Power Scaling is Flawed?

[Jag]

I have to agree with Gary, I am constantly referring to the TUT series. I've gotten a huge amount of knowledge from that series.

Steve, why don't you email Kevin directly at London Power? Just use the email on the site. He sells them direct and, IIRC, I thought import duties were not as bad shipping from Canada to the UK. Of coures, I could be completely wrong.

[Enzo]

A variac won't work , as you know of course, because it also brings down the heater voltage. You have to split them up .

Alf, remember that the reason he wanted a variac was to bring down the current higher mains voltage to something closer to what was common back when. If your amp was designed for 110v and it is running on 124v, then the heaters will be higher than normal right along with the B+.

I would agree that if you are experimenting with dropping the B+ from 450 to 390 or something, the heaters migh object. But in the application originally suggested - dialing 124 back down to 110 or whatever - then a variac does exactly what he wants.

[Steve Conner]

I owe an apology. I just realised that the prices on the London Power website are in Canadian dollars!

http://www.londonpower.com/books.htm

[kg]

A variac won't work , as you know of course, because it also brings down the heater voltage. You have to split them up .

Has anyone tried this , keeping the heatervoltage , and using the variac for the plates and OT ?

Alf

the specter of "cathode poisoning" from insufficient heater voltage is largely overblown and overstated imo and ime. perhaps it would adversely affect tube life if the amp were run for tens of thousands of hours with reduced heater voltage, but for fooling around and/or testing proof of concept, i've been doing it for years.

i certainly wouldn't worry one bit about doing it to a preamp.

fwiw, i have separate preamp and power amp chassis and i find better variac results on the power amp. no futzing around with heater voltage--it just tracks the plate and bias voltage as expected.

remember, there are some tones and distortions that come from starved cathodes that can't be reproduced any other way.

[Paul P]

I owe an apology. I just realised that the prices on the London Power website are in Canadian dollars!

That might have meant something a year or two ago but not any more.
Current exchange rate is 1 $US = 1.0025 $CDN.

Paul P

[Alf]

the specter of "cathode poisoning" from insufficient heater voltage is largely overblown and overstated imo and ime. perhaps it would adversely affect tube life if the amp were run for tens of thousands of hours with reduced heater voltage, but for fooling around and/or testing proof of concept, i've been doing it for years.

That's very interesting Ken. I have always read that it would be harmful to tube to overrun or underrun the heaters more than 10 % . But I for one don't believe that doing this for a short time would damage a tube. I am aware of the fact that you can't believe everything you read on the net.

What I meant is, that is would be possible to set an amp up with seperate filament and plate supply to be able to bring down the Plate voltages to get a whole set of starved plate sounds.

But what Kevin O'Connor is doing is certainly a much cheaper method .


How low can you go with the filaments before they affect the emission and sound ?

Alf

[BackwardsBoB]

Well, I guess that settles it.

The arguments presented here against power scaling are convincing, and I think that history bears out this conclusion.

It's a little known fact that in the late 60's and early 70's, bands of long-haired hippies with wild names like "Jimi Hendrix Experience", "Led Zepplin", "Cream", "The Rolling Stones", etc. performed very loud music in front of huge, adoring crowds. These artists are now lost in obscurity. The Fletcher-Munson curve and volume dependant speaker response, coupled with the low 5W-12W power output of car stereos of the day, rendered their music unlistenable on radio, and without airplay, the general public remains unaware of these musicians' very existence.

All one has to do is turn one of those silly volume knobs on a car stereo, home stereo or personal music player to see that the technology just doesn't work. Music that sounds good when played at its recorded volume becomes unrecognizeable and annoying with the slightest change in level.

But imagine if things were different! Concert halls could have more than one row of seats, and dance clubs could hold more than four people! Musicians wouldn't have to sell 20 versions of their releases for playback at various sound pressure levels! We wouldn't need an SPL meter and wouldn't have to look up the recommended level setting on a CD in order to play it! We could appreciate the chirping of a cricket without sticking it in our ear! But alas. Such is not the nature of the physics of sound.

Since power tube distortion only sounds good at levels that cause ear (and apparently brain) damage, the only solution is the traditional solution: Eliminate power tube distortion entirely when the amp's output level is reduced. Based on the pro-power scaling arguements, you'd have to come to the ridiculous conclusion that pre-amp distortion and power tube distortion are almost the same thing, not totally different phenomena that deserve separate treatment.

Perhaps in a few million years, our ears and brains will evolve and gain the ability to adjust to different sound levels in the way that our eyes adjust to different levels of illumination. Until then...

[mooreamps]

Based on the pro-power scaling arguements, you'd have to come to the ridiculous conclusion that pre-amp distortion and power tube distortion are almost the same thing, not totally different phenomena that deserve separate treatment.

They are not. Besides, one does not need to power scale the pre-amp for pre-amp distortion. In fact, most designers "don't" power scale the pre-amp for there "crunch" and "over-drive" channels. Besides, lowering the plate voltage on the pre-amp is for harp amps, not guitar amps. Thus, the reason why I designed my electronic power brake to scale only the power tubes.

Kevins circuit and the vrr or vvr, what ever the hell it is, are other ways to do it. I believe those circuits involve scaling the plate voltage on the power tubes, which I just do not personally agree with. I'm not suggesting it's the absolute wrong way to do it. I'm just saying by virtue of my background and experience working with vacuum tube transmitters, and my formal training as an ET in the US Navy, I seem to have a different school of thought how power pentodes behave. The fact that I have a working power brake would suggest that as well.


Secondly, a well designed system, in my view, would include keeping the filaments regulated at 6.0 volts.


-g

[Chuck H]

Gary,

Your "power brake" system doesn't involve reducing power tube plate voltage?

When reading another post on "power scaling" I remember reading one of your replies where you indicated that you had done the same thing without the use of solid state devices. But if your not reducing plate voltage then what your doing is not the same thing as "power scaling". Don't take offense, notice the quotation marks. I'm not saying your system doesn't scale power. Just that it's not "power scaling" as KOCs copyright applies to the description.

I haven't heard your version yet. But a couple of people who have say it's very good. So why is everyone so hot on reducing plate voltage if it isn't necessary? Seems like it's just more complicated. I've heard "power scaling" MP3s done by virtuoso players that really make that system shine. But I'd be interested in the truth. How does your system work? Do you just reduce screen voltage?

Chuck

[mooreamps]

I just reduce the screen voltage.
I keep the plates hot, because I think if you drop the plate voltage too low, it makes the tone too dark sounding. Also, I'm only regulating maybe 10 mills of current verses up to about a 100 mills of plate current. That's why my brake doesn't have the same issues of dealing with waste heat as the plate rail regulators. It's the I squared R factor involved here. It still regulates the amount of plate current flowing through the tube, and I also use it as a way to put the power tubes into a stand-by mode during power on / power off.

I don't think Kevin has any problem with it. I sent him a print, just to be sure.

-g

[BackwardsBoB]

You're not taking me seriously, are you?

Steve-

I'm sorry. I thought the part about the cricket was funny. Soon the only sin will be telling jokes badly.

I re-read this old thread today. It was clear to me that most of the anti-power scaling arguments also applied to car stereo volume knobs, and I'm pro-volume knobs. When I turn down the volume on the car stereo, the speaker tone changes with drive level, my ears percieve the tone differently, and I hear more road noise. I think AC/DC sounds better loud, but sometimes wifey's in the car, and you need to make compromises. If it wasn't so, we'd have the world I jokingly described. I just want power tube distortion when my wife is sleeping at the other end of the house, and I see lots of people attempting to get there with a single-ended EL84 that's still way loud and sounds bad to me. I hate those bee-cans. We must all be getting old.

My sarcastic point about pre-amp vs. power tube distortion was that if one argues that power scaling is a bad thing, and it is therefore appropriate to eliminate power tube distortion when you turn down an amp's volume, as is usually the case in a non-power scaled amp, whatever argument one is using (other than simplicity) should also apply to pre-amp distortion, and the pre-amp distortion remains as we turn down the master volume on non-power scaled amps with the usual drive/master control pairs. In other words, if, for some reason, it is best for the power tube distortion to go away when you turn the amp master down, why isn't it best for the pre-amp tube distortion to go away too?


There's no reason to powerscale the pre-amp and PI if it preceeds the master volume. This would cause major tone changes. If you place the master volume post-PI, you only need to scale the power tubes and "anti-scale" any NFB into the PI.

You can get the audible effects of power scaling by scaling just the drive, screen and bias voltages, but if you scale the plate voltage too, you dissipate much less heat in the power tubes (but perhaps more in the scaling circuit, depending on implementation).


Going way back in time, you lamented that there should be a non-proprietary power scaling solution. O'Connor's effort is to put the whole concept in the public domain so that he can use it without fear of future patent infringement. He doesn't document any one circuit arrangement. As part of his anti-patent protection, he describes a whole range of methods to achieve the effect, along with a bit of history. He does trademark "Power Scaling". You can use his circuit ideas on a one-off for free, and he only wants a pittance per unit if you want to mass-produce it and use the "Power Scaling" name. He makes money off the books and kits along the way. If you are worried that he's infringing some valid patent, I'm sure he'd be glad to hear about it and discuss it. He's into tube patents too. He even sells a small book that summarizes the results of his patent research, and he believes that most of them are prior art and that they are indefensible. He's about as "open source" as you can get.

[mooreamps]

He does trademark "Power Scaling".

If he does have a trademark for power scaling, then it must be canadian, because the document I have from the US Trademark Office has deemed both "power scaling" and "electronic power scaling" as generic terms ; and thus not eligible as a US Trademark. Sort of like, if you wanted to trademark the term "tone stack", or something like that. Perhaps thats why some of the major amp builders have come up with special names for their circuits ; like Air Brake, and yes Power Brake, and such.

If so, that's good, but a Canadian trademark is only valid in Canada.

-g

[BackwardsBoB]

His books have a list of known trade-marked terms they use. "Power Scale" and "Power Scaling" make the list. He doesn't put a mark when he uses the terms in his books or web site, though he may include a statement somewhere.

O'Connor explains his position here:

http://www.powerscaling.com/community/index.php?page=9

"Note that a "trade mark" and a "trade name" do not have to be registered to be recognized and protected. "Power Scale" and "Power Scaling" are trade names and trade marks of London Power and Kevin O'Connor, despite not being registered with the USPTO. That organization would like for you to believe that unregistered marks are without merit and protection. The only mark that is not legal to use is the circle-R for "registered". The "TM" symbol can be used for both registered and nonregistered marks alike."

I neither question nor defend his position.

Basically, if you're a boutique amp manufacturer, you learn to power scale from his work, and you make money including his circuits in your product, you have to buy him a pizza now and then, and he'll throw some business your way. Sounds fair to me. If you've developed your technology on your own, and you choose to read his books, I'm sure he'll consider the price of the book fair compensation.

[Steve Conner]

Backwards Bob, your argument is kind of mischievous

First of all, when those little-known bands recorded their albums, the amps were cranked pretty damn loud in the studio. Then when the album was mixed, everything was EQ'd so that it would sound good at nominal listening levels: in other words, compensating for the Fletcher-Munson thing. Also, back in the 70s, people had loudness buttons and tone controls on their stereos, and weren't afraid to use them to compensate different listening levels.

Second, I really wish that artists would bring out different versions for different listening levels! The standard multiband-compressed-to-hell version for Joe Public to listen to in his car, and another one with the dynamics left in, for audiophiles to enjoy on a high-end stereo in a quiet listening room.

Thirdly, if we assume that power tube distortion is somehow different and better than preamp tube distortion, why shouldn't we also believe, say, that the distortion of power tubes running at full rated voltage and current is different and better than the distortion of power tubes scaled down to 1/20 of the power they were designed for?

This is my own personal opinion. I think that the only difference between power tube distortion and preamp distortion is that power tube distortion happens at thrilling volume levels that shake your whole body. If we scale power tube distortion down to half a watt, no matter whether we do it by reducing B+ or using a power brake, it just becomes another kind of cheesy fuzz box.

The only thing that it does that preamp tubes don't is the compression and pumping effect that comes from sagging B+, grid blocking, and cathode bias pumping up. But a savvy designer could easily simulate these in the preamp. I'm sure I remember seeing Marshall added a pair of diodes to a post-PI MV, to tick most of these boxes.

Finally, people like single-ended guitar amps, even though they don't sag B+, and the response of a SE power tube under overdrive is identical to a preamp tube. In other words, in a SE amp, preamp tube distortion *is* power tube distortion: there's no distinction. And yet THD Univalves and Valve Jrs fly off the shelves. This is proof that the compression effects in a Class-AB P-P power amp aren't fundamental to guitar tone.

[BackwardsBoB]

Now that's an argument, not just a series of contradictions!

I don't agree with much of it, but I think most of our differences are subjective.