Pre Amp to PI connection Question

[Rocketrob]

I thinking of building some thing like this attached drawing; AB763 type single channel with reverb and no vibrato. My question concerns joinning the pre-amp and phase inverter. I made some notes on the drawing to explain my question. any help understanding this would be great.
Thanks, Rob.

[Chuck H]

"What's this .1 cap for?"

As drawn the .1uf cap prevents B+ from loading on the signal load resistor. But...

There is no problem implementing the split plate load as you drew it provided the .001 cap is still in place. In fact I do it this way in one of my designs and it works great.

[Rocketrob]

We've got to stop meeting like this, Chuck. As you know quite well, I used the split plate load resistors on my last build, following your advice, and it did indeed work very well. So I thought I would do the same with this build. Simple solution and it makes it easy to adjust the amount of gain going to the PI to get the level of drive that works for me. Thanks again. Rob.

[J M Fahey]

Hi Rob, your solution (Chuck's suggestion?) is the sensible way to go .
You can pad down the excess gain/signal at will.
The California Dreamer .1/50K/220K net is nonsense.
It only pads down V3b signal by loading it with way too low 50K.
What were they thinking?
It reminds me that Internet is chock full of "designs" which are not more than Frankestein style body parts (well, *schematic* parts ) clumsily grafted together .
Even AX84 (which is a serious site) concedes that

This started as a kit from Drifter Amps. However, Gino hadn't quite got the bugs out so we worked through those

... and now the Project is not on the "Recommended" section any more.
Oh well.
PS: to be fair I must recognize these guys can make up some *great* model names:
"California Dreamer" .... WOW !!
Should hire some of them for that job.
PS2: I *loved* the

tweedy growl of the classic blackface amps.

part.
Last time I checked, Tweeds were Tweedy and Blackfaces were .... Blackfacey?

[Rocketrob]

J M Fahey: well yes, "Blackfacey" of course; everybody knows that. The AX84 site is very good, lots of creative ideas and good info there. I was looking for a schem for a blackface type amp with single channel, no vibrato, with reverb, and found the California Dreamer. Although I won't be building it as drawn, it gets the wheels turning for me, and for that I am thankful. Watch out for a bunch of dumb questions comming as I move through this build. Thanks, Rob

[J M Fahey]

AX84's original ones are *great* sounding.
I'm not up to the latest ones or the doubtlessly many variations that surely exist, but I've helped a couple friends build theirs and results were gorgeous/astounding (take your pick).
Boutique amp class. They just deserve a great speaker.

[jpfamps]

The California Dreamer .1/50K/220K net is nonsense.
It only pads down V3b signal by loading it with way too low 50K.
What were they thinking?

Looks like they were trying to recreate the loading effect of the tremolo intensity pot in the absence of tremolo.

[J M Fahey]

Maybe. Probably. Why not?
Don't know what sonic effect they expected to achieve, beyond the quasi-religious intention of "doing things just like Leo did ... whether I understand why or not".
In fact, if that *was* the intention, the network is incomplete: they need to add another 220K mixing resistor (otherwise the one they used is nonsense, it's "mixing" to nothing) in series with a .047uF capacitor, in series with 40K to ground. (The effective impedance of the other channel they are not using).
Only in that case the channel they built behaves exactly as the tremolo/reverb channel of, say. a Twin or some of its brothers.
If they don't add the network I mentioned, the preamp output will be too high, almost 6dB higher than expected .
Not something that can be dismissed.
That's whay I applauded ChuckH's suggestion which would have solved both problems at the same time (the 50K padding and the 220k/220k padding) by judicious choice of split-load resistor values.

[jpfamps]

Maybe. Probably. Why not?
Don't know what sonic effect they expected to achieve, beyond the quasi-religious intention of "doing things just like Leo did ... whether I understand why or not".
In fact, if that *was* the intention, the network is incomplete: they need to add another 220K mixing resistor (otherwise the one they used is nonsense, it's "mixing" to nothing) in series with a .047uF capacitor, in series with 40K to ground. (The effective impedance of the other channel they are not using).
Only in that case the channel they built behaves exactly as the tremolo/reverb channel of, say. a Twin or some of its brothers.
If they don't add the network I mentioned, the preamp output will be too high, almost 6dB higher than expected .
Not something that can be dismissed.
That's whay I applauded ChuckH's suggestion which would have solved both problems at the same time (the 50K padding and the 220k/220k padding) by judicious choice of split-load resistor values.

Indeed.

When I've built this circuit without the other channel I've always attenuated the signal using a split load on the stage before the PI (which Fender did on the Brown Tolex Deluxe).

My guess is that once you have included the 50k resistor to ground, then the impedance of the 220k mixing resistor + the output impedance of the valve from the other channel which is in parallel with the 50k resistor won't have much effect.

Regardless, it's an inelegant solution.

[Chuck H]

(which Fender did on the Brown Tolex Deluxe)

Huh... Good catch. I got the idea to use split plate loads in my designs after building a 6g16 Vibroverb. Fender used them in a few other designs as well. I don't know how much audible difference there really is between the split plate load or other attenuation methods. Probably not much as it relates to guitar amps. But it is an eloquent circuit that is usually easy to lay out on an eyelet board.

[J M Fahey]

My guess is that once you have included the 50k resistor to ground, then the impedance of the 220k mixing resistor + the output impedance of the valve from the other channel which is in parallel with the 50k resistor won't have much effect.

Well, it depends on what point you are referring to.

If we are talking about the signal at the triode plate itself, yes, its plate load is the 100K resistor, in parallel with the 50K resistor straight to ground .
Let's not forget that it also has its own plate impedance (around 70K) in parallel with any load.
It all comes down to around 22K
Compared to that, whether it also sees in parallel either 220K+1M (next grid) or 220K+220K+40K (if there was another channel connected) , as you say, won't make much difference.
We are working with a 22K generator impedance, quite unusual in the Tube world.

Now, if we are interested in the signal that the PI receives, we start with the plate signal we had before, whatever it is, and we further attenuate it with the 220K/260K network, almost 6dB.
In fact, if earlier we had, say, 50V RMS available, now we'll have around 26 or 27V RMS, quite a difference.

[Rocketrob]

This is why I keep comming back to this forum. You ask a simple question and get an answer as well as some discusion as to why. Thanks for continually advancing my limited kn owledge of guitar amp circuits. Hopefully I can help someone else down the road. Rob.

[jpfamps]

Now, if we are interested in the signal that the PI receives, we start with the plate signal we had before, whatever it is, and we further attenuate it with the 220K/260K network, almost 6dB.
In fact, if earlier we had, say, 50V RMS available, now we'll have around 26 or 27V RMS, quite a difference.

Obviously there are two effects here, the loading effect of the 50k pot on the triode, and the voltage divider effect that attenuates the signal going into the PI. Possibly the designers of the California Dreamer were only aware of the former.

[J M Fahey]

Yes, I fully agree.
After posting I kept scratching my head: "why, oh, WHY had they put that network there, with no visible "good" function?"
Specially since loading an amp stage with such a low value resistor is bad practice.
If necessary (as when driving tone stacks or line outs), a cathode follower is added.
Then I remembered the tremolo pot. Although there was no tremolo here.
I didn't want to sound harsh, maybe I did involuntarily, but it makes me sad when I see somebody designing, in good faith of course, and taking parts from different schematics and joining them in a way that was not originally intended, and the desired effect is lost.
In this particular case, if, for example, that preamp overloaded in some specific way which was found agreeable to the ears of the maker (I have *nothing* to say on personal taste, that's personal (duh)), that might be achieved (as was correctly stated) just by putting a 50K resistor to ground, after a .1 cap , period.
If it was getting the correct signal level hitting the PI, sorry, that was not achieved.
I repeat, don't want to sound harsh, but on the contrary, want to help a budding designer to reach the sound he's trying to.