Wattage and headroom

[txstrat]

Hi there,

still tinkering with my B-15 clone.
I installed a 135 watts ultralinear OT recently but I expected more change in headroom compared to the 60w OT I had in there before.
The amp runs on 423 volts on the plates 'only'.
Could anyone give me an estimate how much power I will lose compared to a voltage of around 500 volts?
Or in other words is there a way to calculate how many watts the OT puts out at max with my given 423 volts?

thanks

Matt

[eschertron]

I'm not familiar with the material in your other threads, maybe a link?

To get the power available from the output tubes, not only the plate voltage, but screen volts and bias point is needed along with the valve complement. The UL transformer may give you more bandwidth, more headroom, less distortion, or a combo of all three. depends on the power amp as a whole. How does the OT primary Z for the new tranny compare to the old one? If it's different, that may be the most crucial change in amp performance.

edit:
http://music-electronics-forum.com/t17328/
http://music-electronics-forum.com/t41033/

[Leo_Gnardo]

Could anyone give me an estimate how much power I will lose compared to a voltage of around 500 volts?

What Eschertron said ^^^, and

Back of the envelope calculation shows me you will have about 70% of the power with 423V B+ compared to 500V. Turned on its head, you'll have about a 40% boost in power by going to 500V from 423V. Note these are B+ voltages at full power, not at idle.

Or in other words is there a way to calculate how many watts the OT puts out at max with my given 423 volts?

Better measured, and we don't know if you have 2 or 4 (or more) output tubes. I'd give it a WAG (wild ass guess) of 60W with a pair of 6550 or KT88, near double that with 4 output tubes.

Moving to a bigger OT doesn't mean you have more power, it means you could have more power if other factors were adjusted. Like a higher B+, more output tubes. Analogy, if you have 1 liter of water in a 1 liter bucket, that's what you have. If you put your 1 liter of water in a 2 liter bucket, you still have 1 liter of water.

All else being equal, an advantage of your larger OT would be clearer low frequency production, iow less distortion as you send low frequency notes thru it.

Let's get some more info, and better tech minds than mine will help you dial it in, I'm sure.

[pdf64]

Plug the numbers in to nickb's calculator Interactive Valve Data Sheets
The voltages need to be as measured at full power, not idle.
Looks like you've only got 2 x 6L6GC?
The OT primary will likely be around 2k, rather too low for best power output from them.

[Dave H]

The amp runs on 423 volts on the plates 'only'.
Could anyone give me an estimate how much power I will lose compared to a voltage of around 500 volts?

A very rough estimate would be with 500V you will get about 1.4 times the power compared to 423V. e.g if it's 30W at 423V it will be about 40W at 500V. That's only 1.45dB. I don't think you'd notice the increase in headroom.

Edit: I see the question is already answered, must learn to type faster.

[eschertron]

Plug the numbers in to nickb's calculator Interactive Valve Data Sheets

nickb's website? That's awesome!

[txstrat]

Ok I think I left out some information on the amp.
It WAS a two 6L6GC poweramp with a 60W OT.
It is now a four 6L6GC with a Twin Reverb OT made by Hammond (1750WR) with 2Kohms primary. The PT winding is 300v ac at 500mA. Which is about 423volts after the rectifier.
With the added parts the amp just doesn't feel like it's power was more than doubled.
What the heck?

I tried to beef it up a little by changing to a long tail PI (look here: http://music-electronics-forum.com/t41033/)
It got tighter but could still be better.
I even have the feeling that I have to turn it's volume up more to get the same headroom (as before I changed the PI).

As for the PI I've used 1M grid resistors, 680 ohms cathode resistor and a tail of 22k and 100 ohms to ground. The NFB resistor is 10K.

[DRH1958]

Let's assume you did exactly double your power. That is only an increase of 3dB which is just about enough to notice a difference. As you configure the power tubes from pure pentode mode to closer to triode mode, which is what a UL tranny does, you lose some power, so UL OTs sacrifice some power for having less distortion. So you probably didn't even double your power which results in possibly not being able to notice much improvement. A 300 volt output from your PT is what is holding you back some as the above posts correctly say.

[Dave H]

Let's assume you did exactly double your power. That is only an increase of 3dB which is just about enough to notice a difference.

Never truer words spoken, I have a 15W 2 x EL84 amp and a 30W 4 x EL84 amp and through the same speaker the 30W amp is just about loud enough to notice.

[txstrat]

Before I changed to the long tail PI the amp was (well, at least it felt) louder. That was with the PI of the Ampeg schematic. What I mean is, for the same amount of headroom I didn't have to turn the volume pot as far up.
The drawback was the 'sort of' compressed sound.
Would it make a difference if I disconncet the UL wires and feed the grids from the plate wires?
Regarding the 300 volts of the PT. I know a voltage doubler circuit. Is there a way to lift the 300 volts to 350 volts or so?

[DRH1958]

Have you tried more efficient speaker(s)? Or add an extra cab? So many people look for solutions in the amp and forget about speakers. Let's say you get a speaker with 3dB more efficiency if you are just using one. This is equivalent to doubling your amp's power. So now you've got some extra power with the extra tubes, cleaner sound with your UL OT. Now maximize it with either a more efficient speaker or add another cab.

If you connect your screen grids to the plates, you have a triode and you will reduce your power even more. Here's a good tutorial on the subject: https://www.ampbooks.com/mobile/amp-...y/ultralinear/ You will be eliminating the purpose of the screen which is to accelerate the electrons better than a triode can with just one grid thus losing more power.

Sorry, I can't advise you on the voltage doubler. I've never done one. The other option since you are thinking about not using the UL taps would be to create another node in the power supply and configure it the normal way. You can run the screens at a higher voltage. This way you can get back some of the power you lost with the UL OT and with the more efficient or extra speaker, you can make use of this power without speaker breakup. To experiment with less cost, you could use a dropping resistor instead of buying a choke.

[txstrat]

Sorry I wasn't precise enough. What I meant was to connect the grids the normal way with a dropping resistor.
A second cab would surely add some headroom but I'm still on my way about checking the power of the two 6L6 compared to four 6L6.
OK I'll try to not using the screen taps and change the values of the PI.
Then I'll report back.

[txstrat]

I changed the PI to 5F6A values including the 27k NFB resistor. Put a 4K7 restistor to ground instead of a pot.
Disconnected the grids from the UL taps and hooked'em to the b+ after a 1K resistor (which I use instead of a choke).
The amp isn't much louder (headroom wise) but I can turn it up to 3 o clock with no perceptible distortion.
That way (and with a 4x10 Glockenklang cab) it was loud enough for a club filled with 80 People (tested last saturday night).

[frus]

I changed the PI to 5F6A values including the 27k NFB resistor

Which output tap? 5F6-A had a 2 ohm secondary only, and 27k/5k works for those values only. Using this calculator:

https://www.ampbooks.com/mobile/ampl...ck/calculator/

we get that 5F6-A had around 54% closed-loop gain as a percentage of forward gain.

Now enter your values, and you will get your NFB resistor.
For instance, if you enter 2k plate-to-plate primary impedance (since you have 4 tubes OT) and, for instance, 8 ohm tap, your NFB resistor would have to be 88k for the amp to have the same amount of NFB as 5F6-A. Otherwise, if you just put 27k you would have much more NFB

[Chuck H]

WRT experimenting with the PI, pay no attention to the actual number on the volume knob!!! Why should it matter? The power amp is the limiting factor. How much preamp signal you need to put into the PI to drive that power amp into clipping is moot as long as you can, indeed, reach clipping. If one PI design sets the amp to clipping with the volume knob on 3 that doesn't mean that PI design offers more headroom than a different PI that requires the volume knob to be at 5. The "compressed" sound and other elements are a different matter.

Already inferred... I probably wouldn't bump to 500V and keep the UL connection. The screen grids will hate it and it won't offer more power. UL was actually developed as a way to get more triode like linearity and distortion figures from tetrode and pentode valves without the associated loss of power that from conventional triode operation of those tubes. It wasn't intended to increase power. And it doesn't. I'll speculate that UL only has that reputation because of it's use being limited to higher power instrument amps (Dr.Z not withstanding).

Regarding a voltage doubler... How would you intend to re tame the now too high voltage? Even if you have an idea for that, it isn't free. Your PT would probably need to be two and a half times greater than needed in the amp now to pull this off without stressing the transformer. The actual volume advantage you'd get by upping the volts is tiny. Also already covered. But it would take doubling the power to make a notable difference. A 15% or 20% increase will go unnoticed. I wouldn't bother.

If you REALLY want to get the most volume out of that amp I think the speaker advice given is good. You could also put a serious eye to the frequencies the amp is expected to make. If, say, the circuit is capable of, and making generous amounts of power down to 40Hz (probably more like 30Hz), but the speaker system is rolling off hard at 50Hz (probably more like 55Hz), well... That LF is using a lot of power that the tubes now can't use for frequencies at 50Hz and up. Which would be your useful range. Wasted watts.