Calculating Output Impedance for a 6G15

[tubeswell]

Following the sub-dedate on this thread;

http://music-electronics-forum.com/s...ad.php?t=10431

with Sir Cuitous about output impedances for the 6G15 in order to come up with a suitable impedance for a theoretical 1:1 isolation transformer for the 6G15 output (is this a trick question? - I thought 1:1 meant that the reflected load was the same as the input impedance of whatever was following the earlier device? So I could be barking WAAAYYYY up the wrong tree - please tell me if this is so), I decided to start another sub thread too help me work out the answer. So..... (big breath in - )


To my simple mind, the 6G15 has two stages that provide source impedance leading into the mixer (I think that much is correct - but Ah Foohey - I could be wrong about that too). These stages are in parallel and one of them is a cathode follower. Does this mean that output impedance is the product of paralleling those last two source impedances? And since the cathode follower has really low output impedance, does that make the total output impedance at the mixer really low too? Or does that depend on how the mixer is set?


Formula for bypassed gain stage; Zout = Ra || ra

Anode resistance for a 12AX7 triode is 70k

Therefore, for the recovery stage Zout = 100k x 70k/(100k + 70k) = 41k (I think that's right)


Formula for a CF stage Zout = ra/mu (I'm not sure about this tho')

Assuming the gain is about 0.93 for a typical CF stage, therefore for the 12AX7 CF stage; Zout 70k/0.93 = 752k (That doesn’t sound right - what have I done wrong? The formula?)


Anyway even if I didn’t get the CF Zout impedance correct, if the two output stages were paralleled, then does that make the output impedance:
Zout = normal stage Zout x CF stage Zout/(normal stage Zout + Cf stage Zout)? (= really low?, or does that depend on how the mixer is set, as I pondered earlier)

[Steve Conner]

The original 6G15 has an output impedance of about 220k, IIRC, set by the mixing resistors. This is higher than any off-the-shelf transformer can cope with and still maintain a flat frequency response.

The Torevibe circuit has a cathode follower on the output, so the output impedance is just 1/gm of the tube. So about 500-1k ohms for a 12AX7.

If you want to add a transformer to the 6G15, you should add a cathode follower after the mixer and then use the highest impedance 1:1 transformer you can get, which is typically 10k.

[tubeswell]

Okay thanks Steve (and thanks for the impedance formula for the CF stage)

[Sir Cuitous]

That's a lot higher impedance than I expected. It seems VERY high since the next stage is a guitar amp. Does that compare to the impedance coming out of a guitar?

So, I'm now interested in getting the impedance down with or without transformer isolation. Can I do this with 3 tubes? Would 1/2 12AX7 work instead of a 12AT7 for the first stage before the 6K6? Would that free up a triode for a cathode follower at the output?

[Steve Conner]

Well, I could be wrong. Does anyone have the original 6G15 schem so we can check?

It is comparable to the output impedance of a guitar, though. Even with the volume up full, the output impedance of a pickup is much higher than its DCR because of the inductance.

[tubeswell]

6G15 Schematic

Here 'tis

The CF stage is 1/2 12AX7 - so (BTW) assuming that the gm is 1.4mS, then does that make the output impedance of that stage 1/1.4 = 0.7k? Did I understand that right? (or should I be using 1400umos in that equation? - the answer to that would be way to low wouldn't it?)

[imaradiostar]

just a thought- if a lower Z transformer is the only thing available or none is available this might be the ideal place for a mosfet based source follower. It should be able to drive a much lower load than a 12ax7 without using extra filament current and should still be linear when the triode in front of it starts to clip.

You could bias up an irf810 or similar mosfet with a few high value resistors (1m-2m2) and add it at end output of the original circuit with very little effort.

jamie

[Enzo]

The entire point of using the transformer was to provide isolation so as to prevent ground loops when connecting other equipment. If you don't care about isolation, then just build the thing stock.


I wonder, what was the last time anyone worried about the impedance of his guitar versus the input impedance of his Fender amp?

[Sir Cuitous]

The original 6G15 has an output impedance of about 220k, IIRC, set by the mixing resistors. This is higher than any off-the-shelf transformer can cope with and still maintain a flat frequency response.

The Torevibe circuit has a cathode follower on the output, so the output impedance is just 1/gm of the tube. So about 500-1k ohms for a 12AX7.

If you want to add a transformer to the 6G15, you should add a cathode follower after the mixer and then use the highest impedance 1:1 transformer you can get, which is typically 10k.

Steve, 220k is probably right. Do you have a schematic for the Torevibe or can you tell me how they implemented the CF after the mix pot?

[tubeswell]

torevibe schematic

Here 'tis

As you can see, the CF stage before the output is at the tail end of the LFO circuit. There is also a Cf stage bypassing the reverb driver/recovery befor ethe LFO circuit

[Enzo]

Uh...

The output is V4B, The LFO is V5. The LFO output is a sort of CF paralleled to V4A cathode resistor.

Or do I misunderstand?

[tubeswell]

My mistake - V5A is the oscillator - I looked at V4A quickly and thought it looked like an oscillator (what with the network of caps and resistors off the plate) but on closer inspection I see that those caps and resistors are (somewhat unique) frequency shaping and padding before the CF stage in front of the output. I must confess to a bad habit of hastily reading schematics and jumping to the wrong conclusions. My shameless habits always get me into trouble, and I can't seem to shake them. I feel like Yogi Bear.

[Sir Cuitous]

Tubeswell, thanks (again) for the schematic.

Just so I can tell myself I'm keeping up wtih you guys. The LFO circuit impacts the cathode of V4A in order to create the tremelo effect? Since I do not want tremelo, can I dispense with that part of the circuit?

V4B looks exactly like V1A, but what is the circuit between V4A and V4B? It looks like a tone stack without pots. Can I use a more concise circuit here like the tube that drives the Bassman 5F6 tone stack? Do I have to be concerned about the phase of the signal being inverted or not? I understand that a cathode follower does not invert the signal.

[tubeswell]

Yep what Steve said (if you don't want the tremolo) involves having the 6G15 scheme with toreT V4B stage tacked on after the 250k mixing pot (and before the isolation tranny). You would keep the 2 x 2M2 from plate to grid and grid to ground (R23 and R24 in the tore schematic) to bias the CF stage, and I guess you should have a coupling cap between the 250k pot and those resistors to stop DC getting back to the pot.

[Sir Cuitous]

OK, here's what I came up with to use an additional tube:



I just draw these in Corel so I realize it's not art. It's just a drawing of the modification. The rest of the plan is still a 6G15. Will this work?