Power section & bias checks

[Wonder2121]

Deciding to Build Fender style amp in donor chassis which already has transformers and multi can cap installed.
Robrobinette RR763V Blackvibe 6V6 https://robrobinette.com/RR763_BLACKVIBE.htm#BLACKVIBE_6V6 seems to fit the bill with my Transformers.



Rectifier
To gain more voltage and reduce sag, replacing GZ34 with Full wave to hit 400V from my xformer. Can anyone confirm the rectifer looks correct with 2 diodes in series, I've seen 1 diode and even 3 in line.
Will I get 400V?

Choke
4H 50mA guessing this is Ok...The 32/32 can-cap wired in parallel before the choke, will this cause any issues? Reason I'm not splitting can-cap between plate/screens cus wanting stiff, clean sounding quiet amp..

Bias
Implementing dual bias adjustment pots with safety resistors. I have no way calculating voltages, just working off similar schematic I've found online. It would be a real help if someone could give the A/OK green light on this...The 1M saftety resistors are worrying me a little...

Presence
From what I've learned, feel confident my presence circuits good, unless anyone can spot any faults..

Other than that, if anyone sees anything else of concern your post is welcomed.

[tubeswell]

290 x 1.4142 = 410

50mA choke will work

Re bias - Some of the figures on the schematic are a little hard to read.

[Wonder2121]

Hey tubeswell thanks for input..
Heres the power & Bias section

[tubeswell]

Try that and see. A good rule of thumb for developing a fixed bias voltage for octal output tubes is that you want the voltage to be about 1/10th of the B+ voltage (albeit the bias voltage has to be a negative voltage). So in this case aim for something that’s -40v at the output tube grids with an adjustment range plus or minus 10v.
Adjust the (68k) supply resistor if need be to get the range right once it’s all up and running. (I assume you know you want to measure output tube current to work out Pdiss?)


(And don’t plug the output tubes in until you can measure -ve voltage at the grid pins.)

[pdf64]

The diode type for the HT rectifier isn't shown, so it's not feasible to work out if series pairs or triples are needed :)
The diodes will be exposed to the peak of the full winding voltage, so 290x2x1.414=820.12
Then increase by 20% to accommodate unloaded operation (eg no valves fitted eg when testing) and high mains wall voltage.
So 820x1.2=984V
Hence even worst case, a single pair of 1N4007 should never have their 1kV rating exceeded.

[Wonder2121]

The diode type for the HT rectifier isn't shown,

Using Vishey UF5408, 700RMS, feel Im good with 2 in series. Just realized why some amps have 3 in series, it's to totem the voltage..

Try that and see. A good rule of thumb for developing a fixed bias voltage for octal output tubes is that you want the voltage to be about 1/10th of the B+ voltage

First time building fixed bias, will test voltages and adjust first resistor accordingly...Probably use JJ 6v6s DM...If the circuit looks good will drop it into DIYLC.

Does the presence control look good? Think I copied Robs other 6v6 amps with 8ohm tap.

Also if you guys were building this amp for stiff, cleans is there anything you would add/change?
I heard SRV tech also removed tube rectifier on his Vibroverb and up'ed the xformer...

Full JSchem attached*

[pdf64]

I don't have any apps to open sch files, still can't see component values, so difficult to comment 
Personally I can't envisage a presence control being beneficial with such an amp, and I like a really bright tone.
A type 4 or 2 master volume seems much more useful, as they allow the preamp to be set to its preferred sweet spot, then the master controls the room SPL.

[Wonder2121]

Best can do using PicResize as per "How to add images in the middle of your post"
Pondered a MV, but first time building a presence circuit, would like to keep it..Could always pull it later date, I'll leave space...Also MV with the bias circuit might prove tricky, the 220 grid leaks, I'll look into it...

[Wonder2121]

Hey pdf64
After a bit of reading think your right, a PPIVMV might be beneficial.

Wouldnt this alter the bias, the 220k grid leaks are now being altered with sweep of pot?

[pdf64]

Hey pdf64
After a bit of reading think your right, a PPIVMV might be beneficial.

Wouldnt this alter the bias, the 220k grid leaks are now being altered with sweep of pot?

No, grid current and hence V DC across the grid leak resistance is negligible.
In regard of the idle operating point, it makes no difference whether the grid leaks are 0R or 220k, idle anode current won't be affected.     

To make a LarMar / type 2 master volume work with dual bias outputs, decoupling caps at both 'x' nodes may be beneficial. Perhaps get rid of the second smoothing cap, so that there's only 3 bias supply caps total, the reservoir and the 2 x caps.

[Wonder2121]

"Decoupling caps" Something like this...

Or do you mean smoothing caps to ground?

[AlNewman]

You can't put capacitors in series between the bias supply and the grids.

[pdf64]

Decoupling = smoothing = bypass.
Alternative terms for removing AC from a circuit node.

[Wonder2121]

So like this, but with say 16uf electrolytics?

[pdf64]

So like this, but with say 16uf electrolytics?

Yes, anything more than a few uF at the x nodes would be fine.

The component values are readable, yay!
Yes, 1M safety resistors on the bias trimmers seems too high, I suggest 100k, 220k at the most (ie about 10x the pot value).

[Wonder2121]

Ok safety resistors lowered to 220k..
Regarding the bias, now the 2nd smoothing caps removed should the 4.7k resistor be removed and the 1st filter cap be increased to 47uf? 

[Merlin]

Will I get 400V?

I'd estimate about 430V no load, sagging to about 370V if you hit full load. So 400V sounds pretty likely at idle.

[Wonder2121]

Anyone, out of thse 2 bias circuits which would you choose for type2 MV..

1st one has 150k grid leaks with 500k strereo pot.
2nd one with no grid leaks with 10uf caps from 250k stereo pot

I'm confussed which one to use...

[Merlin]

They're both OK, but your MV might work better with the second one, since it has a cap to ground from the bottom of each pot.

[Wonder2121]

Whats the benefit using grid leak resistors in the first one, is it safer in the event the MV wiper loses contact..
Cap to ground in second one, I can't understand its purpose. Could this be adopted into the first one?

[Merlin]

Could this be adopted into the first one?

Yes, very easy

[Wonder2121]

The above circuits not working for me..
Each pot interacts voltage with one another. Turning pot A alters pot B and visa versa...For some reason they're not acting individually.

For sanity it was rebuilt exactly same with different components to rule out a malfunctioning part..

UPDATE:
I went wrong...
I initially lowered the 47k's to 10k's to achieve the voltage range sweep, This lower value must leak voltage into opposite pot...Raising back to 47k stops this happening...Its now working as should..

[shampeon]

I don't think there's any way of them not interacting slightly, since the resistance is in parallel.

Here's a simulation of a dual bias circuit that shows it:
https://tinyurl.com/23fawvnl

[tubeswell]

The changes in bias voltage on each side won't just be caused by how much you dial the pots in - each output tube's cathode current affects the plate voltage of each output tube, which in turn will affect the load on the B+ rail, which in turn will affect the other output tube. So it's all a bit interactive.

[Wonder2121]

Quick update for future folks reading in ~

Set both pots -40v without tubes.
Fired it up with tubes installed, raised to -60, panicked, no sound...
Adjusted pots down to -36v, around -40 started hearing hiss, happy days..All working.

Big shout out to me main man Merlin for post (Reply #20) it worked great, hardly any interation between pots..MV works great..
Thanks to pdf64 for chipping in and Tubesie as per usuall on hand...
Oh I did change pots to 50kL, below is schem might come in use for voltages..