Will Pre-PI master volume work on Tweedle-Dee?

[dannieh42]

I just finished my Tweedle-Dee build. I can't turn it up enough to start breaking up before it is too loud for my house. Will a Pre-PI master volume work on a Tweedle-Dee. I saw one for a 5E3 on Rob Robinette's website, but since the Tweedle -Dee has local negative feedback from plate to grid, I was not sure if it would work or not?

[pompeiisneaks]

I would think you can add one just fine, but you'd have to mod the board a bit. The connection of the coupling cap from the previous stage connects to the grid, and the LNFB. If you just took the output from the .022uF coupling cap, off to the MV pot, and then returned it back to the 1M resistor down to the cathode, and the grid, and the NFB, it should work.

Actually, looking at the layout, you may just be able to remove the under board jumper (the blue wire), and connect the MV to that output and return it back to where the blue wire connects at the top end of the board, that way you get it in circuit pretty easily. All a MV there does is allows you to send some of the signal to ground instead of on to the next stage. I don't know how much that may impact the tonality of the amp, though. So let us know how that works, if it's nice I may try it on mine.

~Phil

[dannieh42]

Well I still don't know if the Pre-PI master volume will work...but I know the Post PI works. It sounds good but it seems like all of the adjustment in in the first quarter of the pot. I used an audio taper pot, maybe a linear would have worked better...any thoughts?

[pompeiisneaks]

Well I still don't know if the Pre-PI master volume will work...but I know the Post PI works. It sounds good but it seems like all of the adjustment in in the first quarter of the pot. I used an audio taper pot, maybe a linear would have worked better...any thoughts?

That's basically the common expected fix yes. If you don't like the adjustment for audio, then try linear or log (audio is an approximation of log but not exactly log, I think log is often more expensive, audio is usually two slopes that change shift about mid point) . Some of it is just personal preference.

~Phil

[dannieh42]

I checked and the pot I have is audio taper so I couldn't understand why all of the control was in the first 25%...Then I did a little math and realized that putting the pot parallel to the 220K resistors linearizes it! I am going to remove it and replace the 2 220K resistors with a dual gang 200K pot. Since putting the 1M pot in parallel with the 220Ks reduces the resistance to 306K, changing from 220K resistors to 200K pots should keep the circuit values closer to original at full volume. Any one have any thoughts on this?

[pompeiisneaks]

OH yeah that's required for PPI MV you need to basically replace those 220k with a pot. Also, though make sure to add some 2M to ground from each pot as backups in case the pots lose their track or go bad so you keep consistency in that chain.

~Phil

[dannieh42]

Do you mean something more like this?

[pompeiisneaks]

exactly! That helps the tubes ensure they don't lose negative bias and can basically go into shutoff instead of runaway.

~Phil

[tictac]

The Tweedle Dee uses a splitload inverter. The best master volume for this type of inverter is to replace the 1M resistor on the grid of the splitter with a 1M audio pot with the wiper of the pot connected to the grid.

With the pot full CW you retain the original, unmodified circuit.

TT

[dannieh42]

Are you talking about a Pre-PI MV? I saw this one on Rob Robinette’s website (Thank you Rob! Your site is awesome. It's what made me believe I could actually build one of these). This is what I wanted to try first, but wasn’t sure how it would work with the LNFB. If you’re saying it will work with LNFB then I will solder it up as soon as I get home from work. (shhhh don’t tell anyone I spend all day on ampgarage instead of working)

[pompeiisneaks]

Yes that shot is missing the LNFB but that won't stop it from working. the LNFB takes the output at the anode, puts it through an RC filter and back into the grid of the same tube. that just happens to also be the input off of the wiper of the MV you'd install.

[dannieh42]

Man, am I glad I read your post. I was going to leave the LNFB connected like it is, rather than attach it to the wiper. But, after reading your post I see that would have been changing the feedback circuit as I adjusted the volume. THANKS!

[pompeiisneaks]

Man, am I glad I read your post. I was going to leave the LNFB connected like it is, rather than attach it to the wiper. But, after reading your post I see that would have been changing the feedback circuit as I adjusted the volume. THANKS!

oh you can connect it anywhere, I just meant it's at the same potential as the wiper. Leave it connected to the grid, (which is also connected to the wiper) apologies for the confusion. Basically that's a single wire. I think connecting the LNFB is always best very close to the grid, short leads, etc. You don't need to run a second wire all the way over to the wiper of the pot.

I'm sure I could have made that more clear. My apologies.

~Phil

[dannieh42]

No, you did a great job of explaining it. Being a NUBE I just had trouble explaining how I previously thought it was supposed to be done. These pictures explain it way better than I can. I had planned on doing it the wrong way until you told me about the right way. The way I had planned to do it was definitely the wrong way. Because, now I see that when the volume was turned down the the 3.3M resistor would no longer have been at the same potential as the grid. I just had not thought of connecting the LNFB to the wiper, which is definitely a better way. And, I will make sure to keep the feedback resistor close to the grid.

[tictac]

Actually wiring the MV the "wrong" way might be just fine since you're turning down the MV to distort the preamp anyway...

As you turn the MV down the FB resistance gets higher which in theory would increase distortion as well.

In actual practice you'll likely won't be able to hear much difference since you're only varying the FB by 500k to 900k.

You'd need to change the FB resistance by many Mega-ohms before there'd be much change in tone....

TT