RFT CM7151 reconstruction - plate voltage? Frequency response? Any info?

[ln76d]

Slowly i'll be reconstructing my VEB Funkwerk Leipzig CM7151.
Any info, tips and others are highly appreciated

Does anyone knows what's the DC plate voltage value?

 

[ln76d]

In attachment is simplified schematic - without PSU and filament connections.

Anyone knows why frequency response is limited to 10Khz?
It's not transformer by itself, because it's pretty linear in 20Hz-20kHz range.
Connection of transformer one side to the plate potential (not ground) or is there anything else which can cause LPF?
Does anyone see anything suspicious on schematic?

 

[Matt Nolan]

It might be the Miller Capacitance of the second Pentode working with the 1 Meg. 15pF is all it would need to be. Try a larger resistance there.

You know that mic produces pretty much line level output.

I think the plate voltage is around 137V. If I remember correctly, the capsule is run at something insane like 105V. So, you really want to change that polarisation divider to bring it down some. At most 85V, preferably less.

There's a nice Andreas Grosser alternative circuit for this which I believe he has made public. It uses the two tubes. Though, you might also want to keep it original.

I had one of these a while ago. It really made me chuckle how much hum there was. The mains transformer is right there in the base of the mic. Something else you might want to consider is moving that out to an external box so you don't have mains wires going into your microphone. Safety first and all that. The mains transformer is in a mu-metal can with ventilation holes. So they went to the bother of trying to reduce the hum that it induces in nearby wires by putting it in a magnetic shielding can. Then, they ran the audio output cable *through* the shielding can via a couple of the vent holes. Hilarious!

 

[ln76d]

Hey Matt!

Thanks

I have schematic from Andreas Grosser, even newer version than this online.
It's really interesting circuit, but first i want to make something close to the original with some mods.
PSU of course will be in the outside box and filament voltage will be regulated.
After you replied i looked on Andreas schematic, and he also put higher grid resistor on the second stage, so probably you are right, that the main problem is the Miller capacitance.
Replacing 30k/100k divider with 100k/100k polarisation voltage should be less than 70V.  It is a good value for the start, then i will empirically choose the best value for my ears

For now i want to make something like in attached schematic.
Any comments appreciated

 

[ln76d]

Warning pictures!

That how it looks when some idiot wrong connected mains voltage to the signal output pins!

 

[ln76d]

And other idiot bought it "as is" previously not asking about the pictures of the inside...

These are the residues of the burnt output transformer.

Currently cleaning the mess.

 

[TLRT]

Hey Matt!

Thanks

I have schematic from Andreas Grosser, even newer version than this online.
It's really interesting circuit, but first i want to make something close to the original with some mods.
PSU of course will be in the outside box and filament voltage will be regulated.
After you replied i looked on Andreas schematic, and he also put higher grid resistor on the second stage, so probably you are right, that the main problem is the Miller capacitance.
Replacing 30k/100k divider with 100k/100k polarisation voltage should be less than 70V. It is a good value for the start, then i will empirically choose the best value for my ears

For now i want to make something like in attached schematic.
Any comments appreciated

FYA,This schematic is not correct it won't work.First tube needs external bias,(from heater) the second amplifier tube the cathode bias is ok.

 

[Menschmeier]

Hi, I'm revisiting this old topic because I'm currently restoring an RFT CM7151 and am dissatisfied with the original circuit diagram. I know that Andreas Grosser had developed a heavily modified circuit at the time, where the first EF12 is operated with a bias voltage from a negative heater voltage, the second EF12 compensates for the limited frequency response of the original output transformer and where the anode voltage is 210 volts and the capsule bias voltage is only 80 volts. When I spoke to Andreas back then, I didn't yet have a CM7151. Now I can no longer ask him.
Does anyone have this circuit diagram?

Best wishes from Berlin and a happy new year!

 

[Menschmeier]

It's a pity that nobody replied. In fact, Andreas had not built the second stage as a cathode follower, as I had first assumed, but he had built in feedback from the output of the transformer in order to linearise it. I don't yet know the values of the components used. But you can certainly try it out and measure it yourself.

I am restoring my microphone using the original circuit with an external power supply, but I have revised it. Only a few changes are important here:

The two cathode electrolytic capacitors must be replaced with modern ones with 100 µF 25 V. Then you do not need to bridge them with a film capacitor due to the better ESR. The voltage divider for the capsule bias voltage must be dimensioned so that 80 volts are applied to the capsule. With the original 140 volt anode voltage, this results in an 80 kOhm resistor. Furthermore, the primary side of the transformer must be connected between the capacitor and earth and not via the anode resistor. The grid leakage resistance of the second stage must not be increased to 3.9 MOhm. The 1 MOhm is correct. The treble weakness of the microphone is due to the transformer and not to the Miller effect. Increasing the resistance is actually counterproductive.

If you want to improve the CM7151 beyond that, you have to form the bias of the first stage firmly from a negative heating voltage (voltage divider as with the Neumann CMV5/b) and not with a cathode resistor. Furthermore, a feedback loop from the output of the transformer back to the second stage must be installed. This linearises the frequency response while at the same time reducing the gain. As we have very high levels anyway due to the additional second stage, this doesn't interfere. Andreas had also converted the microphones to 210 volts anode voltage. However, this requires a recalculation of some resistors and a different power supply unit.

 

[Menschmeier]

A few words about the Miller effect:

The grid leakage resistance of the tube forms a low-pass filter together with the Miller capacitance. To ensure an upper cut-off frequency of 20 kHz, the grid leakage resistance should be selected so that the resulting cut-off frequency is above this. For an EF12 tube with 6-8 pF as pentode, the recommended maximum value according to the data sheet is 500 kΩ if the EF12 is used as an amplifier and not just as an impedance converter.

So if you want to improve the circuit, you should not increase the grid leakage resistance of the second stage to 3.9 MΩ, but you can reduce it to 500 kΩ. However, the EF12 only has 3.5 pF in the triode circuit, so the 1 MΩ is also fine here.

I now strongly doubt that the circuit published here in the forum at the time was created by Andreas. He would not have made such a mistake and he would never have installed film capacitors in parallel with the old electrolytic capacitors.

 

[jeremynothman]

@Menschmeier, obviously you've done your own research, but I'm also fairly sure that Andreas Stenzel now does the circuit change a la Andreas Grösser; Whether you can get him to share his secrets are another question. He's on the forum as Ericbazaar . It would seem that he's restoring/modifying the circuit on a few of these each month or so.
Otherwise, maybe you can get in contact with somebody who had their mic modded by Grösser?

 

[Menschmeier]

Yes, I know Andreas Stenzel. But he remained silent, even though I had always shared everything I knew with him. I did, however, receive a version of the circuit from someone else. However, it is surprisingly different in the second stage and does not match the photos of microphones that Grosser had converted himself. The feedback in the photos connects the output of the transformer to the second stage. But not the circuit diagram. I first have to find out. There were obviously several versions that Grosser used.

 

[jeremynothman]

Well, as implied, maybe you can get in touch with someone who had their mic modded by Grösser. Maybe you’d have to pay them for an hour of studio time in exchange for a schematic or something? I don’t know if they would have schematics, but I do remember Francesco Donadello telling me that he likes his modded RFT + M7 as much as his U47, just that some voices liked one more than the other.

 

[Recording Engineer]

Here’s mine and Chris Prutcher’s (Barbaric Amplification) take on the 7151. He finished just yesterday and I’ll pick it up tomorrow.

 

[rock soderstrom]

Here’s mine and Chris Prutcher’s (Barbaric Amplification) take on the 7151. He finished just yesterday and I’ll pick it up tomorrow.

Delicious, looks great . I like the very stylish 5693 tubes you use instead of the EF12K. Too bad they are so rare on our side of the pond.

 

[Recording Engineer]

Delicious, looks great . I like the very stylish 5693 tubes you use instead of the EF12K. Too bad they are so rare on our side of the pond.

Thanks! It is an 8-year idea, finally come together. From the beginning, that was from a note filed in the back of my mind, with Ben Sneesby saying years ago, with his findings in his mics and opinion, that the 5693 is a EF12K, only quieter.

 

[rock soderstrom]

Thanks! It is an 8-year idea, finally come together.

From the beginning, that was from a note filed in the back of my mind, with Ben Sneesby saying years ago, with his findings in his mics and opinion, that the 5693 is a EF12K, only quieter.

Are the 5693 really quieter than the EF12K? That may well be the case, I would really be interested in reliable measurements in direct comparison, which would require a clever socket adaptation.

Cheers

 

[Recording Engineer]

Are the 5693 really quieter than the EF12K? That may well be the case, I would really be interested in reliable measurements in direct comparison, which would require a clever socket adaptation.

Not a clue… Would also require multiple samples of both tubes.

I can tell you my original 7151 I had was quiet, hiss-wise, but anything that might have been there, would be greatly overpowered any attention by the hum generated by the power supply in it! Using it as the M mic in an M/S setup, your S mic would greatly pick it up! Thanks, Obama. Ha!

 

[rock soderstrom]

Would also require multiple samples of both tubes.

That is correct. In order to generate truly reliable data, you need a corresponding database.

I once tested (not particularly scientifically) some modern EF86 against a few EF12K in the exact same circuit The best EF86s were superior to my best EF12Ks in terms of noise level. But that also has something to do with the condition and age of the EF12s that were available to me.

YMMV, take this with a grain of salt.

Edit: one of my test adapter, you have probably seen it before...

Edit2: EF12K were more microphonic than EF86 in my tests. How microphonic are the 5693 tubes? They are optimized for this, just like the EF12K. The K at the end stands for 'klingarm', i.e. less microphonic.

 

[Menschmeier]

Such sweeping statements are nonsensical. When I compare my many EF12, EF12k, EF40 and EF86 (which are basically all the same tube), I find huge differences. And I'm only interested in their noise when installed in a microphone in a triode circuit. I don't use them as pentodes. And the noise depends on many other factors that can vary from microphone to microphone. Circuit details, bias, cathode follower or bias voltage generation from negative heater voltage. There are so many variants. You actually have to adjust the bias for each tube to make the noise comparable.

My experience (triode circuit with fixed bias in the Neumann CMV/5b - my reference):

Only about every 10th tube is low-noise enough for my requirements. Most of them are unsuitable for this purpose. And it is hardly ever a Telefunken tube that is among the lowest-noise ones. And it doesn't matter whether the EF12 has a k on it or not. Only a few tubes are microphonic.

In short: You have to test the tubes in a microphone and select them individually. "A is better than B" is a nonsensical statement, as A1 is already completely different from A2.

Neumann always selected tubes individually and stamped them with an M. I know this at least from the VF14 and EC92. However, I have not seen an M stamp on the EF12. The only tubes where the manufacturer had already selected for suitability in the microphone were the pure microphone tubes from Hiller and the Telefunken MC1, AC701(k) and RFT AC761.