Bassman AA864 mods and musings

[Bieworm]

Hey y'all

A friend came to me yesterday asking me to build him an amp with particular specs. He is doing a project where he plays bass, but at the same time lead guitar on that bass. He's got some processing devices where he can put out 2 signals. 1 dry bass guitar and 1 pitched and processed signal that would give a guitar sound. Sorta kinda Royal Blood'ish.

Now, I was thinking of a 2 channel amp in the spirit of an AA864 bassman. Bass channel standard, but normal channel modded to a lead channel. I would like to implement the unused triode to add a gain stage. That and some component values tweaked. Here is the modded schematic as I would take as a starting point.
Both channels would be used simultaneously.
I aim for a lead channel that still is pretty beefy and good low end, not too bright and tight..but still

anything I overlooked? Thanks!

[dwinstonwood]

Cool idea!
Having just built the Valve Wizards medium gain amp, which I think sounds better than my Musing 40 and 15 (which are also modded AA864's), I would put the Bass channel's extra triode before the tone stack in the lead channel.

[Bieworm]

Why another extra triod? I don’t want a high gain plexi thing😉

[dwinstonwood]

Sorry, I wasn't being too clear (it was like 5:30am). 
I wasn't suggesting adding an extra triode, just using the AA864's unused bass channel triode in the lead channel like you plan to, but placing it in front of the tone stack, instead of after it.
I probably shouldn't post before coffee.

[Bieworm]

That is what I did. Check the unaltered schematic 😉

[sluckey]

The lead channel will likely be much louder than the bass channel. Probably need a voltage divider to knock the signal down a bit. A 470K on top of the gain control may be all that's needed.

[Deric]

anything I overlooked? Thanks!

Coupling cap between the first triode and the gain control on the Lead channel?

[Bieworm]

anything I overlooked? Thanks!

Coupling cap between the first triode and the gain control on the Lead channel?

Right! good call! thx

[stratomaster]

The lead channel will likely be much louder than the bass channel. Probably need a voltage divider to knock the signal down a bit. A 470K on top of the gain control may be all that's needed.

Or put it in series with the wiper of the treble pot to keep all the available saturation/gain from the added stage. 

[sluckey]

The lead channel will likely be much louder than the bass channel. Probably need a voltage divider to knock the signal down a bit. A 470K on top of the gain control may be all that's needed.

Or put it in series with the wiper of the treble pot to keep all the available saturation/gain from the added stage.

But he doesn't want a high gain amp.

[Bieworm]

The lead channel will likely be much louder than the bass channel. Probably need a voltage divider to knock the signal down a bit. A 470K on top of the gain control may be all that's needed.

Or put it in series with the wiper of the treble pot to keep all the available saturation/gain from the added stage.

That area will be a point where some tweaking will take place. I anticipate on putting a full on 470k/470k vontage divider before the gain pot. Anyway, I will be wiring the amp next week, so we’ll soon find out. I trust in Steve’s statement where the 470k in series with the coupling cap before the gain pot might be all it takes.

[stratomaster]

The lead channel will likely be much louder than the bass channel. Probably need a voltage divider to knock the signal down a bit. A 470K on top of the gain control may be all that's needed.

Or put it in series with the wiper of the treble pot to keep all the available saturation/gain from the added stage.

But he doesn't want a high gain amp.

I missed that part.  I'm that case, and if you're wanting a beefier sound, use the 250k pot that is currently your bass pot on the lead channel over to the gain role and use the 1M pot you currently have listed as the gain pot as the bass control.  Combined with the .68 cathode bypass cap and the reduced slope resistor you'll have less gain (more loss at the 250k gain pot) and a bigger, tighter lower end.

Finally, my personal preference is to minimize low end as early on in the circuit as possible.  I'd be inclined to put the 0.68 cap on the first triode and the 4.7 on the 2nd. 

Lots of voicing options. Play around and have fun. 

[Bieworm]

I'm a little puzzled with this one. I finished the amp, but the lead channel sounds a little too dark. I have to completely roll up the treble pot, close the bass pot and roll up the mid pot to get a mediocre decent brightness, which I am trying to get brighter. I changed a lot of stuff , but the amp keeps being too bassy and lack treble. any ideas where to go from here? Should I also put in a smaller cathode bypass cap on the 3rd amplification stage of the lead channel? See attached schematic where the red annotations are what I have done to add brightness. The bias is set at 65% dissipation of the 6L6WGC tubes

Also, there is some annoying hum which seems to decrease substantially when I put a 1k resistor on the PI tail, instead of the 100R. It's even gone when I disconnect that resistor, but the PI is no longer referred to ground when I do that. The hum disappears when the PI tube is pulled.
The ground goes to power amp ground, and not to preamp. Preamp is grounded at the input
Will post pics later today

[stratomaster]

Let's split this up

-PI noise-

Eliminate the lead channel as a possible culprit.  Lift that channel's 220k mixing resistor.  Build the PI per schematic, and listen for the hum.  If it's still there then the problem isn't your new lead channel.  By changing the tail resistor you're changing the negative feedback ratio.  Too much negative feedback will strangle the tone.

-Lead channel-

All the of voicing choices you've made except the coupling cap are emphasizing the full frequency. With the added gain stage this means muddy tone.

No bypass cap: full frequency amplification through the triode but at reduced gain vs a fully or partially bypassed triode circuit.  Also some circuits rely on large cathode bypass caps to reduce power supply or heater noise.  Leaving out your bypass caps may be adding noise.  Easy to test for, put a 10uF + cathode cap on all stages and see if the hum goes away.  If so, consider heater elevation and/or additional power supply filtering depending on the cause.

Reduced slope resistor:  tilts the balance of the tone stack to the bass side.

Reduced mids cap: Big emphasis on lows and low mids

Follow my advice above.  Try 0.68 bypass cap on first triode with 820 to 2.2k cathode resistor (play around to see what you like).  Between this and the smaller coupling cap you should have a nice bright tone driving the added gain stage and the tone stack component value changes you've made will make more sense. 

Put the 4.7uF on the 2nd gain stage for higher gain, or leave it off entirely.

Also, just noticed the placement of your 470k resistor off the gain pot. Right now it's acting as a grid stopper and not a voltage divider.  The corner frequency of the filter formed by that resistor and triode capacitance is taking quite a bit of high end too.  This resistor should be moved to the input lug of the gain pot (between the pot and the coupling cap).  Leave out the 100pF cap for now.  You may not need it.

[PRR]

> hum which seems to decrease substantially when I put a 1k resistor on the PI tail, instead of the 100R. It's even gone when I disconnect that resistor, but the PI is no longer referred to ground when I do that.

1K there in the NFB loop makes a major cut in gain. No wonder hum (and guitar?) is less. Disconnected, PI still returns to ground via 820r and OT secondary.

[Bieworm]

Let's split this up

-PI noise-

Eliminate the lead channel as a possible culprit.  Lift that channel's 220k mixing resistor.  Build the PI per schematic, and listen for the hum.  If it's still there then the problem isn't your new lead channel.  By changing the tail resistor you're changing the negative feedback ratio.  Too much negative feedback will strangle the tone.

-Lead channel-

All the of voicing choices you've made except the coupling cap are emphasizing the full frequency. With the added gain stage this means muddy tone.

No bypass cap: full frequency amplification through the triode but at reduced gain vs a fully or partially bypassed triode circuit.  Also some circuits rely on large cathode bypass caps to reduce power supply or heater noise.  Leaving out your bypass caps may be adding noise.  Easy to test for, put a 10uF + cathode cap on all stages and see if the hum goes away.  If so, consider heater elevation and/or additional power supply filtering depending on the cause.

Reduced slope resistor:  tilts the balance of the tone stack to the bass side.

Reduced mids cap: Big emphasis on lows and low mids

Follow my advice above.  Try 0.68 bypass cap on first triode with 820 to 2.2k cathode resistor (play around to see what you like).  Between this and the smaller coupling cap you should have a nice bright tone driving the added gain stage and the tone stack component value changes you've made will make more sense. 

Put the 4.7uF on the 2nd gain stage for higher gain, or leave it off entirely.

Also, just noticed the placement of your 470k resistor off the gain pot. Right now it's acting as a grid stopper and not a voltage divider.  The corner frequency of the filter formed by that resistor and triode capacitance is taking quite a bit of high end too.  This resistor should be moved to the input lug of the gain pot (between the pot and the coupling cap).  Leave out the 100pF cap for now.  You may not need it.

Thx! You gave me a lot to chew on. Will roll up my sleeves tonight😎

[DummyLoad]

Also, just noticed the placement of your 470k resistor off the gain pot. Right now it's acting as a grid stopper and not a voltage divider.  The corner frequency of the filter formed by that resistor and triode capacitance is taking quite a bit of high end too.  This resistor should be moved to the input lug of the gain pot (between the pot and the coupling cap).  Leave out the 100pF cap for now.  You may not need it.

Corner freq is 3.1KHz with JUST the 470K - 100pF into Rgrid is hi-pass with pole of about 26KHz; assuming around 70K Z-input. Scrap 470K grid stopper or replace with 68K or lower - Yes, a grid stopper does help there, however, 470K is grossly excessive - More is not better there, just enough is.

Also, you have a lot of active elements hanging off the one PS node - With all those high(er) gain elements of the Normal & Lead channels tied together, you'd probably experience unwanted signal bleeding through the PS - To minimize that as much as possible consider another PS R-C network node for just the Lead Channel elements. Adding a .1uF in parallel with that filter (47uF?) will keep HF bleed-through in check for Lead channel.

--Pete

[Bieworm]

Thx for chiming in Pete!

Here's the schematic I have so far in regard to the tweaking.
I'm going to try the extra node thingy for the lead channel. I hope I'm able to do this with the underboard connections on that D node.

[stratomaster]

I don't think you need the grid stopper you added to the LTPI.  What were you wanting to address with that?

[Bieworm]

I don't think you need the grid stopper you added to the LTPI.  What were you wanting to address with that?

Nothing in particular, but it can’t hurt

[stratomaster]

Nothing in particular, but it can’t hurt

It might. The LTPI is a rather complicated circuit despite the low parts count.  The impedance and effective capacitance of the circuit are not immediately obvious. It is not a basic triode gain stage where the grid stoppers are desired for stability and high freq roll off.  Check out the analysis by Valve Wizard and Aiken's white papers for two takes on the circuit.

If you're going to do tone shaping here it's usually accomplished by adding a cap in parallel to one of the plate resistors or a cap across the plates.  Not using added grid stoppers.  One other thing to look at is the 500pf input cap.  On the Blackface era combos this cap is 2x higher. It affects the low end roll off into the power amp.  If you're using this as a bass amp, then it might make sense to increase this cap and rely on the bass cutting you did in the lead channel to keep that channel from getting too boomy.

Adding/changing components in an amp circuit with no specific goal in mind will get you into trouble. You should have a plan for the changes you make and at least a rudimentary understanding of the underlying circuit to make informed choices on your modifications. 

[Bieworm]

Nothing in particular, but it can’t hurt

It might. The LTPI is a rather complicated circuit despite the low parts count.  The impedance and effective capacitance of the circuit are not immediately obvious. It is not a basic triode gain stage where the grid stoppers are desired for stability and high freq roll off.  Check out the analysis by Valve Wizard and Aiken's white papers for two takes on the circuit.

If you're going to do tone shaping here it's usually accomplished by adding a cap in parallel to one of the plate resistors or a cap across the plates.  Not using added grid stoppers.  One other thing to look at is the 500pf input cap.  On the Blackface era combos this cap is 2x higher. It affects the low end roll off into the power amp.  If you're using this as a bass amp, then it might make sense to increase this cap and rely on the bass cutting you did in the lead channel to keep that channel from getting too boomy.

Adding/changing components in an amp circuit with no specific goal in mind will get you into trouble. You should have a plan for the changes you make and at least a rudimentary understanding of the underlying circuit to make informed choices on your modifications.

I did it to prevent any possible occurring oscillation or blocking. Merlin advises to put a grid stopper on every tube which is fed by a long wire. In the HiFi world it is normal, but us guitar players haven't adapted to that idea (according to Merlin)

"Another reason to use grid stoppers is to prevent local parasitic oscillation (which may be ultrasonic and fiendishly difficult to diagnose). This tends not to be such a problem with low-gm triodes like the ECC83 / 12AX7, whereas high-gm valves like the ECC88 / 6DJ8 will oscillate quite easily, especially if there is a long wire leading to the grid. A grid stopper of a few hundred ohms should really be used on every valve, just to be safe. This is normal practice in hi-fi amps, but a lot of guitar amp builders still haven't cottoned on."