Clean vs Distorted Wattage

[pdf64]

I would expect a Marshall to be louder than an Express, even given the same static VB+.
As the Marshall OT primary impedance is about half of that normally used for an Express
Try plugging the numbers into the calculator to see for yourself http://bmamps.com/Tech_tds.html

But Marshall may have a higher VB+, and almost certainly a stiffer screen grid node.
So it may well have a noticeable degree of extra grunt.

[vibratoking]

RMS wattage is mathematically defined and makes good sense to me. Much more relevant than peak wattage IMO. I'm not sure why RMS wattage doesn't exist?

Peak power. Continuous average power. At what frequency? The term RMS watts was dictated by the FTC to rein in power claims by amp manufacturers (Hi-Fi, not MI). But RMS is a voltage related term, not a power one.

The definition of RMS voltage is mathematically independent of frequency. That's one of the reasons it is useful in a strict sense. It's a time integral. I don't know the history of the FTC. I'm pretty sure that RMS voltage and corresponding RMS wattage (as a function of the RMS voltage and load) existed long before the FTC was created. You are right, there is an issue related to the frequency dependence of an amplifier and what frequency standard should be used. This issue is what many manufacturers have exploited to their own benefit and why, I believe, the FTC stepped in.

[Gaz]

Having a true RMS meter for this purpose is not really that important. I have a cheap $20 averaging meter that gets within 1/2W measurement of my Fluke. What I mean is that if the Fluke measures 50W, the averaging meter may read something like 49.5W.

[Firestorm]

RMS wattage is mathematically defined and makes good sense to me. Much more relevant than peak wattage IMO. I'm not sure why RMS wattage doesn't exist?

Peak power. Continuous average power. At what frequency? The term RMS watts was dictated by the FTC to rein in power claims by amp manufacturers (Hi-Fi, not MI). But RMS is a voltage related term, not a power one.

The definition of RMS voltage is mathematically independent of frequency. That's one of the reasons it is useful in a strict sense. It's a time integral. I don't know the history of the FCC. I assume that FTC was a typo? I'm pretty sure that RMS voltage and corresponding RMS wattage (as a function of the RMS voltage and load) existed long before the FCC was created. You are right, there is an issue related to the frequency dependence of an amplifier and what frequency standard should be used. This issue is what many manufacturers have exploited to their own benefit and why, I believe, the FCC stepped in.

Federal Trade Commission. 1974.

[vibratoking]

FTC it is. I'll fix my previous post to avoid misinformation.

Here is a link that discusses this a bit.

http://www.angelfire.com/electronic/fun ... Power.html

[pdf64]

Having a true RMS meter for this purpose is not really that important. I have a cheap $20 averaging meter that gets within 1/2W measurement of my Fluke. What I mean is that if the Fluke measures 50W, the averaging meter may read something like 49.5W.

Is that the case for both sine and square wave signals?

[Littlewyan]

Well this has started quite a discussion.

I wouldn't say measuring wattage is meaningless. Its something you need to know to determine what speakers you need to use and also in my case how rugged you need your attenuator to be. Mine will be rated for 100W which is plenty for the Express, as for the Marshall, we'll have to see!

According to the interactive valve data sheet my Express should kick out 41W and the Marshall 86W!!! However it obviously doesn't take into consideration the extreme sag of a Marshall Power Transformer.

I haven't got an RMS meter so I can't say for sure if my readings are correct. I know Gaz said his cheap meter is very close to his Fluke, however I haven't got a true RMS meter to compare my cheap meter to. Although I do know an Express should push out 35W so I'd say my meter can't be that far off.

[Scumback Speakers]

I've copied the power specs page from a JCM800 manual and turned it into a jpeg. It shows the 50w 1987 amp making 50w at a 3% clean signal, and over 90w with a 10% distorted signal into 16 ohms. Pretty sure those go up slightly as you go to 8 and 4 ohms.

The 1959 100w model is 115w clean at 4% distortion, and over 170w at 10%.

http://www.scumbackspeakers.com/jcm800powerspecs.jpg

[img1159]http://www.scumbackspeakers.com/jcm800powerspecs.jpg[/img]

[pdf64]

According to the interactive valve data sheet my Express should kick out 41W and the Marshall 86W!!! However it obviously doesn't take into consideration the extreme sag of a Marshall Power Transformer.

To account for sag, the HT and screen grid node Vdc should be measured under the 'max sine' & 'max square' conditions, and then used in the calculator.
The static voltages are only relevant if the amp operates in class A.

For a clean sine signal of lowish frequencies, eg up to 200Hz, all meters should give closely similar results.
The results will likely diverge as the the harmonic content increases, due in part to the frequency response of the ac sampling, and the ac-dc conversion / signal level averaging.

Jim, thanks for making those Marshall specs publicly available.
I've not seen them for a while.
Just to note that the 6dB range for the presence control is under the conditions of a resistive load.
With a real speaker, it's rather more than that, due to the inductance of speakers, and the significant output impedance of the power amp (once the global NFB is lifted at high presence settings, for high frequencies at least).

[Littlewyan]

Thanks for that Jim. Very helpful. Those old Marshalls definitely pack a punch!

I do wonder what a Marshall would be like with a TW Express Power Transformer, i.e. no sag. Probably a weapon of mass destruction.

When I get myself a load box I will retake the wattage measurements with a lower frequency. As I said earlier in the thread I used a 1Khz signal. I'm actually going to make my new attenuator a load box when there is no speaker plugged in.

[Scumback Speakers]

Thanks for that Jim. Very helpful. Those old Marshalls definitely pack a punch!

You are all welcome!

Duncan Boniface, former lead speaker engineer at Celestion, made me aware of that JCM800 specs manual page over a dozen years ago. I've been trying to make as many as I can aware about it so they don't think I'm just blowing smoke when I talk about clean and distorted power output from their amps.

There are quite a few amp makers I've found aren't even aware of it. They just quote the "clean, no distortion" power rating and say that's as much power as an amp makes.

Obviously the amp building gene pool here is a few notches higher than other forums I've posted this at (no names, sorry, I'm not trying to start a bunch of crap online).

I've got tickets for Spectre in 45 minutes, gotta go. Hope you all have a good weekend!

Jim

[Firestorm]

Something Aiken said in justifying his 70% max figure is interesting: a full power square wave results in the least plate dissipation because voltage and current are out of phase. But how does this relate to power delivered to the load? Transformers behave in ways that I 100 percent do not understand.

[pdf64]

Something Aiken said in justifying his 70% max figure is interesting: a full power square wave results in the least plate dissipation because voltage and current are out of phase. But how does this relate to power delivered to the load? Transformers behave in ways that I 100 percent do not understand.

It should result in the max power being delivered to a resistive load.
I'm not sure what % THD of a sine wave would describe a square wave of the same peak voltage - may be 41%, as the rms voltage ratio is ~1.41?
Whatever, way more that the 3-10% noted in Marshall's overdriven power info, and so we can expect that the actual square wave power from the Marshall will be even higher than that.
However, with speaker type loads that have a significant reactive component, the current and voltage may not be completely out of phase anyway, which will complicate things.

[Firestorm]

Something Aiken said in justifying his 70% max figure is interesting: a full power square wave results in the least plate dissipation because voltage and current are out of phase. But how does this relate to power delivered to the load? Transformers behave in ways that I 100 percent do not understand.

It should result in the max power being delivered to a resistive load.
I'm not sure what % THD of a sine wave would describe a square wave of the same peak voltage - may be 41%, as the rms voltage ratio is ~1.41?
Whatever, way more that the 3-10% noted in Marshall's overdriven power info, and so we can expect that the actual square wave power from the Marshall will be even higher than that.
However, with speaker type loads that have a significant reactive component, the current and voltage may not be completely out of phase anyway, which will complicate things.

Seems reasonable. How easy is it to get tubes to pass square waves from overdriven "sines?" Seems they would fight that. A little clipping is one thing, but...

Plus many Marshalls have greater than Fender feedback. Something else I completely don't understand (kind of like Dr. Michael Mann).

[Littlewyan]

Yes the Bassman used 27K on the 2Ohm Tap and I'm guessing it wasn't until Marshall went to 47K on the 8Ohm Tap that they matched that. Only guessing though, I can't remember how you work it out.

I need to read Aiken's article on the 70% dissipation

Jim - Hope Spectre was good. I'm going to see it tomorrow night!