power scaling

[daveS]

Hello amp experts. Sorry for the dumb and perhaps repeatative question but . . . .

What is power scaling ?
How does it play into the sound of the amp ?
What are some amps that have this ?
Is this someting that can be designed into an 18W clone ?

Thanks guys !

-d

[JonFrum]

Go here:



http://londonpower.com/pscaling.htm

[jersey_aaron]

there's a couple of threads discussing this here. The bottom line is nobody has had desire to actually purchase his $120 kit and build one.

[chainsawmillerman]

I mentioned this before a couple days ago in another thread. What I did in a laney clone (that was built using Kevin Oconnor's schematic from tut#3) was I used a seperate filament transformer to run the heater requirements for 4 6550's and 4 12ax7's( 2 el34 and 2 6v6 are what I'm curently using and like the best tone wise).The filament transformer uses a constant 120 vac.The power transformer is plugged into the variac.When used this way there no threat of cathode stripping.There is a tone change only at the lowest ac setting say 10 vac from the variac.If your starting from scratch it is not hard to plan for this.Variacs can be had for around 60.00 off ebay(mine was purchased for that amount, a Staco 10amp).Filament transformers can be had for 5.00. This is a form of power scaling and not as expensive or as difficult to implement.

[zaphod_phil]

But yes, you can buy the kit from KoC and add it into an 18W amp. But IMO, it's not the same amp anymore. I'm just using an attenuator from Rat Shack instead. A fraction of the cost and does essentially the same jo0b at the end of the day - and without having to mod the amp.

[jaysg]

Eventually, I will be building power scaling into a 36W. That said, it takes some space and I don't see it fitting into a standard 18W chassis. Something like an SLO or JCM800 era chassis with lots of extra room would be preferable.

[chainsawmillerman]

But yes, you can buy the kit from KoC and add it into an 18W amp. But IMO, it's not the same amp anymore. I'm just using an attenuator from Rat Shack instead. A fraction of the cost and does essentially the same jo0b at the end of the day - and without having to mod the amp.

I found the amp behaves closer to a cranked volume amp , feedback and all at signifacntly less volumes ie bedroom volumes. At least thats the experiance I had with the amp I did this to. It, like so many things is personal preference. The power scaling mod is likley out of the realm of the novice like myself, although I hav'nt actually seen the kit.It is a mod to the power supply right?

[chainsawmillerman]

Another site worth checking out is http://www.stephensonamps.com/40watthead.htm
He uses power scaling and Nickleback uses these amps I beleive.

[rjgtr]

So Power Scaling is really Power Starving of the B+ supply. This might easily be accomplished post power transformer with some kind of lpad on the B+ line. I know O'Connor uses a bunch of stuff, probably to scale voltage and current, but if a variac does the job pre-power tranny. then something as simple post PT should also be effective.

However, this does bring up an important issue, and why using an attenuator is probably still preferable. What about the filter caps? How does power scaling take into account the fact that filters tend to settle into the voltage they are being run at and don't like higher voltages after a year or so. So lets say a guy runs his power scaled amp at 10 watts for a couple years and then gets in a band. He goes to the first rehearsal, cranks his amp and poof! Lets face it if changing the recifier in an amp from a tube to SS can blow the caps due to age, wouldn't using power scaling possibly pose a similar threat?

[chainsawmillerman]

I think that would apply to using the variac method, if one was not aware of the point you made.I can not comment on the actual power scaling mod But I suspect there's more involved than what I proposed with the variac. One could shoot off an email to Kevin from london power to find out.

[loverocker]

But yes, you can buy the kit from KoC and add it into an 18W amp. But IMO, it's not the same amp anymore.

Same old propaganda, Phil, and still no actual explanation of the physics behind this claim (and I think I've asked you about 5 times now?)... And you've still not actually heard an amp being PSed.

When you say the amp is not the same amp anymore, that's nonsense - it patently is the same amp. Way moreso than, say, the legions of 18W Lites, TMBs, Trex's, ECL82/86 efforts, PI-mods and other variants are 'an 18Wer'. In particular, if one really wanted to be sure you had the same amp again, you just dial the PS up to regular B+ and there you go. Duh...

And, let's be perfectly clear here: the reason PS seems to cost a lot, is that the price everyone quotes is taken from the cost KOC puts on his kit, because folk don't understands the circuit outside of KOC's schematic and kit. If one knew the circuit and sourced the parts yourself, it's maybe $10-15.

As for worries about filter caps, I've not heard any stories of PSed amps blowing them. In effect, if and when you dial your PS control up to higher B+, I guess it's rather like reforming an amp's caps with the amp running on a VariAC? Maybe there's a rate at which you can do this safely/unsafely - I don't know. On my PSed amp, I sometimes run it at high B+ (only way you'll get headroom even at low levels!), sometimes at medium settings, and sometimes at low settings. Depends on my mood

Actually, without consciously thinking up any particular power-up/power-down strategy, I do tend to turn the PS control back up (and the Limit control down) when I power off. So my PSed amps always starts up with full B+.

[jaysg]

If one knew the circuit and sourced the parts yourself, it's maybe $10-15.

A minor quible on cost. You really do want to use RV4 pot(s). Those run about $10 each. Scale and Sag.

[jersey_aaron]

Jeez, I'm ready to buy one just to end the freaking curiosity.

[zaphod_phil]

Same old propaganda, Phil, and still no actual explanation of the physics behind this claim (and I think I've asked you about 5 times now?)...

Not propaganda. I have to cut into the B+ line of my amp in at least one place and install some alien circuitry. And I have seen the schematic of the basic power scaling kit. I'm personally just *not* very keen on doing that. And as for the science, it's a simple and well-known fact that when you change the voltages in an amp, you change all kinds of things about the amp's characteristics. KoC tries to compensate for that by installing yet more circuitry in other places. Hence my comments about it not quite being the same amp any more, regardless of whether that amp is a Tremolo 18W, Lite, TMB, JTM 45, Twin Reverb, or whatever. I certainly don't wish to cause offence with any fans of Power Scaling. However I'm just slightly skeptical of claims that appear to set it almost on a par with the discovery of cold fusion or perpetual motion.

[loverocker]

Not propaganda. I have to cut into the B+ line of my amp in at least one place and install some alien circuitry. (my emphasis)

Well, at least your perjorative words are showing that you've got something of a bias here. You could have simply said 'extra' instead of 'alien'.

PS is the same as dialing down the B+. Nothing more or less (except for the Sag control add-on if you want to add it). Making PS 100% bypassable is as simple as adding a SPDT switch.

Now, to state the bleedin' obvious, we all know that changing B+ has an effect on tone - that's the*whole* point! Problem is that when I see your propaganda, I'm not smart enough to do the maths on those valve response graphs to determine whether the behaviour of the amplification stages changes in an unwanted fashion as B+ is lowered (your claim). Curiously, I know you can work out this stuff... but you choose not to. I wouldn't mind so much if you kept your superstitions about 'alien circuitry' to yourself, but you keep repeating the same claim in several forums without ever explaining it. I'd genuinely love to see an explanation of your position...

This maths doofus can only go on what my ears tell me on the one PSed amp I have (so far), and PS does what it says on the tin. As chainsawmillerman says from his VariAC experiment, at ultra low voltages (we're talking whisper levels), there's a change in tone. Whether this is due to speaker non-linearity, the fact that the amp is now operating right down at levels close to the heater voltages, etc, I can't say.

You can listen for yourself at Bruce Clements site (the far more complicated addition of PS to a SuperLead). Of course, it's physically invasive to put into in an existing amp's chassis - because the hardest part is finding space - but in a new build, it's trivial (way simpler than adding reverb, for example).

jaysg - I sit corrected! You're quite right - the pot's the most expensive thing in the kit.

chainsawmillerman - when you used a VariAC, did you use an extra isolating transformer? As I understand it the VariAC's primary and secondary are connected on one side - not a good idea?

[zaphod_phil]

I'm sorry that your response is so emotive. I could say a few things about how bias points are affected (which they are), but feel disinclined to, after having been subjected to this tirade.

[rjgtr]

ANY attenuation method, including power scaling, won't sound exactly the same because the speaker will never operate the same. If you're using a JBL or EV the change in speaker response will be less noticable, but a Celestion will sound different at different volumes due to the nature of the beast. That part of the equation isn't compensated for with either a Hot Plate or Power Scaling. The other missing ingredient will be how a guitar reacts physically to the sound of the amp. Feedback techniques would surely be affected too.

Now I'm sure that power scaling is great and that the compensation circuitry helps to fool the ear. I found the clips interesting. However, clips can be misleading. Go to Weber's site and listen to the clips that a customer made of the Mass recorded at different volume levels. When recorded and the playback levels are compensated it sounds almost exactly the same at all but the most extreme settings. However, in the room you'd percieve a difference in tone. So clips are great but don't always represent the real experience in the room with the amp.

And I agree with zaphod_phil - it isn't the same amp. It is working in a different way than the non-modified circuit. That isn't necessarily a bad thing if you like the results. But I still wonder what the long term effects of modifying the voltages in an amp are to cap forming and cap memory. It could be none, but I'm not sure.

I still think the best way to achieve bedroom volume is with a simple smaller amp. The more complex the circuit the more opportunity to lose tone.

[chainsawmillerman]

[quote="rjgtr"]ANY attenuation method, including power scaling, won't sound exactly the same because the speaker will never operate the same.

Yes. With an attenuator thats hooked after the OPT it much harder on your tubes and OPT. Where if one were to use power scaling your tubes will actually last longer and it is much easier on the transformers. It's funny how somthing new that works is always rejected by the people who should know better.

[rjgtr]

I agree that it is generally easier on the tubes (though in some cases under powering the tubes could be a problem) and the transformer than running a normal amp on 10. Then again not everyone uses an attenuator with an amp set to 10. Many use it to adjust to different rooms and have their amps set lower. So the tube life/transformer argument is not always a large part of the equation.

However, you bring up another issue - the loss of transformer saturation. On some amps this can play a big part in how they sound when turned up. AC30s or an older GA-5 are a couple examples of amps that get a lot of their feel from how the OT is saturated. With power scaling you would lose this aspect of the amp's sound.

All of these attenuation methods will have side effects. Everyone loved the Tom Sholtz Power Soak, until they started blowing transformers siz months or a year later. I'm just wondering if power scaling might also have some issues. I don't know.

Hey, I'm fully ready to accept that power scaling is the best way to modify an existing amp to get a convincingly good sound at lower volumes. I'd just like to hear an amp modded with a kit in person to decide for myself. As I said, clips can be deceiving.

[KevinOConnor]

Hi Guys!

You totally expect me to say this, but it's true: Power Scaling does exactly what we claim it does. There is no voodoo; no black magic; no big secret; just simple science.

As we state on the site, and in great detail in TUT4 due out later this year, the only thing you have to do to maintain tone as you dial power down is to maintain the shape of the transfer curve of the output stage. This is the input/output signal response and is unique to the tube combination. This itself sounds like voodoo to the nontechnical, but it is ridiculously simple to achieve.

Past attempts to have variable power were partial solutions or cost-limited executions - not that the circuitry required costs much at all. Proper execution requires changing all voltages at the same time, maintaining their proportions, while keeping heater voltages at their correct value. As one poster stated, this is easily done with an auxiliary heater transformer and a variac to control the main transformer. This was also in TUT2, along with numerous other power control methods.

Although the power supply IS the other half of the signal path, you can make bigger changes here than in the audio path proper without having noticable tone changes. Adding a Power Scale Kit to a Marshall amp is no different than changing to another type of diode for rectification. In most cases, the diode noise is filtered out and the tone is a bit more articulate.

You can implement Power Scaling to exactly mimic what a variac would do, or set it up to add distortion, or set it up to just reduce maximum loudness without tone change. These options involve Scaling more or less circuitry, and sometimes adding one more control for increased versatility.

The ultimate limit to how quietly you can play is your own hearing and your speaker's dynamic linearity. You can't change your hearing, but you do have options with the speaker. Our detuned cabinet was designed to extend dynamic linearity downward while improving over-all efficiency. So, you can play wickedly loud or dial down lower than most other cabinets but still have full tone.

The comments regarding output transformer saturation are valid to a point. There are two approaches designers take with OTs: one is to minimse the influence of the OT over sound, the other is to maximise it. In the first approach, which I use in my amps, The OT can provide full power at all frequencies, therefore does not restrict or shape the freqeucny response of the amp. Dialling up or down, the tone is the same.

With the restricted, under-sized OTs typical in many designs, including Marshall, you get full response at low levels and narrower response at higher power.You can skew the drive signals as you dial down to mimic the OT effect but it will be a bit different. If this effect must really be a part of your sound, then intuitively an attenuator between the amp and speaker would be a better option than Power Scaling. However, passive attenuatorts have science working against them and will never fulfill the promise of consistent tone at any level.

Active attenuators are another option. They contain an amp and ideally some EQ-ing options. Unfortunately, the active attenuators that have been built thus far fail to meet the transparency level required and alter the sound as they dial down. They use incorrect circuits which happen to be the simplest possible. There are other circuits that work much better and utilise the power output of the driving amp instead of simply wasting it in a resistor. This is getting a bit off topic, but it is an option you will see more of in the future (and in TUT4, along with passive attenuators). It has the same disadvantage to tube life as any load box or passive attenuator.

Power Scale circuitry takes very little space. There is no problem installing it in any Marshall chassis. Fender afficianados should have it so easy. Bill Clement's initial problems were a part of getting the ealry PSK. At the time, I was still developing the notes that went with the kits, trying to find the balcnce of information required for successful installs. In the end, he and many others have variab;e power amps they are happy with.

Things are simpler with cathode-biased amps.

Have fun
Kevin O'Connor