Finally

[Mike McLane]

Just completed my own blackface Fender preamp as well as one of Sushi's DIY'ers. Want to pair them with the Class D amp when/if PPCB ever gets one put together. I noticed the Finally in your pedal offerings. For those of us wanting to optimally condition the output of our tube preamps to play nice with a SS power amps is this the way to go?

 

[vigilante398]

Probably not. Finally is just a Space Heater with a D.I. transformer on the output, you only need that if you're looking at going into something that wants a balanced XLR signal. The 1/4" output of Finally is exactly the same circuit as Space Heater.

If you want to optimally condition the output of your tube preamp to play nice with a solid-state power amp, the best way to go is to hit it with a nice hot output preamp and then have a nice low-output-impedance buffer between the preamp and the power amp to make sure the power amp's input isn't loading down your preamp's output. I used to offer a standalone buffer pedal called Vampire Slayer, which was based on the Effectrode Glass A, but used a 12AU7 instead of their miniature tube. I don't offer it anymore and I don't plan on designing it as a DIY project, but I can dig through my files and find the schematic, I would be happy to share them if anyone wanted to take a look.

 

[Mike McLane]

Luv to see the schematic.

 

[vigilante398]

Sure thing. This is it, not much to it.

 

[Mike McLane]

Simple is good. Thanks!

 

[Mike McLane]

Do you require both sides of the triode to make this work properly? The cathode followers I've seen in the past only involve one side of the tube. These are in parallel so I wonder if you used both sides simply because you had both available or is there something more substantive going on?

 

[vigilante398]

Running tube stages in parallel is a hi-fi trick for better noise performance. Effectrode also uses it in their LA-1A compressor, this is what they say about it:

"To that end we based our Leveling Amplifier on a parallel tube plate design—the type of tube circuitry only found in audiophile tube phono-stage preamps for high-end turntables. Parallel tube design reduces the LA-1A’s noise floor to vanishing point. In practical terms this means best possible signal to noise ratio for your guitar. It’s an expensive way of doing things, but it does yield stellar sound quality, without the compromise of noise gates or ‘cutting’ certain frequencies using equalisation. The parallel tube design has never been built into a stompbox or even a professional studio leveling amplifier/compressor before, making the LA-1A a truly special compressor."

Effectrode uses parallel stages in a handful of pedals, and I really like the results so I do it where it makes sense to do it as well. You could of course do a cathode follower with a single stage, but the parallel stages help both with the noise floor and with lowering the output impedance of the circuit.

 

[Mike McLane]

Why the "dual" 47K's & 680n's? Could those just as easily been single components? Or perhaps there was a necessity to obtain fairly exact values (like 23.5K) that couldn't be had in a single component?

 

[vigilante398]

I'm not really sure, they could have easily just used single components. This is how Effectrode did it and it works well, but I fully expect using a single resistor and single capacitor there instead of two in parallel would work just fine.

 

[Feral Feline]

Parallel Capacitors sum, so the 680n and 680n together is 1µ36.

So maybe Effectrode is achieving its exact target value by running parallel caps;
maybe they use that cap's value in other circuits and buying in bulk keeps the economies-of-scale balanced...

At any rate, C2 + C3 is straightforward.


What's interesting to me is that parallel resistors do NOT sum...

Do the math:


Rtotal = R5×R6/(R5+R6)
23k5= 47k x 47k / (47k + 47k)

OR

use an online calculator = 23k5


Fine, but R3 (1M) is supplying some voltage and R4 (1k) with R5 (47k) has formed a voltage divider that feeds R6 (47k)...

So while R5 and R6 are parallel to each other going to ground, R3 and R4 are messing with the formula in ways I can't fathom. (I'm mathematically "illiterate")...


For power filtering a 100µ cap and 100n cap are sometimes used in parallel, why not just one big 1µ1 or 1µ2? I've read that the smaller cap handles small fast transients better (or something like that).

In the case of C2 + C3 above...
... I'm guessing that while you could just shove in one huge capacitor in place of C2 + C3;

I don't think you can just use a single resistor in place of R5 + R6, 'cause bearing in mind R3 and R4 — that would change the voltage being fed to C2+C3, no?

 

[fig]

Parallel Capacitors sum, so the 680n and 680n together is 1µ36.

So maybe Effectrode is achieving its exact target value by running parallel caps;
maybe they use that cap's value in other circuits and buying in bulk keeps the economies-of-scale balanced...

At any rate, C2 + C3 is straightforward.


What's interesting to me is that parallel resistors do NOT sum...

Do the math:

View attachment 39616
Rtotal = R5×R6/(R5+R6)
23k5= 47k x 47k / (47k + 47k)

OR

use an online calculator = 23k5


Fine, but R3 (1M) is supplying some voltage and R4 (1k) with R5 (47k) has formed a voltage divider that feeds R6 (47k)...

So while R5 and R6 are parallel to each other going to ground, R3 and R4 are messing with the formula in ways I can't fathom. (I'm mathematically "illiterate")...


For power filtering a 100µ cap and 100n cap are sometimes used in parallel, why not just one big 1µ1 or 1µ2? I've read that the smaller cap handles small fast transients better (or something like that).

In the case of C2 + C3 above...
... I'm guessing that while you could just shove in one huge capacitor in place of C2 + C3;

I don't think you can just use a single resistor in place of R5 + R6, 'cause bearing in mind R3 and R4 — that would change the voltage being fed to C2+C3, no?

...and electrons can flow in more than a single direction.

 

[Mike McLane]

OK. . .OK. . .OK. . . then how about this? . . . .

2" x .4" Use the Sushi HV PSU to heat up the tube? No plate resistors? The caps don't need to be HV? Or do the 680n's need to be HV?

 

[Mike McLane]

On second thought. . . this is about 1" x 1". Probably could fit in-line with the HV PSU on the same side of the 125B box and give more clearance for the tube.

 

[vigilante398]

OK. . .OK. . .OK. . . then how about this? . . . .
View attachment 39632
2" x .4" Use the Sushi HV PSU to heat up the tube? No plate resistors? The caps don't need to be HV? Or do the 680n's need to be HV?

The caps don't need to be HV, they won't see more than single-digit volts.

 

[giovanni]

Parallel Capacitors sum, so the 680n and 680n together is 1µ36.

So maybe Effectrode is achieving its exact target value by running parallel caps;
maybe they use that cap's value in other circuits and buying in bulk keeps the economies-of-scale balanced...

At any rate, C2 + C3 is straightforward.


What's interesting to me is that parallel resistors do NOT sum...

Do the math:

View attachment 39616
Rtotal = R5×R6/(R5+R6)
23k5= 47k x 47k / (47k + 47k)

OR

use an online calculator = 23k5


Fine, but R3 (1M) is supplying some voltage and R4 (1k) with R5 (47k) has formed a voltage divider that feeds R6 (47k)...

So while R5 and R6 are parallel to each other going to ground, R3 and R4 are messing with the formula in ways I can't fathom. (I'm mathematically "illiterate")...


For power filtering a 100µ cap and 100n cap are sometimes used in parallel, why not just one big 1µ1 or 1µ2? I've read that the smaller cap handles small fast transients better (or something like that).

In the case of C2 + C3 above...
... I'm guessing that while you could just shove in one huge capacitor in place of C2 + C3;

I don't think you can just use a single resistor in place of R5 + R6, 'cause bearing in mind R3 and R4 — that would change the voltage being fed to C2+C3, no?

R5 and R6 are equivalent to a single 23.5K resistor. It doesn’t matter what else is going on in the circuit. What matters is that the two resistors are in parallel. IIRC this follows from the principle of superposition.