# Phase inverter, 12AT7 or 12AX7?

Discussion in 'Amp Central Station' started by JohnnyCrash, Jun 15, 2019.

1. ### JD0x0Poster Extraordinaire

Age:
31
Posts:
5,919
Joined:
Feb 22, 2009
Location:
New York
I think NOS Milspec JAN are like \$2 or \$3 more than something like a JJ. One of the rare occasions where I'll opt for NOS, because the increased cost is essentially negligible.

2. ### Greg_LTele-Holic

Posts:
969
Joined:
Jun 27, 2018
Location:
Texas
Can you go into a little more detail on this please? I'd like to get a better understanding of what you're tweaking and why. If you don't mind.

3. ### LeicaBossTele-Holic

Posts:
699
Joined:
Oct 16, 2015
Location:
New Jersey
I agree, NOS phase inverter is one place where I don't mind a few bucks extra

4. ### 11 GaugeDoctor of Teleocity

Posts:
12,040
Joined:
Mar 21, 2003
Location:
Near BWI Int'l
Sure, or at least I'll try.

I started making adjustments to the PI's tail resistor after reading what Randall Aiken said about it here:

https://www.aikenamps.com/index.php/the-long-tail-pair (scroll down to The "tail" resistor section).

...What Randall writes is super-verbose IMO, and it becomes clear (at least to me) that you have to understand how a singular part of the circuit works in conjunction with everything adjacent to it.

Anyway, the critical stuff seems to be encapsulated in this paragraph:

"The tail resistor also "bootstraps" the stage, resulting in a higher input impedance, due to the local feedback action, as described in the grid resistor section above. Note that the bias resistor, R5, sets the current through this tail resistor. The amount of current set by the bias resistor, along with the value of the tail resistor, determines the DC voltage dropped across this resistor, which, in turn, partly determines the headroom of the circuit. If no global negative feedback is used, the tail resistor should be made as large as practical, with respect to the amount of current being drawn, and the desired headroom of the amplifier. This will give the best balance to the PI outputs. This resistor has little effect on gain, but a major effect on balance and headroom."

I made other changes after reading about preferences for BF vs. SF, and how some folks would actually leave some of the PI stuff as is in a SF amp that they'd otherwise BF'ed. It was also somewhere within this guy's website - http://tone-lizard.com/ (I don't recall the exact page, as it was quite a few years ago) - that I saw mention of using a lower value for the power tubes' grid bias leak resistors. The SF amps already typically have 100K resistors for that, and I had a Marshall years ago, that originally had 150Ks, and (originally) used 6550 power tubes.

...It was probably about a decade ago, or maybe longer, that my dad picked up a really rough AB165. I recapped it for him, and did the "half-backdating" or whatever you want to call them changes to it, to give "the best of both worlds" of it and its predecessor. Ultimately, that amp still dirtied up too fast for my father, so I ended up getting it from him a couple of years ago.

Originally, I didn't mind how easily the AB165 seemed like a natural candidate for playing old AC/DC stuff (especially for getting Malcolm Young types of tones thru it), but I wanted as much of a clean with the tight low end as I could get. So, along with modifying its "extra" gain stage (with the local negative feedback) to overdrive significantly less, I changed the power tubes' grid bias leak resistors to 100K.

There are other articles and suggestions to do things like decrease the output coupling caps from the PI down from .1uF to .047uF or smaller, and I've done that too, but it seems to have less effect than the tweaks mentioned above. It's something I'll typically do to a (SF) DR, but otherwise, I usually will pass on it.

Another tweak to consider, or to simply "not BF" in a SF amp, is the value of the PI's plate resistors. My '71 VR had 47Ks in there, and I originally BF'ed them to 82K/100K. They're now back up to 68K or 56K. I think this has more value than reducing the value of the PI's output coupling caps.

...Since a lot of folks find value with at least BF'ing a SF amp so that the bias on the power tubes can actually be adjusted, and since there is typically the benefit of decreasing the value of the PI's input cap (which Randall Aiken discusses, too), I simply will add the other 2-3 changes while I have the chassis pulled.

Last edited: Jun 16, 2019
telemnemonics and Greg_L like this.
5. ### bftfenderPoster Extraordinaire

Posts:
5,271
Joined:
Dec 21, 2017
Location:
York PA
I understand it..but why not just get a jtm45? it does what you are asking from the door ac/dc & big cleans) & stays big clean...master up & chan volumes down..it's chimey clean..much more so that people give credit for. but goes into ac/dc land even if you cant play it, it almost gives that tone soon as you strike a chord, back guitar vol to 5 & big fat clean..

6. ### bftfenderPoster Extraordinaire

Posts:
5,271
Joined:
Dec 21, 2017
Location:
York PA
absolutely selling my tweed deluxe head now,,it will never give the goods..i feel a kt66 bluesbreaker in my near future

telemnemonics likes this.
7. ### 11 GaugeDoctor of Teleocity

Posts:
12,040
Joined:
Mar 21, 2003
Location:
Near BWI Int'l
...Because I want more than the AC/DC thing. I actually don't play that kind of stuff as often, any longer.

I also didn't mention that the bass channel in the AB165 also got modded for more of a JTM'ish tone stack, along with some smaller cathode bypass caps.

...Since I didn't outright bypass the AB165's "extra" gain stage, but only made it clip much less, I've actually arrived at what's pretty close to my dream amp, at least for what I mostly play right now.

...I can actually get "crunchier still" on the AB165's non-bass channel by adding a fairly mild drive box. That makes it the equivalent of a super-versatile 3 channel amp.

I've already got way more amps than I can play, so it comes down to having enough variations with what I'm not planning to let go of. For that reason, I don't intend on either getting or building a JTM45 or anything like a tweed Bassman.

...If anything, all of those amps are simply brutally loud, and not very portable. But good lord does that AB165 sound unreal, especially thru a 2X12 open back cab w/a pair of WGS ET65 speakers.

bftfender likes this.
8. ### Greg_LTele-Holic

Posts:
969
Joined:
Jun 27, 2018
Location:
Texas
Wow, very interesting thank you. As a noob amp builder myself, it's a lot to digest but I'm gonna chew on that info for a little while and read the links you provided. Thanks for your time.

Age:
61
Posts:
26,715
Joined:
Mar 2, 2010
Location:
Maine
For years I always swapped the 12at7 for a 12ax7, pulled V1, and removed or altered the neg feedback to try to get those Fender amps to sound more Marshally.

Finally just switched to old Marshalls and let the Fenders be Fenders, though I didn't really swap all those changes back, because they really didn't seem to do all that much. Stayed with them for the idea that they were more in the direction of impolite sound.
Heard the most from neg feedback change.

Super Reverbs get more impolite than most 6l6 BF/SF without losing bass clarity as badly as say a Pro Reverb, or it seemed that way.
But the other conundrum was that the old CTS speakers sounded better but lost that tighter bass a Super could get.
Tighter newer ceramics got the bass tighter but the tone suffered compared to old CTS.
I'm pretty sure I had those Emi tens (blue) in the Bassman RI and ran a Super through them, actually went back to four old Jensen and Utah 12s for fatter less strident sound with solid bass, obviously not all in the BFSR cab.

Overall I've had literally tons of great sounding old amps, but none have been perfect.
Marshalls got closer than Fenders, but the old Bassman heads are contenders.

I'm not familiar enough with the various BF/SF Bassman circuits though, and they seem endlessly tweakable, which I've also not really done because they work pretty good stock. I suppose all these amps are tweakable, but the Bassman is cheaper.
My last SFSR is overly tweaked and ought to be re tweaked for the 21st C!

I have a Tweed Bassman like Holland, had a Fender Bassman RI, and had an aluminum chassis Park with GZ34 that would have been a 45/50 but had no model designation.
All those very similar circuits seem to have a pretty huge range of sound, despite being so close to the same amp.

Seems like the biggest difference is PT voltages.
But of course after moving away from the stock circuit any can be moved in other directions.
That's another thing with buying vintage amps and Marshalls in particular.
We can have different experiences with the same model because techs over the years took liberties with the circuits but we didn't get the memo.

For all the preamp changes possible, I think I like lower power section voltage and inadequate filtering for the touch response range available, but don't really need the tube rectifier as much. All anecdotal though, those specs seem present in amps that respond how I like.

Just remembered I found putting a 12at7 PI in a 1974 year (not model) Marshall 50w Bass made a bigger difference than any time I changed a Fender 12at7 PI to a 12ax7. Seems like a bigger loss of gain in the Marshall, compared to not much of an increase/ change in gain to the Fenders.

bftfender likes this.
10. ### bftfenderPoster Extraordinaire

Posts:
5,271
Joined:
Dec 21, 2017
Location:
York PA
Great, thats the beauty of the circuits..you can hone inn ones particular style.et65 are great speakers in open back, nice set up ya have !

11. ### 11 GaugeDoctor of Teleocity

Posts:
12,040
Joined:
Mar 21, 2003
Location:
Near BWI Int'l
My pleasure.

Yeah, it is a lot to digest, and stuff like the Randall Aiken articles are super tech-oriented, so it's very hard (IMO) to just casually read.

I actually got pretty well-versed in the technical aspects of tube amps a few decades ago. Then my daughter was born, and I stopped for a bit. Then I got a decent high voltage shock/burn on my hand, despite being extra careful. Since then, I've just limited all the tinkering to primarily maintenance, but I'll do (well researched and planned) tweaks as needed. I figure that will help improve the odds of no lethal accidents happening.

Greg_L likes this.
12. ### Greg_LTele-Holic

Posts:
969
Joined:
Jun 27, 2018
Location:
Texas
I've been scouring Rob Robnette's pages for about year now. I love it. Even if it's over my head, I read it and often some time later it will click and I'll get it as my knowledge base grows. So yeah, I'll read anything, even if I don't understand it right this minute.

bftfender likes this.
13. ### archetypeFiend of Leo's

Posts:
5,880
Joined:
Jun 4, 2005
Location:
Williamsville NY
I'm no expert, but a lot depends on depends on the PI circuit you have. You've gotb

I've got an SCX2 with a cathodyne phase inverter in a Princeton circuit. It comes stock with a 12AX7 and the first triode is a driver, second triode is the PI. With a 12AX7 I get a bit more drive into the PI than lower gain 12A** family tubes. I've used a 12AT7 and got a bit cleaner results, but am rolling tubes in that position. Currently it's got a 5751.
Tube Depot bought a ton of NOS JAN Philips 12AT7 tubes. They run a sale on these at least twice a year and I bought 2 of them at about 8 or 9 dollars each. They'll outlast me.

14. ### 11 GaugeDoctor of Teleocity

Posts:
12,040
Joined:
Mar 21, 2003
Location:
Near BWI Int'l
I got a complete 5E3 from Boot Hill Amps last year - this was my first new tube amp in years, and was intentionally purchased with the idea of just being able to play it.

...It was only a few weeks after I got it that I was snooping around sites looking for tricks and tweaks, and IMO, RR's stuff is probably as good as it gets. It would have been great to have something like that years ago.

The other great thing about RR's stuff is he's pretty clear at conveying the likelihood of tweak X being the kind of thing that any given player might find appealing, and might bond with for the long haul. IMO, that makes it easy to cut to the chase. I'm all for tweaks, but not if they lead to a rabbit hole with diminishing returns.

Greg_L likes this.
15. ### Greg_LTele-Holic

Posts:
969
Joined:
Jun 27, 2018
Location:
Texas
Yeah right! I like how he explains the tweaks, what they do and why, and also shows you where on the schematic and layout that these changes take place. It's been very helpful in learning how to compare the symbols on a schematic to the physical location on a layout or actual physical board. I successfully built my very first amp last year - an AB763 Deluxe Reverb - and the info at Robinette's site was pure gold.

bftfender likes this.
16. ### Rick330manTele-Afflicted

Posts:
1,536
Joined:
Jan 9, 2011
Location:
Florida Keys
I've gone to using the 1980s production JAN GE 5751 in the P.I. slot almost exclusively. In amp after amp, it brings out a clarity and detail that no other P.I. tube can.

17. ### 11 GaugeDoctor of Teleocity

Posts:
12,040
Joined:
Mar 21, 2003
Location:
Near BWI Int'l
Congrats on that!

...And again, I would have loved if RR had his stuff online years ago, because one of my earlier tube projects was rebuilding a DRRI with a (populated) Hoffman turretboard, for someone. In hindsight, it would have been a much easier build if just starting with an empty chassis and all of the necessary pieces. Retrofitting for all of the non-RI bits was an extra level of a challenge. I had to relocate the reverb tank send & return to the backside of the chassis, and/or the footswitch jacks (too long ago to remember clearly now). I had to do "something clever" with the filter caps, because Fender changed a few things with the RI, and it wasn't the same, and I didn't know that until I got the doghouse cover off. And I clearly remember the pain of having to ream out all the pot holes in the front of the chassis, because the bigger CTS ones didn't fit.

...In retrospect, drilling the existing chassis for the turretboard's nuts/bolts/standoffs was probably the easiest part.

By the time I got it all done, I think I had a really good idea of all of the right and wrong ways to rout and solder wires for a "nontrivial build" such as the AB763.

So - on the bright side, after someone has built something like an AB763, it makes a lot of other projects much simpler and straightforward!

Greg_L likes this.
18. ### Greg_LTele-Holic

Posts:
969
Joined:
Jun 27, 2018
Location:
Texas
Thanks!

That's cool. That actually sounds like fun to me. I'm a weirdo I guess!

Yeah, I've got an old SS combo out in the garage doing nothing. I'm thinking about gutting it and turning it into an 18W Marshall TMB. The Marshall schematic seems like child's play compared to the Fender.

19. ### archetypeFiend of Leo's

Posts:
5,880
Joined:
Jun 4, 2005
Location:
Williamsville NY
The JAN GE 5751 is what I currently have in my SCX2. I'm liking it.

bftfender likes this.
20. ### bftfenderPoster Extraordinaire

Posts:
5,271
Joined:
Dec 21, 2017
Location:
York PA
my tech recommended it for my Laney Ironheart 60. It radically changed the amp, i dont use it for metal but for clean , the channel is pristine but it can be set to slight OD, fantastic with a strat

IMPORTANT: Treat everyone here with respect, no matter how difficult!
No sex, drug, political, religion or hate discussion permitted here.