6922 6DJ8 or 12 AX7 tubes better for preamp?


I have found conflicting information regarding these 2 tubes for preamps. One camp, including Roger Modjeski in this very detailed article with testing data http://store.electron-valve.com/suof6dforau.html
feels the 6922 is superior. Tim Paravicini (EAR) uses the 6922 in the 88PB Phono preamp and is used in the Manley Steelhead. My local retailer, whose opinion I value, strongly believes that 12AX7 tube is superior, which is used in the Aesthetix IO and in the less expensive EAR phono preamp, 834 P. The 12 AX7 was reportedly specifically made for audio. Detractors of the 6922 state it sounds sterile. The IO and its 12AX7's has been criticized for being noisy, unless one takes extraordinary care with cables, tube selection, and tube matching, which would appear to support Roger's thesis. Am I sinking into the depths of audiopilia nervosa in evaluating tubes as part of my selection criteria for a great phono preamp? It goes without saying that audition is essential in the final decision. I have heard a modified EAR 834P on Quads Unlimited 57's and liked everything about what I heard, except would have liked a somewhat larger soundstage, which was likely more related to the speaker. I hope to audition an IO soon.
tompoodie
Ummm.... this is all baloney. The designer chooses, or should choose, the tube best suited to the job. The 12AX7 is a very high mu (gain) tube that is best suited as all or part of the input voltage amplifier in a phono stage. IMO, one never ever needs that much gain for a linestage. However, a linestage does need to have a very low output impedance and the capacity to deliver current. (The two go hand in hand, actually.) Thus you can say that the 6922 and its congeners are best suited to use in a linestage. However, they can also be used to make a very nice phono input stage, by using a pair in cascode, for one example. The 12AX7 has a very high plate resistance and tolerates only a very small amount of current. The high Rp gives it a high output impedance and very little current is delivered, which makes it a lousy choice for a linestage. If 12AX7 is used in a linestage (as is often the case in commercial products due to its cheapness and availability), then one almost certainly needs an additional tube to be used as a cathode follower, to achieve a useably low output Z. The 12AU7, which has the same electrical parameters as a 6SN7, is frequently used for this purpose. (Worst of all is when a 12AX7 is used as CF, which I have seen in some products.) However, IMO, the 6SN7 sounds way way better than any 12AU7 (I have done the listening tests, FWIW). Among medium mu, high transconductance tubes (like the 6922), there are a few other choices, like the 5687 and its relatives, that I personally like better. Then also there are low mu triodes, like the type 27, etc, that sound great in a linestage. Once the proper tube type is chosen for the job, then sound quality will vary depending upon other things, like the linearity of the tube at operating points determined by plate voltage and current, etc, the nature of the plate resistor, coupling caps, etc. SO, it is folly to talk about how a tube sounds without considering these many other factors.
Thank you for the link, Dave...I will digest that in the course of the week.

Lewm...if I understand you correctly, for a phono stage such as the EAR 834p with three 12AX7's and a pair of output transformers, is that a reasonable compromise to use 12AX7'S in that context? When modified, it is quite impressive at its price point to my ears. I haven't auditioned phono preamps and preamps with phono stages above $5K, both of which it seems to compete against. I would like to avoid using a preamp, and create a phono only system. Thank you for such a detailed and knowledgeable response.
Dear Tom, Yes, IMO the 12AX7 is fine in that application. If I understand you correctly, then the 834P is using the output transformers to reduce the output impedance that would pertain if the output was taken directly from the 12AX7. So the transformer substitutes for a cathode follower. In no way did I intend to question the work of Tim de Paravicini. He's one of the best.
The 12AX7 is great for noise and low distortion, but is bad at bandwidth, making it a poor choice if you want passive EQ. The 6922/6DJ8 has good bandwidth, but the design of the tube is not for audio: while very good linearity, the design is prone to excessive microphonics. IMO anything designed with them will exhibit a sonic signature that has a lot to do with microphonics.

12AT7s are also designed for audio and have less microphonics. They are also higher mu yet allow for wide bandwidth and so seem to me a good compromise of gain vs bandwidth vs coloration compared to the other two types, assuming that you want passive EQ.

So in this case we see that the design of the circuit says a lot about what the tube choice is, for example if you are running feedback for the phono EQ the 12AX7 can do quite well.
I have an Aethetix Rhea that is noisy. Mentioned in the initial question is that extraordinary care needs to be taken with cables. Could someone expand on that requirement. I am going to retube the unit and perhaps I need to deal with cables also.
thanks