How to determine acceptable plate load for power tube and output transformer?

[joulupukki]

In the Bassman Micro LTP design (https://robrobinette.com/Bassman_Micro.htm) with a 12AU7 as the power tube @robrob shows using the Hammond 125B output transformer and a plate-to-plate load of 22.5K for an 8 ohm speaker. Where does the 22.5K value come from (or at least how was it calculated)?

And maybe more directly, if I wanted to use a single output jack with a switch that goes between 4 and 8 Ohms, I could share one common wire between the two if, for 4 Ohms the plate-to-plate load is 27K…

4 ohm secondary wires: 4 - 5, 27k load
8 ohm secondary wires: 2 - 4, 22.5k load

I’ve looked at different threads & articles trying to understand this stuff and I guess it’s still beyond me. How would I calculate whether 27k is a suitable load for the 4 ohm option? I see that I could uses wires 5 & 6 for 4 ohms for a 22k load but then I don’t think I’d be able to use a switch (SPDT) to easily switch the impedance of the speaker out jack.

 

[peteb]

Where does the 22.5K value come from (or at least how was it calculated)?

The plate to plate impedance should be the speaker impedance multiplied by the turns ratio of the OT.

the primary has many more turns than the secondary, to convert high voltage low current to high current and low voltage.

a single ended amp may want to see 4000 ohms of impedance. 4 ohms times 1000 turns ratio is 4K ohms, for example.

22500/ 8= a turns ratio of 2,812.5 to one.

8 ohms times 2,812.5 = 22.5k ohms of impedance, seen by the tube.

 

[joulupukki]

Thank you @peteb, that much makes sense. Do you know (or anyone) how I look up or understand what is "ok" or "healthy" impedance for the plates? i.e., is it okay to have the plates on the 12AU7 have 27K ohms of impedance? ...or should it stay near 22.5k (or in this case 22k with the violet and white wires)?

 

[Ten Over]

22500/ 8= a turns ratio of 2,812.5 to one.

The turns ratio is 53.033:1

 

[andrewRneumann]

27K looks totally fine. One could even make the argument that it’s better than 22.5K for a B+ of 400V, but I won’t. I seriously doubt anyone could tell the difference. @Ten Over is right about the winding ratio.

 

[peteb]

The turns ratio is 53.033:1

Thanks ten over.

I was just going from memory.

The impedance ratio is 2812

the turns ratio is the square root of that.

 

[joulupukki]

27K looks totally fine. One could even make the argument that it’s better than 22.5K for a B+ of 400V, but I won’t. I seriously doubt anyone could tell the difference. @Ten Over is right about the winding ratio.

The schematic shows a B+ of 300V.

I know I should know this, but this is what I’m trying to learn…

Knowing that info and the info about the output transformer, how do I read the information from the 12AU7 to know what range of impedance values on the primary winding of the OT / tube plates are acceptable? …so the amp will be healthy, safe, efficient, all that?

 

[andrewRneumann]

The schematic shows a B+ of 300V.

I know I should know this, but this is what I’m trying to learn…

Knowing that info and the info about the output transformer, how do I read the information from the 12AU7 to know what range of impedance values on the primary winding of the OT / tube plates are acceptable? …so the amp will be healthy, safe, efficient, all that?

Oh, I just saw 400V on the layout diagram. I guess it's inconsistent... but 300V makes more sense with the PT he has specified and a tube rectifier.

Still, 27K works just fine.

To grok these concepts, one could do a lot worse than read these web pages ... slowly:
http://www.valvewizard.co.uk/se.html
http://www.valvewizard.co.uk/pp.html

Essentially, the load impedance presented by the OT will control the relationship between voltage and current in the power tube(s) and OT. As a general rule, higher load impedances lead to more voltage swing, but less current swing. Lower load impedances lead to less voltage swing, but more current swing. To see this, you really need to see the load line plotted on the valve characteristics for the 12AU7. This next link is a handy tool for doing that without too much trouble. (I still like to print graphs and use a pencil and straight-edge!)

https://www.vtadiy.com/loadline-calculators/loadline-calculator/

 

[rdjones]

You need to keep in mind that an 8Ω speaker is really 8Ω at only one or two frequencies within the guitar spectrum. It's all just approximations and compromised choices beyond that.

Here's an impedance vs distortion and power curve for a typical output tube.
This doesn't apply to your tube, it's just an example to show what happens with impedance.

Do you want to optimize for distortion or power ?
Even harmonics or odd ?

A load line is a good start but it will give you a calculated value for a single point on the speaker's impedance curve.
Available transformers are based on the most popular ratios for the most tubes.
If it were me I'd optimize the load for the most powerful tube you expect to use (12BH7 ?) and accept any other slight mismatch as a compromise value along the curve.

 

[NTC]

A rule of thumb for triodes is to pick a load impedance that is 2 to 4 times the value of the plate impedance. A triode connected 6L6 has a plate impedace of about 1700 ohms. About 3400 ohms (2x) on the graph above gives the lowest distortion. That graph probably ISN'T for triode mode, however. A 12AU7 has a plate resistance of about 7700 ohms with 250V plate to cathode. 2x is 15.4kohms, 3x is 23.1k, 4x is 30.8k. So 22k or 27k should be ok.

Note this is for a single-ended amplifier.

 

[LOSTVENTURE]

Don't the tube manuals cover all this data. They did back in the 70's.

 

[joulupukki]

Awesome info everyone. Thank you @andrewRneumann. I’ll do some slow reading and studying with those two links you sent!

 

[Bendyha]

For a short rule of thumb, from an RCA app. - "Short-Cut Method for Determining Operating Conditions of Power Output Triodes." Using the " The Power Output Rule, or, Distortion Rule"

If you like this sort of stuff, I can post a lot more for you. (see it as an entrance dose, but be warned, it is easy to become addicted)

Edit; I see I attached the wrong AN, I've changed it to the correct one now.

 

[2L man]

I recommend to use this excellent loadline calculator because it is very fast to use. When you select the tube there comes "default" setting. For 12AU7 it is resistive and single ended so just tick to reactive and push pull boxes.

When tube is cathode biased the B+1 target is a sum of grid bias/g1 and V1/anode voltages. For fixed bias the V1 is B+1.

For push pull the loadline can safely go to almost double the max power curve because when it does the other tube cools Sometimes it is not good for sound and sometimes it isn't. When you see schematics which have voltages (currents you can calculate) you can test what their loadlines look.

Universal Loadline Calculator

This calculator allows you to trace the anode characteristics graph, and to compute and tune loadline and operating point of various vacuum tubes.

www.vtadiy.com

This is default 12AU7 PP/reactive loadline:

 

[andrewRneumann]

Just a note of context here. You want to learn this stuff...

…so the amp will be healthy, safe, efficient, all that?

The power tube(s) and OT have limitations/maximum ratings in 3 areas: voltage, current, and power. (Note that instantaneous power is voltage times current.) A load-line analysis shows you the relationship between voltage and current (and if you multiply them, instantaneous power). By checking the extremes of the load-line (max current, max voltage, max power), we can reasonably assure ourselves that we are being nice to our equipment. Guitar amps don't always follow the rules though...in fact, good sounding amps break some rules on purpose. These are musical instruments, not scientific instruments.

The load-line shouldn't be taken too literally vs reality, but don't let anyone convince you they're worthless. This is where design starts--we always use some simplifying assumptions to get ourselves going.

 

[joulupukki]

If I wanted to use a single output jack for the speaker and have it switchable, I think this would be the correct wiring, right?

Above shows the option if I were to use the 5 & 6 lines for the 22K plate-to-plate load for 4 Ohm output. This would use a DPDT switch.

Below shows the option if I use the 4 & 5 lines for 27K plate-to-plate load for 4 Ohm output. This would use a SPDT switch.

Also assuming a headphone output could work for either option?

 

[2L man]

Top wiring seems to change phasing between selections but should work when you cross change other pair between OT and Switch.

 

[robrob]

Oh, I just saw 400V on the layout diagram. I guess it's inconsistent... but 300V makes more sense with the PT he has specified and a tube rectifier.

I just reviewed all the Bassman Micro layouts and schematics and they all seem to say 300v B+1. Can you link to the 400v B+1 layout?

 

[andrewRneumann]

I just reviewed all the Bassman Micro layouts and schematics and they all seem to say 300v B+1. Can you link to the 400v B+1 layout?

https://robrobinette.com/images/Guitar/Bassman_Micro/Bassman_Micro_LTP_Layout.png

Note by the EZ81

 

[mictester]

Don't the tube manuals cover all this data. They did back in the 70's.

The simple reason they don't is that they assume that everyone is mathematically stupid, and scared of formulae....

The output impedance of a bottle is given by:

Zout = Vcc²/(2 X Power Out)

So if you have a 20 Watt amplifier with a 300V anode supply, you'll see an anode impedance of 2250Ω. That 2250Ω then has to be converted to 8Ω for the loudspeaker... You'd need a Primary of (say) 1200 turns, with a secondary of 72 turns.....