Bias tap

[Paul1453]

I'm looking at the Hoffman Plexi 6V6 PS - bias tap.

I have one of those metal FWB rectifiers and my HV tap has a CT.

Could I use the - side of the FWB for my - bias supply as shown.
I think I would need to adjust the dropping resistor as I'd now have -400VDC.
Would this just waste power and create heat?
On the plus side, this would have 2 diodes supplying - bias
and would not lose it's bias if a single diode opened like the 1st circuit would.

If I wanted to use my metal FWB rectifier
should I just use 1/2 of the FWB and tap the - bias like the Hoffman.

[mresistor]

take a look at this Ampeg V4 power supply      it might help   


http://el34world.com/charts/Schematics/files/ampeg/Ampegv4poweramp.pdf

[Paul1453]

Thank you mresistor,

So, that one doesn't have the CT but it still pulls the - bias off the same place as Doug does.
Kind of odd with no bias adjustment pot.

I could try to use just the + side of my metal rectifier and the 1N4007 diode.
I could also pull out two of my heavier duty type diodes and save the metal rectifier for something else.
It's probably best to just do it like that.

I'm kind of trying to regain my understanding of a PS that uses both a positive and negative voltage rail.
It's been a very long time, but I remember some of the Army gear I repaired had a PS like that.   

[HotBluePlates]

I'm looking at the Hoffman Plexi 6V6 PS - bias tap.

I have one of those metal FWB rectifiers and my HV tap has a CT.

Could I use the - side of the FWB for my - bias supply as shown. ... should I just use 1/2 of the FWB and tap the - bias like the Hoffman.

Mresistor pointed you to an alternative circuit from Ampeg (which is "same but different"). I'd advise you to stick with what is shown in the original Plexi 6V6 diagram you posted.

Even if you could do what you're proposing, all your B+ current would also have to flow through the bias circuit to define that ground point. That's because each cycle, 2 opposing diodes are conducting in the bridge. However, you can't ground your center-tap as well, because that will short 2 of the diodes in the bridge. (EDIT: Sluckey pointed out the case where this can be done, which I overlooked at first)

... On the plus side, this would have 2 diodes supplying - bias and would not lose it's bias if a single diode opened like the 1st circuit would.

Most every fixed-bias amp I can think of (certainly guitar amps) use a single diode in the bias supply because the bias circuit is low-current and easy to filter. There are untold numbers of amps produced since the late 50's whose single diode has survived (selenium rectifier replacement doesn't count, since you're unlikely to install one).

When I've fixed amps with fixed-bias issues, I've seen leaky coupling caps which negate the bias somewhat and bias filter caps which lose capacitance over time (causing hum at the output tubes). Maybe I'm lucky but I haven't seen a failed bias diode. If it did fail, I'd expect it would be likely to fail as a short-circuit and make the bias filter cap blow up. Hopefully, a fuse also pops with the extra output tube current draw.

Altogether, that might motivate me to install a B+ fuse to open in that scenario. If I really wanted extra rectifiers for a more bullet-proof bias supply, I'd be looking to having a dedicated bias winding (perhaps in the form of a small transformer, as London Power sometimes uses with a 120vac:6.3vac transformer, with the 6.3vac winding attached to the filament winding of the main transformer to develop 120vac to be rectified & filtered in the bias supply).

[HotBluePlates]

... I could try to use just the + side of my metal rectifier and the 1N4007 diode. ...

You know better...  That voltage is already rectified to be positive relative to ground.

That's why the Princeton, Marshall amps, and some others take voltage from the raw high voltage a.c. and divide it down for the bias supply.

... Kind of odd with no bias adjustment pot. ...

Lots of amps out there with "fixed fixed-bias" rather than adjustable fixed bias. If you need adjustment, you can modify the resistance from bias filter cap - to ground to allow for adjustment.

... I could also pull out two of my heavier duty type diodes and save the metal rectifier for something else.
It's probably best to just do it like that. ...

Yeah, this seems like a good option.

I'm kind of trying to regain my understanding of a PS that uses both a positive and negative voltage rail.
It's been a very long time, but I remember some of the Army gear I repaired had a PS like that.   

Solid-state often needs dual-rail power supplies of equal and opposite voltages. Like +15v/-15v, etc.

These have a PT, a bridge rectifier, and filter caps in series from + rail  to ground to - rail. The power transformer center-tap (if one exists) can be grounded or otherwise connected to the junction of the series power supply filter caps, because this is effectively the same thing and results in correct voltage & reference relationships everywhere.

[Paul1453]

Thanks HBP,

I knew I couldn't tie the - side to Gnd with the CT grounded.
I also knew that this would change the ground reference and the current flow,
but couldn't recall exactly how that all worked.   

It was mostly another one of my silly ideas to try to use parts that might fit better into a re-purposed chassis. 

Following chassis size and layouts exactly is so much easier.
That is what convinced me to try to learn how to make my own proper chassis'.   

[sluckey]

Could I use the - side of the FWB for my - bias supply as shown.
I think I would need to adjust the dropping resistor as I'd now have -400VDC.
Would this just waste power and create heat?
On the plus side, this would have 2 diodes supplying - bias
and would not lose it's bias if a single diode opened like the 1st circuit would.

Yes, You can do that. Just ground the center tap and use the bridge positive for B+ and the bridge negative as B-. It's called a bipolar power supply, or dual polarity power supply. The voltage on the negative side will be same amplitude (but opposite polarity) as voltage on the positive side. And it would be full wave rather than half wave provided by the more common single diode circuit.

The down side is cost. You'll have to use filter caps rated for 400V (or more) whereas, the more common circuit uses much lower voltage caps (much cheaper). It's a clever  but costly idea.

If I wanted to use my metal FWB rectifier
should I just use 1/2 of the FWB and tap the - bias like the Hoffman.

You can do that if the FWB can handle the voltage. Most metal block bridge rectifiers I've seen were rated low voltage high current.

[Paul1453]

Thanks for the feedback guys!

Just another one of my silly ideas.

[HotBluePlates]

Be careful with R4 on your PSU sim. This is to approximate a load. As drawn, you input stages (connected to C5) draw 223v/5kΩ= 44.6mA. I don't think that's what you intended.

Instead, every filter cap should have a resistance to ground to approximate its load on the power supply, so the PSU designer can calculate individual and total current draws.

[Paul1453]

Thanks HBP!

I'm not super knowledgeable about this PSU program.

I usually just go with the default values it puts in, unless I know for sure the correct values.

Yes, I've always been quite suspicious of that 5K load it likes to tack on the end.

I'll try changing it to like a 5mA current tap and see what that does.   

Oh my!  A huge difference in voltages results.

I think that was probably why Dummyload told me to enter current taps before.   

[Paul1453]

Modifications added.  Stepped load for the output tubes after 3 sec.
Used pretty much max current values from data sheet.

5mA of current for 1 12AX7 in the C,D, and E nodes, no steps.

I know my choke is 210 ohms but not how many Henries.
It's pretty big, similar in size to the Triad one I have.

This puts the 6V6s right about at their data sheet values listed.
Negative bias is listed at -19 for this voltage.
I might need a slight tweak to that circuit too.

I modded resistors to hit the C,D, and E node values as near as possible.
Original values: C=276, D=238, E=223
I hit those numbers almost dead on before the current steps up.
That is how it should be, and shows max sag right?

[HotBluePlates]

... Used pretty much max current values from data sheet.

5mA of current for 1 12AX7 in the C,D, and E nodes, no steps. ...

The Plexi 6V6 schematic has voltages, so you should use them.

V1A shows 0.77vdc across 820Ω, or 0.94mA.
V1B shows 1.56vdc across 2.7kΩ, or 0.58mA.
That's 0.94mA + 0.58mA = 1.52mA total for V1 and Node E. This cross-checks with 238v (Node D) - 223v (Node E) = 15v / 10kΩ = 1.5mA

The implication is if you're trying to hit a voltage target, you can make R41 bigger (and closer to stock), which improves filtering and decoupling (resistance to power supply oscillation).

V2A shows 0.84vdc across 820Ω, or 1.02mA
V2B shows 135v across 100kΩ, or 1.35mA
1.02mA + 1.35mA = 2.37mA.
Cross check: 276v (Node C) - 238v (Node D) = 38v / 10kΩ = 3.8mA - 1.5mA (for V1's series current) = 2.3mA. So that checks good-enough.

For V3, there's a misleading voltage measurement for the grids. Ignore it.
Total resistance to ground is 470Ω+10kΩ+5kΩ = 15470Ω. 33v / 15470Ω = 2.13mA
Cross check 1:
91v across R20, or 1.1mA
106v across R21, or 1.06mA
Total current is 2.16mA (within combined tolerance of resistors)

Cross check 2:
Total V1 and V2 current is 1.5mA + 2.3mA = 3.8mA
400v (Node B) - 276v (Node C) = 124v  / 20kΩ = 6.2mA - 3.8mA = 2.4mA (close enough)

Plug those values into your PSU circuit and use the stock power supply resistances. When designing preamp power supply nodes, you don't want to over-estimate current (which might be useful when deciding component power ratings) because of the tendency to push decoupling resistors too low.

[Paul1453]

I often stare at my pile of tube junk and say to myself:

"What could you possibly make with the stuff you have here"?

Most of the time, after hours of playing with chassis/parts/schematics/layouts,
I finally realize I can't make what I have, work for a specific design.

Even when I think I could possibly use the parts I have to make a design,
it usually still requires layout and circuit modification.

That is what I am attempting to do with my parts and the Hoffman Plexi 6V6 design.

I have a PT that can power this not at Doug's design specs, but at voltages listed on the data sheet.
So here is a pic of the 1st proposed layout in my Hammond chassis.
I know that is a weird placement for the OT, but it seems to work in my mind.
Input jack would go in front of the 1st preamp tube.
Output jack would go on the back somewhere near the output tubes.
So far I think I would only need to drill a hole for one more pot on the front, and the jack on the back.
I seem a have all the pots and major components to make this Plexi design work with my parts with just a slight PS mod.   

Maybe not?   

[HotBluePlates]

... I have a PT that can power this not at Doug's design specs, but at voltages listed on the data sheet. ...

Using that PT is no problem. It has way more 6.3vac current capability than you need, but that doesn't matter. Rectifying 440v CT into 220 * 1.414 = 311vdc is fine, you'll simply not make all the output power you might have with a higher voltage.

The challenge is the chassis/layout you're proposing.

... So here is a pic of the 1st proposed layout in my Hammond chassis.
I know that is a weird placement for the OT, but it seems to work in my mind.
Input jack would go in front of the 1st preamp tube.
Output jack would go on the back somewhere near the output tubes. ...

I don't know which of the three 9-pin sockets will be your input tube.

Let's say it will be the one furthest left. That's not good because the power transformer will have big a.c. and induce hum in the weak audio signal at the V1 stages.

Let's say it is the one furthest right. That's no good because you have the output transformer sitting there will have big a.c. on its primary which will likely couple into the high-impedance, low level input circuit of the V1 stages.

So it appears that while using the PT is fine, using that chassis for this project is a non-starter. You can use parts you have available, but they should still go together the way a guitar amp is typically laid out.

Do you happen to have a different chassis with a more conventional layout?

[Paul1453]

Could you take a look at this pic?

After many long hours of staring at this chassis,
and thinking what it might work with.

I agree that even with all it's socket holes everywhere,
It could not be used with anything with 4 or 5 preamp tubes.
I'm not convinced 3 preamp tubes can't find a reasonable home on it yet.

I had to rotate the PT, and I've identified the leads it will use.
I've also considered where the rectifier and bias components will go,
and placed those parts near where they will be.

12 and 13 are the AC primary on the PT.
They are as close to the chassis wall as can be.
The power switch and fuse will go underneath in the lower right corner.

I put the input jack on to show where it will be.
It is 7" from the PT AC, and only about 1" of shielded wire to V1.
V1 is almost 6" from the nearest AC, and almost that far from the rectifier components.
This has a very short shielded input wire to pick up signals where it is proposed.

If this chassis is no good for this I am SOL.   

[Paul1453]

I also made those changes to the PSU you suggested, HBP.

Here is a pic of what I came up with.

This seems very close to the original preamp voltages, and the 6V6 data sheet voltages.

Output power near 14W instead of 18W would be fine for me.

[HotBluePlates]

I also made those changes to the PSU you suggested, HBP.
...
This seems very close to the original preamp voltages, and the 6V6 data sheet voltages.

Output power near 14W instead of 18W would be fine for me.

Agreed, everything looks good with the 2.7kΩ dropping resistor.

... I'm not convinced 3 preamp tubes can't find a reasonable home on it yet.
...
I put the input jack on to show where it will be.
It is 7" from the PT AC, and only about 1" of shielded wire to V1.
V1 is almost 6" from the nearest AC, and almost that far from the rectifier components.
This has a very short shielded input wire to pick up signals where it is proposed.

If this chassis is no good for this I am SOL.   

Well, let's look at a built Plexi 6V6 in a Stout chassis.

This is a somewhat compact chassis, 12.5" x 6.5". Your donor chassis is bigger, no doubt. The spec sheet for the Heyboer PT gives the chassis footprint dimensions as 3.77" x 3.14".

Now look where the example's PT & input jack are: at literally opposite corners of the chassis. So why is your jack in the middle of the chassis apron, closer to the PT?

To get a sense of scale, those chicken head knobs are about 1.2" long, and so the pots are spaced on 1.5" centers. The input jack is about 6 pots away from the edge of the PT, or 9" away. You're looking at placing your input jack about 3-4" away from the closest edge of the PT. You should ask yourself why.

The whole circuit lays out easier, and is less prone to oscillation, if the tube sockets are in a straight line from input tube to output tube. The circuit will also be laid out roughly on this same line, as will the pots. If the tubes are closer to where the pots are mounted on the chassis apron, you have a natural place for putting the OT & choke.

I think you're hung up on using the existing holes in the chassis. But those holes and the original component placement worked well for the circuit originally in that chassis, now long gone. You should really use as much of the space available to you as is reasonable/practical. And copy as much of the example Plexi 6V6 layout as possible. Unless you're starting with an audio amplifier, you'll likely have holes to punch/drill to assemble a project.

You can probably get there with that chassis; it seems large enough in overall size. But it will probably have an air of "Fromage Suisse" once you're using a good, functional layout.

Way back, I offered my opinion that when purchasing old tube gear, it was usually best if you leave an audio amp's complete power supply, OT and phase inverter intact on the chassis. Then you could modify the preamp as needed to do what you want. Also, that when buying old tube gear, buy something which is already very much like what you'll transform it into. I don't know what items donated the chassis, choke & transformers, but perhaps you're seeing the difficulty of getting disparate parts to play nice together.

[Paul1453]

To the rescue again.  Thank you!,  HBP   

This site has so much information that I missed that full build documentation.

I had seen the picture of that amp a number of times.

You are completely right, I am hung up on the former placement holes!   

I'm from WI and like cheese.  Who cares where the parts were?

I need to start asking myself: "Where do these parts need to be now"?   

As you said "This chassis is certainly bigger than that one".

Let me try again, take a pic, and see what you think of that.
I think I see the better way now.  Thank You!   

[Paul1453]

Mental block out of the way, here is a better idea for the layout.

Presence pot and output jack on the back.

I'd like to have the Bias pot and tube test points available on the top.
Put them near where I thought they could go.

This looks much better now, No?   

[HotBluePlates]

This looks much better now, No?

Yep, that's looking more like a guitar amp!

[Paul1453]

 

This looks much better now, No?

Yep, that's looking more like a guitar amp!

  I can see clearly now, that block is gone.
I can see all obstacles in my way.
It's gonna be a bright, bright, Sun shiny Day 

It's flowing now.  Ideas on the PS caps - bias parts/pot/TPs area?

[HotBluePlates]

... Ideas on the PS caps - bias parts/pot/TPs area?

The cap can you had in the photo only has 2 sections, which is unfortunate because there are 4-section caps in the exact same can which would eliminate the problem. And that's also how the original layout is drawn (4-section cap can plus one on-board e-lytic).

I know you're trying to use the parts you've culled, but 20uF 450v caps are available in pretty small radial packages, which could easily be distributed across the main circuit board.

Otherwise, everything's looking pretty good.

[Paul1453]

I've got a bunch of new from China radial 22uF 450V and 10uF 400V caps.
They were very inexpensive.  Hopefully they are not cheap low quality caps.

That is why I used those values in PSU.

When I tried using the 22uF for the 1st 2 caps the AC ripple was excessive.
So I put in my 50/50 uF 450V cap can and got better filtering results.

PRR said 5% or less ripple on B+ was acceptable.
22uF didn't quite meet that standard, 50uF did.

My idea of bias pot/TPs removed from the board and placed there for top-side adjustment?
Does this seem reasonable, and not likely to cause problems?   
It should make it much easier to swap and balance output tubes.

[Paul1453]

Finally making some real progress.   

Layout mock up nearing completion.

Of course I still have a few questions.

I have 2 6.3VAC supplies I planned on using 1 for pre-amp and 1 for output tubes.
Now I don't think I like running more AC wires over to the pre-amp.
I don't see a great way to get to them, and am concerned about creating hum.
Either supply is more than enough for all heaters.

My PT also has 2 other HV CT supplies I will not be using.
I connected all 3 CTs to the PT's Gnd lug.  Good, Bad, or indifferent?

PS caps and bias circuit seem to fit OK where placed.
Any concerns about their placements?

OT and Choke are pretty close to each other.
I thought maybe that is OK?
They will tend to resonate together from plate current flow, correct?
I could move the OT to the left over the big hole.
I didn't really like the idea of moving the OT closer to the input and pre-amp tubes.

My parts board looks like it might be OK there?
There will be some changes needed from removing PS and bias circuits from Doug's layout.
I could actually cut that board in half, to stagger them, or move them to the upper left corner.
That could move them further from the noisy areas.
I don't know if that would help or hurt the layout.   

Feedback is greatly appreciated.
And HBP, thank you again for helping me to remove my mental block!   

[HotBluePlates]

... I have 2 6.3VAC supplies I planned on using 1 for pre-amp and 1 for output tubes.
Now I don't think I like running more AC wires over to the pre-amp. ...

Why?

2x 6V6 draws 0.9A of heater current, and 3x 12A_7's draws another 0.9A (when wired for 6.3vac), so 1.8A total. You have 2x 6.3vac windings which greatly exceed that requirement, so pick one & be done.

And agreed that if you are running more 6.3vac lines around the chassis, you're just creating more opportunity for that hum to be picked up.

... My PT also has 2 other HV CT supplies I will not be using.
I connected all 3 CTs to the PT's Gnd lug.  Good, Bad, or indifferent? ...

I don't think it will be a problem, but since you're not using those windings I would connect their CT's anywhere either. Insulate the ends and stow them away.

... OT and Choke are pretty close to each other.
I thought maybe that is OK?
They will tend to resonate together from plate current flow, correct?
I could move the OT to the left over the big hole.
I didn't really like the idea of moving the OT closer to the input and pre-amp tubes. ...

Recall that a.c. from one conductor couples easiest into another conductor in parallel with the first. But if the 2 conductors are oriented perpendicular, a.c. doesn't tend to couple. Same with transformers & chokes. Don't orient the cores of the OT & choke the same way, and there will be a greatly reduced chance of coupling.

The PT (I assume) will already be oriented to lay down on the chassis, so its core is already at 90-degrees to any orientation of the choke & OT.

Personally, since the choke is so large I'd place the OT closest to the PT, and the choke further away from the PT. This may also alleviate space issues with the cap can. The example Plexi 6V6 isn't arranged this way, but many Marshall amps are (of course, you have a massive, honkin' choke when a smaller one would suffice). As long as the choke wiring is placed towards the back of the amp chassis, opposite the side the pots will be on, you still have plenty of space away from preamp circuitry.

Hopefully, the chassis will still act as some shielding between the choke itself & the board. You may need to consider a cover plate for the big cutout under the board, regardless of which magnetic device is positioned in that area. Again, sorry to bring it up but these are the common issues when reusing a chassis which doesn't have exactly the same form-factor as the circuit you'll build into it.

Other than that, every seems fine so far.

[Paul1453]

I've made some more progress, I'll try to post a pic after work tonight.

I could still rotate my OT 90' over the big hole if that would be better.
It would point the windings directly at V2 and V3.
I do plan to use shielded sockets for all the 12AX7s.

Another question on the bias circuit.
Since I'm using a significantly lower B+ and the target for my bias voltage is -19VDC.
Could I get by with 10uF 50V ecaps in the bias circuit?
I have 2 of those that I tested at 10uF with my new capacitance checking meter.
These are little caps and will fit well in the spot I planned.
I also have a big honking 10uF 350V ecap which won't fit as nicely.
Or 2 new 10uF 400V ecaps which are smaller than the 350V one.
I'd like to save my higher voltage caps for something else if possible.

PRR's tool recommendation for a handheld saber saw was spot on.
With a little metal blade I can easily turn my chassis' into swiss cheese. 

[HotBluePlates]

... Another question on the bias circuit.
Since I'm using a significantly lower B+ and the target for my bias voltage is -19VDC.
Could I get by with 10uF 50V ecaps in the bias circuit? ...

Since your B+ is lower and you only need -19v, why not just cathode bias? It'll work just as well as fixed bias in this output stage.

Total current draw (plate & screens) for the 2x 6V6's is 74mA. 19v/74mA = 256Ω, so your standard 250Ω resistor is perfect. 19v²/250Ω = 1.4w, so a 3-5w resistor would be a good choice.

[Paul1453]

I've only done cathode bias circuits so far.

I was hoping to get some experience with a relatively simple grid biased circuit this time.      

[Paul1453]

Here is a pic of the top and guts so far.

The OT is the same way as the Plexi in a Stout chassis is.
I could still rotate it over the big hole.
My guts pic shows that my OT leads may not reach the tubes if I did that.
As it sits now, all OT primary leads are long enough.

The guts pic shows most of my bias and PS parts where I would like to place them.
The 10uF 50V caps are very small and will easily fit in this spot.
The rest of the PS caps are all in that area.
I could fit them all there, or put the final filter cap on the board as in Doug's layout.
I have selected a single 6.3VAC heater tap.
If I could finalize these sections in these spots,
I could then work on the necessary modifications to the parts board layout.
Those changes could hopefully simplify and shrink the number of turrets needed for my parts board.

After this build I will be trying to make my own chassis' that conform to the design standards.
And purchasing Doug's layout boards to make more professional looking builds from now on.

Your feedback on this proposed layout is greatly appreciated.   
I thank you guys for all your help!

[HotBluePlates]

I think you're on to something workable. Do consider covering the open hole by the choke with a cover plate (to help shield choke hum from the preamp board).

[Paul1453]

Thanks for the feedback HBP.

I tried to use a common sense calculation for my bias circuit's output voltage.
Since my AC input is 73.333% of the spec. I applied that to the bias voltage to get -26VDC.
My target bogey is -19 and I know that if I have too high - voltage my output tubes won't conduct.

I didn't know how to calculate a change to the 1st 220K resistor to get -19.
My next higher resistor is 270K.
I understand that if I don't have enough - voltage my tubes could red plate.
I was leaning towards leaving it alone and seeing if it would adjust down to -19 or lower as is.
Should I stay with the 220K or put in the 270K instead?

I did find a 10uF 100V ecap for the 1st filter cap and a larger 10uF 50V for the 2nd filter cap.
I thought that sounded like a workable substitution.   

[Willabe]

With an adjustable -bias amp you don't put the power tubes in until you check with your meter that you have -bias voltage at the power tube socket pins.

Use a gator clip on your meter probe so you have your hands free so you can measure the -dcv swing from hi to low.

Then if needed you can change a resistor in the -bias supply to get where you need it to be. The -bias draws almost no current so you don't need to worry about measuring without tubes/no load. It won't change enough if at all to make any difference when you put the power tubes in.       

[Willabe]

I did find a 10uF 100V ecap for the 1st filter cap and a larger 10uF 50V for the 2nd filter cap.

You should be fine with those as long as their not dried up.

Modern ecaps at those values are very inexpensive. Really not a need to recycle scrounged ecaps for your -bias.

If an old recycled -bias cap dies you risk burning up your power tubes, penny wise and dollar foolish.

[Paul1453]

Thanks for that feedback Willabe!

You are correct about the e-caps cost.

I have new Chinese made 10uF 400V radial ecaps that I got for $0.12 a piece delivered.
They were way overkill for this circuit, so I went through my recycled stock.
I'm always a little suspicious of the materials used and QC of such cheap Chinese parts.
But they are so inexpensive in comparison that I can't seem to pass them up.
I did test the old ones on my new cheap Chinese capacitance testing meter.
They both tested good and are axial Sprague ecaps.
I don't know how old they are, or if they are just about to give up the ghost.
I thought they might fit and look better being axial leads.

I have made a little more progress, and was planning on soldering some things in place tonight.   

[HotBluePlates]

Honestly, your best bet is (if nothing else) copy the original bias circuit.

You'll end up with too much negative voltage (cold bias), but that's the safe end of things. Build at least until that point, and as Willabe said, power up without the output tubes installed. At that point, you can parallel an existing resistor to drop its effective value. Dial it in until you have the desired bias voltage, then install output tubes and check idle current.

The problem is rectification is a non-linear process, and simply applying Ohm's Law doesn't tell you the correct resistances to use. That's because the voltage before the diode is a.c. while the voltage after is pulsating, half-wave rectified d.c. It's just very much faster to adjust on test than to hunt down the series of calculations to estimate parts values beforehand.

[Paul1453]

I have 1 ohm 1% resistors ordered.

I have made significant progress on my build.
I am hoping to be able to be able to actually make some noise with it this weekend.

My 1 ohm resistors won't be in by then.
I could just connect the cathodes to Gnd and not have bias TPs until they come in.
Or I was wondering if I could use 10 ohm resistors (lowest value I have in stock) for now,
and just shift the decimal point over 1 place on my TP readings, to be able to see the actual performance of my different 6V6's?

I realize that will raise the cathode's Gnd reference.
But my negative bias circuit could be adjusted to account for that if necessary, Correct? 

I'll post a pic of my progress after work tonight.
Thanks for all your help guys!
I couldn't have done this without your help. 

[HotBluePlates]

... Or I was wondering if I could use 10 ohm resistors (lowest value I have in stock) for now,
and just shift the decimal point over 1 place on my TP readings ...

Sure, why not?

Amps used to commonly have 10Ω cathode resistors (or even higher in some cases) because it provided an easier-to-see on analog meter movements.

You're gonna have plate & screen voltage up around several-hundred volts, and bias close to 20 volts. It won't matter to the tube one bit if you have 500-600 millivolts instead of 50-60 millivolts at the cathode.

[Paul1453]

I've got an unloaded B+ of 330, and the - bias circuit ranges from -19 to -25 now.

Finishing up the rest of the PS now.

Need to work on the necessary modifications to the layout from the P2P parts I've installed.

It should actually simplify the layout and turrets needed to complete the circuit.

The smaller board should be easier to find a good home for, I hope.

I'm pretty confident I can get this thing making some noise this weekend.

I really do appreciate the help you have given me!   

[Paul1453]

P2P work complete, and things are looking pretty good to me so far.        

I don't know if it's gonna get done this weekend.  I doubt it now.

Getting closer to firing her up and seeing how she sounds.   

[Paul1453]

Ran out of caps.  Subs OK for now?

[Paul1453]

I should be ready to test fire this 6V6 Plexi tonight.

I ran out of .022uF caps, and just put in .01uF instead.
Will that still sound OK, or should I parallel 2 .01's for each instead?
I can't put this on hold until new caps come in when I'm this close. 

The 2 .68uF cathode bypass caps, just parallel caps until close?
I have some new cheap Chinese little caps but not sure I have that value.

One more question, on the heater filament circuit.
I don't have a CT.  I put in a 200 ohm resistor and the LED like shown.
I also installed an artificial CT with 2 100ohm resistors to GND.
That shouldn't be a problem should it?
I was concerned that if I test that circuit without any tubes in I'll smoke my little 1/4w 100ohm resistors.

She's just about ready for the highlighter doublecheck and then the first trial run.   

[HotBluePlates]

I ran out of .022uF caps, and just put in .01uF instead.
Will that still sound OK, or should I parallel 2 .01's for each instead?

It will probably be fine with 0.01uF. If bass is a little thin, you'll know why and which parts to adjust.

The 2 .68uF cathode bypass caps, just parallel caps until close?
I have some new cheap Chinese little caps but not sure I have that value.

Your call.

Paralleling wastes space & caps. The caps make things brighter, you could use a higher value or no cap (one will sound fuller, the other will be less gain).

Doug could probably have those to you in a few days. Seems a shame to have to redo those parts later; this is now tallying to quite a few caps where the correct value isn't available.

One more question, on the heater filament circuit.
I don't have a CT.  I put in a 200 ohm resistor and the LED like shown.
I also installed an artificial CT with 2 100ohm resistors to GND.
That shouldn't be a problem should it?
I was concerned that if I test that circuit without any tubes in I'll smoke my little 1/4w 100ohm resistors.

100Ω 1/4w is fine.

When you create the artificial center-tap, each resistor sees 3.15vac on one side and ground on the other (half the total winding voltage). 3.15v²/100Ω = ~1/10th watt. I'm sure you accidentally used 6.3v for the calculation and came up with ~4/10ths watt.

[Paul1453]

Made a good noise last night, 1st try! 

Still got some issues to address, and then will need more calculation/optimization guidance.

Thank you for all your help!   

[Paul1453]

I really didn't know what to expect from this amp.
I only knew I had all the parts I needed for it so why not give it a shot.

I really appreciate all the help!

I couldn't have done it without you, especially you HBP.   

So I had parts, but not the exact parts called for in the Hoffman design.
I had to make a number of modifications and calculations.
This really helped me in the long run.

Once HBP got me past my mental block,
I could finally see where I should put my parts to make this design.
And it wasn't where the old holes were, but who cares?
It is where they need to be to function properly is what really matters.

Got a decent layout plan, and decided where my parts could go a function well.
Huddled my PS circuit around a Gnd  that was just a short wire away from the PT's Gnd lug.
I decided to run my heater wires close to the chassis and enter the sockets from the side.
Had the pot lugs pointing away from the heater wires, to give my signal wires some space from the heaters.
That led me to wiring up the pots backwards, but I quickly changed all 3 volume knob's wiring.
Tone pots still backwards but, oh well.
I had to make my own parts layout board too.
I attempted to make my wire runs as short as possible.
I learned even more doing that.

I got it all wired up with a few more parts substitutions.
1st test  didn't smoke anything, and I actually got sound from her.
Stopped to check all the voltages.  Needed a slight adjustment to drop my C-E voltages.
Made that and checked all the voltages again.
I'm happy with a 325V B+, and C-E are within 5% of the Hoffman plans.

Everything looked pretty good except V3b's voltage.  It was far too high at supply levels.
Had to trouble shoot to find C11 was leaky and not letting V3b conduct.
Fixed that, and actually tore apart some more junk gear to get the caps called for, replaced those too.

Even though my tester showed the output tubes to be strong and exactly the same.
They didn't draw the same amount of current, 1 30+% more than the other.
Swapped in an even stronger one for the weakling, and got them to match within 10%.

Well then it was time to really try her out.  I didn't understand the master and 2 other volume controls but learned how they interact quickly.
You can get a really nice over-driven tone even at low volumes.  Crank them all up and she really screams.
I got no hum or oscillation problems, and am very pleased with the sound and adjust-ability of this amp.

I really learned a lot n this build, and did create a better looking finished product than I had before.
Of course it doesn't compare to one of BOSS Scraggs professional looking products.
But I'm happy with my improvement and will continue to get better.   

Thanks again for all your patience and guidance.   

I'll post better pics when it is all cleaned up and housed in something.   

[HotBluePlates]

Congrats! Now time to enjoy it! 

Even though my tester showed the output tubes to be strong and exactly the same.
They didn't draw the same amount of current, 1 30+% more than the other.
Swapped in an even stronger one for the weakling, and got them to match within 10%.

If your tester is an emission-type tester, it doesn't tell you much that's useful for how the tube will perform in-circuit.

Transconductance (Gm, "dynamic mutual conductance") types are better in that regard, as you might be able to assume similar idle current draw for tubes with the same Gm measurement.

Best test is in-circuit, assuming you've verified the tube is not shorted or gassy.

[Paul1453]

This one turned out so well,
I'm already working on a re-design to use the quad 6V6 parts from my Belvedere purchase.

That should be loud enough for a gigging musician.
I'll look to make a professional quality build to sell at a nice profit on it.

I'm going to start a new thread "Quad 6V6 Plexi" for it.
I mostly just need to work out the output section and layout for it.
I'll post a couple of pics with my ideas for continued refinement.   

[Paul1453]

Maybe I can still run this PS with a B+ of +400VDC with my same PT.   

What if I used a voltage doubler design for the PS?

This PT actually has 3 HV taps, so 2 or still free with adequate current capacity.

The other two are slightly lower VAC, but doubled would put me near +450V with a SS rectifier.

Maybe there is a way to use a switch to choose between a +400VDC and the  +325VDC B+ I've already got?
I've already thought of some issues that would need to be addressed to have a switchable B+.
I also need to do more research on the voltage doubler power supply design and implementation.

This thing already Rocks now.
I had our maintenance guy try it out today.
I got a Thumbs up and big grin of approval from him.   
Since my PT has the other HV taps that may work to give me the +400 VDC B+.
I thought I might at least explore my options, and possibly give my 6V6's the Plexi's designed voltage to see how that sounds.   

[HotBluePlates]

Maybe I can still run this PS with a B+ of +400VDC with my same PT.
...
This thing already Rocks now.
I had our maintenance guy try it out today.
I got a Thumbs up and big grin of approval from him. ...

Why mess with success? The amp is working and sounding good now, and probably won't sound materially-different/better with 400v. However, when you tinker your power supply, you'll create new chances for things to be messed up.

Looking at it differently:
You're already using the highest-voltage winding of the 3 HV windings you have. 440v CT/115 * 120v = ~460v CT, which feeds ~230vac to each rectifier. 230vac * 1.414 = 325vdc.

You have 2 more windings marked 364v CT. 364v CT/115 * 120v = ~380v CT, or 190vac to each rectifier for 190vac * 1.414 = ~269vdc. One winding could be run with a bridge rectifier if the center-tap wasn't connected, but you'll have 380vac * 1.414 = 537vdc output from the rectifier.

If the target is 400vdc, that implies 400v/1.414 = ~283vac input to the rectifier (either a total winding of 283vac for a bridge, or 566v CT for a non-bridge full wave rectifier). Your highest voltage winding is about 50vac less than the target, but your lower voltage winding with a bridge is 100vac over the target. So the goal of 400vdc with the existing transformer seems a non-starter.

However, a low-voltage option of ~270vdc presents itself, especially if the two low-voltage windings are in parallel (lower supply voltage implies higher idle current for class A operation). Phasing of the two windings will be critical for the parallel connection, and needs checking. You should also use both primary windings in parallel if you go this route (I can't tell from your pics if you still use only one of the primary windings); phasing dots for those windings are shown on the transformer. You should be able to get 270vdc and at least 300mA with this hook-up.

[Paul1453]

Well this one is kind of my learning/experimenting test platform.
I really like the way it sounds now, but people seemed to not understand why I ran this on lower voltage.
It kind of seemed like I was missing out on something (what I don't know) by going with the lower voltage.
I know I am missing out on a few more Watts of output power.
But I really don't think I am missing out on any sound quality.

If you think that 75 extra volts won't improve the sound, only the output power.
Then there really is no point in me trying to give it's PS a boost.
I'm building it's big brother quad 6V6 for more output now.
I wanted to explore the voltage doubler concepts in the real world.
And maybe figure out how to have a switchable 325 - 400 PS setting.

It's all good!  I'm planning on keeping this one for a while now.
I brought the K10 and 2 other conversion amps I made,
to a local pawn shop owner, who says he may be able to sell them.
He likes the K10 better than the Plexi.
I like the Plexi better than the K10.
He and his customers like a more heavy metal dirty sounding amp.
He admits it helps his son and the other Head-banging players cover up their sloppy playing.
The K10 is always pushing hard, where this Plexi is much more adjustable to the tone you can get.   

[HotBluePlates]

... It kind of seemed like I was missing out on something (what I don't know) by going with the lower voltage.
I know I am missing out on a few more Watts of output power. ...

Power output (RMS) will be 1/2 of peak voltage swing times peak current swing. You have an 8kΩ (plate-to-plate) OT which will look like 4kΩ to one side of the output stage; let's assume the amp stays class A (so the load seen is 4kΩ instead of 2kΩ). We'll cheat and assume the output tube can swing its plate to 0v.

Power = 1/2*(325v)*(325v/4kΩ) = 1/2*(325v)²/4kΩ = 13.2w

Now assume everything stays the same, except you bump the supply voltage up to 400v:
Power = 1/2*(400v)²/4kΩ = 20w

That's a 50% increase in power output, so you will be able to tell it's a tad-louder than with the 325v condition. A "tad louder" seems understated, until you realize hearing is logarithmic, and it takes 10x power to get double-loud.

If you like being able to drive the output tubes into distortion, that will change with the higher voltage. Unless you choose to keep the screen voltage at the same level as the 325v version, then with the increased screen voltage an increased bias voltage will be needed. Plate current will tend to go up with increased screen voltage, and plate dissipation will go up regardless because of higher plate voltage. It will take more drive from the phase inverter/preamp to push the output tubes into distortion. Along with a little volume, you will gain a little headroom because the onset of distortion will be at a higher volume-knob setting.

But you already went whole-hog and doubled-up the output stage. The 400v condition on the original amp will be a mid-point between the original and doubled-output stages.