Princeton clone reverb low output

[Lectroid]

Hey all,

I'm almost finished voicing this BFPR-ish clone with a 1-tube reverb based on a 12DW7.  I'm happy with the amp overall but it really seems like the reverb output is too low, lower than it should be.  I know that part of that could be down to using the single driver/recovery tube.   But can anyone suggest any ways to boost the relative punch of the reverb signal in the mix? 

I don't fully understand the mixing circuit.  The 470K leaving the reverb looks like ity might work with the 470K mix resistor as a voltage divider that balances the gain of the two signals coming together?  But I'm not sure about that, given how the reverb 470K resistor also seems to function as the grid leak of V3B.   

Can I tweak one or both of them to alter the mix?  What are the rules here?

Is there a 'best-practice' ratio of voltages to keep between the clean vs. reverb signals?


Thanks for reading. 

[Willabe]

The tank # in the schematic is a 8 ohm input tank. If you used that tank, it's the wrong tank to use with a capacitor coupled reverb. 

I bought a Swart 2 x 6V6 amp with a 1 tube, 12DW7, reverb. I put a long tank 3 spring, long decay unit in, much better reverb.

And the recovery triode, your feeding it's plate from the lowest B+ node in the amp. I'd try feeding it from the B node.   

[Willabe]

Here, read this, Swart 1 tube reverb, gives tank # at the very end.

How to build your own spring reverb unit

I think this is better than the 2 spring the link above uses, 3 spring, long tank, medium decay. I don't remember if I couldn't find a 3 spring with long decay?

https://www.tubesandmore.com/products/reverb-tank-mod-9fb2a1c-medium-decay-3-spring

Still worth a read even though it's a cap coupled reverb.
 

[Lectroid]

The tank # in the schematic is a 8 ohm input tank. If you used that tank, it's the wrong tank to use with a capacitor coupled reverb.

@Willabe,
Is that right?  My driver appears to be very close to Fender's except for using 1/2 of a 12DW7 instead of an entire 12AT7.  Why do you say it's capacitor coupled?  I don't see it.

And the recovery triode, your feeding it's plate from the lowest B+ node in the amp. I'd try feeding it from the B node.

Thanks, I will definitely take a look at that.  Will using the "B" node for the recovery tube limit what's available to the driver triode? 

Any thoughts on the mixing circuit?

[Willabe]

My driver appears to be very close to Fender's except for using 1/2 of a 12DW7 instead of an entire 12AT7. 

Yes, your right, my mistake, got 12DW7 in my head, didn't see the verb transformer.

Many/most 12DW7 1 tube verbs are cap coupled.

[Willabe]

Will using the "B" node for the recovery tube limit what's available to the driver triode?

No.

Why would it limit whats available going to the driver tube/12AX7? Those tubes are only drawing a few mili amps. It's the PT's B+ ability to put out. And it's already having to supply that tubes B+, your just going to move it from 1 B+ node to another B+ node.

[Willabe]

Any thoughts on the mixing circuit?

No, I'd look at the recovery tubes control grid return R, 220K. I'd temp in a 1M pot and try different settings. Then measure the pot where it's set at, if not full up. And put in the closest R value in place of the 220K grid R.

Throwing away a lot of gain with that lowish value grid R. Problem might be if you use to high of a grid R value there, you'll amplify any noise coming from the verb tanks output. Verb tanks output is less than guitar PUP output.

You might/maybe have to play around with the pots setting to find a reasonable compromise between added noise and more verb?   

[tubeswell]

I doubt there is sufficient attenuation of the dry signal. You probably need to increase the 470k (in the dry side voltage divider) to something like 2M2 or 3M3 to knock the dry signal down so the wet signal can come across more. (Note this will reduce the dry gain of the amp). The alternative would be to add another gain stage between the reverb recovery stage and the dry/wet mixing stage (to boost the wet signal prior to mixing) - but you’d want to ensure this was free from distortion. Check out Allen Old Flame or Allen Encore schematics.

[stratomaster]

I think the 470k in place of the 3.3M of the typical Fender reverb arrangement is messing with the balance of clean to reverb. 

Try a 1M there instead.

I'd also be inclined to feed the driver with the C node, change that first 18k dropper to 10k, and bias the driver a bit warmer with a 1.5k or so.

You're also letting a lot of low end into (and out of) the reverb stage.  It takes a lot of power to push these frequencies. Fender drivers are high passed in the low 300hz range, I'd mimic that by decreasing your 0.001μF coupler to 390-470pF. 

[HotBluePlates]

... it really seems like the reverb output is too low, lower than it should be.  ...

Compared to what?

   The stock Fender circuit uses a 100kΩ linear taper pot, but folks usually complain of too much reverb too early.

   Your amp uses a 1MΩ audio taper pot, so "7" or "8" on your pot is equivalent to what the stock part has at "3".

   Turn your Reverb pot full-up and ask whether that is "enough reverb."  If it is, you don't need to tinker further, only change your expectations.


Yes, you can increase the proportion of Reverb to Dry by changing the relative-size of the 470kΩ from the Reverb pot (bigger is more-Dry) to the 470kΩ from the 0.022µF coupling cap (bigger is more-Reverb).  But you ought to evaluate this in context of "Reverb control full-up" to know whether a change in mix-resistors is the right move.

[Lectroid]

@Willabe, tubeswell, stratomaster,

A lot of great advice here, thank you.  Some small changes, some bigger.  I think I understand the mixing area better now, how the pieces interact.  I hadn't really looked at that 220K grid blocker before but I will now.

@HBP,
Even with the Intensity at 10, the reverb was certainly weaker than Fender BF amps, and muddier.  I agree that the mixing was maybe not right.  Thanks for explaining how to alter the mix resistors, that confirmed my own guess.

So, start small:  stratomaster's suggestion to reduce the reverb input cap sounded good.  I dropped it to 470pf.  Someone I read here recently (HBP?) said how they have used cathode bypass caps for tone shaping.  So I got wild and crazy and changed the driver's cathode cap from 25uf down to 4.7uf.

Of course, changing the caps lowers the reverb bass frequencies' gain but that's inescapable I guess.

Bingo!  Big change--those two new caps cleaned up the bass muddiness and brought the reverb to life.   At full intensity the reverb is lush and full.  No mud, even when I increase the Bass control. 

Overall, I like this 12DW7 reverb.  I never built one with the long tank before.  The sound is very Fenderish now.  The 1M Intensity pot works smoothly over the whole range.  Turning it up leads to a brighter sound, dialing it drops the tone back to more midrange.  The amp generates a smooth harmonic overdrive at higher volumes.  I can certainly recommend the circuit using these new values.  The 470K/470K mixing resistors I left alone for now. I'm pretty happy with it now.

Posting an updated schematic.

Thanks for all the help.

[Willabe]

I hadn't really looked at that 220K grid blocker before but I will now.

If your talking about V2b, the 2nd 1/2 of the verb tube, it's recovery, then, it's a grid return resistor, not blocker, and I think you meant grid stopper.

[Lectroid]

Yah, I mis-wrote.  What I would have called that 220K resistor is a 'grid leak resistor.'  Is that what you meant or is a 'grid return resistor' something different?  I don't think I've heard that before.

[Willabe]

Yes, same thing, grid leak/grid return.