5F2A Power Scaling Using Multiple Transformer Secondaries - Help Please

[sgbrewing]

Hello, hope you are all well! I would greatly appreciate any help given.

I am planning a 5f2a build with power scaling on the power section by using a custom wound power transformer with multiple secondaries to change the voltage and using a rotary switch to select between them, I dont like the idea of using mosfets. I will also be using a master volume to maintain the drive level upon dropping voltage to the power section.

The cornell romany 12 uses this idea and has 4 settings with 3 different secondary voltages and 1 with a dropping resistor, they use a single 6l6 running SE with solidstate rectification. From researching these are the secondary voltages and resulting wattage:

- 300V = 10w
- 200V = 2w
- 100V = 0.25w
- 100V dropped through 270R = 0.05w

With the build i'm planning I would like to have similar wattage seleciton but i'm planning on using an RCA 6v6 SE with a GE 5y3 rectification, output transformer 7k Primary with 8 ohm output. Would be good to have wattage selection around:

-5w
-1w
-0.25w
-0.05w

But I have a few questions on calculations and equations to be able to plan this out, I understand that measuring upon completion and then adjusting would be the most accurate, but as i have to get a transformer custom wound I would like the secondary voltages to be as close as possible so I can just use a small dropping resistor if required.

1) How do I calculate the b+ from the transformer secondary voltage?

2) How can I calulate the estimated output watts from each of the secondary voltages and hence b+ values so that I can estimate what secondary voltages I would need on the transformer to get roughly the wattage selection I am after?

3) How do I calculate what size dropping resistor is required to drop the voltage to a certain amount, say from 100v to 60v for example?

Thanks a lot for the help and I applogise for the possible silly questions, I'm still relatively new to the theory behind tube amps and electronics.

I will also keep this updated with my build progress so other people can learn from it also. I have some nice paisley upholstery fabric that I am going to cover the cabinet with and then poly, so we shall see how it all turns out :D

[shooter]

figure out the turns ratio for the OT
use ohms law
surf up "understanding transformers", read the stuff with lots of math 

[sgbrewing]

How can I find out that ratio without having the transformer and applying voltage myself, would the manufacturer be able to tell me?

[shooter]

maybe, never asked, never needed to know.
I "match" the OT to data I do know, like
B+ = 300vdc
primary = 5K
max current = 67mA
secondary = 4ohms
stuff like that
then I just match it to available OT's
once everything works and i'm a happy clam, then I put on the mad scientist hat, grab meters, scopes, sig-gens and smoke 'em if I can 

[PRR]

... ... ...Would be good to have wattage selection around:
-5w
-1w
-0.25w
-0.05w

How do you know these are the "good" steps?

IMHO, it is arbitrary. And you may be better served with a continuous trim instead of steps, so you can find the "just right" power for any combination of room, drummer, and genre.

To a first approximation, tube amp power goes by square of voltage. So to go from 4W to 1W, cut the voltage in half. From 5W to 0.05W, 1/10th. This suggests 350VAC, 175VAC, ???, and 35VAC.

[sluckey]

You're gonna spend more money on the PT than the rest of the amp! I'm gonna offer a couple of less expensive ideas for consideration, or not...

1. L Pad

2. Da Geezer’s Bypass Cap attenuator, see page 6 of my Amp Scrapbook.

[kagliostro]

If you are planning an SE amp the use of a tube rectifier isn't necessary, a SS rectifier will be enough (in SE amp you don't ave SAG benefits from the Tube Rectifier use)

About voltage step .... if you are planning to use Cathode Bias instead of Fixed Bias (usually for an SE amp is so) use a VVR and if you like more, instead to vary the B+ voltage turning the shaft of a pot, you can arrange a rotary switch with different resistor, each with the value that permit to the B+ voltage to achieve the desired Power Level step by step

Franco

[sgbrewing]

You're gonna spend more money on the PT than the rest of the amp! I'm gonna offer a couple of less expensive ideas for consideration, or not...

1. L Pad

2. Da Geezer’s Bypass Cap attenuator, see page 6 of my Amp Scrapbook.

Theres this small UK company and I dont think its going to be too bad :P
This amp is going to be for home use/tv volume so im really looking for that low wattage mode.
Thanks for your suggestions I shall look in to them!

Is there any way to work backwards from output wattage desired to the B+ required?

Thanks again

[kagliostro]

I hope I posted the correct Schematic and the correct Board (measures are in mm)

I have several versions scattered around the pc

Franco

[shooter]

amp is going to be for home use/tv volume

Is there any way to work backwards from output wattage desired to the B+ required?

the way you are "understanding" volume and wattage, they are not really interchangeable.
I use a tube 10W PP amp for my audio, the volume control works great for nighttime, low volume listening,, or Friday night, rockin the room out.
the wattage never changes. 

[bmccowan]

I don't like the idea of using mosfets

Not knowing your reasons, I'll question that, and echo Franco's suggestion.
I have installed VVR in a few amps, both cathode and fixed bias. Works well in cathode bias amps. More complicated in fixed bias.
In cathode bias amps it works well and does just what you want. It can also be setup to vary the voltage in just the power tubes.

[HotBluePlates]

Is there any way to work backwards from output wattage desired to the B+ required?

That's the usual "design process":  start at loudness-from-speaker and work backwards to figure everything else.

This amp is going to be for home use/tv volume so im really looking for that low wattage mode.

You will want to consider "desired speaker/sound" and the sensitivity of the chosen speaker.

     One source says TV-volume might be up to 70dB

     If you decide you like the Celestion Blue Alnico for your amp, its 100dB sensitivity (for 1w input) means you need to drop amp-power to 1 milliwatt because:
          10 log (Amp Power / Reference Power) + Speaker Sensitivity
       = 10 log (0.001 watt / 1w) + 100dB = 10 x -3 + 100dB
       = 10 x -3 + 100dB = -30dB + 100 dB = 70dB SPL

     There are speakers that have lower sensitivity (like the Jensen C12R), mostly due to their smaller magnets.
        It takes ~6 watts applied to the Jensen C12R to be as-loud as the Celestion Blue Alnico with 1 watt applied.

        Amp makers in the 1960s used smaller-magnet speakers in some of the lower-end amps.  The tweed Deluxe often had a P12R, while Ampeg's Reverberocket had a C12R.  However, many folks find they prefer the sound of the bigger-magnet speakers so personal preference weighs heavily.

You would normally use the thinking above to:

   1. Determine a power output requirement. 
   2. Think of a class of tube that can likely deliver that range of power output.
   3. Juggle the tube's ability to pull plate current against available output transformer primary impedances.
   4. Decide on an operating point & class of operation for the tube(s).
   5. Settle on a supply voltage based on available output & power transformers, chosen supply voltage, and tube operating point/capability to pull current through the OT primary.


OR... you could skip all that, and decide it's easier & more flexible to have an amp that's clean plus some pedals to deliver whatever flavor of dirt you desire.  You don't necessarily need to have output tube distortion (and it's almost always gonna be loud when you do).

I know I've played an AC30 in an apartment at 2am without complaints from neighbors because I simply turned the guitar volume very low. 

[sgbrewing]

I don't like the idea of using mosfets

Not knowing your reasons, I'll question that, and echo Franco's suggestion.
I have installed VVR in a few amps, both cathode and fixed bias. Works well in cathode bias amps. More complicated in fixed bias.
In cathode bias amps it works well and does just what you want. It can also be setup to vary the voltage in just the power tubes.

I'm worried about failure over time. Is the unwarranted?
Can I install a VVR and use it down really low so its conversation voume for 1-2 hours a day and not worry about failure for decades?
Cheers

[shooter]

only the datasheet knows, and maybe some heavy math, possibly AI??
one of my systems used lots of power mosfets, they broke on average once a year.  they operated close to design max about 12hrs a day


build the F2, make it work, play it for a month, while you're playing have a BB set up with a "test bed" for your VVR, I use the rule of thumb, "If it lasts a month, I don't need the extended warranty" 

[sluckey]

Can I install a VVR and use it down really low so its conversation voume for 1-2 hours a day and not worry about failure for decades?

That's unrealistic! You can't say that about any amp, especially if you built it. In fact, you should expect it to fail.

[bmccowan]

Hmm - I have not had one fail, but then I do not play that long at low volume. If I was going to, I would build a different amp just for that, but I'm an amp junkie. And I'm not trying to crack wise, but do capacitors and resistors last for decades?
The mosfet expert here seems to be tubenit. Maybe he has some views on longevity when used in a VVR circuit.
Typing this when Sluckey's reply appeared - I agree, amps fail - that why we learn to fix them.

[kagliostro]

Think to the VVR schematic I posted previously

in your application you have say 350v B+ and want to scale the voltage as to achieve a power reduction, but if the Mosfet fail and the whole B+ voltage arrive to the amp you only have a transition from a lower volume to a higher volume

In an amp that uses GU50 tubes that, as many other tubes, can't run the same B+ level of plates, or you use a specific builded transformer that has a section (or an intake) dedicated to G2 supply or you can use a VVR, in this case if the Mosfet fail it will be not very pleasant for the G2 of the tubes

So, with a friend, some years ago we developed this simple solution that uses two in series Mosfet instead of a single Mosfet, using two Mosfet the dissipation is splitted in two and if one Mosfet shorts, the other continue to feed the correct B+ level to G2, note that if no voltage arrives to the G2 because of a fail, the tubes won't burn, problem is only if B+ is too high

Obviously if there is a problem that shorts B+ to ground you'll be happy to have a fuse on the path

Franco


p.s.: Note that the presence of a resistor between B+ and the Pot (in the previous schematic it was present, here not) is to give a limit to the highest voltage that the exit of the circuit will supply while the resistor that is between the pot and ground set the lower output voltage