voltage dropping in the power supply related to VVR

[wittyjeff]

I'm working on a Marshall style 50W amp based on Steve's Switchable Dual Preamps build.  I've added an "Ultra High Gain Preamp" based on Merlin Blencowe's design on page 260 of his designing guitar preamps book.  Since it's another fresh build on a blank slate, I'm trying out a host of mods and features. 
One goal I was hoping to achieve was to be able to use different power tubes.  Of course the EL34s are what the amp is based on, but it would be interesting to see the changes in tone and performance when using a range of tubes from 6V6 up to KT88s.  This, of course, requires some ability for the amp to supply the current and voltage in a wide range.  I'm thinking that being able to adjust the main B=/HT voltage will be useful to tone things down for 6V6s and crank things up for the KT88s.  On the schematic attached, I have two variants for how I'm thinking of setting up the Variable Voltage Regulator. 
In the first variant, the VVR would just adjust the overall B+/HT of the entire amp where increasing voltage for the power tube section would also increase the voltage for the preamp section. I am aware that any change in the HT voltage in this design will require re-biasing (since I have not included the bias circuit in the simple VVR). I'm ok with that, I'm not going for full power scaling and using the voltage like a Master Volume... just for adjusting things when changing power tube types.

My concern is that I'm not sure what an increase or decrease in the preamp voltage too far outside of ideal is going to do to the preamp performance or tone.  I know there's some flexibility in voltages, but if changing the HT by 100V... I imagine there's going to be consequences in the rest of the amp.

To deal with that, I proposed a second variant (noted on the schematic) where the VVR only affects the power amp section.  My question in this scenario is how to drop the voltage efficiently from the 474V down to the 350V I need to send to the preamp.  I know that several amps include a resistor to ground like the one I've circled in red labeled "1.".  I know it's elementary electronics, but I've never figured out how to calculate what those do and end up with useful results.  Does that resistor to ground take some pressure off of the 5W voltage reducing resistor?  Seems rather primitive to make that one 20K 5W resistor do all the voltage reducing work (I just guessed at the 20K value).

Finally, when increasing the voltage significantly for, say, KT88s... will my bias circuit be able to provide enough bias voltage?  Are there modifications to the bias circuit that I will need to implement in order to give it more range?

I'm sure I have other whacked things on this schematic and probably some errors also, so feel free to peruse and comment as you wish.
Thanks in advance for your helpful suggestions, advice, and comments.

[sluckey]

I like the alternative circuit better. There's a big difference in bias voltage requirements between 6V6 and KT88. If you make your bias voltage adjustment range wide enough to cover both types all you'll need to do is readjust. Of course when you widen the adjustment range, the pot becomes touchy to adjust. A 10 turn trimmer would be very useful.

[wittyjeff]

I like the alternative circuit better. There's a big difference in bias voltage requirements between 6V6 and KT88. If you make your bias voltage adjustment range wide enough to cover both types all you'll need to do is readjust. Of course when you widen the adjustment range, the pot becomes touchy to adjust. A 10 turn trimmer would be very useful.

Great suggestion.  To I know to increase the bias range I need to lower the value of the Bias Range Resistor which is currently in this schematic quite high (220K).  I'm sure I'll need to tinker with it, but do you think 100K is low enough or perhaps too low for a starting point?  I've got 350V coming off that half of the winding.
I have some additional minor questions noted on the Bias schematic attached.

[sluckey]

This is stuff that you can just experiment with before you put any tubes in the sockets. If you can get a negative voltage on pin 5 of the output tube sockets that's adjustable from about -30 to -70, that should cover any tube you use. The 220K and the value of the pot will mostly determine the range. A bigger pot will give you a bigger range. I would connect the lower pot lug to the wiper as a safety precaution.

[wittyjeff]

Thanks and will do.  That's the info I needed on the bias circuit.  Any thoughts on:

My question in this scenario is how to drop the voltage efficiently from the 474V down to the 350V I need to send to the preamp.  I know that several amps include a resistor to ground like the one I've circled in red labeled "1.".  I know it's elementary electronics, but I've never figured out how to calculate what those do and end up with useful results.  Does that resistor to ground take some pressure off of the 5W voltage reducing resistor?  Seems rather primitive to make that one 20K 5W resistor do all the voltage reducing work (I just guessed at the 20K value).

[sluckey]

Just increase the size of the dropping resistor to get the desired voltage at the preamp node. No need for a high watt bleeder. This is also something to experiment with after the amp is up and running.

[wittyjeff]

Got it.  Just didn't know when and where those "bleeders" were needed.  I can do the calcs (and, of course, see that I did them wrong when I fire up the amp later) for the dropping resistor.
Thanks again for your help.  It's been immensely valuable in my learning process.  My most recent amp (a 3-channel AC30) turned out VERY well (thanks to your help and other members of the forum).

[d95err]

The R96 resistor is just there to bleed the voltage out of the preamp filter caps when the amp is turned off. It is not intended to adjust the preamp voltages.

[d95err]

Some other notes/thoughts on the schematic:

First - Simplify! This is just too much. The chances of getting all this to work without issues, hum and noise is slim, even for a very advanced builder.

I tend to get the feature overload syndrom myself when desinging. What I do is to set hard limits, e.g. max 7 pots and two switches. That forces me to prioritize and focus on what really matters.

One specific note - you are switching on the input to the channels. That has a high risk of causing pops or noise. You can just tie the input to all channels and switch later in the circuit.

For the high gain channel you should have a mute switch point somewhere, or there is a high risk of bleed to the other channels.

Good luck!

[wittyjeff]

Some other notes/thoughts on the schematic:
First - Simplify! This is just too much. The chances of getting all this to work without issues, hum and noise is slim, even for a very advanced builder.

On this we are agreed, and it's great advice.  Amps like this one are a nightmare to build and troubleshoot.
I've only been building amps for about 10 months (it's my pandemic hobby).  Where I'm at right now in my amp building adventure is trying to push the limits of my building skills and trying out as many features as possible.  I don't really play guitar well so I won't really be the end-user of the amps I build.  I send them to my cousin who is an accomplished guitarist who performs regularly in a variety of bands and does studio work.  He uses my builds for a few months and then provides feedback on what features work and are useful.  Usually he pares the feature list down a lot and ends up using only a limited selection after working with the amp awhile.  When (or rather IF) I build a final version of a particular style of build, I would only put in the features that are recommended keepers.
My most recent amp was an AC30 style build with 10 pots, 7 switches, and 8 footswitchable relays controlling most of the switched features.  It ended up being very quiet and one of his favorite amps (he's currently gigging with it).  Mind you, that's more luck than skill on my part. 
This amp here has 13 pots and 9 switches but feels somehow less complex.  I suppose without that VOX Trem/Vib complexity there's more room on the board and there are 4 less tubes in this build, so that helps open up some space.  I hope my luck holds out and this one turns out as good as the last one... but you never know.  It's a high stakes game since these things are not cheap to build.
I appreciate your sound advice and encouragement for sure.