Modifying the Collins 356A Mic Pre for variable Negative Feedback Control

Hi y'all! New to the forum, but very glad to find such a place! I'll dive right in:

I rebuilt a Collins 356A mic preamp and I would like to modify the circuit to be able to vary the amount of Negative Feedback from the output transformer (see shcematic and note). I received this tip from maestro Jakob Erland:

"Off the top of my head, I would move C205B and R208 to be permanently coupled between V201 cathode and ground. I would then connect output transformer pin4 directly to ground, and take pin5 through a variable (or switched) resistor AND a 100uF/35V electrolytic (minus towards transformer) to V201 cathode. This way, you'd be able to go from full nfb (as it is now, with variable resistor at zero ohms), towards gradually lowering feedback. i.e. increasing gain and "sound"..."

But he suggested I get on here to ask you fine folks. Any thoughts on this? Thanks!!

To make that scheme work you would need to leave out C205 all together but otherwise it seems like a sensible approach.

Cheers

Ian

I've done it differently.  100R 2W wirewound pot in series with 7R6 across the feedback winding, 7R6 to pin 4, wiper connected cathode.  There are enough connector pins to add this once you free up the input transformer center tap connection.  This will give you 40-60 dB range, approximately. 

The 100R value provides a bridging load to the feedback winding, so as to not disturb loading, while also minimizing overall cathode resistance change. 

Fully removing NFB to get 68 dB gain destroys response, both top and bottom.  You can lessen by 20dB satisfactorily, with loss of 1.5dB or so at 30Hz.  The 7R6 stop resistor limits you to about 20dB. 

The method above may limit useful gain to a range of 34-54 dB from the loss of the cathode bypass cap; I haven't tried it.  That may be a more desired range.

Raising the cathode bypass caps to 100 or more will help linearity on the bottom end as well. 

Noise goes up with gain, to be expected. 


Who can tell us exactly what R205/C204 achieve?  I have never seen anything like this in any other amp.  A bit of modeling once suggested to me that it helps the extreme bottom end, but that modeling was likely flawed, and I'm not lending it credence. 

emrr said:

Who can tell us exactly what R205/C204 achieve?  I have never seen anything like this in any other amp.  A bit of modeling once suggested to me that it helps the extreme bottom end, but that modeling was likely flawed, and I'm not lending it credence.

Click to expand...

Isn't this just a standard cathode resistor with bypass capacitor, with the feedback winding in series, just drawn weirdly? Something like e.g. http://gyraf.dk/schematics/tubetech_mp1a.gif ?

Jakob E.

gyraf said:

emrr said:

Who can tell us exactly what R205/C204 achieve?  I have never seen anything like this in any other amp.  A bit of modeling once suggested to me that it helps the extreme bottom end, but that modeling was likely flawed, and I'm not lending it credence.

Click to expand...

Isn't this just a standard cathode resistor with bypass capacitor, with the feedback winding in series, just drawn weirdly? Something like e.g. http://gyraf.dk/schematics/tubetech_mp1a.gif ?

Jakob E.

Click to expand...

It is.  The resistor provides the dc bias for the tube and the capacitor decouples it for ac. The winding is in series with tehm both  and provides the negative feedback.

Cheers

Ian

R205/C204 is 5.6M/0.1uF between V1 plate and V2 grid. I would guess too that it is shaping the low end along with the other elements to give 20Hz-20Khz or whatever bandwidth it is rated for.

Does anybody know the transformer ratio?

5M6/0.1uF is something like 0.3Hz - does not compute.

Schematic error and/or smoke screen?

Jakob E.

16k5:600 plus tertiary which I've measured roughly, notes not in front of me now.  About 40 dB lower than the primary.

RC values correct and real; I've seen lots of them.  Not mentioned in lit. 

Then with relatively low primary inductance loading V2 that RC filter may be needed to party form the roll-off rather than the transformer winding and 20uF capacitor. It may minimise low frequency distortion created by the transformer in that case.

One output I checked out of circuit with a hand held meter read:

16k5: 35.74h
600: 2.15h
Tert: 31mh

Raising of cathode cap values would seem to begin to mask the purpose of the filter, possibly. 

very uncommon design. How does this sounds?

Sounds like Nashville 1962.  Actually 95% the same as Langevin 5116 and several General Electric. 

gyraf said:

5M6/0.1uF is something like 0.3Hz - does not compute.

Schematic error and/or smoke screen?

Jakob E.

Click to expand...

I'm not sure it's a smoke screen or an error. It changes the VLF phase response, probably in order to extend the phase margin and prevent motorboating (which can happen at a fraction of Hz).
Using overall NFB over a transformer can be (and usually is) tricky. One has to be "trickier".

That makes sense.

A random Collins example measures -5dB at 8Hz, with 150 source and 600 load. 

Comparing the GE and Langevin versions again, the Langevin is the most different (not much), and doesn't use this network.  The GE is part for part the same except for the output transformer having a 25K primary.  The GE type may be the first version, with a 9/51 date on the manual cover. 

I would love to see these Langevin and GE schematics you are talking about. I have the 5116B, but it is a PP design, and resembles nothing about this Collins. Would you post those?

Emmr wrote:

Sounds like Nashville 1962.

Click to expand...

'nuf said.  8)


Rafafredd wrote:

I would love to see these Langevin and GE schematics you are talking about. I have the 5116B, but it is a PP design, and resembles nothing about this Collins.

Click to expand...

Google is your friend.
http://collinsradio.org/archives/manuals/356A.pdf
http://www.technicalaudio.com/pdf/Langevin/Langevin_type5116_original_version_schematic_correct.pdf
(Perhaps Doug can add the relevant GE.)


Single ended bliss. 


Henk

http://groupdiy.com/index.php?topic=55375.msg706973#msg706973

Great thread. I have a couple of GE BA-1-F modules that I was about to attempt this on. Thanks for the info.

Paul

I had a GE a few years back (http://analogaudiorepair.com/2012/08/05/ge-tube-preamp/) and I used the Langevin schematic since I didn't know Doug back then  ;D

That's pretty funny, I'm using the same GenRad power supply for mine!