Adding a master volume to a PR before the PI driver stage, thoughts?

[guitylerham]

Hey everyone!

I'm in the building/planning stages of a new amp. It's a beefed up (bigger iron) PR with... Power Scaling! And cathode/fixed bias switch as well as 6v6/6l6 compatible (hopefully).

I have a question regarding which components to scale. As I've read pretty much every thread on the matter, I'm lead to believe in this particular amp, I want to scale the output tubes, cathodyne phase inverter, and because the NVB is injected into the cathode of the the gain stage just before the PI I'll want to scale that as well. This leads me to the issue of where to put the "drive compensation" or master volume control so that I can adjust the drive to the newly scaled stage.

My thought is to tack on a 1m volume pot just in front of the last stage before the PI, which is also right around where the reverb signal is fed back into the dry signal. One other change I made is to use Tubenit's take on the One-Tube Reverb, which is tapped and returned to the dry signal much the same way as a stock PR would (just with a 100k mixing resistor instead of 3.3m/10pf and some other value changes).

I'm curious if I can just add the volume like so in the drawing below. Take note of the other changed values in red in the reverb circuit because I'm thinking it might come into consideration when figuring out how to implement the "master" volume.



Alos, if someone more experienced than me sees a reason to power scale in a different way, please feel free to enlighten me! I am here to learn/I'm obsessed at the moment. Part of my goal is to control the power scaling and this new master volume with a dual-ganged pot so I need only turn one knob (also I only have one pot space on the completed faceplate so that's my self-imposed build challenge!).

The other change from a PR is I'm using the bias vary trem from the vibro champ on a yet-to-be-determined preamp stage... I wanted the trem to be non-scaled so I figured wiggle a preamp tube instead of output tube. So problems at every turn! Any way I can add another "transparent" triode stage after the Reverb but before the driver just to wiggle that one instead? I can change the cathode follower in the trem oscillator to a transistor to free up half a tube! 😬

I look forward to hearing from anyone who can shed some light on this. I'm just getting to the point but not quite where I understand impedences and the effects of voltage dividers and whatnot but it's just not clear yet.
Thanks in advanced!

[tubenit]

I think you can do it the way you are suggesting and that it will work.

Having done a VVR on numerous amps, I'm thinking you may not need the master volume between the 2nd and 3rd gain stages?  However, you may like having it there?

Regarding using the 100k resistor instead of a 3.3M,  try it and see if you like that.  In the PR idea that you're using, I might be inclined to try
a 470k there?

When you can,  I'd love to see an entire schematic of the amp design.  I've attached an ExpressSCH editable schematic of an original PR for you. There is a "2nd page" on the schematic that shows an example of a VVR set up for scaling the power amp and the phase invertor

With respect, Tubenit

[guitylerham]

Thanks, Tubenit.

I only assumed I'd need a master volume anytime you weren't scaling the whole amp. I wanted to scale justvthe power section and realized doing just the output tubes and PI would change the nfb loop since its injection point at the driver stage would remain unscaled. If this isn't a concern then that certainly simplifies things. I'm about to dive into KOC's TUT4 tonight.

I'll definitely try that 470k Reverb mixing resistor!

Thank you for that schematic. I'm on a Mac and have yet to use .SCH files so I'll have to figure out how to open that. I'll be able to get to it tonight after work. Thanks a million!

[Willabe]

I only assumed I'd need a master volume anytime you weren't scaling the whole amp.

Yes, many do use a MV with Power Scaling, because it takes less signal to drive the output tubes as you turn the power scaling down. Some prefer the MV after the PI so you can get the PI to distort.

And you do want to power scale the bias trem for the same reason.

[guitylerham]

I only assumed I'd need a master volume anytime you weren't scaling the whole amp.

Yes, many do use a MV with Power Scaling, because it takes less signal to drive the output tubes as you turn the power scaling down. Some prefer the MV after the PI so you can get the PI to distort.

And you do want to power scale the bias trem for the same reason.

Agreed. The bias trem is what led me to not bother with wiggling the output tubes because the tracking circuit for the scaled trem was yet another level of complexity I was hoping to avoid. Figured non-scaled preamp bias vary trem was the answer. 

[tubenit]

J-schem is the program to use to open the files on a Mac.  It's a free program.

http://jschem.bplaced.net/

[2deaf]

Attached is an alternate idea for a Master Volume before V3.

[2deaf]

I don't think it's any big deal if the Master Volume is inside the NFB loop, especially if there isn't a presence control involved.  I also don't think that lowering the supply voltage for V3 is going to do much since the gain will still be about the same.

Because cathodyne inverters don't sound good when overdriven, I would put the Master Volume before it so that the output tubes clip before the inverter does and I wouldn't scale the voltage on the inverter.  I would also put the Master Volume after the V3 mixer stage so that you can get a little bit of hair out of it.

I think the vibrato would be better after the reverb so that the reverb doesn't smooth the vibrato out.

If you reduce the plate voltage to the output tubes, you will need to reduce the fixed-bias voltage, also.  If you go cathode bias you won't need to fool with the bias voltage.   

[guitylerham]

J-schem is the program to use to open the files on a Mac.  It's a free program.

http://jschem.bplaced.net/

Got it downloaded and am working on the schematic tonight! Thank you for those that you provided. Clears up my questions on providing separate power supply chains for the scaled and non-scaled portions. When I receive the London Power kits I'll see what's up but I wasn't sure if the kit allows a stock power supply chain to simply splice in the PS (i.e. onboard voltage regulators). I'm planning to also clamp the voltage on the unscaled portions too.

[guitylerham]

I don't think it's any big deal if the Master Volume is inside the NFB loop, especially if there isn't a presence control involved.  I also don't think that lowering the supply voltage for V3 is going to do much since the gain will still be about the same.

Because cathodyne inverters don't sound good when overdriven, I would put the Master Volume before it so that the output tubes clip before the inverter does and I wouldn't scale the voltage on the inverter.  I would also put the Master Volume after the V3 mixer stage so that you can get a little bit of hair out of it.

I think the vibrato would be better after the reverb so that the reverb doesn't smooth the vibrato out.

If you reduce the plate voltage to the output tubes, you will need to reduce the fixed-bias voltage, also.  If you go cathode bias you won't need to fool with the bias voltage.

It's ok that the NFB loop extends from the scaled output section to the non-scaled PI driver stage? Wouldn't scaling down the output section disproportionately vary the NFB that is injected at the full on V3B?

I agree about MV after V3 to get what we can out of that stage. Also about the vibrato being post reverb.

The London Power kit I purchased will track the fixed bias, when in that mode. Good call.

Thanks for those schematics too! Especially the tremolo circuit that I had in mind!

[guitylerham]

Still working on building that schematic for y'all (and myself, smart).

Now I'm thinking an LTPI in this thing would make a lot of sense! Could easily scale the PI, have nfb right there, single MV into the PI.  Would clear up a lot of issues. I love the Princeton Reverb and how nice it is but maybe a DR with power scaling can still be versatile AND plenty powerful. Decisions...

[Ed_Chambley]

A Deluxe Reverb is a fine amp and if you use a inefficient speaker, like a 93 spl, it will tame a lot of volume.  What a Princeton reverb has going for it is the smaller output trans which saturates easily.  The PI overdriven tone can be adjusted by balancing to sound better, but each tube swap in the PI position takes balancing again.  I simply replace the 47K resistor with a Kermit 100K trimpot.  I will balance using a scope, but after I turn it to taste while playing.


Power Scaling of the output tubes can work, but the tone changes.  I have never been a fan of master volumes, but are needed for around the house.  For a Gig if a Princeton is too much, take a Champ or a GA5.  It is ok to build more than one amp.


A Deluxe is not near as spongy feeling as as a Princeton and is quite a bit more articulate and pounds 6V6 tubes fairly hard.  Consider a Fixed Bias and switching to cathode as an option too.

[Willabe]

Power Scaling of the output tubes can work, but the tone changes. 

One, perhaps main reason for this, is as you turn down the power scaling/volume, it unloads the PSU more and more as you turn down.

This raises dcv on the preamp tubes, and PI if not scaled.

KOC makes/sells a voltage clamp mod/kit that stops this from happening to the small bottle tubes.

I think he also bought the voltage clamping kit with the scaling kit?

I'm planning to also clamp the voltage on the unscaled portions too.

 

   

[guitylerham]

A Deluxe Reverb is a fine amp and if you use a inefficient speaker, like a 93 spl, it will tame a lot of volume.  What a Princeton reverb has going for it is the smaller output trans which saturates easily.  The PI overdriven tone can be adjusted by balancing to sound better, but each tube swap in the PI position takes balancing again.  I simply replace the 47K resistor with a Kermit 100K trimpot.  I will balance using a scope, but after I turn it to taste while playing.


Power Scaling of the output tubes can work, but the tone changes.  I have never been a fan of master volumes, but are needed for around the house.  For a Gig if a Princeton is too much, take a Champ or a GA5.  It is ok to build more than one amp.


A Deluxe is not near as spongy feeling as as a Princeton and is quite a bit more articulate and pounds 6V6 tubes fairly hard.  Consider a Fixed Bias and switching to cathode as an option too.

Thanks for that description. Since I've already upgraded the transformers and will use a 12" maybe I'm just closer to building a DR anyway and have already changed it from the sweeter tone of the PR. Also, important note, I'm using SS rectification instead of a tube, so already a stiffer amp.

I accept that the tone may change with power scaling I'm mostly interested in having some power adjustment for different gigs where I don't need quite as much stage volume (most gigs these days) but still want power on tap for those gigs that need it. I think the proper MV will be essential with power scaling.

This amp will have a cathode/fixed bias switch. I'm excited to mess with that! Thanks for the PI balancing info, I'm curious about that effect on the cathodyne.

This is probably my fifth amp build so I whole-heartedly agree with you on building more than one! I'm hoping I can sell my fender PR and this can replace it as my number one amp.

[guitylerham]

Power Scaling of the output tubes can work, but the tone changes. 

One, perhaps main reason for this, is as you turn down the power scaling/volume, it unloads the PSU more and more as you turn down.

This raises dcv on the preamp tubes, and PI if not scaled.

KOC makes/sells a voltage clamp mod/kit that stops this from happening to the small bottle tubes.

I think he also bought the voltage clamping kit with the scaling kit?

I'm planning to also clamp the voltage on the unscaled portions too.

 

   

You got it. I was concerned about the rising preamp voltages so I ordered a clamp kit and will use it for that reason. Thanks for the affirmation!

[2deaf]

It's ok that the NFB loop extends from the scaled output section to the non-scaled PI driver stage? Wouldn't scaling down the output section disproportionately vary the NFB that is injected at the full on V3B?

A Master Volume inside the loop will foul-up the NFB, but a lot of people do it including some production amplifiers.  Scaling the supply voltages is a little different.  Scaling the output tubes doesn't make the amp any quieter until they clip.  Before clipping, the gain is about the same just like the scaled preamp tubes and PI so that the NFB is operating as usual. 

[guitylerham]

It's ok that the NFB loop extends from the scaled output section to the non-scaled PI driver stage? Wouldn't scaling down the output section disproportionately vary the NFB that is injected at the full on V3B?

A Master Volume inside the loop will foul-up the NFB, but a lot of people do it including some production amplifiers.  Scaling the supply voltages is a little different.  Scaling the output tubes doesn't make the amp any quieter until they clip.  Before clipping, the gain is about the same just like the scaled preamp tubes and PI so that the NFB is operating as usual.

So to clarify, the supply voltage to the output tubes allows a window of clean amplification that the phase inverter presents a signal within. The PI (earlier stages of the amp) can deliver a small signal or a large one as long as it fits within the window that the output tubes provide based on a given voltage. Lowering the output tube voltage doesn't lower the volume until the voltage/ceiling is low enough that runs into and clips the PI signal. You would then lower the PI signal or continue lowering the output tube ceiling and creating more clipping and lower volume.

Is this correct? It makes me think of my vague idea of Class A where the output is run at full power regardless of signal presented to it. I knoooow I don't quite have any of this accurately in my head so forgive my lack of understanding...

[guitylerham]

Here's that updated schematic with some of my ideas. So far, I've switched over to a LTPI but haven't yet figured out how I want to go with the master volume. Now I'm contemplating a tripple-gang potentiometer for the power scale/drive comp control since it looks like a dual-gang is needed for a decent master volume. Many ideas swirling around so I'll have to think on this for a while. Still wondering about inserting another triode stage between the last preamp stage and the driver stage. This would give me a spot that's outside the NFB loop to inject some tremolo too. Not sure how adding another stage would effect the sound though. I'd want it to basically be invisible as far as gain goes.

[tubenit]

Can you please post the schematic in a GIF format for all of us to view?   I'd love to see what you're doing?

Most of us don't have J-schem.  However, on ExpressSCH,  I save the file as a Bitmap and then open that and then resave it as a GIF.  Not sure how it's done on J-schem.

With respect, Tubenit

[tubenit]

Many ideas swirling around so I'll have to think on this for a while

 

This amp will have a cathode/fixed bias switch

Well,  as long as your pondering ideas ………………..               I drew this up a couple of yrs ago and hope some day to build it.  I am NOT suggesting you "should" do this. I think you should build what you want.  But perhaps you (or someone else on the forum) may be interested in the idea of the D-style tone stack in a Fenderish design.

This idea uses some Dumblish approaches with 5uf cathode caps and a Dumble style tone stack with PAB (preamp boost) & midboost.  IF you've never tried the PAB or midboost switching before, you can get some pretty amazing tone and overdrive using those features.  Something you could do with either a mini-toggle  OR  relay switching.

As an FYI,  I almost never use the 22uf or 25uf caps on cathodes anymore.  I find the 5uf cap to be sufficient and MUCH smoother to my ears.  Having said that,  I have experimented with a 22uf only on the first gain stage and then 5uf caps in the following gain stages prior to the LTPI and found this gives more of a Fendery tone but with added smoothness and less "grit" to my ears.

There is a SCH layout attached that goes with this schematic.  The layout uses a parallel turret design which is easy to convert to something else IF you're experimenting around.

With respect, Tubenit

[guitylerham]

Can you please post the schematic in a GIF format for all of us to view?   I'd love to see what you're doing?

Most of us don't have J-schem.  However, on ExpressSCH,  I save the file as a Bitmap and then open that and then resave it as a GIF.  Not sure how it's done on J-schem.

With respect, Tubenit

Yes!

Remember, whoever comes across this in the future, NOTHING in my edited version is verified!

[guitylerham]

This idea uses some Dumblish approaches with 5uf cathode caps and a Dumble style tone stack with PAB (preamp boost) & midboost.  IF you've never tried the PAB or midboost switching before, you can get some pretty amazing tone and overdrive using those features.  Something you could do with either a mini-toggle  OR  relay switching.

As an FYI,  I almost never use the 22uf or 25uf caps on cathodes anymore.  I find the 5uf cap to be sufficient and MUCH smoother to my ears.  Having said that,  I have experimented with a 22uf only on the first gain stage and then 5uf caps in the following gain stages prior to the LTPI and found this gives more of a Fendery tone but with added smoothness and less "grit" to my ears.

Interesting! Thank you for sharing your ideas. I don't often play with overdrive really so I've never looked into designs that feature it. I do plan on this build to have a 3pdt switch that changes the V1a voicing from the Fender style to a Lead style ala RobRob (2k2/0.68uf cath, 220k plate load, and v1b coupling cap to 0.0047uf). This should give me some other voicing but not sure if it'd achieve anything close to what you're suggesting.

I am very curious about updating the habit of always using 25uf bypass caps. Might be worth buying a bunch. I think your circuit looks pretty good. That master before the PI. Because you have that 220k on the grid/wiper, turning down the master shouldn't effect tone too much right? Can that 100k value be anything else, like 250k perhaps? It'd make using a dual-gang easier since the PS pot is also a 250k.

Do you find that there is a good effect when you have the first preamp on it's own, smaller power supply cap node E instead of all preamps being fed by a single node D?

[2deaf]

The 25uF bypass cap on V3a is too large and will allow too much of the LF vibrato signal to shunt to ground.  Try 5uF.

[guitylerham]

The 25uF bypass cap on V3a is too large and will allow too much of the LF vibrato signal to shunt to ground.  Try 5uF.

Done and updated! Thank you.

[guitylerham]

Potential epiphany (pun mostly unintended), I just ordered a dual-concentric potentiometer and will swap out the wafers for whatever values I need once I decide. This way maybe the base knob will be Power scale and the top knob will be the master volume. One hole needed on the front panel and I get to retain some tweakability and still with ease of turning both controls at once in equal amounts. Just wanted to share!

[Willabe]

I just ordered a dual-concentric potentiometer and will swap out the wafers for whatever values I need once I decide. This way maybe the base knob will be Power scale and the top knob will be the master volume.

That may be asking for trouble.

B+ directly next to AC signal.

[guitylerham]

I just ordered a dual-concentric potentiometer and will swap out the wafers for whatever values I need once I decide. This way maybe the base knob will be Power scale and the top knob will be the master volume.

That may be asking for trouble.

B+ directly next to AC signal.

Very true but the Power Scaling uses low watt pots for scaling which leads me to believe the pot doesn't have B+ on it but a small control voltage of some sort.

[tubenit]

Reply #19 schematic

V3a should have node D not node C. 

I think having the node E added may make for a quieter amp at idle.

Yes, the master volume would have to be experimented with & I think 250ka might be a good place to start?

Experimenting with the cathode caps on the first 3 gain stages is pretty easy to do.  Maybe use a 22uf or 25uf in V1a & install 5uf in the 2nd and 3rd gain stages.  IF you don't like the tone, then parallel a 10uf and see IF that moves it the direction you want & if it does, then remove the 5uf and install 22uf.  My overall experience is that 5uf caps give me a tone the seems smoother with less grit. Your experience may be different.

With respect, Tubenit

[sluckey]

Very true but the Power Scaling uses low watt pots for scaling which leads me to believe the pot doesn't have B+ on it but a small control voltage of some sort.

There is full B+ on the pot.

[guitylerham]

Very true but the Power Scaling uses low watt pots for scaling which leads me to believe the pot doesn't have B+ on it but a small control voltage of some sort.

There is full B+ on the pot.

Oh really? Dang, I thought I was safe. You can put 400vdc on a small 16mm pot? I still need to learn how this technology works.

[Willabe]

The new(er) power scaling circuit is able to use any pot.

The older circuit he said you had to use a mil spec pot that handled 500dcv.

Just call Kevin and ask him if you can dual-concentric pot for scaling and MV.

[guitylerham]

I'll look into the feasibility of the dual-gang pot and signal proximity to the PS. Kevin's great at email correspondence and advice.

Ok, looks like now I'm building a Deluxe Reverb with that master volume (250k-1ma) before the PI, power scaling V4-V6 while voltage clamping V1-V3. One tube reverb and hybrid triode/mosfet tremolo injected at PI driver cathode (thisc will ensure the trem is consistant when power scaled and when in cathode bias mode. I'll experiment with the preamp cathode bypass caps as well.

Thank you for the awesome help and discussion everyone! Now I can populate the board and wire it into the chassis that's been sitting idly on my bench for a couple weeks now.
Stay tuned!

[ac427v]

I usually see and use the 820/100r feedback resistors in a circuit connected to the 4 ohm tap of the OT. It appears that your transformer does not have a 4 ohm tap. Connecting it to the 16 ohm tap as in your schematic would double the feedback voltage. You could change to 820/47r resistors and connect to the 8 ohm tap of the OT for the traditional Deluxe Reverb sound.--AC427

[guitylerham]

I usually see and use the 820/100r feedback resistors in a circuit connected to the 4 ohm tap of the OT. It appears that your transformer does not have a 4 ohm tap. Connecting it to the 16 ohm tap as in your schematic would double the feedback voltage. You could change to 820/47r resistors and connect to the 8 ohm tap of the OT for the traditional Deluxe Reverb sound.--AC427

Good catch, I wouldn't have fixed that if you didn't mention it. I'll use that 820/47r setup on the 8ohm tap like you suggested. Thanks!

[d95err]

Very true but the Power Scaling uses low watt pots for scaling which leads me to believe the pot doesn't have B+ on it but a small control voltage of some sort.

There is full B+ on the pot.

Oh really? Dang, I thought I was safe. You can put 400vdc on a small 16mm pot? I still need to learn how this technology works.

If it's the newer style called "DC Power Scaling", the control voltage on the pot is only around 15V. There should be no problem having both the control voltage and AC signal on the same dual pot.

The older more primitive versions of Power Scaling (and copies like "VVR")  put full B+ directly on the pot. In that case, you a need mil spec heavy duty pot and you should keep it awat from AC signals.

I would use a post-PI master volume and not worry too much about the reduction of NFB. However, you'd need a dual pot just for the MV (single-pot "cross-line" post-PI MV is useless, IMHO).