Listening: trying to discern differences in sound between various 12AX7

Morzh

Morzh

Active Member
Hi everyone,

So. I finalized my second stereo system after having restored two Eico HF-22 monoblocks, and having acquired a C22 (1960-s version) in rebuilt condition (recapped+power) plus SA-14S1 SACD player (used but fairly new - I picked it because I had one I bought new and I liked it, and the price was good) I put it to the test.
The speakers I have are a pair of Klipsch Cornwalls, the ones with vertically oriented horns.

The C22 which is the item in question here had originally JJ 12AX7 in it mixed with Marshall ones. The Marshalls tested a bit on the lower side (still good, but one of them borderline) and they were, I think, in V5 and V6 places.

Well. I listened to some things I have in my phonotheque -
- Nazareth (Malice in the Wonderland)
- QUEEN (II, NITO)
- Vivaldy Stravaganza (Rachel Podger)
- Tom Petty (the best of, mostly using the "Free Fallin'").
- Lynyrd Skynyrd (Sweet Home Alabama).

The system sounded great the way it was. However knowing a couple of tubes deserved retirement I ordered the following setup from Jim McShane (I only say that to anticipate the questions about what the source of the tubes was - I think with Jim it is as good as it gets); my choices were pretty much the result of consulting him:

- Two Sovteks for the Phono input (that I was not using in the audition as I am not playing the phono at this point, so I used the placeholders - the tubes that are not necessarily good but will provide the needed filament interconnect)
- Four Genalex Gold Lion for the V3-V6 position
- Two Tung Sols ECC803S to try in V3&V6 positions (described as "more energetic")
- Two good original used Mullards ECC83, that Jim had for sale.

So the first arrangement was all (except the noted V1&V2 phono tubes) Genalex.
Then I tried the Tung Sols, and then - Mullards, getting back to all Genalex.

What can I say.
Before you guys start bashing me, I have to say that as a child I was diagnosed with "Golden ear" condition :) which, I believe, still persists (I recently started learning playing classical guitar and my teacher is quite amazed with what I have, though my playing...well....sucks :) but this is the hands, not the ears).

After putting the Genalexes in, I noticed.....no perceived change. I did sound great before, it still sounded great, but when I went back and forth, those details that I noticed did not change with the tubes.

Then I put the Tung Sols in V3 and V6. I tried to listen to it after warm-up, tried to see the "more energetic" changes....I failed to discern anything different.

And then there were the Mullards. With very much the same results.

I spent quite a bit on the extra tubes. I do not regret it, it is good to have a supply of good tubes if one owns tube equipment. But......here's what I want to ask:

Is there any setup you guys use when doing that type of comparative listening? It is not like I intend to go the whole hog and buy two C22 amps and a switch. (I mean, I do have a second one, but it is a re-issue so not quite exactly the same). Obviously comparing sound after you played one set, then turn it off, swap the tubes, turn in on, wait 'till they are warmed up, is not ideal at all - you will forget the sound of the first set by that time.
I even think maybe what would help is building a small chassis with several tube panels and extention cables, and a switch that would keep the tubes hot and switch only the non-filament electrodes (has to be thought out well, as that could result in popping/transients/etc), so different sets could be plugged and evaluated without long wait, but.....if there were substantial differences, wouldn't I noticed at least something?

I'd like to hear from folks with experience of doing this kinds of things.

Best,

Mike.
 
I run a Jolida 502CRC tube amp. It came with TungSol 6550 tubes, and I got some Genelec Gold Lion KT-88s from Jim McShane last year.
Jolida specs the Bias setting for the output tubes at 0.500v - but a friend recently recommend that I
"reset the bias to 38-ma (aka. 380-mv -or- .38-volt reading) you'll find that mid-range opens up and the sound is more open overall. The added benefit is better tube life"
Sound like good advice to me - tho I'm not sure I hear the difference, extended tube life is a plus!

So, I think the original "preamp level" tubes were Russian Electroharmonix, which I swapped for NOS GE brand tubes -
12 AT7 JAN (GE Joint Army Navy)
12 AX7 (marked GE Made in USA)

And yes, it IS awkward to AB when do some listening, power down the amp, swap the tubes, rebias the amp, then do some listening ...
 
Last edited:
A couple things

With any given set of tubes installed, with no input signal, turn the volume waaayyy up and listen for how much hiss you get. Some tubes are much quieter than others with regard to noise floor.

Now, with the volume lower, but still no input signal, tap on the tube with the rubber eraser of a pencil--this will give you an indication of how "microphonic" a tube is, and this is generally indicative of internal construction (mica isolators). MIL-spec tubes generally have greater isolation and are far less microphonic. Microphonics can cause audible distortion (basically feedback) at louder listening levels if in close proximity to speakers due to vibration (kind of like a TT).

12AX7's come in several variants--long plate vs short plate, and black plate vs grey plate--some are even ribbed. The long plate provides greater surface area for electron transfer and heat distribution. I have no idea as to the significance of grey vs black, but I am a fan of long black plate 12AX7 tubes--maybe someone else can explain it for us both--I believe it has to do with material of construction, but I don't know the how or why of the difference.

Finally--pick A track--just ONE--something that you are very familiar with and with good dynamic range. Listen over and over, focusing on the high end "sparkle", low end extension and tightness, and midrange presentation. I am not a fan of Mullard tubes--many are, but I refer to them a "Dullards" in that IMO, they lack high end detail and extension, but many prefer their warm, laid-back presentation. The russian tubes tend to have a "brighter" top end, but a "thinner" bottom end--again, just my opinion--others may feel differently.

My favorite tubes for phono stages that utilize 12AX7's are Baldwin "greenies"--they are Sylvania long black plate labeled for Baldwin for use in their organs. Very quiet, high gain tubes. You will see other Baldwin labeled tubes with white or orange screening, but they were made by Raytheon and RCA IIRC and are not as good.
 
Last edited:
My Raphaelite DP84 amp and modified Paradigm 7SE speakers are quite revealing, and I have some memory of sound.

I prefer the 12AX7 RCA long black plates over many short and long plates tubes in this setup.

But sound depends more on setup than only tubes themselves.

To improve for sure a setup, try to add decoupling caps to each electrolytic filtering cap, from main B+ filter, to B+1 , B+2 , to cathode bypass caps...
 
It is interesting that you did not notice any differences when you had the Mullards installed as they always seem to add a British character to the sound. I mean they seem to be more polite. Your C22 is a McIntosh tubed preamp? Since other makers like to 'borrow' numbers to try an elevate their product. It may be that your preamp could also use some maintenance. Before doing a listening test it is good to make sure that the equipment has clean rca input and output jacks, pots and switches lubed, and tube sockets also cleaned and pins tightened. Hopefully it was recently overhauled with new power supply caps and coupling caps. Also the C22 original tubes were Telefunken 12ax7 if i am correct. I always feel that it is best to have the original type tubes in the piece to use as a baseline. While i am not a big fan of tube rolling, i would expect there to be subtle differences between vintage and modern tube types.
 
One thing I would be interested in seeing, and this is just for my own technical reasons, is some comparison of how the various brands of tubes bias in an otherwise identical circuit. The reason I'm curious about this, I've recently run into a situation where the Sovtek tubes act differently enough to cause less than ideal behavior in a circuit that works absolutely perfectly with any old production tube I can stuff in there. I'm just curious if any of the new tubes bias closer to old ones for touchy circuits like that one.

and yeah I know most of the time when we talk about tube biasing we're looking at output tubes, but all tubes do this. Most small signal tubes use a cathode resistor and just take care of themselves. The circuits are usually forgiving enough that basically anything works on. Except when they don't.

I also suspect that this plays into how different tubes may sound, but its probably one factor of many and not enough to draw any solid conclusions from.
 
Binary Mike:
I haven't noticed a deep feedback in C22. Some corrections like in V4 via C35, but not the overall gain setting one.

Primosounds:
Yes the amp has been serviced recently. I wish they left the Bumblebee caps in so I could restuff them but they replaced them.
The amp is the Mac C22 from 1960-s. Not someone else's C22 (not sure who makes a C22 other than McIntosh?)

One mistake I made: I was told to switch tubes in V3 and V6; now I realize changing V3 was not effective as it is a tape tube. Though V6 should have the effect if it is warranted by the differences.
What I also should try is V4 and V5 as those are the line tubes whch is what I am using for the CD input.
Then again, when I yanked the old JJs out and put Genalexes in, these tubes were changed, so at least that was correct.
 
What kind of caps did you put into the c22? The original BB were a PIO type, so replacing like with like keeps the sound the same. I rebuilt one maybe 5 years ago and used sprague black beauty and when i plugged the preamp into my system i was pleasantly surprised at the sound quality which was quite good. I did not do any tube swapping as it had a brace of Teles installed.
 
When swapping preamp tubes, I like to audition a tube (or pair, depending upon amp design) for a session of varying types of music lasting about 40 minutes. Then I will switch 1 (or the pair as noted previously) and listen to the same music. I try to listen beyond timbral and "watch" the soundstage for width and then depth. Tubes with wider and deeper soundstage win, excepting tubes that have harsh treble or dim/dopey bass response.
people that call the Mullards "Dull sounding" may be missing the thing I find them best at- soundstage depth. I noticed it first playing through an old Marshall guitar amp in MONO. The sound had a 3d quality that you just don't get with most guitar amps.
 
primosounds said:
What kind of caps did you put into the c22?
Click to expand...

I was not the one who rebuilt it. I bought it off of eBay and the seller said this:

"The preamplifier has been professional service by 40 years experience Marv Electric at Houston TX". I do not know them - if it were my amp I'd rebuild it myself, never went to service folks before. I just used a service for the first time - but this is Willy Hermann and the reason for the job (although he will go through the whole thing) is primarily mechanical.
However I still have the eBay photos: from those the caps seem to be "Audioclassics".

s-l1600 (1).jpg s-l1600 (2).jpg


This said, myself I am not a strong believer into interstage / freq correction capacitors affecting sound much (I do see how a power decoupling caps can). But then this is an EE in me talking. He is the one who I am trying to defeat while writing all this :)
 
Fr
Morzh said:
I was not the one who rebuilt it. I bought it off of eBay ......


This said, myself I am not a strong believer into interstage / freq correction capacitors affecting sound much (I do see how a power decoupling caps can). But then this is an EE in me talking. He is the one who I am trying to defeat while writing all this :)
Click to expand...
The photo shows that the preamp was rebuilt recently. The choice of coupling caps is rather generic, being plastic type caps, probably polyester or poly propylene and metallized conductor. Nothing wrong with that but probably slightly different sonic signature from PIO type. To me a classic like the C22 should be rebuilt as close as possible to the stock unit. Then you would know what the sound engineers at Mac were trying to achieve.
 
PIOs....are they even made anymore? The only axial PIO I am seeing today are the Soviet made. I had gobs of them while growing up in there (for that matter those were the most common if not the only axial type we had around, like "KB, KBM, BM, BGM, MBM" plus some large cube-shaped ones, like "KBG-I, KBG-MP" etc).
But from my experience with the PIO, they all go bad eventually. In old radios I restore every single one has leakage and the value drifts up 2-3 times. I have no experience working with the later versions, though the Mac's caps were 1950-s - 1960-s, so fairly recent compared to 1930s I deal with, and they still failed. By using those, doesn't one expose himself to the same time-delayed issue, especially considering those Soviet caps one could find today are quite vintage as they were no longer made since 80-s?
 
But Mundorfs are not Paper caps. They are Polypropilene type. Yes, in oil. Oil is a way to achieve higher breakdown voltage.
 
Where did you see that ?

I doubt that they put polypropylene in oil ?

Edit, ok, I readed that, but they do sound excellent anyway.

The only risk, is that they last for longer than paper.
 
Last edited:
I have no doubt they sound excellent. BTW so do those I have in my amp: I do like the sound.
But as Primosounds noted, if you want to do a restoration as close to authentic design as possible, the PIO caps have to be used. Which I am not sure where to go for those, as the only source I am seeing so far is the old Soviet-made caps, which I could get if I wanted to (I have my connections in the former USSR) but then every single one of them has to be thoroughly tested and then after another 10-15 years it's a good question if they will perform the same, as they are 30-60 year-old now even when NOS.

Again, from the experience with the antique radio sets, I could see that the originally used PIO caps started being replaced as early as 10-15 years from the date of manufacture. With PIO caps made then. Which all got bad eventually, and I am sure it was as fast as it was with the originals.
 
My personal goal is always to have the better sound, not the "original" sound.

I hardly keep sonething not modified at least a little bit.
 
I believe the Russian PIO caps are hermetically sealed like some of the really good USA made caps. That will make a huge difference in the life of them if no oxygen can get inside and degrade the oil and paper.
If they test as new when installed they will probably last a lifetime.

BillWojo
 
You must log in or register to reply here.

Similar threads

ssmith3046
Genalex 12AX7 in the Fisher 500C
Replies
6
Views
1K
miner
miner
vonclod
Buying tubes any difference from guitar amp to home audio
Replies
8
Views
455
vonclod
vonclod
F
12ax7 reissue..mullard vs tungsol...
2
Replies
22
Views
5K
fabchef
F
twist72
Going from a 12ax7A to a 12ax7
Replies
2
Views
599
twist72
twist72
dosmalo
Power tube pin sockets
Replies
4
Views
398
jaymanaa
jaymanaa
Back
Top Bottom