SA-39B Gates - PSU Stabilisation Section

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

opacheco

Well-known member
Joined
Mar 16, 2006
Messages
934
Location
HONDURAS
Hi,

I have taken the time for drawing the power supply stabilisation section for a SA-39B in a more clear fashion and I would like to know what kind of topology or type of stabilisation circuit is that??.

The +B is the voltage after of LI Choke Coil; here is a very clear to see that +V is the regulated power supply for the V1, V2, V3 and V4, the preamps tubes and the +42 Volts is the pre-adjusted voltage for the Threshold control fixed by Gates in this unit. We can do some very interesting things for more controllability in this unit, like Threshold variable control by a Pot.

I have draw the Release and Attacks sections too but I would like to discuss or receive some idas about the stabilisation funtion. I would like to get more about the electronic performer of this stabilisation section but I can't recognise this circuit sincerely!!.


Thanks a lot in advance,
Opacheco
 

Attachments

  • SA-39B Power Supply Stabilization Section.pdf
    289.3 KB
> recognise this circuit

Redrawn in logical form.

A very standard tube regulator.

6(S)J7 cathode stands on a gas voltage-reference tube. A fraction of the regulated output is fed to 6J7 grid. The error is amplified to a 1meg resistor at 6J7 plate. This is fed to 6L6 pass-tube grid.

6J7 screen grid is fed by the regulated voltage through a small dropping (or protection?) resistor.

That leaves the diode, 4K7, and a 5meg. I bet the 5Meg is a start-up trick, to get the gas-tube going reliably. I have no idea about the 6X5 and 4K7.... total 6J7 current will cause very small voltage drop, and the exact voltage at this point is not very critical.
 

Attachments

  • SA-39B-PSU.gif
    SA-39B-PSU.gif
    4.6 KB
PRR said:
> recognise this circuit

Redrawn in logical form.

A very standard tube regulator.

6(S)J7 cathode stands on a gas voltage-reference tube. A fraction of the regulated output is fed to 6J7 grid. The error is amplified to a 1meg resistor 6J7 plate. This is fed to 6L6 pass-tube grid.

6J6 screen grid is fed by the regulated voltage through a small dropping (or protection?) resistor.

That leaves the diode, 4K7, and a 5meg. I bet the 5Meg is a start-up trick, to get the gas-tube going reliably. I have no idea about the 6X5 and 4K7.... total 6J7 current will cause very small voltage drop, and the exact voltage at this point is not very critical.

PRR,

You are right like always!!....I redrawn following your instructions and I verified the Diode 6X5 connections in the SA-39B schematic for any error in my original drawing but everything look ok, no errors  in this section!

I am not sure of the 6X5 Diode function too in this circuits....I could think like "protection" unit when the +275 drop down to certain level but I am speculating!!.

May be someone have a idea about!

Thanks a lot,
Opacheco

 

Attachments

  • SA-39B Stabilisation Section v2.pdf
    236.8 KB
Nice re-draw. 

The first version doesn't use R50, and had a 10K load resistor on the output in parallel with C10.    Many differences in the amp too.

The sound is the transformers, the limiting action is secondary.  5 different era of SA-39 all 'sound' somewhat different. 

The 6X5 answer might or might not show up by taking the 6X5 out of the amp, see what changes. 

I'd guess the double filtering of the 6X5 creates a more stable sample for the 6SJ7, with R40 creating a more specific filter against it.  Just a guess.  That might be considered an edge in an era of tighter enforcement of over-modulation regulations. 

Many broadcast limiters don't go so far with the PSU, BA-6A is just a OD3 regulator for the GR tubes alone, SA-39 is for stage 1 and 2. 
 
emrr said:
Nice re-draw. 

The first version doesn't use R50, and had a 10K load resistor on the output in parallel with C10.    Many differences in the amp too.

The sound is the transformers, the limiting action is secondary.  5 different era of SA-39 all 'sound' somewhat different. 

The 6X5 answer might or might not show up by taking the 6X5 out of the amp, see what changes. 

I'd guess the double filtering of the 6X5 creates a more stable sample for the 6SJ7, with R40 creating a more specific filter against it.  Just a guess.  That might be considered an edge in an era of tighter enforcement of over-modulation regulations. 

Many broadcast limiters don't go so far with the PSU, BA-6A is just a OD3 regulator for the GR tubes alone, SA-39 is for stage 1 and 2.

emrr,

Thanks for your comments and interesting notes.

According to your comments I noted in the schematics there is 0 volts between Cathode and Anode for the 6X5 diode, it is conducting all the time, suppling voltage Screen to 6SJ7 by R38!!....I would like to check the 6X5 Curves and see what is the voltage drop (Cathode - Anode Characteristics) for this tube!!

I can be wrong with this analysis maybe...Please correct me if I'm wrong!

Opacheco.

 
> check the 6X5 Curves and see what is the voltage drop

You need to know the current. The drop will be many volts at 100mA but much less for lower current.

The 4K7+6X5 feed the 6SJ7 plate and screen. The plate current is just-enough to hold the 6L6 under control. Say the 6L6 drops 375V to 275V, 100V across it. 6L6 triode Mu is about 10. So the 6L6 needs 10V grid to cathode. This is developed across 1 Meg. The current must be about 10V/1Meg or 10uA or 0.010mA.

6SJ7 screen current is about 1/3rd of the plate current. Now we have 13uA.

You will not find 13uA on the 6X5 curves. It is essentially zero-zero voltage and current.

The 4K7 can pass the whole 13uA with only 0.06V drop. There is certainly nothing about this 275V circuit which cares about 0.06V differences.

I wonder if the designer got confused, or got a bonus for every tube he designed-in, or if there has been a transcription error in later re-drawings.
 
The 6L6 has about 425VDC on it's plate, if I recall correctly.  In the first 10 years of production, that tube is a 6Y6.  No changes around that.  The 6X5 was always there, from the first drawing.  As cheap as Gates could be, I'd expect it to have gone away if anyone thought about it being redundant in 25 years of production.  They did spend the time to change it from an 8RU limiter to a 7RU limiter, with a cheaper case.  I really can't say, don't have one around anymore to pull and check. 
 
PRR said:
> .......You will not find 13uA on the 6X5 curves. It is essentially zero-zero voltage and current.
The 4K7 can pass the whole 13uA with only 0.06V drop. There is certainly nothing about this 275V circuit which cares about 0.06V differences.

Exactily(open circuit for the 6X5 diode)!...the 6X5 diode will come in scene when the necessary conditions of the designer be established; a change in the 275 V going up maybe!!!....in that conditions the 6X5 diode "drain" some current (the necessary quantity only!), changing the 6SJ7 plate and screen voltage, in turn changing the 6L6 grid bias and returning to the stabilised 275 V!...Look like the error voltage is R40/ 6X5 for me.......again, may be I can be wrong!!

A couple of measurements will be fine in this point.

I appreciate any comment
Opacheco.
 
Back
Top