Power Scaling is Flawed?

[d95err]

maybe it's because i've been playing out of a truly large musical instrument tube amp for close to 10 years now that i've come to terms with what i'm talking about--there is SO MUCH DIFFERENCE between room response, cabinet response, speaker response, ear response, output transformer response, output tube response, etc, with different power levels, that trying to achieve a consistent tone across all of it will inevitably end in failure and disappointment.

Again, that applies to every method of reducing volume, so it is not specific to Powerscaling. If anything that is not the sound of a 100W cranked tube amp is disappointing, then I guess you're facing a lot of disappointments as a musician, because for the vast majority of musicians, it is rare to gig somewhere where you could actually get away with being that loud...

But, I think it's more interesting to start from the other end, and discuss how to make the best of the limitations of a specific situation. Let's say we have a really small club gig, and your favorite tone is a cranked 100W "Plexi". How can we get the best possible tone given the limitations?

Is it running the Plexi with the volume on "1"? Is it using an attenuator? Is it using Powerscaling? Is it by using a POD into the PA? Is it by using a small 15W amp that tries to sound like a Plexi? Is it by using the "Crunch" channel of a modern multichannel mastervolume amp?

I agree that Powerscaling is complex - when added to an existing amp. When designing and building an amp from scratch, it's fairly trivial and very cheap to implement.

And for the record, I'm not endorsed by Kevin O'Connor...

[Steve Conner]

Well, the reason I've never read the TUT series is because I think they're somewhat overpriced and overhyped, especially since I need to import them from the USA.

I did a lot of work on power scaling when I was an EE student back in the nineties, and came up with my own system to boost a small tube amp with a large solid-state unit, while reflecting the actual speaker impedance back onto the tube amp.

But around that time Korg/Vox brought out their Valve Reactor technology for their Valvetronix hybrid amps, and when I read the patent, it looked like my system was just a reinvention of the Valve Reactor, so I kind of lost interest. It's not really research if someone's done it before!

I've A/B tested the "Herzog" dummy load approach, and the difference isn't anything special. It sounds a little flat because the solid-state amp doesn't react to the speaker resonances and voice-coil inductance, which give extra bass and presence with a tube amp. But that can be compensated if needs be.

I bet that anything Kevin O'Connor invented that isn't a "Herzog" will be covered by the Korg patent too. I haven't figured out a way around it yet.

Oh, by the way, KG, I take it the BAGA doesn't sound anything like as good with 46 of its 48 power tubes pulled? The point I've always tried to make is that for a tone to *sound* consistent across volume levels, it has to change radically with volume in terms of its measurable parameters. I like to think that this is why many guitarists use lots of gain and scooped midrange to rock out at home, but for the gig, they use less gain and let the midrange rip out. It's instinctive power scaling.

[kg]

Again, that applies to every method of reducing volume, so it is not specific to Powerscaling. If anything that is not the sound of a 100W cranked tube amp is disappointing, then I guess you're facing a lot of disappointments as a musician, because for the vast majority of musicians, it is rare to gig somewhere where you could actually get away with being that loud...

you need to play heavier gigs.

and yes, that absolutely does apply to every method of electronic power scaling.

the best power scaler i've heard was a purpose-built isolation cabinet that completely encased a 1x12" closed back cab. an sm57 was buried inside with a rather trick arrangement that allowed one to adjust the position of the mic while keeping a hermetic seal. it was very quiet and allowed one to get some pretty killer miked tones while being barely audible.

But, I think it's more interesting to start from the other end, and discuss how to make the best of the limitations of a specific situation. Let's say we have a really small club gig, and your favorite tone is a cranked 100W "Plexi". How can we get the best possible tone given the limitations?

set an upper spl limit and design the entire amp accordingly.

it is hardly a new concept to build an amplifier "backwards," ie, from the speakers forward.

Is it running the Plexi with the volume on "1"? Is it using an attenuator? Is it using Powerscaling? Is it by using a POD into the PA? Is it by using a small 15W amp that tries to sound like a Plexi? Is it by using the "Crunch" channel of a modern multichannel mastervolume amp?

the lower you set your acceptable SPL the harder the task will be, and the more the amp will deviate from the original design.

I agree that Powerscaling is complex - when added to an existing amp. When designing and building an amp from scratch, it's fairly trivial and very cheap to implement.

agreed.

Oh, by the way, KG, I take it the BAGA doesn't sound anything like as good with 46 of its 48 power tubes pulled?

c'mon now. there are only 12 output tubes. i guess if you count the sv83 cathode followers as power tubes then there would be 16.

The point I've always tried to make is that for a tone to *sound* consistent across volume levels, it has to change radically with volume in terms of its measurable parameters. I like to think that this is why many guitarists use lots of gain and scooped midrange to rock out at home, but for the gig, they use less gain and let the midrange rip out. It's instinctive power scaling.

absolutely. an eq that sounds good soft does not sound good loud. and as you stated earlier, this is exactly what the FM curves predict. this is why if you're designing an amp for a particular purpose, with a particular output power level, build it that way from the beginning, and don't expect it to "scale."

ken

[David Schwab]

Many people are aware of the fact that our ears don't respond linearly to LF input. Snobby know it alls at The Gear Page casually throw this fact around all the time when they profess that there exists no good attenuator.

Interestingly, the ability of a TS ported array to produce bass response is measured/spec'd with small signal outputs, but the bass response of TS ported arrays falls off markedly when reproducing large signals.

Don't forget your Fletcher-Munson curves!

I don't know about snobby know-it-alls at the Gear Page, but the Fletcher-Munson curve has been known for a long time, and is why they have "Loudness" buttons on stereo systems. It boosts the bass response at low volume levels.

[d95err]

Well, the reason I've never read the TUT series is because I think they're somewhat overpriced and overhyped, especially since I need to import them from the USA.

Kevin has some resellers in Europe. I got mine from Owe the "Amp Healer" in Sweden where I live. I don't think they're overpriced. It's about the price you'd expect for a book printed in such small series.

I did a lot of work on power scaling when I was an EE student back in the nineties, and came up with my own system to boost a small tube amp with a large solid-state unit, while reflecting the actual speaker impedance back onto the tube amp.

But around that time Korg/Vox brought out their Valve Reactor technology for their Valvetronix hybrid amps, and when I read the patent, it looked like my system was just a reinvention of the Valve Reactor, so I kind of lost interest. It's not really research if someone's done it before!

Sound like a cool idea. I've always thought that the Vox approach is the most clever hybrid I've seen (using the valve for poweramp simulation, rather than preamp distortion).

I'm not an EE so, admittedly, the Super Scaling chapter of TUT4 makes my head spin. The basic approach seems to be to use an output transformer "backwards" to scale the speaker signal up to higher voltages then boost the current to the appropriate power level using tubes or solid state amplification.

[d95err]

I don't know about snobby know-it-alls at the Gear Page, but the Fletcher-Munson curve has been known for a long time, and is why they have "Loudness" buttons on stereo systems. It boosts the bass response at low volume levels.

...and this is probably the most misunderstood button in all of amplification history! Loudness? Yeah! We want loud! Push the button!"

[David Schwab]

...and this is probably the most misunderstood button in all of amplification history! Loudness? Yeah! We want loud! Push the button!"

It's actually short for "equal-loudness contour". The Fletcher–Munson curves is one of the equal-loudness contour curves used.

The sound system in our new Nissan does this automatically. As you turn down the volume, the bass doesn't decrease as much. It's actually too boomy for me turned down, so I keep it cranked up!

[Steve Conner]

That's pretty cool. According to that description, the Super Scaling stuff in TUT4 is the same as what I came up with independently.

Originally I started with the idea of just boosting the current, without increasing the voltage. So, you'd put your amp on its 16 ohm tap, plug it through some kind of booster box that multiplied the current by 4, hook up 4 ohms worth of speakers, and get 4 times the power. If you kept the original 16-ohm speaker, the reflected impedance would look like 64 ohms to the amp, no extra power would be added, and something bad would happen.

Later, I cottoned onto the idea of using the transformer. If you were building a super scaling amp from scratch, you could just add extra taps to the O.T., in fact the 70 volt line tap that some old amps had would have been great to drive a current booster for 1kW into 8 ohms

This "something bad" gets on to the reason why I never actually tried promoting this stuff or taking it off the drawing board: stability. Since the system has feedback in it, Murphy's Nyquist Law says that there must be some condition somewhere that will make it go unstable, and feed hundreds of watts of ultrasonic crap back into your OPT, frying it. A kind of audio version of what the General Electric Hardiman did to its first test pilots. I'd much rather Kevin took the blame for this than me!

Obviously the Herzog approach can't cause instability, since it doesn't involve any feedback. So this makes it more foolproof.

I didn't understand the Loudness button when I was a kid either. All I knew was that it added great thumping bass to my Metallica records at any volume.

[mooreamps]

Power scaling gives the benefit of dialing down the plate current on a power tube. External attenuators do not. Thus, the reduction in plate current will drop the total plate dissipation in watts. This is where the obvious benefit occurs. As for the change in tone with respect to the change in plate voltage, that's above my range in hearing perception. Besides, even if an amp does have a power scale knob on the front panel, it does not mean one "has" to use it.

-g

[kg]

Power scaling gives the benefit of dialing down the plate current on a power tube. External attenuators do not.

variac?

[mooreamps]

variac?

Yes, a variac can ; provided it is isolated from the filament power, and pre-amp power. It's an expensive option compared to some circuits of electronic power scaling which could only requires less than 10 dollars in parts.

-g

[Alf]

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

[d95err]

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

To this you'd need to add a separate transformer for the filamemts and/or the preamp voltages. Then add an expensive and heavy variac. It simply makes no sense, when you can get the same (or better) result using a MOSFET and only a few bucks worth of parts.

[mooreamps]

Well, the reason I've never read the TUT series is because I think they're somewhat overpriced and overhyped, especially since I need to import them from the USA.

With all due respect, I think that's a mistake. If you are really serious about learning how a tube amp works for the electric guitar, both the theory and math, I think you should reconsider your position on this one.

-g

[WholeToneMusic]

On page 3-94 of the TUT4 book KOC explains that speaker breakup does not give a player his signature sound due to the misperception of speaker breakup at higher SPL's. As the late geat Ken Fischer stated " Beauty is in the ear of the beholder."