Adding a presence control to a Deluxe AB763 type circuit that didn't have one

[Dave H]

I love on line calculators, but I've never been convinced that the ampbooks PI calculator is accurate, I think it may be missing something because if you plug in a tail of just 100R, while the balance is skewed horribly, of course, you get an average gain of about 25. Then if you plug in a 100k tail you get about the same gain!?!

With a 12AX7 and the range of tail/bias resistor values usually used the gain is 25 to 30. Ideally the current in the tail is constant so the voltage across the bias and tail resistor doesn't change and therefore they can have no effect on the voltage gain (I think).

[Dave H]

I think I have a vague understanding of the voltage divider part, but not so much the whole phase inverter.

I don't have any idea how to size the cap. Im assuming that the cap would be sized based on the guitar's frequency curve, rather than the bass, although my Brother (who is a fine guitarist) says that lots of his friends use the Bassman as a guitar amp.

First to tackle was how to setup the voltage divider to work on the Deluxe circuit. I tried to think of it as two circuits: one with the pot fully CW and one CCW. The voltage divider is the same since the lower resistor is across the pot. But at one point, the FB signal is fed about directly into the cap, and the other rotation, the signal is first fed through the approx 5k resistor, then to the cap. One is like a RC in parallel, the other way is like the cap is almost shorted. I tried to draw it but didn't come out very good, will try again and post a picture to get around my poor wording.

Formulas would be great if you happen to find them in a file someplace, someday. I would love to learn to use Spice someday. (long list of things to learn!). Specific formulas are sometimes hard to find.

By itself the PI has a voltage gain of about 25 but it is inside the output stage negative feedback loop. It's the voltage divider which sets the gain of the output stage (from PI input to speaker)
When the Presence is fully CCW the cap is out of circuit and the (closed loop) gain is set by the divider ratio. When the Presence is fully CW the cap shorts the higher frequencies to ground so there is no HF negative feedback and the gain is the open loop gain of the output stage which is usually about 2 to 3 times the closed loop gain.

A cap value of 0.1u is used with a 5k pot because that value affects the 3kHz 'Presence' frequency range. If you change the resistor from 5k to 1k the cap has to be changed to 0.5u to keep the RC time constant and frequency response the same.

I've Spiced it up below. It shows how the frequency response changes as the Presence pot is rotated from CCW to CW. You can see that at low frequencies the gain is 17dB and at higher frequencies (3kHz) with the Presence fully CW (green trace) the gain is 25dB i.e. it has a maximum Presence boost of 8dB.

[pdf64]

I suspect that you may have used a resistive load in the sim?
With a real speaker inductive load, the degree of available presence boost may be a fair bit more than that.

[Dave H]

I suspect that you may have used a resistive load in the sim?
With a real speaker inductive load, the degree of available presence boost may be a fair bit more than that.

OK here you go

[eschertron]

With a 12AX7 and the range of tail/bias resistor values usually used the gain is 25 to 30. Ideally the current in the tail is constant so the voltage across the bias and tail resistor doesn't change and therefore they can have no effect on the voltage gain (I think).

I'd think so too, that the gain isn't affected by changes to B+ (in a reasonable range of voltages). However,the available HR is going to be different. The load line plot will have the same slope, just shifted R-L based on the B+ to tail voltage (I think ).

[J M Fahey]

Let´s take a shortcut.
Simulating is great, online calculators are the best human invention since the hot water bottle, sadly real World parts refuse to work like their models

My approach here is:
1) start with the "original" schematic , in this case the AB763 (or any other you like), built as God/Leo intended.
Meaning with the original 820/47 ohm NFB.

2) inject enough 1 kHz tone (won´t link to my favorite tone MP3 download site for fear of being accused of trying to sell sports jerseys cheap or something ) so you´ll have to google by yourselves.

3) rise volume until you have a nice round value at the output, say, 2V RMS , call it V1 .

4) open FB loop, in this case by lifting one end of the 820r resistor.
Output will rise by a definite amount, because now you turned the power amp into an open loop one.
Measure new value, call it V2 .

5) Now calculate feedback factor, or how large a gain margin you have to use for Presence (or Resonance) duties.
You can use a calculator fo V2/V1 , throw values into a simulator or just click wedding cakes cheap .

Or do what a fascist misogynist bigot retrograde guy would do: use pencil and paper

6) you will typically get between 6dB and 12dB .

Calculate the value of a cap which will have the same impedance as the smaller resistor at 400 to 800Hz, that will give you your Presence Boost.

7) since very low resistance pots are hard to next to impossible to get, consider using a Presence **switch**.

Presence is not subtle at all, you may have it OFF (which is nothing more than the amp in its original form) or ON ; you can back down a little the Treble control as a fine tuning one.
End result is good and quite practical.

[Justin Thomas]

Juan,

I tried clicking on your calculator links, but they didn't work. Can you upload them again?
Especially the second one, about cheap cake?

Justin

[Chuck H]

I love cake! How can I get cheap cake?

[Chuck H]

Let´s take a shortcut.
Simulating is great, online calculators are the best human invention since the hot water bottle, sadly real World parts refuse to work like their models

My approach here is:
1) start with the "original" schematic , in this case the AB763 (or any other you like), built as God/Leo intended.
Meaning with the original 820/47 ohm NFB.

2) inject enough 1 kHz tone (won´t link to my favorite tone MP3 download site for fear of being accused of trying to sell sports jerseys cheap or something ) so you´ll have to google by yourselves.

3) rise volume until you have a nice round value at the output, say, 2V RMS , call it V1 .

4) open FB loop, in this case by lifting one end of the 820r resistor.
Output will rise by a definite amount, because now you turned the power amp into an open loop one.
Measure new value, call it V2 .

5) Now calculate feedback factor, or how large a gain margin you have to use for Presence (or Resonance) duties.
You can use a calculator fo V2/V1 , throw values into a simulator or just click wedding cakes cheap .

Or do what a fascist misogynist bigot retrograde guy would do: use pencil and paper

6) you will typically get between 6dB and 12dB .

Calculate the value of a cap which will have the same impedance as the smaller resistor at 400 to 800Hz, that will give you your Presence Boost.

7) since very low resistance pots are hard to next to impossible to get, consider using a Presence **switch**.

Presence is not subtle at all, you may have it OFF (which is nothing more than the amp in its original form) or ON ; you can back down a little the Treble control as a fine tuning one.
End result is good and quite practical.

Eh.?. Low value pots aren't THAT hard to come by. I'm not talking 47R or anything. But 100R (might need to use wire wound 3W), 500R (Alpha makes them, I buy them from Mouser and use them all the time) or 1k (also Alpha/Mouser). Those values are low enough not to kludge any PI circuit operation significantly and cap values that correspond to these pot values aren't tricky either.

I'm not too hot on a presence "switch". It's one of those tonal refinements that feels better 'dialed in'. Probably because of the "feel" factor involved with NFB levels at ANY frequency.

JM2C

[Dave H]

Let´s take a shortcut.
Simulating is great, online calculators are the best human invention since the hot water bottle, sadly real World parts refuse to work like their models

Measuring the gain of a real amplifier is cheating
I can't do that when I'm sitting here with my feet up watching TV.

How about this?

The sim I posted was for a 6V6 amp driving an 8k : 8ohm OT.
LTP PI gain is about 25
The data sheet says 6V6 transconductance is 3.75mA/V. 6V6 voltage gain is therefore 3.75 * 8 = 30
OT impedance ratio is 1000 therefore voltage ratio is 1000^0.5 = 31.6
Open loop gain is- PI gain * 6V6 gain /OT ratio = 25 * 30 /31.6 = 23.7 = 27.5dB

27.5dB is nearly the same as the open loop gain in the sim. Is the calculation valid or did I just get lucky?

I repeated it for a similar EL84 amp and it also worked out to be the same as the sim.

[bob p]

Or do what a fascist misogynist bigot retrograde guy would do: use pencil and paper

I like pencil and paper. Does that make me a fascist misogynist bigot retrograde guy???

But seriously, both methods have their own merit.

This thread is posted in the Theory & Design sub-forum, so pencil and paper methods aren't out of line.

If your interest is limited to one particular amp then measure the open-loop gain and closed-loop gain and make changes as appropriate. that's the most efficient way to address the problem with an amp on the bench.

If your interest is in understanding the math, or gaining an understanding of a principle so that you can apply it universally across all of your future amp builds, then the most efficient way to address the problem is with pencil and paper; the pencil and paper method takes longer in the beginning but saves you bench time in the long run.

Different strokes for different folks.

[Sarah]

So from the discussion Sarah, do you think this is a simple or complex case?
Can you recommend any of those hooking manuals?
I find it interesting that for your first post you have chosen to link to an external website.

Hy g1 Sorry for not adding it: Please check this: https://robrobinette.com/How_The_AB7...verb_Works.htm

[Chuck H]

Again with the vaguely related post and no links to any weirdness!?! I'm having trouble figuring this out with so little incriminating evidence.

[g1]

Hy g1 Sorry for not adding it: Please check this: https://robrobinette.com/How_The_AB7...verb_Works.htm

Thanks Sarah. Your input is appreciated.

[Sarah]

I clicked it but didn't stick around long once I saw it was just another on line calculator.?. But the bots are becoming more strategic. They play it cool and then try to subtly slip in their wares. It never works here because:

A) No chicks post here (ok, I suspect one or two, but they never came off like a bot)
2) That calculator is only vaguely related to the subject by virtue of being electronics related. Clearly just an insertion maneuver.

Thanks for your comments