Help with Gates tube mic preamp???

[lassoharp]

Ever since reading this thread nearly a year ago there's been a question in the back of my head that I wanted to address before attempting a build.

What's been bothering me is the loading on the 6C5 as the output tube for the SA-70.  From Dougs post above I take it the stock OT is as was likened to the LS-27 meaning a 15Kish pri.  If that is the case then you would have the 6C5 running at 2.2ma and a plate resistance of ~19K.  Rp is 56K so the 6C5 plate is loaded with ~12K.

Are there any significant problems with loading the 6C5 that far below its plate resistance?  I ask this because many of the vintage mike circuits I've seen have the output stages loaded at or slightly above matching condition.  The 6J5 (8K) loaded with 100K \\15K on the earlier circuit would seem more like what I'm used to seeing.

I think this question arises out of not being entirely clear on what is most important for the output stage - voltage or current gain.  I think it is current though some circuits like the BA-2c have the output tube loaded at 2-4 times rp like a small signal stage.

 

[PRR]

Load a triode at 2X to 4X plate resistance.

When you go to 1X, THD at max output rises. Interestingly, you can go as far below 1X and THD at max output does not rise much more.... but max output falls.

Going from 1X to 4X will give roughly similar Max Output, but THD may fall from near 10% toward 2%. This seems like a good idea.

BUT: general-purpose triodes have plate resistance well in excess of 10K ohms. Electrons do NOT like to flow in vacuum. OTOH, general purpose audio transformers have optimum impedances below 10K ohms. It takes a bunch of wire to choke bass, and that heap of wire's capacitance sucks the treble off.

So you pay more for fancy iron, or you pay more for a big low-Rp tube. Maybe a little of both. It is a problem in cost compromise.

But what is Max Output here? Evidently not huge: else it would not resistor-couple the iron. (R-couple does cut iron cost; another cost tradeoff.)

This is a broadcast studio for well-modulated un-raised voices on dynamic mikes. The normal input would hardly be more than a few mV. The normal output would generally be 100mV. As I read it, the MAX output is 5V. So it normally works at 1/50th of maximum output. And if heavy loading caused 5%THD near Max output, it would be 0.1% at normal output. All pure sweet 2nd harmonic, unmeasurable 3rd, negligible 4th. This is Not A Problem.

If you feed a hot condenser a foot from a Fender Twin, you are not using it as the Designer intended. Certainly any old-tyme studio engineer would wire a pad in front; large consoles had spare pads in the patchbay for ad-hoc padding and sometimes screwed to the wall inside for long-term level solder-in correction.

 

[lassoharp]

Thanks for the good info PRR. 

That all makes good sense - I didn't realize that fact about triodes  being able to go that far below rp without a major penalty in THD.

Now that I think more about this - I've seen this more or less when using a 2nd stage grid resistor(as a pot) that was equal to or 1X lower than the 1st stage triodes plate resistor and was always kind of surprised at how cleanly it performed - only a little less output but better bandwidth.

 

[emrr]

What's really curious about this Gates design is the larger UTC output transformer is rated for 20 mA DC, yet they don't use it that way.  The smaller Triad is a 0 mA DC type. 

The use of volume control within a NFB loop is screwy, yet it works fine, being not much NFB. 

The input stage headroom is, in practice, greater by 6 dB or so than most every other preamp of this era that I've tried, so you can get away without a pad in more instances. 

In the proper Gates console, the output is connected for 200 ohms, and it feeds a 250/500 ladder.  And consider as well that 150/200 and 500/600 are considered more or less the same Z. 

 

[lassoharp]

What's really curious about this Gates design is the larger UTC output transformer is rated for 20 mA DC, yet they don't use it that way.  The smaller Triad is a 0 mA DC type. 

The use of volume control within a NFB loop is screwy, yet it works fine, being not much NFB. 

The input stage headroom is, in practice, greater by 6 dB or so than most every other preamp of this era that I've tried, so you can get away without a pad in more instances. 

In the proper Gates console, the output is connected for 200 ohms, and it feeds a 250/500 ladder.  And consider as well that 150/200 and 500/600 are considered more or less the same Z. 

I've wondered about that vol control inside the NFB as well.  I've boarded up this circuit a few times with 15K/600s and a 25K/600 and the tradeoffs are hardly noticeable.

The input stage headroom you mentioned - does that apply for both the UTC and Triad iron?

 

[emrr]

I've wondered about that vol control inside the NFB as well.  I've boarded up this circuit a few times with 15K/600s and a 25K/600 and the tradeoffs are hardly noticeable.

The input stage headroom you mentioned - does that apply for both the UTC and Triad iron?

Headroom same for both with the input stage. 

I think the gain versus turns comes out in the wash for the most part. 

 

[emrr]

Got to look at a rare variation of the SA-70A, the SA-73A.   It sticks a 6SL7 in the 6C5 slot, adding another stage and giving 60 dB total gain.  Downside is input headroom is reduced, as max output remains the same.  Here's the reverse engineering photoshop job.  Some info gleaned from the 1950 Gates catalog.  

 

[shoonbassjp]

Hi all
I build a single channel SA-70 first, loved so much. then build another with 2ch.

Very easy build and noise wise very quiet.
I made power unit separate from main box  that helped low noise too.

Usually use with ribbon mic, or put in front of Federal 864 clone sounds great.
Used Edcor 0.5w OT 15k/600 as output, Ampex for input. 

Thank you all for great info.

 

[rmburrow]

The old Gates tube console mix buss consisted of Daven ladder or "T" attenuators wired for constant impedance.  Normally the mic preamp input was strapped at the input transformer for 250 ohms for a dynamic or ribbon mic.  (A condenser mic may overload the preamp so you may need to calculate resistor values for a 12 or 18 dB 250 ohm attenuator.  I use 1/4 or 1/8 watt film resistors and build the attenuator in the Switchcraft S3FM barrel assembly which can be plugged in series with a high output mic.  Note the S3FM costs approximately $33 unless you have some of those in the parts box.)  The preamp's output transformer was strapped for 600 ohms, to feed a Daven pots (2 db/step and off) which controlled the preamp output level to the mix buss at 600 ohms.

The program amplifier in the console raises the output level of the mix buss to +4 dBm or other suitable level for other processing equipment or a peak limiter (to avoid over modulation of a transmitter).  With a Gates tube console, the program amplifier may include a frequency limiting capacitor to roll off the high end for AM broadcast.  The program amps in the Dualux console did.  Lifting one side or removal of the cap makes the board's frequency response flat.

The HT supply normally was 300 volts, regulated.  If you are racking the preamps only, a suitable power supply can be built using a 6X4 or 6X5 tube rectifier and filter in the usual manner.  (The 6X4 or 6X5 is indirectly heated so there will be minimal surge voltage.)  AC or DC should work for the preamp tube heaters...you may get a few dB improvement with hum suppression with DC heating.  Some of the console power supplies floated the AC heater voltage and used a hum balance circuit.

Doug Williams:  Do you have any loose Gates M5215 mic preamps (the 5879 version) in your parts box?  If so, would you be so kind to place the preamp PCB side on a copier and make a 1:1 copy and send it over?  Either the 5879 or EF86 versions are nice preamps and a nice DIY build.

 

[emrr]

I do not.  Perhaps there's a PCB layout in the manual?  I haven't looked at that in awhile. 

The earlier Gates adhered to RCA standard of 250 ohm output, 250/500 ladders.  The SA series PGM amp did not have a rolloff cap, but shows a selectable hi boost cap that can be added at a cathode. 

 

[mjrippe]

It's almost impossible to put an M5215 on a copier -  there are transformer wires, bolts, and solder tabs everywhere!  I took a snapshot of one with a tape measure along the long edge (4 5/8") and attached that here.  Obviously C5 is removed from the board.  This is the 5879 version.

 

[rmburrow]

Sorry for the late response.  Thanks for posting the Gates M5215 PCB board picture.

 

[Dmichel123]

We have one of these early 31B consoles with 6F5/6J5 preamps and ADC transformers. It sounds very cool, but I feel distorts much too easily. We use ribbons and dynamics almost exclusively and would like to get a little more input headroom on this thing. I think I'm going to try Doug's UTC circuit with mods and see where that gets us. I'm pretty sure we have some 6C5s and 6J7s around, at least one of each to try a channel. We even have a UTC transformer pair in one channel already(had a couple dead ADCs when we got it).

 

[emrr]

You tried the first mods i suggested?

 

[Dmichel123]

You tried the first mods i suggested?

Pad and pot? We had pads installed for a while. I think that did help, somewhat. Putting in a volume control pot isn't a possibility, we don't want to drill anything in this chassis... Would changing the first stage tubes to triode-strapped 6J7 with smaller Rp gain me some headroom?

 

[emrr]

Change to triode 6J7 or 6C5/6J5 would certainly be less noisy, meaning i've almost never heard a quiet 6F5.    Some less gain, but a lot of theoretical gain difference disappears under the actual load present. 

Take out a 1st stage cathode bypass as I suggested and see how it feels. 

You will have to use a 20dB input pad many times, it's the nature of these things. 

You could parallel the input secondary to knock 1st stage drive level back some more but that's a balancing act for tube noise.  Success depends on the range of your typical input levels. 

 

[Dmichel123]

I've been simulating this circuit in SPICE and it appears the distortion happens in the output stage.  Probably not much room to tweak values to get headroom there?

 

[emrr]

You could replace the 100K plate with a choke.  Even doubling B+ as-is doesn't buy a lot.  These things aren't meant to have modern headroom as-is, they're designed to drive -22dBm mix busses.  The kinda similar 61 standalone pre is rated for 1% distortion at 0 VU. 

 

[beezer4]

Hi folks,
Newer to tube builds and recently built a two channel Gates SA70. Stayed true to the schematic Doug shared earlier in this thread. After quite a bit of troubleshooting I'm getting a pretty decent 60hz hum in both channels. I've attempted moving the output transformers around, moving cabling around the unit to no avail. I've including some pics of the inside. I know the wiring layout isn't the most pretty, could this be a contributing factor? I'm using a single buss bar for all the grounds that attaches to that chassis at one location. One thing I haven't done is lift grounds for any of the shielded wire used through the unit. Wherever the shielded wire is used, it's only grounded at one end.

The hum is around -40dbu at full gain in both channels. I feel like I may just be missing something.

All voltages check good and within spec around the unit. Tubes are nos RCA. Chicago transformers wired 600:50k / Edcor output transformers 15k:600.

Any suggestions or advice would be super helpful.
Thanks!!

 

[lassoharp]

Hi folks,
Newer to tube builds and recently built a two channel Gates SA70. Stayed true to the schematic Doug shared earlier in this thread. After quite a bit of troubleshooting I'm getting a pretty decent 60hz hum in both channels. I've attempted moving the output transformers around, moving cabling around the unit to no avail. I've including some pics of the inside. I know the wiring layout isn't the most pretty, could this be a contributing factor? I'm using a single buss bar for all the grounds that attaches to that chassis at one location. One thing I haven't done is lift grounds for any of the shielded wire used through the unit. Wherever the shielded wire is used, it's only grounded at one end.

The hum is around -40dbu at full gain in both channels. I feel like I may just be missing something.

All voltages check good and within spec around the unit. Tubes are nos RCA. Chicago transformers wired 600:50k / Edcor output transformers 15k:600.

Any suggestions or advice would be super helpful.
Thanks!!


View attachment 91505View attachment 91504
View attachment 91508
View attachment 91506View attachment 91507

Nice work. You may have the PSU a bit close to the 1st stage tube. Maybe unsolder and pull it outside of to see if the hum drops.

In general, AC filaments in a DIY tube pre will often leave you with a hum level of around -70 dbish which can be right on the cusp of acceptability depending on your applications. Tweaking the lead dress can sometimes improve it a few db. -40 sounds like something other than filament hum.