Gates 3638 preamp

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

shuchoco

Well-known member
Joined
Mar 20, 2009
Messages
89
Hi,

Bought this Gates preamp on scaryBay recently. The label says it's a Model 3638.  It uses a UTC A-10 ;D, a Triad R-3A power transformer, and the following tubes:  6J7,6SN7,6SJ7, 6X5.

I powered it up (scary) and plugged in an RCA BK-5B ribbon mic, and it sounds pretty damn good.  A little hum,  but really not bad at all, lots of gain, and, i think, really nice sound (not good at describing these things).  I feel like it has alot of potential to be awesome.

Is anyone familiar with this thing? 
Anyone have a schematic I could bum? ;D

Any help/opinions/flames would be most appreciated. 

30dh5pv.png

343p4yb.png





 
That is officially an SA-134, though that's almost never found on an actual unit.   Gates was weird that way; you usually only find the model and part #s together on the price sheets, and some of the catalogs.  It should actually be part number M-3638 or MO-3638.   The best of the Gates remote amps, IMO.  The A-10 is a replacement; might not be as good as what they replaced.  That depends on the era it's from.   I have the schematic somewhere.  Replace those electrolytics, and get a few 1620 tubes to select from for the input stage, and it'll probably have no hum at all.  

I use the beefier rackmount version daily.   Here's the Gates SA-94, which is another variation with an additional stage:

http://www.groupdiy.com/index.php?topic=28450.0
 
Thanks very much Doug!  Sounds like it's a keeper.

You mentioned in your post that you use yours as a DI.  Where would I tap in to the circuit to add the DI input?

I'll get to replacing those caps and play with the tubes and report back.

Thanks Again!

Brad


 
i see.

will the options you implemented into your SA-94 work on this circuit?  :
"- Mic or hi-Z input switch; hi-Z skips the first two stages and puts input into the attenuator.
- Internal high/low gain switch that bypasses the 2nd stage 6C5; cuts gain to about 65 db.
-  Optional 500 ohm input and 5 ohm output wired to original terminal strips."

is the load impedance of this amp well suited to older ribbon mics (like the RCA BK-5B i mentioned), or do i need some kind of mod to make it more ideal?

is there a trick to spotting 6J7s that are fraudulently relabeled as 1620s?  from what i've read the 1620s are identical to the 6J7 but have tested to lower noise specs.  i'm worried i'll get duped.  i guess i should just search the meta for a reputable source. 

sorry if these are dumb questions, i'm clearly new to this stuff.

  i wish i could find an enclosure like the one you used, looks so awesome!  maybe i'll hit Apex electronics salvage and look for something cool. 

thanks so much for your input!

 
My chassis is the correct Gates type, and it took me several years to find. 

the first option applies, the next two don't; there is no 2nd stage before your volume control, and the output doesn't have multiple taps like the UTC versions do. 

6J7 = $5, 1620 = $10.  I don't see any sort of pricing room to bother counterfeiting in the first place.  1620's say 6J7 on them also.  Buy some of both, try them all, ignore the labeling and use your ears.  Be sure to burn them in a good while before listening to noise. 
 
I too have the same preamp but it is in bad shape.  I think some parts are missing and I would love a schematic if someone has one.

 
emrr said:
I have the schematic somewhere.

Hi! Pleaese let us know if there is any chance to find it. Looks like here is a several of us looking for schematic for mo-3638 amp.
thanks a lot!

Mike
 
> 6J7s that are fraudulently relabeled as 1620s?

My $2 KenRad 6J7s are just as quiet as any of the many 1620s I have used.

Get several 6J7. Try them. If one is hissier than the others, don't use it in the first-stage socket.

> i'm worried i'll get duped.

No shame on you. And at $5-$15 for generic 6J7/1620, no great financial disaster.
 
mike22 said:
emrr said:
I have the schematic somewhere.

Hi! Pleaese let us know if there is any chance to find it. Looks like here is a several of us looking for schematic for mo-3638 amp.
thanks a lot!

Mike

If nobody has the original Gates schematic I have a hand drawn schematic of this preamp I have tracked myself from an actual unit. If there is still interest please let me know and I will try to scan it and post it here. Better than nothing...
 
tubologic said:
If nobody has the original Gates schematic I have a hand drawn schematic of this preamp I have tracked myself from an actual unit. If there is still interest please let me know and I will try to scan it and post it here. Better than nothing...

Yes please do!
 
I've had this circuit boarded up for the past several days and am running into oscillation (motorboating type) problems when attempting to use a 100K lin pot as the interstage vol control.

Gates Radio Company SA-134 / MO-3638 schematic

My circuit is using a 6N7 in the output position in place of the 6SN7 - paired with a 25K:600 OT. 6N7 is biased a little colder than stock (2K vs 470ohm).  Everything else in the main circuit is the same.  The PS I'm using uses the following chain:

300-0-300 - 5Y3 - 10HY/515r choke - 40uF - 1K - 40uF - 1K - (plus the additional 3 stages of decoupling as shown on the schematic)

I only mention the supply here in case there's an issue with overall impedance or resonance. The rough DCR of my supply and the stock supply are ~ 26K.  The 1st cap and the choke look to resonate at ~ 8Hz

I ran 2 sets of tests each using a source signal of 5mV @ 1K. 

(1) -  With 15K:20K as the input wired for essentially 0db gain

(2) - With a 150:51K approximating the stock input transformer.

In each case the circuit would begin to oscillate at ~25% of the 100K pot's rotation. This was equal to roughly 71db of actual circuit gain (16db of OT loss was added to the measured value). After this point the gain fell sharply through the end of the pot's rotation.  71db + the net 10db transformer gain of the stock unit would seem to put the total gain at the correct value of 80-82db.

Note - when using the 150:51K as the input, the actual circuit gain (sans transformers) was ~ 55db before oscillation/cutoff. This would still net ~71db of gain from input to OT pri (+19db w/input iron added)

So after all this (hopefully useful info to someone attempting to build on of these) I'm left with the decision as to how to handle the oscillation issue with the 100K pot - specifically, what to best use at that position in the circuit - fixed resistor or switch?  Still, there is the issue of why the circuit is reaching the oscillation(and apparent max gain) point so early using the stock 100k value?

I realize there's plenty of open doors for motorboating in the typical breadboard spaghetti.  I've fiddled with wire positions a good bit - esp keeping input away from output - to no avail.  However the circuit seems to be already at it's limits  when the oscillation begins - But why at only 25% rotation?  Even with a log pot you're still given the possibility of full rotation.

If there's anyone who can spot what's going on here, I'd greatly appreciate any suggestions.

Thanks
 
With the output tube and load conditions different, I'd consider whether or not the feedback path needs adjustment.  Original output here is 12K:600.

I remember trying to up the coupling cap into the pot on the SA-20, and found it oscillated if you went any higher.    In that case it is 0.1 feeding a Daven 250K pot.  You do want to use a log pot here. 

Be sure your input signal ground is tied in correctly; sometimes that can be an infuriating variable that solves everything, once correct.
 
> plenty of open doors for motorboating in the typical breadboard spaghetti.

My vote: the spaghetti isn't wired according to plan.

> Be sure your input signal ground is tied in correctly

Yes, one of many possibilities.
 
Thanks guys  - Fiddled with the ground placement and lowered 1st stage coupling cap to .05.  This seemed

to push the onset of oscillation a little farther in the right direction.  Looks like best thing for now is to call it

a night and reboard everything tomorrow and see what that nets.
 

Latest posts

Back
Top