Understanding transformer requirements

[Tomahawk]

I am working on a my first design and I am working backwards from the power supply to the preamp section. The goal is a Single Ended 6V6 with two channels (EF86 and 12A_7) and an effects loop. I understand the base concepts, but I am unsure of my maths. I am using the data sheets for each of these tubes. All tubes are JJ Electronics: 5Y3S, 6V6S, EF806, 12A?7. I do plan to run this as a Capacitor Input vs Choke Input.

My understanding is if these tubes are run as Class A the "Typical Characteristics: Class A1 Amp" would be the values I need to review.

Rectifier 5Y3S: Limit values 350v per plate, 20uF capacitor, and 120mA @ 360v DC
My question here: Does this mean I can choose a transformer up to 350-0-350? I do see it has a hard limit of 360v DC, which would mean my input transformer should be less than that to accommodate the voltage gain, post rectification?

Power 6v6S (Pentode): Limiting values Plate 500v, Screen 450v, and a max dissipation of 14w. The current draw is 45mA on the plate and 5mA on the screen (50mA total).

Preamp/Effects 12Ax7 (not sure if wanting different ECC8x tubes): 2.4mA per tube @ 3 full tubes would be 7.2mA

Preamp EF806 3mA for the B+

The math I have leads me to 60.2mA of B+ current, 1.55A of 6.3v heater current, and 2A of 5v rectifier heater current.

If that holds true, what is the max voltage transformer I can use given a single ended 6V6 and 5Y3 rectifier? I see the Fender Champ uses a 325-0-325 transformer and ~354v at the plate of the 6v6 and 360v at the output transformer. I am going for a dual channel preamp, similar to Marshall where I have a clean channel and a dirty channel and a single ended 6v6 because I am designing this for home use.

Thanks
Tomahawk

[thetragichero]

just as a reference this is convenient for getting ballpark power transformer specs (i often use this as: i have this transformer, can i build what i wanna build with it): https://thesubjectmatter.com/calcptcurrent.html

what output transformer are you using? i take the primary impedance and look at the datasheet, see what my plate and screen want to be at and then work from there (sometimes will look at 'prior art' for ideas as well). but yeah, you need to know where you want your tubes operating at before you know what transformer to use

this is another handy calculator: https://www.vtadiy.com/loadline-calculators/power-stage-calculator/

another piece of advice for single-ended amps: capacitor input supply is standard but i will place a choke and then another capacitor (called a pi filter) before the plate supply node. you don't get the benefit of a lot of hum rejection that you would in a push pull amp.

that all being said i built a "big boy" vibrochamp (used a 6l6g i pulled from old test equipment, solid state rectification, and a nice 10" speaker) that i loved playing at church until i built something a bit bigger and "vox-y". fun little amps with surprisingly great tone if you've got a nice speaker

[PRR]

FWIW: 325VAC makes 459V peak or nearly 459V of DC which suits nominal 450VDC capacitors. Anything over that for a single 6V6 is just foolish.

"500V" is far past original 6V6 specs. If you need a replacement on a Saturday night, are you sure the junk you can get at the Banjo Center will stand it? Anyway SE power does not go up much with voltage (current has to be reduced to stay within Pdiss). Better to stay near the 315V-350V rating.

Plagiarize plagiarize plagiarize!! What you want is a Champ. The second channel matters little to the B+ compared to the suction of the mighty(?) 6V6. Heater demand may be an issue with more small tubes. But both the Champ and the DeLuxe derive from fancy parlor radios with a lot of small tubes (for radio). Today you tend to get the same part for any of these applications. These transformers will not usually be strained.

[Tomahawk]

FWIW: 325VAC makes 459V peak or nearly 459V of DC which suits nominal 450VDC capacitors. Anything over that for a single 6V6 is just foolish.

"500V" is far past original 6V6 specs. If you need a replacement on a Saturday night, are you sure the junk you can get at the Banjo Center will stand it? Anyway SE power does not go up much with voltage (current has to be reduced to stay within Pdiss). Better to stay near the 315V-350V rating.

Plagiarize plagiarize plagiarize!! What you want is a Champ. The second channel matters little to the B+ compared to the suction of the mighty(?) 6V6. Heater demand may be an issue with more small tubes. But both the Champ and the DeLuxe derive from fancy parlor radios with a lot of small tubes (for radio). Today you tend to get the same part for any of these applications. These transformers will not usually be strained.

Can you explain the math on that? I know the champ uses a 325-0-325 transformer and they show a much different B+. My resources state the B+ should have a value of 320V DC.

[thetragichero]

i estimate by taking the voltage of the full transformer winding (650vac) / 1.4 (full wave rectification with diodes) - voltage drop of tube rectifier (might be as low as ~20v since you're not pulling much current). that gives me ~444vdc on average. close enough for punk rock (if there are more in depth way i'm all ears)

can't always trust voltages listed on fender schematics

[Tomahawk]

I was under the impressions a 5Y3 rectifier tube has a DC voltage gain of about 1.1. Additionally, I thought you didn't add the positive peak and negative peak. My understanding is when using a tube rectifier you flip the 0 to negative voltage positive and that is how you get 120hz vs 60hz from the AC socket. This would be 120v AC 60hz to (325v * 1.1) @ 120hz.

[PRR]

....5Y3 rectifier tube has a DC voltage gain of about 1.1.....

Read the datasheet.
https://frank.pocnet.net/sheets/049/5/5Y3G.pdf
page 5
"Gain" (wrong word) can be 1.414 at no-load to about 1.0 at maximum rated (~~128mA) for 325VAC.

[d95err]

AFAIK, old data sheet values are often very conservative. Guitar amp manufacturers often abuse tubes, going well over those limits.

My recommendation is to look at guitar amp schematics with similar design and tubes to see what voltages and transformers they use.

[tubeswell]

Yes a 320-0-320 winding will result in a good tweed Champ or Princeton running SE 6V6

[Tomahawk]

Is my math right for a single ended 6v6 (planning for a JJ 6V6S):
PT: Edcor XPWR012: 350v-0-350v @ 100mA, 6.3v @ 2A, 5v @ 2A
OT: Edcor GXSE 10-2.5K: 2.5K ohm @ 4 ohm, 10 watts output (using as 5K ohm @ 8 ohm speaker)
Input Choke: Edcor CXC100-7H-150mA: 7H at 150mA
5Y3 rectifier
B+1 voltage: 431.5
Point A: B+1
Point B: B+1/Ra
431.5v/5000 ohm = 86.3mA
9:1 plate to screen current: 4.89mA

Bias set to -20v.
Q: 260v @ 36mA
-95 to +95V swings (155v to 265v)
171.5v dropped: 43k resistor .83watts dissipation - 2watt resistor

[PRR]

> 431.5v/5000 ohm = 86.3mA

This is 35 Watts plate dissipation. About 3X what a single 6V6 can take.

Use "Champ" iron. It works.

[Tomahawk]

Thanks for the double check. It appears I could change the output transformer with an Edcor GXSE10-4-8K. This is 8,000 ohm @ 4 ohm, or in my case 16,000 ohm @ 8 ohm. This would bring the 26.97mA. Or, 11.64w, which is at the edge of the 12w power handling. This is more an exercise in learning to build an amp vs building an amp instead of buying one.

[PRR]

I do not understand your computations. I fear you may be going off the path.

Also compare to the tube manual suggested conditions. They may not be the very-most the tube can do (some hold-back to sell you a higher price tube) but they are not far out.

[dbishopbliss]

I have found the Duncan PSUD very useful for determining how power supplies will behave. I like to use the "current" sink modeling. That is... plug in the current draw for each stage of the amplifier. Then you can play with the capacitor values and resistors (or chokes) to determine the voltage you will get at each stage.

[thetragichero]

your choke is a bit overkill as well. the 90ma types for a 6l6 fender-types are more than enough (could probably get away with the 50ma deluxe reverb type but i am too lazy to do the maths)

[Tomahawk]

So I think I figured it out. The only thing I don't get is how to select the screen resistor and if I understand the bias Q point.

From my research I found that 1k seems about right for a 6v6 as it matches the Champ, but I couldn't find the math on the why. I somehow mathed the number for 20v dropped/ X = 2mA. Is that correct?

Is the mA picked for the Q based on the heat dissipation curve and how much power you want out of the amp?

The input cap was sized per the JJ 5Y3 rectifier sheet. The cathode resistor was the value in the transformer calculator from above and the cap was sized to roll off at 10hz.

[shooter]

it matches the Champ, but I couldn't find the math on the why

Vk/Rk = I total
(Vp-Vk) * Itotal = tube dissipation in watts

[tubeswell]

What is the 6V6 cathode voltage (at R1)?


Also, I presume you have another filter cap at '1' (where you have the OT primary supply after the choke)?

[Tomahawk]

Maybe I need to take a step back:
I've been reading Guitar Amplifier Electronics Basic Theory by Richard Kuehnal and Designing Tube Preamps for Guitar and Bass by Merlin Blencowe trying to learn the math behind guitar amps in an effort to learn to design my own.

I get the concepts and the ideas, but the examples are what is tripping me up. I want to start with a simple 6v6GT single ended output stage. From what I was reading it was better to start at the power stage and work your way back. So I found the datasheet for a Tung-Sol 6v6GT. They used a class A example as 315v for the plate voltage. I heard good things about Edcor so I found a transformer that could work.

The Edcor XPWR221 provided 250v-0-250v and rectified via a 5Y3 you would get a B+1 of 312.5v. 306.5v with a choke, which is planned since I would like a low noise floor.

The Datasheet for the 6v6 showed a Load Resistance of 8500 and I found Edcor made an 8k@8 ohm single ended output transformer, the Edcor GXSE10-8-8k. This should be a close enough match to the datasheet. This gives me the purple load line attached. The datasheet then states a max plate current of 35mA. So setting the Q at 306.5v and 35mA and drew a load line in red.

The next step is where I am getting hung up. The datasheet shows a grid 1 voltage at -13, but the load line has it on around 19. Is there something in my math that is wrong?

Then the next step would be to calculate the AC load line, correct?

Power Transformer https://www.edcorusa.com/xpwr221
Output Transformer https://www.edcorusa.com/gxse10-8-8k
B+1 calculation https://thesubjectmatter.com/calcptcurrent.html

[PRR]

6v6 Load Line.png is not a line but a dot.

> The datasheet shows a grid 1 voltage at -13, but the load line has it on around 19.

Are you correcting for screen voltage? FWIW, in real life that is too much work, we find bias experimentally.

[sluckey]

Yeah, but it's a big dot! 121 pixels.

[Willabe]

Look at your load line attachment.

[Tomahawk]

Sorry, I uploaded the wrong file. Here is the correct load line.

[tubeswell]

Sorry, I uploaded the wrong file. Here is the correct load line.

That load line shows a cathode bias voltage of about 19V. Seeing as how you haven't actually measured the cathode voltage and assuming your cathode resistor is exactly 330R* (as per your earlier schematic), the tube current will be 58mA, which is way too hot for a 6V6 with a plate voltage of 305V (305V x 0.058A = 17.56W). Try a 470R or 560R cathode resistor

[Tomahawk]

How comes the tube data sheet says at that voltage and output transformer impedance it should be around 13v?

[Williamblake]

Maybe this is not the right problem to your solution but you could gut a radio with a EL84 or 6V6 so you have matching iron that you can toy around with and draw load lines later? Sorry, i dont know anything about load lines, but neither do you, yet, but my suggestion aims at getting some practice on the cheap and learning the theory on the way. If this is you first build there will be so much back and forth no matter how many books you read you will be glad you did not put yourself in the corner to soon. Downside being you will not only occupy your bookshelf but all your area available.

[thetragichero]

this may be helpful for loadlines for output stages: https://www.vtadiy.com/loadline-calculators/power-stage-calculator/

but i would recommend considering previous suggestions wrt 'borrowing' from known good circuits. you basically wish to build a champ; look towards champ (black/silverface suggested but tweed if you wish)-speced transformers while adding the pi filter before the plates (i can see why fender would wish to save a couple bucks on a practice amp by leaving out the extra cap and choke but in a diy build i would highly suggest, along with more standard heater wiring with an artificial center tap. include a better speaker and you'll get plenty of low end without 60hz hum that probably wouldn't be noticed on a champ anyway)

my philosophy with tube amps is that it's pretty much all been done before, so you could try to reinvent the wheel but it's not particularly worth the time

[PRR]

How comes the tube data sheet says at that voltage and output transformer impedance it should be around 13v?

Are you correcting for screen voltage?

[Tomahawk]

I did some research and was less focused on specific numbers and let the numbers lead me to the values.

The only thing I don't fully get is how to calculate the second smoothing capacitor, the one used for the screen.

I know the 5Y3 rectifier has a hard limit for the reservoir cap of 20uF.

Lastly, the screen tap on the transformer will not be used.

[PRR]

> the second smoothing capacitor, the one used for the screen.

Uh, that's the plate.

And also the screen. Rule of thumb is the plate may be 10X dirtier than the screen. But here they are the same. Actually I would not think to "calculate": dozens of amps have done well with 16u-10H(or FC)-16u to plate and screen.