Gates SA-134/MO-3638 Issues

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
used transfer curves for 6SN7 to guess operating point,
we know that 470 is the cathode resistor
plate current through that resistor will generate self bias volts, draw a redline representing all the possible bias voltages due to plate current through the cathode resistor, find out where it intersects the estimated plate supply voltage, and read the corresponding plate current on the left axis>
6sn7 bias.png
 
Alright, new output transformer (thanks to the absolute GOAT magnet wizard CJ) installed and only a few holes were needed to get this beast in. It still smells like lacquer (the good kind).

I used the original OPT hole and one of the filter cap can holes for the leads. (I replaced this can with a floating cap) Marked the holes for screws.

IMG_0743.jpeg

IMG_0744.jpeg

I plan to install a balanced TRS output jack, but for now just connected the OPT secondary to the terminal jacks.

IMG_0745.jpeg

The mic signal is now being passed! However... there is a crazy oscillation on the output signal. See attached video.

I did measure about 9V coming out the hot lead without any signal with the volume all the way up. What do I check now?


View attachment IMG_0746.mp4
 

Attachments

  • IMG_0744.jpeg
    IMG_0744.jpeg
    132.3 KB · Views: 29
Cant figure it out. Everything seems to trace properly, all the tube sockets are connecting correct ( i manually checked each pin of the tube connecting with the pin on the back of the socket with the tube mostly inserted).

One interesting thing. I had a mic connected and was yelling into it when I powered off the amp, the oscillation went away quickly but the mic amp kept pumping the mic signal through my mixer for a second or two without the oscillation. Might be something with the power supply?

I happen to have a 369JX I was going to use for another project I could try. The power transformer seems to be outputting the correct voltage however.
 
Last edited:
try moving wires around, sometimes altering the lead dress can help.

i doubt it has anything to do with the output transformer s the NFB loop comes off the tube plate and not the transformer secondary. and even if it did come off the secondary, the transformer tested very good for phase shift all the way to 100 K Hz, you can connect a 500 pf cap or 0.001 uf , ground one end and connect the other end to various points along the circuit, if it stops at a certain point, that would be a good area to check. and check. all the grounds, that is also a common cause for instability.

you might try loading the transformer secondary with a 1 K resistor and see what happens.

BTW, what is the voltage on the 470 cathode resistor with the transformer installed?
 
Is the oscillation present at all gain settings ? Is the gain potentiometer working ? Did you replace the capacitor C12 between filament ct and ground ?
It's possible you have some issues with the grounding. Make sure all the 0V points have a solid connection to chassis with a reading as close as possible to 0ohm. Definitely below 1ohm
In this preamp some of the terminal strips are used for grounding and perhaps oxidation or a loose rivet is preventing a solid bond to the chassis.
Make sure the soldering on the 0V bus bar is solid, you need a powerful iron to do a proper job with that.
Try moving wires around with a chopstick with the amp powered on and see if the oscillation disappears.
Personally I would twist the opt primary wires and route the secondary further away from the mains wiring.
Did you also check the input xlr wiring? It was mentioned before the stock wiring is not standard, make sure pin1 is tied to ground.
It would certainly help probing the circuit with a scope, it's the best tool for this kind of troubleshooting.
The new opt looks great, it 's massive, the power transformer is getting envy.
 
Alright. I solved the problem! TL DR: Negative feedback resistor was dropped too low and caused motoroboating in 6SJ7.

Here's where I am dumb: In my haste to DO SOMETHING I preemptively lowered the NFB resistor anticipating the output gain of the amp to be too high. In doing this, I think the NFB increased too much and thats what caused the motor boating. I still dont understand why motor boating happens but an EE friend of mine told me this is common with pentodes and even more common with 6SJ7s in particular.

Anyway, I learned two things:
  • NFB values have a range before changes can produce unintended consequences
  • BE PATIENT, only change one thing in a circuit at a time while troubleshooting.
Thanks for everyone's help on this, (especially Dr magnet face himself, CJ).

NOW TO THE STUDIO FOR AUDIO SAMPLES. I also need to decide whether to use Cherry or Walnut for the box...
 
Last edited:
Check out some audio! Forgive the playing... I only had time to do some guitar. But I have a feeling this thing is going to smoke on bass or drums.

It's super clean and warm. Very low noise, even when compressed. It has IMMENSE gain.

American Jazzmaster, 70s champ, Stager SR-2 ribbon right up on the grill into Gates pre, into Burl ADC:

Preamp gain lower
Preamp gain maxed out with 1176 to attenuate input into converter (warning its intense)

Old accoustic guitar, cheap C12 clone, Gates pre, into DBX 165a (2-4dB GR) into Burl
 
Alright. I solved the problem! TL DR: Negative feedback resistor was dropped too low and caused motoroboating in 6SJ7.

Here's where I am dumb: In my haste to DO SOMETHING I preemptively lowered the NFB resistor anticipating the output gain of the amp to be too high. In doing this, I think the NFB increased too much and thats what caused the motor boating. I still dont understand why motor boating happens but an EE friend of mine told me this is common with pentodes and even more common with 6SJ7s in particular.

Anyway, I learned two things:
  • NFB values have a range before changes can produce unintended consequences
  • BE PATIENT, only change one thing in a circuit at a time while troubleshooting.

it's cap coupled NFB so you need to compensate the cap too. The same output stage amp in another gates amp is 0.5 and 56K, versus 0.05 220K here. You lower the R, you increase the low end bump = motorboating at some point.
 
it's cap coupled NFB so you need to compensate the cap too. The same output stage amp in another gates amp is 0.5 and 56K, versus 0.05 220K here. You lower the R, you increase the low end bump = motorboating at some point.
Interesting. I will experiment with this. The amp sounds really good right now and the noise floor isnt a problem even at lower gain, but I would like to be able to get a bit more color and drive the amp a bit harder.

Thanks!
 
I would like to be able to get a bit more color and drive the amp a bit harder.

Don't change anything then. More NFB = less color and more nasty distortion. The only thing here you can 'drive' IS the output stage, with the gain control.
 
For me the easiest is real time analysis of the amp. There's math, but I'm far slower at remembering what math and why. The quick math in my head says the other way is 12dB more feedback. If the current cap is sized for exact linearity, or even a bass bump, the more feedback the greater the bass bump, so the cap has to get correspondingly larger. In a simple case 220K to 110K is gonna be roughly 6dB more NFB. Double the cap size. All that makes the two versions of the amp look like a 12dB difference, roughly. There are frequently other associated factors that can skew this. The 0.5 of the other may have wanted totally linear bottom, with this one wanting slight bass boost.
 
you can put a pot in place of the NFB resistor, with maybe a limiting resistor on one end, then you can play around and find the sweet spot. this also works on the API mic pre.

freq response was sometimes limited for radio purposes, so the new OPT may have better response than the original, but it should not be much as a 4:1 transformer usually runs pretty flat up to 100 K, especially if there is some power feeding it, as the power charges the winding capacitance no problem, so as to cancel out the effects, thus extending the freq band.
 
you can put a pot in place of the NFB resistor, with maybe a limiting resistor on one end, then you can play around and find the sweet spot. this also works on the API mic pre.

freq response was sometimes limited for radio purposes, so the new OPT may have better response than the original, but it should not be much as a 4:1 transformer usually runs pretty flat up to 100 K, especially if there is some power feeding it, as the power charges the winding capacitance no problem, so as to cancel out the effects, thus extending the freq band.
That is something I had though, the new OPT is probably the factor making this thing so sound rich and luxurious. At this point, this amp sounds great and has very low noise. There does seem to be a bit of a low freq emphasis but it’s pleasing and not extreme at all.

I don’t think I’ll mess with it. I’m extremely happy with the way it’s sounding.

Im building a wood faceplate to rack this in and plan to have the power switch, the super cool nuclear looking indicator lamp, and a big gain knob.

Thanks again for all the help fellas!!
 
Finished the face plate/rack for this thing! Made of sapele (mahogany type). I made a rack for it as well, but just going to put it in the rack at the studio.

Used the same power switch, super cool nuclear symbol pilot light, and the gain pot.

Looks pretty cool! Now if only i could find an old Gates badge to mount on the front....

Thanks for all the help!
 

Attachments

  • IMG_1120.JPEG
    IMG_1120.JPEG
    153 KB · Views: 23
  • IMG_1115.JPEG
    IMG_1115.JPEG
    144.3 KB · Views: 25
  • IMG_1114.JPEG
    IMG_1114.JPEG
    174.1 KB · Views: 24
  • IMG_1103.JPEG
    IMG_1103.JPEG
    178 KB · Views: 23
Last edited:

Latest posts

Back
Top