RCA 41b Preamp

[dmp]

I started trying to bring this preamp back to life this weekend.
I thought I'd post some pictures as it goes along.
First off, the schematic.

 

[dmp]

A fixed gain amp (42 dB) designed especially for the RCA 44 ribbon microphone.
Huge transformers on the input and output.

 

[dmp]

The input transformer:

 

[dmp]

And the output transformer: a combined inductor and transformer, for the inductive coupled output stage.

 

[dmp]

Replacing the old caps in the can. It was all filled with wax. It was brittle enough that it was easy to dig out. The tubular electrolytics for the cathode bypass get replaced
The square black ones are probably not electrolytic and I didn't replace them.

 

[emrr]

Cool, I hadn't seen one of those cap cans open yet. 

You may want to strap out the c6/r6 treble boost.  It was designed to compensate the 44-A treble droop, and NBC issued a mod note in the mid 1940's to reduce treble for the 44-B and other later mics. 

 

[dmp]

You may want to strap out the c6/r6 treble boost.  It was designed to compensate the 44-A treble droop, and NBC issued a mod note in the mid 1940's to reduce treble for the 44-B and other later mics. 

Can you post the mod note?
So the plan is to get this working and bring it over to a studio in town that has an original RCA44 and try them out together.
But I think it is a later 44, not 44-A. It is a pretty dark mic as is. Maybe it is darker than it should be and needs some rehab.
Current status is that when powered up the preamp oscillates severely. With both grids grounded with clips, the voltages look reasonable and the tube currents are pretty close to spec (1: 0.52 mA, 2: 4.2 mA).
With just the first tube grid grounded, oscillation still present but less severe. I'm thinking the coupling is happening through the tube grounds, positive side of R7. Maybe C4 or  C1 are not doing their job or there is a bad connection somewhere.
Wiring is in really poor shape - very brittle. Not sure it will be reliable without re-wiring.

 

[emrr]

The NBC mod note doesn't make much sense today.  It's meant to keep gain the same by adding a cap shunt to ground after the cap boost.  Better to just strap out the high boost. 

Actually, your wiring looks great compared to most.  I can't tell that you have any plastic coated wire on this one.  That kind are all falling apart. 

 

[G-Sun]

So cool to see a restoration like this :thumbsup:

 

[emrr]

any reports?  I still haven't heard mine. 

 

[dmp]

I haven't gotten back to this since I did the powered it up and found it was oscillating.
Will have more time coming up and will get back to it.
Finding tubes is a real challenge, especially the 1603 that is primary spec'd.

 

[emrr]

1603 is a special 77. Just use 77's. That's what's in the earlier version.  1603's are stupid expensive for little difference.  You also generally find 6C6 in 77 and 1603 sockets, the next closest thing available along the way. 

Check out the partially shared cathode resistance.  That's not a move I can think of seeing many places.  I suppose that would be a bit of feedback in a sneaky place?  Not obvious the way it's drawn with the current jacks.  I believe it's also a move to reduce noise, details of which I forget ATM. 

 

[dmp]

Finally got back to this project. I got some 77 tubes to try out. First measured all the Resistors and Caps to check they were in spec and they were.
With the retube, seems to be working but had some buzz (can't remember what exactly was going on before that made me think it was oscillating).
This was a power supply problem, adding 100 ohm resistors to the heater taps to ground got rid of it.
A little hum now, but not bad. Maybe a ground connection problem or transformers picking it up from the surroundings. Have to play with it a little more, but it's close.

 

[emrr]

Cool.  I just fired up a 41-C, which is a pretty similar circuit.  The treble boost is big!  But sounds pretty great, not at all like what you'd imagine seeing a plot.  I need to try bypassing it, and also with new EQ caps to be sure the old haven't drifted. 

In those days, even with all that heavy shielding, they always specified power transformers 3 feet away from input transformers, minimum.  It could easily be from something else nearby.  I didn't have any hum issues in the 41-C. 

 

[dmp]

Could have been - as I was testing it, I shut off the TV screen nearby and the buzz dropped significantly. I'll try it in a better situation and see if is good.
I don't see anything abouta 41c - is it also a heavy 3U beast?

 

[emrr]

41-C is mostly unknown, cost a fortune compared to rest of era RCA line, and not many out there.  3 discrete channels in one, 8RU, 61lb monster. 

There's also the 41-A, probably the first amplifier specifically sold as a 'microphone preamplifier', introduced along with the first RCA ribbons.  Before that we were all using condensers and carbon mics that put out much higher levels. 

 

[dmp]

So I've still been trying to get this quiet enough for recording. I tried it with a Heathkit IP-17 power supply, and it was dead quiet.  So it seems it has a problem with the psu I made for it.
I adapted a General Radio power supply - it's dropping a lot of voltage to bring the B+ down to 180v - so should be very well filtered. I drew out the schematic. I'm not sure what could be wrong with this. Any ideas?

 

[emrr]

The GR supplies generally don't have any filament ground reference, but I see you drew it and mention it in the older post.  Be sure all the grounds are indeed there and connected to one another at some case point, including the preamp.  I've never had noise problems with GR supplies.  You are dropping a LOT of voltage to get this down to 180 at minimal current.  I assume this is the 50mA unregulated supply, as opposed to the 70mA regulated supply. 

 

[dmp]

Yes, I added the 100 ohm resistors to ground for the heater supply. It is the 1203 psu - stock it is 300 vDC @ 50 mA, so yes, dropping a lot of voltage.  Bad idea?
Interestingly, I just looked at the schematic for the IP-17 high voltage supply and the heater voltage does not have a filament ground reference as far as I can tell.
I'll check and rearrange the grounding in the GR supply if needed.

 

[emrr]

Yes, I was thinking there's no ground reference drawn on the IP-17, yet I use it as well with no problems.  It must be there, undrawn. 

The only 'problem' with the large drop in the GR is all the excess heat generated in the supply.  Filtering should be better.