Variation of Type 3 Master

[jordan86]

Hey guys,

I’m about to rewire an amp and will take my first stab at a type 3 master volume. It will go into a cathode biased power section fed by a plexi styled preamp, Vox style PI, no NFB. I’ve read about some of the quirks of this type of MV, how it really quickly jumps to full volume with a 1M pot. So I was thinking maybe could use a smaller value pot? Not sure how low one could go before causing other problems like killing output signal. Obviously a switched pot could work to remove it from the circuit, but was wondering if a no load pot would do the same? With it’s no load at full rotation, would it essentially remove itself from the circuit (disconnect the two plate signals) or would it just short the two signals coming off the phase inverter? I’m just trying to experiment with variable resistance well below 1M to see if the master sweep is more useable….but with the option to bypass the MV as well.

Note: I am also familiar with the other trainwreck type masters and the larmar. I’m not necessarily looking for a recommendation to do those instead, although I may try those at some point.

[Dave]

I've played around with a lot of those PPIMV's. In one fashion or another, they tend to utilize phase canceling to achieve their goal. In my experience, my mental note is that the resistance across the PI outputs is usually down to somewhere around 350k before any noticeable volume is reduced. That being said, I have also discovered that I really don't like any of them. Unless you use them just a little bit, they have a huge not-so-nice effect on the tone.

[jordan86]

I have also discovered that I really don't like any of them. Unless you use them just a little bit, they have a huge not-so-nice effect on the tone.

Thanks, Dave. I’ve noticed the same. One of my Dr Z amps has a really nice master but it doesn’t do much until you get down to about noon on the dial. It’s a cathode biased 18w EL84 based amp. At noon it takes just a bit of edge off the power amp distortion below that it starts attenuating volume more noticeably.

How do define a “little bit”? Mine will be a 20W cathode biased amp. I’m not trying to get bedroom volumes, just knock down a few dBs - in the basement to keep the fam happy - or on stage to make the sound guy less fussy.

Still unsure too of how the No Load pot responds. I guess if I’m replacing the pot a 250-500K push pull would be more guaranteed success? Pull it to defeat the master.

[Dave]

"Little bit" = what you said, just take a little bit of volume off. Any major swing in the volume causes bad tone from my experience.

Since 1 Meg , IMHO, is way too much, in order to get the master volume to be effective throughout the sweep, something more like a 500k is better.

I have no opinion about using a no load pot. I don't have any experience with that other than I used one on a guitar once in an experiment and it only served to annoy me.

I also decided, for a time, that a Larmar was much better. It seems to work more smoothly. But, at the ends of the day, smooth plus crap still equals crap. Like lipstick on a pig.

On the other hand... I have an amp that has a pre PI master volume. I added a LARMAR to it and I have noticed that the pre PI doesn't sound very good either, but the right combination of pre PI and post PI can yield some decent results. For whatever that is worth.

Dave

[sluckey]

I used that crossline MV in my Matchless Lightning. Hated it! Probably a big reason why I never bonded with that amp. That was the shortest lived amp I ever built. Gone in a flash!   

[shooter]

^^^ what he said;
I built off the Matchless, but with my twist, left the cross-fade, "brilliance" knob there.  the guy that wound up with the amp loved it, but said that pot wasn't any help, offed to cut it out, in the end it shipped with the pot/cap still in-circuit.
NOTE: schematic is a work-in-progress, never documented the final version, mostly small tweaks to R's n C's for different stages

[passaloutre]

Not to pile on, because you seem determined to use the Type 3…

Recently tried the Type 3, with a few different pot values, and did not find it to be particularly useful. It’s fussy to dial in, and doesn’t seem to sound the same, even if it’s just barely active.

Very satisfied with the LARMAR. easy to dial, sounds fairly similar throughout its range, and full up it’s not even in the circuit. Most recently installed it in a cathode biased Bassman/Plexi-type circuit, but I’ve also enjoyed it in a Twin Reverb and a Silvertone 1484. Downside is it’s definitely fussy to build.

[sluckey]

left the cross-fade, "brilliance" knob there.

I find your "brilliance" knob to be very useful in my Vox AC-15 amps. Vox calls it "Top Cut".

[pdf64]

Over most of its range of rotation, I think it’s best to regard a type 3 as operating by loading down the signal from the LTP anodes.
Similar principle to a 5E3 volume control.
Yes, in the type 3, that occurs due to a signal cancellation effect at the virtual mid point of the setting of the master. Creating a null point, 0V AC.
The same action as a 1M type 3 could be achieved by using a dual 500k pot, with both its tracks wired as variable grid leak resistors for the output valves.

That analogy fails at low settings.
 
When the master resistance is set much below about 10k, the LTP operation fails and the common grid triode’s anode gets pulled up and down with the common cathode triode’s anode.
In other words, the outputs flip from balanced to common mode.
As they’re feeding a push pull output stage, the audio mutes, even though the valves are still pumping out a bit of power. I guess it’s dissipated by the OT 

Here’s some good debate on the topic
https://music-electronics-forum.com/forum/music-electronics/920070-master-volume-and-treble-loss#post923281

[jordan86]

Not to pile on, because you seem determined to use the Type 3…Recently tried the Type 3, with a few different pot values, and did not find it to be particularly useful. Very satisfied with the LARMAR. easy to dial, sounds fairly similar throughout its range, and full up it’s not even in the circuit.

Alright, you guys all have my attention...Appreciate everyone's input. I'm not as obstinate as I may have sounded in my first post. I hope  Just didn't want to get sidetracked, but here we are :)  Your feedback is legitimately helpful and appreciated.

The same action as a 1M type 3 could be achieved by using a dual 500k pot, with both its tracks wired as variable grid leak resistors for the output valves.

^^^^I'm open to this! And I already have a 500K dual gang pot in the drawer. But I believe what you're describing is a type-2/LARMAR, no? A few questions...

1) The dual 500K pots at max....from a circuit standpoint, with the Master at MAX, it would be IDENTICAL to just having hard wired 500K grid leaks?
2) In a cathode biased setup with no negative voltage on the grids, I could forgo the second set of caps on the back side of dual gang pot

[pdf64]

… I believe what you're describing is a type-2/LARMAR,

No, it’s simply just swapping the output valve grid leak resistors for a pot wired as a variable resistor. It would probably be best to a dual 220k pot, as you’ve got 500k one handy just wire 470k resistors in parallel, so the combined value comes out close on to the typical 220k.

1) The dual 500K pots at max....from a circuit standpoint, with the Master at MAX, it would be IDENTICAL to just having hard wired 500K grid leaks?

Yes.

2) In a cathode biased setup with no negative voltage on the grids, I could forgo the second set of caps on the back side of dual gang pot

Correct, but it’s the type 1 that usually an extra pair of caps for use with fixed bias amps.
The type 2 and its LarMar variant doesn’t have extra caps with either fixed or cathode bias.

[dude]

Food for thought:Why not just skip the Trainwreck MV's and try a VVR, not the whole amp just the power section keeping the pre-amp voltages high when cutting the B+..? I've done this on my 18 watt cathode biased amp and works great. Very good response turning down the B+ just a little, I found the tone at a little lower voltage unchanged.  Thought's

[jordan86]

...it’s simply just swapping the output valve grid leak resistors for a pot wired as a variable resistor. It would probably be best to a dual 220k pot, as you’ve got 500k one handy just wire 470k resistors in parallel, so the combined value comes out close on to the typical 220k.

So I am thinking on this again, and may be the route I go. I was hoping actually to experiment with 100K, 220K, and 330K grid leaks. This variable resistor would give me all those options, right? I thought 100K might be cool to keep the 6v6's from driving too hard, but 330K would be better for EL34's.  I would like the flexibility to be able to use both tubes.

So PDF, I've never tried this but theoretically a 500K dual gang with 1M resistors gives me 330K at max. Things clean up as I roll her down to noonish(100K range)? Below that, I am basically bleeding the post PI signals to ground due to absence of gird leak resistance?

{EDIT-- fix broke quote --PRR}