A couple of 6G15 Reverb "whys"

[vanslycken]

So I'd like to build a 6G15-like circuit into my tube amp project (as the very first stage). I did some preliminary fiddling around and I already got some good sounds out of that pan with some basic circuitry.

Now I'm looking at the "real" Fender schematic and I have some questions:
- What's with the use of those 2 halves of the 12AT7. They tap 1/10th of the signal of the first triode to the "dwell" pot and send that on to the next tube. Why didn't they work with just one half, without the voltage divider?
- If this has to do with polarity of the signal (which is my best guess), couldn't you just reverse the polarity of the send or return of the pan? It's just a matter of reversing the connection of the 6K6's output transformers secondary, right?
- Has anybody experimented with the value of the 250pF cap between the tone and mixer potmeter? Seems like an awfully small value...

Any other thoughts you might have are welcome; I'm making some changes to the circuit to accomodate it to the stuff I have lying around: I have a 4AB3C1B pan; the transformer I use to drive it is a Visaton 100V system transformer driven by an EL84 with Ua about 220V and Uk about 10V - for experimenting I used the 2,5 and 1,25W taps.

[tubeswell]

- What's with the use of those 2 halves of the 12AT7. They tap 1/10th of the signal of the first triode to the "dwell" pot and send that on to the next tube. Why didn't they work with just one half, without the voltage divider?

They work together to drive the tetrode with a clean, strong signal - remember that this reverb is designed to sit in front of the amp and gets its signal straight from the guitar. However the characteristic 6G15 reverb starts to dirty-up a little when you turn up the dwell. Some peeps like this and others don't - You can moderate this by altering the split load on the first triode. Changing the tranny's output polarity won't make any difference to the end result. There is no NFB in this circuit, and the human ear can't detect the difference.

- Has anybody experimented with the value of the 250pF cap between the tone and mixer potmeter? Seems like an awfully small value...

I find the stock tone set-up more than adequate for the function it has of making subtle changes to the wet reverb signal. The main tone controls are in your amp.

[vanslycken]

Thanks for the answer; I would've expected that the phase of wet and dry signal would matter when you mix them?

(edit: I had posted something wrong here.)

[Enzo]

The reverb signal has no phase. At least not with respect to the dry signal. The dry signal was pounded into the end of a couple long springs, and whatever made it o0ut the other end is amplified as your reverb. That pretty much dissolves any phasal relationship.

[tubeswell]

Thanks for the answer; I would've expected that the phase of wet and dry signal would matter when you mix them?

No it doesn't matter - The signal(s) you get from the end of the reverb recovery stage are not exactly 'in phase' with the signal from the CF cathode anyway, whether you have the tranny secondary wired one way of the other. The reverb recovery happens after the springs in the tank, which do weird things to the signal phase, and that's the whole point. Your ears won't tell the difference.

(Edit) Dang - I see Enzo pipped me at the double-post :-)

[tbryanh]

According to Adam’s Amps Triode Calculator (http://amps.zugster.net/tools/triode-calculator), the gain of the first stage is about 50, and the gain of the second stage is about 25.

With the 1/10 attenuation, there is about a gain of 125 (50/10 * 25) instead of 1250 (50 * 25).

With a single stage, there would only be a gain of 50 or 25 depending on whether the cathode resistor is bypassed or not.

Since a gain of 125 was chosen, it appears gains of 1250, 50, or 25 are not correct amounts, hence the need for 2 stages with one stage throwing 90% of its gain away.


As far as phase goes, as mentioned in a previous post (sorry can’t find it), because of the delay through the springs, no matter how many gain stages are used, it is near impossible to make the wet and dry signals be in phase. (Sorry for the duplicate of Ezno's and Tubeswell's responses. We responded about the same time.)

Not absolutely sure, but phase might only be an issue when two signals being mixed are exactly 180 degrees out of phase.

[tbryanh]

Since a gain of 125 could be had with both cathode resistors bypassed or with neither cathode resistor bypassed, why was it chosen to have one bypassed and one un-bypassed instead of both bypassed or both un-bypassed?

Also, since it was chosen to bypass the cathode resistor in one stage only, why was it chosen to have the first stage's cathode resistor bypassed instead of the second stages cathode resistor?

[Enzo]

Without looking at specific schematics, in general, the job of the first stage is to lift the signal away from the noise floor. Later stages then work with the signal. For that reason I would bypass the first stage for added gain, then later stages wouldn;t need it. This is often seen an amplifiers.

Phase is only an issue when it is the SAME signal. The reverb signal looks nothing like the dry signal.

Yes that is an oversimplification. One could argue that a 1000Hz sine wave and a 1000Hz square wave have a phase relationship. But one would have a hard time convincing me that a 1000Hz sine wave and the sound track from West Side STory have a phase relationship.

[tubeswell]

Since a gain of 125 could be had with both cathode resistors bypassed or with neither cathode resistor bypassed, why was it chosen to have one bypassed and one un-bypassed instead of both bypassed or both un-bypassed?

Also, since it was chosen to bypass the cathode resistor in one stage only, why was it chosen to have the first stage's cathode resistor bypassed instead of the second stages cathode resistor?

Having the 2nd stage being un-bypassed allows for some local feedback built into the stage giving better fidelity (at the expense of reduced gain)

[Diablo]

I am one who didn't like all the gain when the dwell pot gets turned up - turned into a fuzz tone. I added a split plate load on V2B (12AT7), 66K and 33K, so I threw away about a third of the signal and that tamed it nicely.

[vanslycken]

Thanks guys, the gain with a single triode is indeed insufficient and changing the phase indeed had no effect. It all makes sense