12DW7 in Princeton Reverb PI...

[lookah]

Hi folks. I'm going to perform the mod that reverses the wiring on V4 of my '71 PR in order for it to accomodate a 12DW7 (1/2 12ax7 + 1/2 12AU7). I stumbled upon the archived post below and tested my amp with a NOS 12AU7 first. Indeed it really does wonders to the bass...cleans it up and makes it nice and tight.

Here's my question: I've ordered the tube but won't have time to mod the amp until the weekend and I have a few gigs before then with my PR. Is it safe to run my amp for a full gig with the 12AU7 in V4? I don't care about the vibrato being temporarily disabled.

Here's the thread post by "Bill"

I would recommend a simple test to see if the 12DW7 mod will improve your amp- simply replace the stock 12AX7 with a 12AU7 and play. In the case of the PR, you will lose the Vibrato during the test (a 12AU7 can't get the oscillator oscillatin'), but your testing the tone, not the vibrato. That is why the 12DW7 is specified.
I was able to score my tubes off of ebay. 2 (one British, one Japanese) for under $20.00 including shipping. New 12DW7 tubes are sold under the EI label (and possibly others).

[tubeswell]

FWIW JJ sells a tube that is already a 12DW7 in reverse 'ECC832' (or 'ECC823'? can't remember which)- to save you doing the mod.

[lookah]

Hmm. Are you sure? I got the mod idea from the Torres website and they sell the JJ 12DW7 tube specifically for this mod and say to switch the wiring

DOWNLOAD The Newest Coolest PRINCETON mod

JJ new 12DW7 dis-similar dual triode

[tubeswell]

Yup - The info is on the JJ ECC832 datasheet

http://www.jj-electronic.com/pdf/ECC%20832.pdf

the reversed triode set is called the 'ECC823'

[lookah]

Ah, I see now. But what's the deal with the reverse pin wiring "on request" bit?

[lookah]

never mind. figured it out. looks like the JJ ECC 823 is what i want

btw, would you know the answer to my original question...if it's safe to temporarily run the 12AU7 in V4 until i receive the JJ tube?

thanks!

[tubeswell]

Probably not 12AU7 in V4. AFAICT the cathodyne phase splitter triode part of it would function fine, but IIRC, you need a minimum gain of 29 for the LFO tube in a 3-stage RC network (a 12AU7 has a gain of 17). A 12AY7 or a 12AT7 might work well tho'.

[lookah]

thanks very much for the info. and if i'm not concerned about using the vibrato at all, will it hurt the amp if i use the 12au7 simply for the PI function? or is the amp expecting to "see" a gain factor of 29 or more regardless whether i engage the vibrato or not?

[tubeswell]

If you're not worried about the LFO, then go for it and see what you think. running a 12AU7 as a cathodyne stage with 56k plate and cathode resistors will work, (although it may work better with 22k plate and cathode resistors, but you could just tack a 47k-56k in parallel with each 56k if you wanted to check out the difference)

[Enzo]

No, it won't hurt anything, go ahead and try it.

[lookah]

great. thanks for your replies

[lookah]

Hi. One more question...if I go with the lower 22k plate and cathode resistors when running the ECC 832, will 1/2 watt Carbon Comps be safe? I found this quote on an other forum regarding this mod that suggests otherwise:

"Be careful It's a very good idea but the bias will be way off on the
12AU7 section and it's going to run very HOT! You will need lower
value plate and cathode resistors on the PI. Those resistors need to
be bigger as well (5W) because they will take a lot more current than
on the 12AX7 PI. (around 10mA)"


Thanks!

[MWJB]

5W would appear to be overkill. A 12AU7 can kill a 1W resistor in a big, 6L6 Fender with a LTPI, but the cathodyne in a PR runs much lower voltage, in a 5E3 w/56K, for example dissipation is under 1/4W per triode, won't be that much more in a PR. It will draw less current with a 22K & larger cathode (you won't use 22K for the cathode) so half watt will probably do, 1W safe as houses.

[lookah]

That's great info...except now i'm confused by your saying "you won't use 22K for the cathode". In your previous reply you mentioned using 22k for both the plate and cathode. Did I misread?

BTW, thought you might like to know that I found out JJ only produces the ECC 823 (reversed 12DW7) for orders of 1000 or more. No wonder no one carries it!

[MWJB]

"That's great info...except now i'm confused by your saying "you won't use 22K for the cathode". In your previous reply you mentioned using 22k for both the plate and cathode. Did I misread?" You're confusing me with someone else, I never suggested changing any resistor values...personally I'd just stick the 12DW7 in there, reverse the wires to the 2 triodes if you want the low mu side on the PI. Cathode load resistor (to ground) can be 22K, but cathode resistor, where the 1Meg grid load?PI coupling cap connects, will be smaller (stock 1K?).

[lookah]

MWJB, sorry for the mixup. Yes, it was tubeswell who said "it may work better with 22k plate and cathode resistors".

Ok, so if I lower the cathode load resistor (to ground) from 56k to 22k, What is the smaller value you suggest I replace the stock 1k cathode resistor with? 820 Ohm? 470 Ohm? And I can leave the plate resistor stock 56k? Also, are there a measurements I can take on the 12AU7 PI side that will help me find the best resistor values?

Many thanks!

[MWJB]

Plate load & cathode load resistors should be the same value & matched, e.g. both 56K or both 22K/'both whatever value you choose'. The 1K biases the tube, the 12AU7 side should dissipate 3.5W...but preamp tubes are typically not run at anything like max dissipation, use the value that sounds best, measure dissipation by voltage drop accross the plate resistor, multplied by plate to cathode voltage.

[lookah]

Great! That really clears things up...thanks again.

[lookah]

Hi again. I performed the 12DW7 "rewiring mod" on my Princeton Reverb and tried to optimize the resistor values for the 12AU7 side. I'm a bit confused with the results I'm getting when I plug the numbers in the dissipation formula above so hopefully someone can help. Here's some readings with 2 different cathode resistor settings. Which one is more on the right path of properly biasing the 12AU7 side?

JJ ECC82 (12AU7) Data Sheet says Max Plate Dissipation is 2.75W


1)

lowered stock 56k plate resistors to 22k
increased stock 1k cathode resistor to 3.9k


Voltage drop across 22k plate resistors = 38v
Voltage drop across 3.9k cathode resistor = 6.8v
Plate to Cathode Voltage = 146V

-------and------

2)

lowered stock 56k plate resistors to 22k
lowered stock 1k cathode resistor to 680ohm


Voltage drop across 22k plate resistors = 63v
Voltage drop across 680ohm cathode resistor = 1.9v
Plate to Cathode Voltage = 80V

[tubeswell]

Hi lookah

When you say 'voltage drop across 22k plate resistors = xyV", are you meaning voltage drop across each 22k resistor?

[lookah]

hi tubeswell

sorry for late reply and thanks for asking. i ended up studying up on this material after my last few posts and now have a much better understanding of the cathodyne PI. ultimately, learning how to create load lines off of the tube data chart was the real break thru. In fact I'd like to share my findings as well as the actually mod I've ended up with. I'm enjoying the sound of the mod which has provided more focused bass response and less harmonic madness when dimed. I would much appreciate any feedback on what I've done...

Here's what I did:

1) "stokes mod" to increase B+ voltage to PI (stock was 230v...now it's 270v)
2) Rotated V4 180 degrees and rewired it to accomodate 12DW7 (JJ ECC832)
3) changed the two 56k Ra+Rk resistors to 12k (each)
3) changed cathode bias resistor (Rb) to 1.5k
4) added large, 1M grid stopper resistor




The blue "X" below is where the 12AX7 was performing





The blue "X" below (a bit hard to see here) is where the 12AU7 side is currently performing (plate to cathode voltage is 162v)

[Rick Erickson]

This reminds me of swatting a fly with a sledgehammer. If you want to tighten up the bass, just replace the .1uF coupling caps between the PI and the power tubes with .022uF's. This will also increase the depth of the vibrato (tremolo).

[lookah]

fair enough but i was namely going after a better cathodyne/concertina PI by using a 12au7. and one of the results is a tighter bass and smoother distortion when cranked...

[lookah]

Hi MWJB and Tubeswell (and anyone else who wants to chip in). Still very interested if you have any feedback for me regarding the load line calculations and subsequent mods I performed on my Princeton (pics and data are above). Thanks!

[TheElectricMoron]

Fly/hammer thing seems true to me too. Way easier to do the Stokes mod and the PaulC mod and get a decent speaker.

hint hint a DeluxeReverb OT will make your day and with a 12" speaker in there you will be in another amp world.

JMO of course YMMV

Charlie

[lookah]

Thanks Charlie. No disrespect to you or any kind folks who have replied but...I did the mod because several folks starting to talk about it on forums once JJ has re-released the 12DW7. I then ran across Lord Valve's free article on Cathodyne PI's, enjoyed it, was curious enough to try it, and went ahead and plugged in some of his formulas in order to obtain the values for the resistors I was changing. So what I'd really love, is hearing if anyone can chime in on whether I applied the theory "correctly". For example, one thing I still haven't figured out, is how the value of Rg should be calculated. Thoughts?

Cheers,

Luca

P.S. I have changed speakers...going back and forth b/w a celestion g10 and a emi copperhead. both great. although I will say that the original oxford blows both of them away in "vintage mojo" up until 4 on the volume

[tubeswell]

Hi MWJB and Tubeswell (and anyone else who wants to chip in). Still very interested if you have any feedback for me regarding the load line calculations and subsequent mods I performed on my Princeton (pics and data are above). Thanks!

If you like it, it must be good

[dang!]

hi tubeswell,

thanks for sharing your extensive knowledge

i have a 65 bf pr that i'd like to use the reverse 12dw7 u mentioned as "ECC823" in V4

do u or anyone else know where can i get these?

i'd rather not mod a vintage amp...

thanks again for your help!

[dang!]

another question:
if i use the 12dw7 or even a 12au7 in V4 of a princeton reverb, do any other parts of the circuit need to be changed to accomodate the difference?
or is 12au7/12dw7 a safe substitue without any other mods?
btw
i was able by chance to track down a reverse 12dw7 on ebay!
thanks again for your help...

[lookah]

"do any other parts of the circuit need to be changed to accomodate the difference"

that's what this whole thread is about...just read the earlier posts.

fyi/ after going through all the mods i listed above for the 12dw7, i've reverted my 71 PR to stock with the exception of a bias pot and a new speaker. happy with it again. it is what it is

[dang!]

"do any other parts of the circuit need to be changed to accomodate the difference"

that's what this whole thread is about...just read the earlier posts.

fyi/ after going through all the mods i listed above for the 12dw7, i've reverted my 71 PR to stock with the exception of a bias pot and a new speaker. happy with it again. it is what it is

sorry if my question seems redundant.

my concern, as was yours in your initial post, was if 12au7 in v4 is in any detrimental to the amp, or if anything else needed to be changed to accomdate this different tube. there were some suggestions about possibly changing resistors. unfortunately, i'm a gtrst but not an engineer, so that part was over my head. i'm asking if those changes are necessary for the safety/well-being of the amp when using 12au7/12dw7 reverse in v4 for pi?

my experience, like you initially wrote, is that the pi sounds better with 12au7 than the 12ax7. since you've reversed the mods, is it possible for you to say how the bias pot affects your sound now? does it give you something similar to the sound of the 12au7 that you liked? also, which spkr do u like now?

thanks for your help

i found a reverse 12dw7 so i wouldn't have to mod the amp. right now i like the sound with 12au7 in v4, but would like to have the vibrato back too.

[tubeswell]

Hi dang, JJ makes the ECC823, go and see one of their agents near where you are. As for a 12AU7 in V4, that will make the PI sound good, but you will lose out on the trem, as the LFO stage ideally needs triode with an amplification factor of around something like 29 minimum IIRC, and a 12AU7 only has an amplification factor of 19 (hence why some guys use a 12DW7)

[dang!]

Hi dang, JJ makes the ECC823, go and see one of their agents near where you are. As for a 12AU7 in V4, that will make the PI sound good, but you will lose out on the trem, as the LFO stage ideally needs triode with an amplification factor of around something like 29 minimum IIRC, and a 12AU7 only has an amplification factor of 19 (hence why some guys use a 12DW7)

yes i know about the trem, so i just bought a ecc823 on ebay. this way i won't need to mod the tube socket, and will still have trem. my only concern is: are any other changes needed in the amp to accomodate using the 12au7 side in the pi? there was mention of changing resistors, but it was not clear to me from reading those posts if changing resistors was necessary or optional. is the increased current of a 12au7 a problem? i just don't want to damage the other components in the amp.

the sound seems a bit cleaner, sweeter, less raspy with the 12au7 to me. is that other folks' experience?

do u use this yourself?

thanks!

[MWJB]

There are no components that NEED changing to accomodate a 12AU7 or 12DW7 in a Princeton PI. the 12AU7 does draw more plate current than 12AX7, IF you were especially concerned you could measure the voltage drop across the plate resistors, square it, then divide by resistor value in ohms to ensure that you were not running them over dissipation. The half watt, carbon comp, plate resistors in old Fender are pretty bullet proof though, plenty off guys have run 12AU7 in long tail PIs at higher dissipations than you'll find in a Princeton, with very few issues.

[dang!]

There are no components that NEED changing to accomodate a 12AU7 or 12DW7 in a Princeton PI. the 12AU7 does draw more plate current than 12AX7, IF you were especially concerned you could measure the voltage drop across the plate resistors, square it, then divide by resistor value in ohms to ensure that you were not running them over dissipation. The half watt, carbon comp, plate resistors in old Fender are pretty bullet proof though, plenty off guys have run 12AU7 in long tail PIs at higher dissipations than you'll find in a Princeton, with very few issues.

thanks very much for your reply! i'll experiment with it more now that i know it's safe.

anybody here tried the 12au7 or 12dw7 in v4 of a princeton? i'm interested in hearing other perspectives...

thanks again for sharing your expertise... (especially considering that i'm rather electronically challenged!)