Preparing for power scaling

[MikeM]

Hey guys

I am working on a BF Deluxe Reverb. I plan to put power scaling in it as soon as I can, but because I have a output tube bias trem and cathode and fixed bias switching, I first want to get the amp working before I add that in.

I read somewhere that when powerscaling, it's beneficial to keep the preamp tubes at their regular voltage, scaling just the output stage. Is this true? Would it be worth running 2 chains of filter caps, one for the output stage and another for the preamp side? What would be the best way to go about this? I am worried a big resistor might influence the current draw, but am also trying to stick to a very tight budget.

Thanks all
Mike

[Amp Kat]

How are you planning to Power Scale the Amp ? If you are just going to use a resistor that is really not power scaling but attenuation. Not a good idea and lots of heat in watts of dissipation. Power Scaling uses a Mosfet regulator and does scale the preamp also however there are versions where it doesn't so I gues it just depends on what and how you want to do it. I would highly suggest getting one of Kevin O'Connors power scaling books and explore exactly what you want to do before diving into something you may not really want as an end result. Just my .002

[MikeM]

I was planning to buy one of Kevins kits (well 2 as I would need to get his bias imedence kit as well).

I was more looking for an opinion as to weather it'd be good to have an independant PS for the preamp so that voltage stays constant while I scale the poweer tubes, or if there were any other considerations I should take while building my amp for his kits.

Mike

[exclamationmark]

I believe there are some versions which do scale both as stated, and some which don't. As far as I'm aware the main advantage to scaling both is the reduction in pre-amp output to the power tubes, which in their low voltage state reach clipping very easily (power tubes that is). Of course this can have other problems because you will be changing the bias point and input sensitivity of the gain stages which would change the tone.

Scaling only the output tubes can be a problem in non-master volume amps because of the giant swing the pre-amp tubes can feed into the sensitive output without the pre-amp section ever coming close to being driven hard. Pretty easy to get around, but will add complexity to the system. Then there's other methods like leaving the B+ as it is and simply dropping the screen voltage of the output tubes.

I haven't bothered with power scaling personally, so I can't really comment on what's best for preserving the 'cranked tone' or whatever you're after.

[soundmasterg]

I'd suggest to email Kevin O'Connor and ask him about his products and what he suggests. I got an SB-1 kit from him for a project amp. I would think something similar should work for your application. Generally people tend to power scale the power amp and phase inverter and let the preamp run as normal, though you can do the whole amp too. The earlier versions of Kevin's power scaling were similar to what Dana Hall is doing now with his VVR. The VVR only works up to 50 watt amps so I needed the power scaling since it works at higher wattages and my amp is 58 watts. The newer versions of Kevin's power scaling work differently and are probably superior to the older versions and to the VVR, though they are a little harder to incorporate into the amp.

Greg

[MikeM]

I think that's the smart thing to do, thanks Soundmaster, will drop him a mail again. Sent him a few mails last year and he seemed rather busy, lets hope I can make some headway now.

[Amp Kat]

He really is a way cool guy and used to come in here and chat with us but some gave him a hard time mostly just jealousy and he got very busy. I have owned one of his amps which is the Studio and it had Power Scaling in it and Scaled the power amp and preamp. One of the interesting aspects was that there were transistors that tracked the bias as the amp was scaled up and down. Kevins knowledge for SS mixed with Tubes is by far the best out there but his books are complex for a person that doesn't know much SS stuff. Volumes 4 & 6 which I have cover Power Scaling and attenuation amonst other stuff too. Volume 6 has a great section on high gain topologies along with Power Scaling and volume 4 has some Super scaling technology in it. Good luck either way and let us know how it turns out.

[Diablo]

If you want to play around with powerscaling by a different name, buy yourself a Variwatt from Skip. They're reasonably priced, and don't require you to purchase a book. You can easily change them over to scale only the power amp or include the PI and preamp. It would take less than 15 minutes with a soldering iron to change the configuration.
I'm using one in a fixed bias amp (Tweed Twin 5E8A), and one in a cathode biased amp (Matchless DC30 clone).

Home

[Amp Kat]

Can't see a problem with that and seems ok. FWIW that is basically Kevin's design and not sure of the legalities but it's almost an exact copy.

[KevinOConnor]

Hi Guys

Power Scaling implemented correctly almost never Scales the preamp - only the output stage and the preamp was never Scaled in any of London Power's amps. There is no issue of this causing excess distortion in the output stage if one installs the Drive Compensation control provided in the kit. Drive compensation can also be done automatically, thus providing a "one-knob" Power Scale solution. However, this reduces versatility of use compared to having both controls.

Dana and Skip both copied early versions of Power Scaling and did not really make good copies at that.

As TUT4 explains, keeping the preamp voltages stable is beneficial in amps where the full-scale idle power is high, including "class-A" amps and ones where the bias is simply set hot. If the supply is too marginal, the sagged full-load voltage will be much lower than the unloaded supply voltage, and this can cause a dead-spot in the sweep of the PS control. Using the VCK eliminates such issues.

All emails get answered regardless of how busy we are. It is important to remember that email is not 100% reliable.

Have fun
Kevin O'Connor

[Steve Conner]

Thanks Kevin. That's interesting to hear. I tried a kind of power scaling circuit on my latest build, and I spent a while scratching my head, trying to decide whether I should vary the preamp voltage too.

Being a SE cathode biased amp, it ended up very simple. I just varied the power tube screen voltage, so I could get away with a smaller FET and no heatsink. The cathode bias seems to adjust itself OK. And since the volume control is connected straight to the power tube's grid, I guess it is also the "drive compensation".

The tone seems to change a lot as a function of the DC current in the output transformer, which of course is necessarily scaled too. I can really hear the bass thin out as the current increases and the core heads towards saturation. I could get a better OPT, but I think it works out in a speaker-saving, Fletcher-Munson-compensating kind of way.

[MikeM]

Well I ordered today, lets hope it goes smoothly

[mooreamps]

The tone seems to change a lot as a function of the DC current in the output transformer, which of course is necessarily scaled too. I can really hear the bass thin out as the current increases and the core heads towards saturation. I could get a better OPT, but I think it works out in a speaker-saving, Fletcher-Munson-compensating kind of way.

The dc current through the primary of the output transformer is scaled with both circuits ; scaling the plate rail voltage or the screen rail voltage.. It was my impression, however, from reading some of the comments in the TUT series some tubes like the EL-34 only sounded best with the plate volts set around 500 volts or there abouts... Secondly, if you have changes in tone with your circuit, perhaps you over looked employing keeping the screen at AC ground in your circuit.... Also, I did put a version of my electronic power brake into a negative dc biased push/pull power amplifier, and I am not seeing any issues with the bias voltage tracking the power brake....



-g

[KevinOConnor]

Hi Guys

Doing partial Power Scaling generally results in some tone change. The AC grounding Gary refers to is important and is easy to overlook.

Reducing the current through the OT cannot push it towards saturation as those are contradictory events. However, reduced current via reduced loading can make it easier for saturation to occur, as is often the case with power transformers.

One benefit of all of the Power Scale copies is that more musicians will have the opportunity to save their hearing.

Be careful using "smaller" mosfets in linear applications. Mosfets are optimized for switching and the dies are made as densely as possible for this function and good economy. thermal resistance of the TO-220 case from junction to ambient air is 62C/W, where the TO-3P is only 40C/W.

Have fun
Kevin O'Connor

[KevinOConnor]

Hi Guys

My actual comments in TUT regarding EL-34s had nothing to do with how they sounded, rather about their reliability.

In 1994 when TUT was written, it was difficult to get a good current-production EL-34, and even Marshall was building amps with 5881s. Poor tube quality and continuous clipping of the output stage is a bad combination, so a lot of Marshall OTs had to be replaced.

EL-84s are a traditionally abused tube, as well. But... this is because of poor circuit design and a lethargy of complacency. With corrected circuit values EL-84s can provide their sound with good reliability.

In a properly Power Scaled circuit, any tube will provide its expected sound from full tilt to a whisper. Amps that do not exhibit this behaviour usually have under-sized OTs, which seriously compromises the full-tilt sound - a compromise chosen by the amp designer to get a specific full-tilt sound.

Have fun
Kevin O'Connor