Variable Supply Voltage Regulation [e.g. Power Scaling (TM)]

[greenm01]

Hello, I'm new to the forum and thought I would re-post a thread I started over at diyaudio; it appears there are more guitar amp related tube enthusiasts on this site:

I took up guitar about a year ago, and like any self-respecting engineer and DIY audio person, I've put a lot of thought into the gear I use.

Without resorting to software modeling (which can produce fairly decent results at low volume), a significant problem for a guitarist is "authentic" power tube distortion/saturation at sane bedroom-household levels (0.5W or less).

Obviously one can achieve nice non-linear distortion from the pre-amp section (both solid state and tubes) and/or effects pedals, although there are various claims that power tube saturation is the "real" deal..... Is this proposition a myth? I don't know....

In my online travels I have discovered London Power Scaling (TM) and other similar products (e.g. http://www.hallamplification.com/main.html?src=%2F#2,2 and http://www.skipzcircuits.com/VariWatt.html). I admit that I purchased the entire TUT series of books - a very worthwhile read - although I find it helps to find perspective from other reputable engineering sources.

London Power claims that Power Scaling enables low volume playing without sacrificing or altering tone. It's my limited understanding that Power Scaling is basically a variable supply voltage regulator (Full disclosure - I have not yet read through TUT4 in complete depth).

I'm a skeptic at heart..... If we change the operating voltage of the power section (lets say 6L6 tubes), won't this in turn change the saturation characteristics, and thus ultimately the tone? The plate characteristics at 300V, on the 6L6, are different from those at say.... 100V. From the purist perspective, this fact alone would imply different performance at lower voltage -> thus a different "tone" and distortion characteristic vs. operating at full scale voltage - which is supposedly the holy grail of guitar distortion to begin with....?

Another thought: are the negative tonal effects of a power attenuator (e.g. THD Hotplate) "worse" than operating the output stage tubes at less than full voltage (as originally designed)?

[John_P_WI]

Don't get too caught up on it. From my experience and many others here it all depends on what you scale. I predominately have built "Marshall" styled amps and have scaled the output plates and screens. Use a good master volume, KOC calls it the improved bootstrapped MV and use a voltage clamp to prevent the preamp voltages from rising as the power amp is dialed down. Check out what he calls the Borris mod in Tut1.

Power scaling is a tool like other voltage modifiers, it works well, but like other methods including attenuation when you stop pushing air with the speakers they all sound like a buzzy am radio.

[bcmatt]

Yes, it does seem to change the distortion characteristic of the tubes distorting. But it is subtle and rather linear throughout the scope of being dialed down.

Nothing is as satisfying as playing loud, and having the speaker respond to lot of power being thrown at it. But, apparently no one wants to hear my amp loud except myself, so to interact with society, I must bring down the volume.

I find that when I turn down the voltage with VVR, the headroom also sightly decreases. So, I find myself also turning down the gain slightly as I go as well in order to get the same tone. Minor voltage adjustments don't require gain adjustments, but dramatic ones do. I don't feel that there is any loss of tone quality unless going to extremes with the voltage reduction, but I do suspect this has more to do with the lack of power the speaker is receiving.
Another aspect is that I feel like the highs are also slightly attenuated with extreme voltage reduction. Again, this could be more to do with how our ear/brain deals with loud treble. Turning up the amp's treble control also seems to perfectly compensate.

To illustrate how well I feel it works, I pretty much never run my amp at full voltage, even when I can, because I notice no advantage. When recording at a studio where volume doesn't matter to the mics, I find myself turning up a bit but still usually only running at about 50% voltage.

I have also found that on some amps I have installed it on that I again start to lose gain at the very bottom extreme of voltage. I could be mistaken somehow, but I felt I was losing saturation at whisper levels. It may have been more the effect of not having the speaker react with the guitar pickups where you would normally achieve a bit of feedback.

VVR is still my favourate method of attenuation. I prefer it over attenuators. The Badcat Unleash is a reactive load (behaves like a speaker on the amp), and then re-amps it with solid-state power. It works much better than typical resistive load attenuators. I do like it in some cases, but I've been finding myself not using it where I have VVR just because of the convenience factor of VVR being already in the amp. Then there are the tube longevity benefits of running at lower plate voltages....
As far as what sounds/feels better, I find it so subtle, that I find myself preferring either for random amp situations. That says a lot for a convenient bit of technology costing less than $50!

[ampedup]

I did it on a couple of amps and didn't really like it. I find a master volume works best coupled with a lower wattage amp. Even my firefly at less than a watt is loud as heck.

[greenm01]

Power scaling is a tool like other voltage modifiers, it works well, but like other methods including attenuation when you stop pushing air with the speakers they all sound like a buzzy am radio.

Yes, playing live in front of a guitar cabinet presents a very important feedback loop for the guitarist during performance...... but when monitoring through headphones or studio monitors, via a miced cabinet located in another sound isolated room or via DAW amp simulation.... we loose that "pushed air" dynamic feedback loop. In this case all we have left to judge is the recorded tone.

It's impractical to record at home with high SPLs. If all attenuation methods (active, passive, and VVR) muddy the tone in significant manner, maybe the only practical solution is to rely on pre-amp distortion for home recording and practice..... or bite the bullet and resort to modeling amps that sound "believable enough" at any volume.

[greenm01]

Yes, it does seem to change the distortion characteristic of the tubes distorting. But it is subtle and rather linear throughout the scope of being dialed down.

Thanks for sharing your real world experience in regard to VVR - very insightful.

I find that when I turn down the voltage with VVR, the headroom also sightly decreases. So, I find myself also turning down the gain slightly as I go as well in order to get the same tone.

This observation makes complete sense if you look at power tube saturation curves such as the 6L6. At lower voltages the transition to saturation occurs at a much faster rate than at higher voltages, i.e. the curve's knee has a sharper bend.

I find myself preferring either for random amp situations. That says a lot for a convenient bit of technology costing less than $50!

Yes, another astute observation from real world experience - thanks. From an engineering perspective VVR is the most elegant solution for reaching saturation at lower volume , but we should recognize that the tone and non-linear distortion will not be the same when operating at lower voltages - tone WILL change significantly for the trained ear - this is not a transparent process.

Is there a good explanation as to why power stage saturation is so much more desirable than pre-amp distortion? Is this another unjustified guitar myth... what is the real difference....?

Square waves distortion will sound buzzy and nasty regardless if it's produced in the pre-amp or power-amp..... in this case we may as well resort to a fuzz-box and be done with it.

[greenm01]

I did it on a couple of amps and didn't really like it. I find a master volume works best coupled with a lower wattage amp. Even my firefly at less than a watt is loud as heck.

Yes, I built the 1W Firefly and she's extremely loud with 100.5dB sensitive speakers..... way too loud for playing around the house in fact.... and this amp has very little clean headroom.

[Milkmansound]

I like the vvr circuit. Tubes don't sound great with 50V on the plates, but if you use it like another tone control it can a very neat addition. Or if you just want to noodle around at home with a little bit of crunch it's fun.

[teemuk]

For "apartment loudness levels" I much more prefer using preamps and headphones than anything that must be cranked in the "final stage" to achieve a certain sound. Yep, the acoustic feedback is lost but playing in an apartment is a compromise anyway. I do like tone that I can "feel" but that's usually impossibility in an apartment. Headphones seem like better alternative than listening to a "whispering" speaker. Perhaps we guitarists start to need "room processors" too.

I do think the detailed "knee" characteristics of the clipping are not as important as the entire circuit architecture and its impact to signal behavior and tone. For that reason I much more fancy alternatives where that behaviour is modelled in low current realm, whether its done by DSP or by analog processing using solid-state or tube circuitry. (Nothing stops implementing solutions like power supplies with variable voltage regulation here either). Such circuitry can be later consistently amplified with headphone amp or a separate power amp if you need higher loudness levels. Plus you can run that signal through ordinary effect chains.

[John_P_WI]

Green, since you are concerned over "real world" application, I gave you the recipe. Using London Power's SB or SV kits, a good master and voltage clamping on the pre and PI. It is the preamp that is held constant through all range of power scaling that you spoke of above coupled with a good MV that minimizes tone change.

The VVR implementation IS NOT THE SAME as London Power's power scaling. Read Tut 4 and 6? which talk about the SB kits, you will quickly realize the placement of Va, Vs and the bias control circuit is different. Additionally, there are MANY ways to implement either Power Scaling or VVR. Some affect the tone more than others. In fact, the London Power kits are easily switched allowing one to link it to channel switching for "instant on and off".

For your "trained ear" what you do lose with either VVR or Power Scaling is the core saturation of the OT as the power amp is dialed down - some "smoothing" is lost here. You also lose back emf from the speakers - affecting the feel as well. It is not only the reduction in voltage on the tubes that causes the change - remember the amp is a system. Lack of "transparency" can not be attributed to only one factor ie, the reduction of voltage on the output tubes or their operating point, there is much more to the "system" than this.

I believe TUT4 talks about the slope or curve of change which can be changed to account for bias shifts as well - I have not played with this.

Your biggest issue in implementing either power scaling or VVR will be lead dress and minding grounds, as VVR and Power scaling exposes the weakness in these two points. Cleaning up the power supply greatly enhances the tone in my real world experience too.

Try it, you will like it, believe me the benefits far out way the minimal tone change when implemented correctly, or build a Dumble and use a Dlator as the MV extensively.

[greenm01]

Using London Power's SB or SV kits, a good master and voltage clamping on the pre and PI. It is the preamp that is held constant through all range of power scaling that you spoke of above coupled with a good MV that minimizes tone change.

Yes, I eventually plan to employ Power Scaling an amp that relies on power tube distortion for its signature sound - the Trainwreck would be a good candidate?

http://home.polstra.com/amps/wreck1/scope/

Read Tut 4 and 6?

In progress; it's a ton of good information to plow through and absorb.

or build a Dumble and use a Dlator as the MV extensively.

Yes, in TUT6, Page 4-15, KOC mentions that the Dumble is a master-volume amp that produces its signature tone in the pre-amp. It's interesting that the Dumble and Trainwreck have such different topologies, yet both produce such sought after tone.

[bcmatt]

Yes, in TUT6, Page 4-15, KOC mentions that the Dumble is a master-volume amp that produces its signature tone in the pre-amp. It's interesting that the Dumble and Trainwreck have such different topologies, yet both produce such sought after tone.

It is interesting, but they are very different styles of amps and feel very different and you end up playing them quite differently.

[greenm01]

For "apartment loudness levels" I much more prefer using preamps and headphones than anything that must be cranked in the "final stage" to achieve a certain sound.

Unfortunately this is a hard truth. Maybe the best we can hope for, in noise constrained environments, is headphone monitoring.

I much more fancy alternatives where that behaviour is modelled in low current realm, whether its done by DSP or by analog processing using solid-state or tube circuitry.

I agree this is the practical solution, although DSP modeling techniques will never fully emulate the non-linear behavior of an analog system. Close - definitely yes and "good enough" to fool most people in the majority of circumstances - but not "authentic" in the classic sense of the 50's and 60's analog technology that gave birth to the electric guitar as we know it today.