AA764 Vibro Champ build (power supply and grounding)

[MichelB]

Good day,

After playing and modding my Korean made Blues Junior clone, its time to build my own amp.

After trying out a 68 custom vibro champ reverb, I’ve realized this might be my dream amp. Then I got a good deal on a SolandFa vibro champ chassis and a Hammond 290AEX PT (I’m in europe on 240V).

All items are ordered and I’ve decided to go with a JJ 40/20/20/20 cap can solution. Later on a lot reading gave some concerns about the common ground on all those caps inside (more on that)

While I know that a SF champ will use 40/20/20 caps, I think I will remain on the 20uF cap for the main B+ node (for sound reasons). Due to better ripple filtering I will add another filter stage up front, with dropping resistors to get to old spec B+ voltage of 355V.

My proposed power supply will then look as follows:

NOS RCA 5Y3GT - 20uF - 470R - 20uF(feeding OT) - 1kR - 40F (feeding 6V6 screens and vibro 12AX7) - 10kR - 20uF(feeding preamp 12AX7)

This will use the existing cap can and provide better filtering than stock, without putting too much strain on the rectifier tube and keeping the tonal qualities of the original intact. Please comment on my logic if you like?

Then the questions about grounding scheme’s. I read a lot about the disadvantages of a cap can being the common ground of all caps, thus being unable to separate power supply and preamp grounds.

One solution is to simply not use the last cap in the cap can and place another on on terminal strips between the circuit board and the pots, creating a nice starting point for the preamp ground bus (ending at input jack).

The other solution which sound completely logical with a star grounding scheme is to isolate the input jacks, and connect the preamp bus to the cap cans common ground. This makes more sense since it doesn’t force the grounds from the complete circuit to travel via the bus, via the input jacks into the chassis (meaning power tube grounds traveling via preamp and input jack creating noise)

My current chassis contains a brass ground plate and with some imagination I would like to use this into a very good grounding scheme. First of all this plate should be isolated from the chassis (including isolating all pots and jacks as well). The physically close to the cap can I need to connect the plate (acting as ground plane) to the cap can common ground (which is then tied to the PT ct and PT main ground).

On this ground plane It’s probably a good idea to make small star grounds connecting cathode resistor and related grid leak resitors etc.

Can your combined wisdom be shared regarding this proposed layout?

Or is this all theoretical and insignificant considering I use center tapped and twister heater filaments and the amp will likely be quiet anyway.

Regards,

Michel

PS: the remaining build will be documented while I go, currently I’m planning my layout first ;-)

 

[tubedude]

I'd use a separate buss for preamp grounds and ditch the brass plate. They are problematic over time and way too much work to isolate.

 

[Phrygian77]

You are over thinking this. The preamp doesn't have enough gain to make doing anything elaborate worthwhile. Run a bus type ground for the preamp and don't worry about it.

There's no reason to isolate the jacks, and certainly not the pots, from the chassis. You would only want to isolate the jacks if you're planning to completely lift the grounds from the chassis, and then use an isolation block to couple it to the chassis.

 

[schmee]

Keep the preamp/inputs a separate ground from the power section for low noise. Cap can grounding works fine as long as it's down at the power end.

 

[MichelB]

There's no reason to isolate the jacks, and certainly not the pots, from the chassis. You would only want to isolate the jacks if you're planning to completely lift the grounds from the chassis, and then use an isolation block to couple it to the chassis.

That's exactly what I'm planning to do, isolating the brass ground plate from the chassis, making one huge brass ground plane, isolated from chassis, connected to cap can ground.

 

[andrewRneumann]

NOS RCA 5Y3GT - 20uF - 470R - 20uF(feeding OT) - 1kR - 40F (feeding 6V6 screens and vibro 12AX7) - 10kR - 20uF(feeding preamp 12AX7)

470R before the OT will lower the idle plate voltage, but it will also cause excessive sag on a Class AB amp when driven hard. Is the AA764 Class AB?

 

[Phrygian77]

@MichelB if you're using pots with a 1/4" (or 6.5mm) bushing length, they won't be long enough. Ditch the brass plate, and use a copper bus wire.

 

[Lowerleftcoast]

My proposed power supply

The 5Y3 can handle the 40uF reservoir cap. Imo, use all of the caps 40-20-20-20 kinda like your plan but start with 40uF. I am assuming it will provide a lower noise floor. (You can try it with a 20uf as a reservoir cap and report if you can discern a difference.)

questions about grounding scheme’s.

Do not use the brass ground plate. There is no advantage.
Either use a one or two point ground scheme.
The choice to use the chassis as a conductor is a choice. The design and gain of the Champ-ish layout seems to keep the amp quiet even when the current flows in the chassis. Iow, the circuit does not seem to pick up and amplify hum.

is this all theoretical and insignificant considering I use center tapped and twister heater filaments and the amp will likely be quiet anyway.

The *power filtering* ground scheme can be considered *theoretical and insignificant* due to the layout design and low gain structure of the Champ.

The heater CT and filament supply is a source of 50Hz hum so it is not related to the 100Hz hum of the power filtering. They are two different sources of hum.

Other stuff:
A fancy ground bus is not a requirement, if it is easier to just use lengths of wire, it is OK to do so.

 

[MichelB]

470R before the OT will lower the idle plate voltage, but it will also cause excessive sag on a Class AB amp when driven hard. Is the AA764 Class AB?

AA764 is single ended one tube 6V6. Sag should not be a problem.

 

[MichelB]

Do not use the brass ground plate. There is no advantage.
Either use a one or two point ground scheme.
The choice to use the chassis as a conductor is a choice. The design and gain of the Champ-ish layout seems to keep the amp quiet even when the current flows in the chassis. Iow, the circuit does not seem to pick up and amplify hu

Seems fair, running ground bus to cap can common ground with isolated input sounds still seems better grounding strategy than power caps to PT grounding and circuit grounding via input and chassis?


And I know the 5Y3 should be able to handle the 40uF, it's just that the potential starving of the 6V6 plates due to the 20uF cap is sort of the idea. The 40uF supposedly sounds cleaner due to bigger reservoir cap. But that will be tried whenever the amp is built.

 

[Linkjr]

I built one of these with a similar setup as I'm in the UK with 240v. I used a Hammond Aex291 PT. I ended up putting a dropping resistor of 470ohm 20w on the PT centre tap, before this it was very slightly red plating 6v6s but sounded great. Putting that resistor there rather than from the 5y3 to board seemed to stop it sounding a bit sterile

 

[King Fan]

The guys are right. Keep it simple. "Starving the plates" is the idea? Why? Do Fender run 20uF on the '68 reissue you liked? 40uF is actually what Fender did on the vintage version -- at least in my BF Vibro Champ and on most other originals you see pictured online. Very few VCs were built to the original VC schematic, which, knowing Fender, was probably pre-production; most were built to the Bronco schematic at 40.

We'd have to see a diagram of your proposed single ground/isolated input scheme. To be fair, there might be ways to make it work, even grounding it at the cap end. But it might not work well, and as noted by @Lowerleftcoast , in these small amps, it won't be better. Also he's right a 'ground plane' is not wanted in amplifiers. BTW, I haven't heard of isolating pots from chassis unless you are somehow soldering their ground lugs to their cases and then linking their cases to a ground bus, which would be doubly unwise. On all these ideas, Google 'Valve Wizard ground.'

A 470 dropping resistor is OK *if* you need it to drop B+ on your 6V6, but will you? Which secondary will you use, 325 or 275? What is your target B+? Again, 'real' VibroChamps ran the AB764 B+ around 410, not the AA version at 355. But to be fair, 410-420 can be hard on modern 6V6s, and some of us actually like the sound a little 'browner' with B+ in the 300s. If you use a dropping resistor, the exact value needed may vary somewhat. And when needed in these amps, it's usually just inserted between the rectifier and the cap can.

 

[MichelB]

Hi King Fan,

Lots of valid questions most of which have crossed my mind. Just to clarify:

I like cleanish sounds with a bit of hair and even some dirt (breaking up on double stops) and I play in my living room, so no huge levels required.

I like the dirtyness of a 5F1, but like BF scooped sounds.

So that is my answer why I propose to lower B+, i dont need the headroom. Also feeding the OT and plated from 20uF rather than 40uF is because of that. However, just rewiring the cap can gives me options to tailor the sound later on.

Next is the use of an additional power filter stage. This comes from the luxury of having one extra and tje wish to get ripple free DC and a quiet SET amp.

Then the grounding.......

Easiest will be to build as Leo designed. Using the brass ground plate, connected to chassis via pots, jacks and random connections between brass and chassis. All with cap can ground direct to chassis.

Knowing about theory somehow forces me to try and upgrade.........

While I like well implemented bus rails, I'm also a bit lazy and like the concept of easy soldering to the brass plate. By just using some shoulder washers between ground plate and pots, some isolation tape between plate and chassis and shoulder washers between chassis and input jacks, I think I've made a floating circuit ground that I can connect to the star wired cap can common ground. Theoretically better then Leo's design and really not much more work. The twister heaters with artificial center tap will be more work then these changes.

For the record, the pots will not be isolated from chassis, just from floating circuit ground.

Then again, I might be overthinking it.

Michel

 

[peteb]

NOS RCA 5Y3GT - 20uF - 470R - 20uF(feeding OT) - 1kR - 40F (feeding 6V6 screens and vibro 12AX7) - 10kR - 20uF(feeding preamp 12AX7)

And I know the 5Y3 should be able to handle the 40uF, it's just that the potential starving of the 6V6 plates due to the 20uF cap is sort of the idea. The 40uF supposedly sounds cleaner due to bigger reservoir cap. But that will be tried whenever the amp is built.

I doubt the plate, there is only one, is going to be starved any way you do it, but I did pick up on your logic, to supply the plate with the biggest cap. So why are you not putting the biggest cap on the OT / plate?

 

[MichelB]

I doubt the plate, there is only one, is going to be starved any way you do it, but I did pick up on your logic, to supply the plate with the biggest cap. So why are not you putting the largest cap on the OT / plate?

I thought about that and there are two reasons.

1: The cap feeding the OT / Plate is the one that I think has the most effect on sound and both 5F1 and BF champs specify a cap on the lower scale of things.

2: The high pass filter created by R and C is important remove riple and 120hz hum. On the node followed by 10k resistor that one has the highest multiplier by adding the double amount of capaticitance. Leading to a cuttoff freq of 0.4 Hz. This should guarantee riplple free and hum free DC to the preamp.

But again the last part is probably theoretical ravings....

I think I'll just build it as 20/40/20/20 order and try both options whenever the amp is finished.

 

[peteb]

1: The cap feeding the OT / Plate is the one that I think has the most effect on sound and both 5F1 and BF champs specify a cap on the lower scale of things.

The 5F1 specifies the biggest cap on the OT / plate.

same with BF and SF. When they use the 40 uF cap it is on the OT / plate. It is only specified on the later schematics like the bronco schematic, but when it was implemented on the earlier units, it was done as it is shown on the later schematics.

 

[Lowerleftcoast]

Starting at post #61, this thread addresses the reservoir cap.

https://www.tdpri.com/threads/is-th...e-ended-tube-amp.1089145/page-4#post-11120535

 

[King Fan]

LOL, @MichelB , I'm no expert. It sounds like you know a lot of theory, starved plates, star grounding, etc. Are you experienced in application, like building these amps? Have you read that Merlin grounding article? It's just I just haven't seen many (OK, any) amps built with an isolated brass plate, a smaller first than second filter, and so on. (I do appreciate your desire for lower B+, but the right tap on your PT might get you there all by itself.)

You asked for comments on your logic. Unless you can share some of the evidence (amp articles, amp examples) for your plan, it's hard for us to tell if your logic is sound, or just theory working in imagination.

 

[MichelB]

Starting at post #61, this thread addresses the reservoir cap.

https://www.tdpri.com/threads/is-there-an-“acceptable”-noise-floor-for-a-single-ended-tube-amp.1089145/page-4#post-11120535

Thanks for the read. It looks like he's only upping the first filter capacitance though, it does work ;-)

 

[MichelB]

@ King Fan,

The theories from Merlin and Aiken are familiar to me, as is for instance Rob Robinette's site.

To be fair I don't have much experience building tube amps besides fixing and adepting a flawed Blues junior clone with PCB layout issues. After performing all BillM mods I've learned that doing things because internet says so isn't always wise ;-).

So for the next build I really want to understand the theory behind what I propose to add to a venerable schematic like this (if any).

And yes I know that my proposed layout is uncommon, that's exactly why I asked the combined wisdom here.

The theory comes basically from www.valvewizard.co.uk/smoothing.html regarding calculating around a smooth and quiet powersupply. That's why I propose an additional filter stage with 470k dropping resistor (2 effects, one action)

Merlins grounding scheme is tough reading though. Some ideas on the isolated from chassis came from that part.