dropping resistor. If everything seems to be as it should but the B+ is too high then, as you asked previously, you should look at a method for reining it in. There are a few ways to do this but you've got some work to do first.

Hello,

So I've changed the resistor and it seems to have solved the problem, I let the amp on for over an hour and no red plate to be seen. Yay! That also explains why the plate stopped glowing the moment I took a reading on the grid, the meter was then acting as the resistor and providing the needed reference to ground.

Now regarding the high voltages, I was just being stupid/ignorant: I took the readings with just the amp part plugged in, but there are a lot of tubes in the preamp/radio part too! The voltages are down to a more reasonable level now that everything is connected, still a bit higher than spec but I'm tempted to leave it like that before doing any serious damage to something... Speaking of, actually one of the ECC83's that was not original and already a bit finicky ever since I got the amp stopped working for good this time, the high voltage probably did it for it. No big deal as replacing these by Telefunkens was next on my list anyway. So I'm now on the hunt for a decently priced pair of matching ECC83's, and then hopefully all will be good for years to come!

Thanks once again, you've been a great help!

 

[RWood]

Nice work - glad to hear you got it going; with those new caps it should work well for years.

Nothing wrong with looking for Tele ECC83s but as you might know they are probably the most sought-after tube on the planet. Luck does favor the well-prepared, though, as they do turn up in otherwise forgotten radios and other equipment. My buddy found a cache of 22 of them in a cheap Estey reed organ!

In the mean time, don't overlook some of the fine ECC83s from Amperex, Mullard, Mazda, et al. Plenty of great ones, too, from this side of la mer but of course they would have a different accent.

Voici un soundcheck je ai fait sur un des Telefunkens je fixe , atterrissant sur la station locale indépendante.

 

[marsupilami]

Looks like you're having fun there! Do you fix these for a living or is it just a hobby of yours?

Regarding the Telefunkens, I knew they made good tubes but the prices on these ECC83's are nuts! I gave up trying to find a set at a decent price and found a good matched pair of Philips Miniwatt's made in Holland, same as Amperex Bugle Boy. Too bad since all the other tubes are Tele's and it would have been nice to keep it consistent across the board but there's no way the price premium can be justified imho. I'm sure the Philips will do the job just fine

 

[gadget73]

I've experiened the bad grid resistor on German radios before. The Nordmende I gave my grandmother did this after a while. It just quit one day, but taking voltage readings would bring it back. Stuck in a new resistor, and it was all better.