"6N3 HiFi Buffer Amp" -- what does it really do? Schematic corrected again

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Blue Jinn

Well-known member
Joined
May 28, 2009
Messages
406
Location
USA Northwest
and how to improve it?

Hello,

I joined a discussion over at another forum about this circuit. I picked on up rather cheap on a lark.

The tubes are 6N3 (Chinese version of 2C51 AFAIK) I haven't measured B+ under load, but I'd guess about 200v. (The psu has 150VAC going in, 2x 270uF/1k RC filters, and a 150k resistor strapped across the output. The heaters are AC, but they also have LEDs for a glow underneath the tube sockets. !

It is supposed to be a buffer, but has some oddball design features. I redrew the schematic from one I was sent by a seller. (which had a glaring mistake on it...) Ignore the tube pin numbers. (They are each one half of two separate tubes.)

The 50k output pot seems to defeat the purpose of a buffer and the 10k resistor on the input seems like it could be higher, or even eliminated.

Any other thoughts about what the designers might have had in mind? And thoughts for improvements?

I had thought about eliminating the pot altogether, and the 10k resistor on the input (R2) or increasing it? Also it seems odd that the output has two caps in series. Perhaps eliminate the .33uF?

 

Attachments

  • 6n3-corrected-again.pdf
    11.1 KB · Views: 62
Seems like only part of a circuit to me. What is the upper stage for? Why the low input impedance? The series output caps?

I would hazard a guess that there is a tone network missing from this.

 
MagnetoSound said:
Seems like only part of a circuit to me. What is the upper stage for? Why the low input impedance? The series output caps?

I would hazard a guess that there is a tone network missing from this.

I omitted the PSU and the second channel in the redraw, but that is all there is.

The upper stage? don't know. It looks *similar* to Bronski's split load phase splitter, but not exactly.

Low input impedance? Don't know why, btw it's an SMD resistor. But it does seem too low, and at some point the output impedance depeding on the position of the of the 50k pot could be *higher* !!!

Also, can't explain the series output caps. Although the schematic I got had B+ intjected between the two of them, and there is an unused pad on the PCB between the caps.

I *could* use a buffer on a plate out preamp I have which has a volume pot on the output like this, so something like this could be useful for that. But I'd think I'd need to make some changes first.
 
that drawing sure looks like a fragment of a larger circuit.  maybe these are leftover boards from a multi-board product, and these are getting blown out. 
 
RuudNL said:
It looks like a unity gain buffer that adds "tube warmth".
(Whatever that may be...)

Well, I did curl up next to a HF transmitter once a long time ago, this won' t get that warm...

I think it is supposed to be something like that, plug it in between your cd player and amp as an effect, (the PAIA tube head, the Berhinger T1950 etc) but the low input impedance, 50k pot on the output, and the series capacitors are rather baffling (even to my novice eyes) and what does the other tube section do?
 
rafafredd said:
save the board with the tubes and sockets and solder some other circuit on it.

Decided to try it out as a simple buffer.  Left the 50k "volume"  pot wide open (will eventually just pull it altogether.) Pulled the 10k resistors off the input, and connected the input straight to the output of a preamp I built a while ago from a circuit in EM magazine.That has a pot on the output and per the article "should drive a fairly high impedance." The preamp is similar to the Nachbauer mctube btw with a tone stack in between two triode stages.

The plate and cathode resistors ( from my math )seem to be OK. Actually as a simple buffer between a pretty high impedance output preamp, and my line ins, it does work, and actually does seem to improve the signal somewhat. Output Z (again from my math) should be about 200 ohms. (Although there is also a 10k resistor in series on the output, not sure why that is there either....)

I'm still wondering about the two capacitors in series on the output though. Next up  I plan to pull one of the .33 uF's and compare with and without, and the output resistor,  then go with what sounds best. Open of course to suggestions. 
 
This thing gets weirder. The schematic is (yet again) wrong. There is a connection between the plate of the upper tube and the two output capacitors.... (I need to double check the exact topography of that. -- and I can't tell if it is on the Left channel or just the Right....) I didn't notice that before until I started swapping out parts and saw another trace on the top side of the board....

However, removing the 10k resistor and replacing the two output caps with a single 1uF, (and bypassing that plate out connection) nets the expected result.

EDIT: I suppose having that plate out connection might be useful for something though....
 
I attached a new corrected schematic. Note, I based the original drawing on the one I was provided by the seller (or one of the sellers) which proved to have several errors. Also, the .33uF caps are actually .47uF.
 
That's actually beginning to make sense. If you move the top tube to the right you can see it is two CC stages with NFB from the plate of the second to the cathode of the first. Overall gain should be about 6 times.

Cheers

Ian
 
ruffrecords said:
That's actually beginning to make sense. If you move the top tube to the right you can see it is two CC stages with NFB from the plate of the second to the cathode of the first. Overall gain should be about 6 times.

Cheers

Ian

Yeah, I wish the original schematic I received had been correct to begin with. The circuit now "makes sense" except what is the design goal of injecting the NFB into the cathode follower?
 
This is a quite known type of buffer, but I can't remember its name for the life of me! I guess PRR knows the name.
It's an interesting combo of a cath-follower with a dual-tube stage; it's a push-pull indeed.
The major benefit is that both alternances of the signal are not limited by the cathode or plate resistor. For positive alternance, the 1st tube pulls towards B+, for negative alternance, the 2nd one pulls to ground. Due to the overall NFB, distortion is pretty good.
Now, there are thousands of "tube buffers" on sale everywhere; they turn your bad solid-state-digital sound into vacuum-tube good. Isn't it wonderful? There's even car audio versions...
 
Aha! I came across this:

http://www.glass-ware.com/tubecad/TC_Tube_Circuits_Descriptions3.html

Constant current white cathode  follower, also labeled as  symmetrical white cathode follower in the link, looks very similar:

http://www.glass-ware.com/tubecad/WCF3.gif
 
Blue Jinn said:
ruffrecords said:
That's actually beginning to make sense. If you move the top tube to the right you can see it is two CC stages with NFB from the plate of the second to the cathode of the first. Overall gain should be about 6 times.

Cheers

Ian

Yeah, I wish the original schematic I received had been correct to begin with. The circuit now "makes sense" except what is the design goal of injecting the NFB into the cathode follower?

There is no cathode follower. The first tube is a common cathode stage with an unbypassed cathode resistor. Its anode output is connected by a capacitor to the grid of the second stage which is also a common cathode stage. The anode of the second stage is connected via C2 and R8 to the cathode of the first stage. This is the negative feedback path; it is not an output. The gain of the two triodes overall is given approximately by (R8+R3)/R3.

The output comes from the anode of the second tube via C3. I have redrawn the circuit is LTspice to hopefully show more clearly what is going on.

Cheers

Ian

 

Attachments

  • Screenshot from 2012-07-12 19:04:44.png
    Screenshot from 2012-07-12 19:04:44.png
    38.7 KB · Views: 49
abbey road d enfer said:
ruffrecords said:
There is no cathode follower.
Indeed. I've been fooled by the first draft. So now, it's just a two-stage amplifier with 15dB of gain.

I think calling it a buffer, and the wrong schematic I got from the seller made me keep looking at it all wrong!

Thanks for the redraw and clarification.  :) It *does* now seem very rational except for the lowish input impedance.
 
abbey road d enfer said:
ruffrecords said:
There is no cathode follower.
Indeed. I've been fooled by the first draft. So now, it's just a two-stage amplifier with 15dB of gain.

damn! I was hoping for some kind of strange SRPP variant. Alas! just an exercise in schematic obfuscation.
 
Blue Jinn said:
@ruffrecords:

What does the NFB do to the output impedance? (Ignoring the 50k pot)

The NFB will lower the small signal output impedance by the amount of feedback. The amount of feedback in dB is the difference between the open loop gain and the closed loop gain. Given the 6N3 has a mu of abut 35 and a rp of about 7K or so then.

1. the gain of the first stage is about 35 * 80/(80+85) = 17 times

2. The gain of the second stage is about 35 * 33/(33+25) = 20 times

So open loop gain is about 17 * 20 = 340 = 50dB

Closed loop gain will be close to (10K+2K2)/2K2 = 5.5 times = 15dB

Therefore the amount of feedback is 50 - 15dB = 35dB

OP impedance without NFB is about 33K//25K which is about 15K

NFB will reduce this by about 56 times to  about  266 ohms

Remember that small signal output impedance is no indicator of drive capability, especially at large signal levels.

Cheers

Ian
 
Back
Top