New build 5F1 & AA764 champ, 2 in 1

[donluca]

Alright, let me see if I'm missing something from your proposed mod:

When in 5F1 mode (tone stack bypassed) all is good: we have the 0.1uF cap at V1A plate output blocking DC so we're set.

When in AA764 and we get into the tone stack things get a bit more complicated: The low pass filter, AFAIK, is unaltered since we still have a variable RC filter made by the 250pF cap along with the treble pot.

Now we don't have the 0.1uF cap anymore so I'm not entirely sure what the bass pot does... it's connected as a variable resistor which, I think, acts more like a control for the mid frequencies.
With the pot completely open (that is, no resistance between its lugs) it is sending the signal straight down to the last part of the tone stack which is generally used to control mid frequencies (but here we have it fixed with a 15k resistor). This means that we're going to get a bass response which is equal to the one we would have with the tone stack removed.

What's interesting is when we actually turn the bass pot, because as it increases resistance, less signal is let through, thus scooping the mids.

I'm going to give this a go although it may stray a bit too far from what I'd like to achieve with this project.
The VHT is entirely on an eyelet board, so making changes is a breeze.

Meanwhile, please correct me if I misunderstood what's happening in the tone stack with the 0.1uF cap removed.

EDIT: also, what's the point of having the 5M resistor now?

[Chuck H]

Actually the tone stack as modified works very much as original. The purpose of the bass and mid caps in a stock tone stack is to 1) Decouple the circuit from DC. The treble cap also does this, but it is not the only one. And 2) they set the pass frequency for the LF and the cut frequency for the mids. The LF pass frequency is the sum of both caps. The mid cut frequency is dependent on just the mid cap value. Using the diagram I posted, the 100n cap is now decoupling DC for the whole tone stack AND setting the LF pass frequency. The mid cap is still passing signal to ground relative to it's value and the remaining signal with the LF is already decoupled so THAT capacitor is no longer necessary. Notice in the diagram that the mid cap is still adjustable solely by the mid pot (or in this case NOT adjustable, but determined by a fixed resistance). Adjustments to the bass pot as shown in the modification will not affect the mids.

I think your mistake is that you are looking at the capacitors strictly as passing signal to the following circuit. The way the standard tone stack works is by selectively DELETING signal at frequency by passing it to ground, removing it from the signal path. This is why there is so great a signal loss by implementing this sort of tone stack.

The purpose of the 5M resistor is to provide an association to 0VDC to sw1a. Without it the switch would make a pop noise. Being decoupled from any DC with the new arrangement would reduce this considerably, but not entirely. You could also hang a 10M resistor to ground from the center lug or tone stack lug of sw1a for the same result since sw2a already has a 0VDC reference on those lugs through the volume pot and the tone stack pots.

EDIT: It occurs to me now that with the bass cap no longer between the slope resistor and the bass pot that IS a 0VDC reference for sw1a. So you're right. You shouldn't need the 5meg resistor anymore. It still might make a tiny noise because there will be a moment of open endedness during switching and sw1a center lug will not have a 0VDC reference, but it also doen't have a path to alter it's charge. So you can skip the extra resistor, the noise will be small enough to ignore I think.

It's difficult to describe how a tone stack works. At least for me. But try to look at it this way. In the mod I posted the bass cap has not been removed from the circuit. It has been relocated.

EDIT 2: Oh crap!!! There was a terrible error in the modification I drew up. I had the coupling capacitor on the wrong side of the plate load resistor. I've corrected the attachment. This was probably where the confusion came from

[donluca]

I think your mistake is that you are looking at the capacitors strictly as passing signal to the following circuit. The way the standard tone stack works is by selectively DELETING signal at frequency by passing it to ground, removing it from the signal path. This is why there is so great a signal loss by implementing this sort of tone stack.

Your explanation is actually spot-on.
My mistake is considering the tone stack as low-pass and high-pass filters working in series, but this is obviously not the case.

I'll definitely do some more research on tone stacks, but your explanation is pretty clear and makes perfect sense to me.

[Enzo]

With the pot completely open (that is, no resistance between its lugs)

I am sure you meant to say with infinite resistance between its legs.

[donluca]

I am sure you meant to say with infinite resistance between its legs.

You're making second guessing myself, but that might be due to my lack of knowledge of how the tone stack works, today I'll just break it down using Kirchhoff to better understand what's going on. Yesterday I gave a look at Rob Robinette explanation but it confused me even more, I need to do it my own way.

Due to the passive nature of the tone stack, it works in a subtractive way (that is, it rolls off high and low frequency, it can't make more as that would take an active circuit), so when we have a pot wired as a variable resistor, which is the case of the bass pot in the schematic, when I say "wide open" I mean that it offers no resistance through its two lugs (middle and one of the outer, depending if the pot has been turned fully clockwise or counter-clockwise).

[Chuck H]

Not sure if this will be anything like Robinette's description (I didn't look ). But here is a (very) simplified description of the standard TMB tone stack we all know. And it may not seem like a simple description either. But read it twice and most will sink in. Also download Duncan TSC. It's a great tool and playing with it will give you great insight into how the tone stack works.

http://www.duncanamps.com/tsc/

EDIT: Opening the attachment at the bottom will make it easier to read the text.

[donluca]

Nevermind my last post, I'm still figuring it out.

[Chuck H]

Ah! I think I see now where the confusion might be. Are you interpreting the physically lowest cap in the diagram as the bass cap? In order from top to bottom on the diagram it is treble, bass, middle. If you look at the operation description I posted it makes this clear.

I'm working something up for you now to show both incarnations (stock and modified) in operation.

[donluca]

Ah! I think I see now where the confusion might be. Are you interpreting the physically lowest cap in the diagram as the bass cap? In order from top to bottom on the diagram it is treble, bass, middle. If you look at the operation description I posted it makes this clear.

I'm working something up for you now to show both incarnations (stock and modified) in operation.

Exactly, sorry, I deleted my post to avoid confusion among other members who might be reading this thread because I figured that out and calculating the filters again from scratch.

EDIT: I give up, because I've done again the calculation but forgot that the treble pot, bass variable resistor and the mid fixed resistors are all in series and the 100k slope resistor is in parallel with the treble pot and in series with the bass VR and mid resistor.

Need to take a break.
Hi-Fi amps tone stacks (when present) are so much more straightforward, sigh.

Here all the controls are interactive with one another and thank god almighty at least the mid pot is a fixed resistor in the Champ schematics. How the hell does it interact with the others? Doesn't it work just like a general "boost"?
I might be talking out of my ass at this moment, I don't know why I'm having so much issues understanding this circuit.

[Chuck H]

It's because of all the variables and interactions, with all three frequency bands in the same passive circuit, that it's much easier to model this sort of thing with Spice.

This may not be your tone stack values, but it demonstrates that it works and there isn't any significant difference in performance.

[donluca]

Nice!

I mean, I'd love to understand why the results are similar, but I guess that for now that'll do.
Ironically, I've been using LTSpice for years (albeit it's been a while since I last used it), don't know why sometimes I still stubbornly decide to do things with pen and paper, the old fashioned way.

I'll definitely import the tonestack on LTSpice, put the correct values in and then play a bit with them to make your solution behave as close as possible to "the real thing" and then post results here.
I've found other threads discussing the topic of taking a blackface AA764 and make a mod to switch back and forth from the AA764 circuit to a 5F1 with the tonestack bypassed.

[Chuck H]

I'll definitely import the tonestack on LTSpice, put the correct values in and then play a bit with them to make your solution behave as close as possible to "the real thing"...

That's something I've done for myself on another project. The trick is finding a way to get exactly equivalent performance with standard values! And then there's the reality that the amp is going to be modified anyway and the existing values may no longer be ideal. And that they aren't going to be the same as any vintage amps anyway. Not to mention pot tapers. I'd probably just roll with the modification for now without refinement and then tweak values by ear and tapers for aesthetics (bass on 5 rather than 7 for example). JM2C from a guy that's done a lot of this.

[donluca]

Today I was wondering... how accurate are Fender's schematics regarding voltages and component values?

I've adjusted the B+ voltage and kept plate and cathode resistors as per AA764 schematic and on the V1A cathode I get 1.385V (instead of 1.8V) and 1.42V on V2A (instead of 1.7V).

All the components in the circuit are the same (1.5k on cathodes bypassed by 22uf on first half of the 12AX7 and 2.2uF on the second half), same B+ voltage.

In order to get in the 1.7V-1.8V ballpark as per schematic I should raise the cathode resistors to ~2k each.

For reference:



Does anyone here have an original Fender Blackface/Silverface Champ AA764?
I know for sure that the voltage reported are lower than they should be now due to higher wall voltages, but still...

[Helmholtz]

Some original Fender schematics show a note stating voltages within +/-20%. So your cathodes voltages should roughly be within spec.
Cathode voltage depends on tube variation (Gm), plate voltage and actual resistor value.

[donluca]

They are at the lower extremes:

1.8V can go from 1.44V up to 2.16V (I'm at 1.385V)
1.7V can go from 1.36V up to 2.04V (I'm at 1.42V)

Maybe I should bump up a bit the B+ in order to get those guys closer to the intended values.

Not even sure if it would be worth it, honestly, if the ampbook's 12AX7 calculator is correct ( https://www.ampbooks.com/mobile/vacuum-tubes/12AX7/ ) the difference in headroom should be minimal.

EDIT: and of course, then there's the Gm variable which I've not considered in my calculation. My modded VHT has its original chinese tubes which I've read around they're not particularly good.

I'll probably scoop some nice NOS Soviet tubes like the 6n2P-EV and do a very minor rewire to have them working instead of spending big bucks on other 12AX7s.