• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Best Sounding Phono Input Tube

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Opinions on the best input tube for a phono preamp with passive RIAA? I've used several integrated amps that use 12AX7A/7025 or 6EU7. Lots of other preamps use 6DJ8 variants. Some use 6SL7. Interestingly, many people seem to feel that a 12AX7/7025 is really not good for a phono amp, that a 6DJ8 is a poor sounding tube altogether, and that a 6SL7 sounds "squawky", whatever that means. So what is really the best sounding tube for a MM phono preamp. I'm planning to build a stand alone unit and don't want to just duplicate something that I already own in another amplifier. At this point I'm leaning towards Bob D's phono amp at http://www.geocities.com/TimesSquare/1965/phono.html but with 6 volt tubes rather than 12 volt since that is what I already have. Any input will be much appreciated.

Merry Christmas
 
There is no one "best." It depends on the cartridge, the topology, the presence (or not) of input transformers, and desired complexity or lack thereof. The sorts of generalizations you've heard are, by and large, wrong. Can you be a little more specific about what it is you want to do? That way, we can pass by the audio legends and zero in on a good solution for you.
 
tubewade said:
Opinions on the best input tube for a phono preamp with passive RIAA?

That's a virtually unanswerable question. What one designer will swear by, another will swear at. It's that much a matter of personal preference.

FWIW: 6DJ8s are at their best when cascoded. Used as a straight triode, they ain't much.

Another consideration: build the phono preamp, and include gNFB. Adding gNFB makes the circuit performance less dependent on both active and passive components. Of course, you're gonna hear from those who will tell you that's "heresy". Are they right or wrong? Not necessarily.
 
Thank you, SY. I value your opinion.

The cartridges to be used are a Shure M97xe on a Linn Axis, Shure M95ed and M75ed-t2 and Ortofon OM20 on a couple of lesser turntables. Topology I prefer would be probably classic two stages or two stage with follower like the RCA circuit in the back of the tube manual. I prefer simple to complex, where possible, in all things. I know that some things are too simple for their own good so I don't want that either.

No input transformer. Directly driven from the MM cartridge.

I want to build a (phono) preamp with the benefit of other peoples experience so I'm not just blindly copying the circuit and sound from an old integrated amp or receiver, unless, of course, there is some truly exceptional one that I need to know about.

Miles, your response is exactly the conclusion I had mostly come to. It's just a matter of preference. So what would you use?

Jeb and Andy, thanks for those suggestions. I'll study up on those valves.
 
tubewade said:
Opinions on the best input tube for a phono preamp with passive RIAA? I've used several integrated amps that use 12AX7A/7025 or 6EU7. Lots of other preamps use 6DJ8 variants. Some use 6SL7. Interestingly, many people seem to feel that a 12AX7/7025 is really not good for a phono amp, that a 6DJ8 is a poor sounding tube altogether, and that a 6SL7 sounds "squawky", whatever that means. So what is really the best sounding tube for a MM phono preamp. I'm planning to build a stand alone unit and don't want to just duplicate something that I already own in another amplifier. At this point I'm leaning towards Bob D's phono amp at http://www.geocities.com/TimesSquare/1965/phono.html but with 6 volt tubes rather than 12 volt since that is what I already have. Any input will be much appreciated.

Merry Christmas


Dude,

The other posters are correct. An absolute best simply does not exist.

An interesting complement is a 6ER5 in the 1st position and a 6GK5 in the 2nd position. Lots of gain is produced, which makes the combo an excellent mate for HOMCs.

The weaknesses of the "classic" RCA circuit are mediocre bass extension and miserable drive capability. Those problems can be addressed easily enough. A tweaked RCA circuit schematic is attached. A plus frequently overlooked about the 'X7 is the availability of a FINE tube, the Sovtek 12AX7LPS. Specimens of the Russian tube, with a superior noise factor, are available at modest cost. Add a nice, "uncolored", tone and the selection nears being a "no brainer". ;)
 

Attachments

  • classicrcaphonoforterry.jpg
    classicrcaphonoforterry.jpg
    57.3 KB · Views: 3,711
Coincidently, that's something I'm dealing with right now- after some years using MCs, I've ordered a promising MM, and need to design a new phono stage for it.

MMs in general, and Shures in particular, are fussy about their loading. If you use a high mu tube at the input, you'll have a very high input capacitance because of the Miller effect. For high mu tubes like 12AX7, trioded D3a, 6SL7, and the like, the input capacitance will run 300pF or more. With reasonably achievable values of cable capacitance, the cartridge will not be happy and the frequency response will look like a roller coaster, but you'll be able to tell everyone you know that 6SL7s sound squawky.

You then have three choices:

1. A lower mu input tube, but the need for two more gain stages.
2. A tube cascode
3. A hybrid cascode with a FET on the bottom

If I were to do one of these three today for a publishable design, I'd probably use an ECC88 variant, just because I know them so well and they're easy to find and inexpensive. Another choice, highly recommended by two guys I trust (and thanks for the tubes, fellas!) is the 6AQ4, and I'll probably end up using that in the one I'm starting next.

FWIW, my last phono preamp, which I used for over 20 years, went with approach 3, using an E88CC variant on top and an NTE458 on the bottom.

And purely IMO, but with some objective reasons, there is not a single "classic" phono stage that's worth a warm bucket of spit.
 
tubewade said:

I want to build a (phono) preamp with the benefit of other peoples experience so I'm not just blindly copying the circuit and sound from an old integrated amp or receiver, unless, of course, there is some truly exceptional one that I need to know about.

You can trust Gary Pimm's ears and designs. Here's the info:

http://home.pacifier.com/~gpimm/phono.htm

Enjoy,

-- josé k.
 
tubewade said:
Miles, your response is exactly the conclusion I had mostly come to. It's just a matter of preference. So what would you use?

First choice: KA5532. Yeah, I know: EWWWWW! SAND! But, seriously, what's another op-amp or two?

If I had to go hollow state, I'd go with a differential cascode up front. For a magnetic cartridge, you're gonna need gain, and lots of it. The usual solutions, a high gain triode or pentode present their own problems indeed. These high-u triodes (12AX7A, 6AV6, 6SF5, and so forth) have some horrendous Cmiller. That's been the big problem using these "up front", regardless. Pentodes have their own problems: noise.

So that leaves the cascode: lots of gain, and manageable Cmiller, and no partition noise. For a phono preamp, I'd go with cascoded 6SN7s. The 6SN7 is quite linear, and, at least the Sovtek version, has very low microphonics. I'd get two of them, and build a differential cascode with an active tail load composed of cascoded BJTs. The magnetic cartridge, being a coil, operates in a differential mode naturally, and a differential front end also eliminates common mode noise. That'll get you your gain. Output stage: Brosky Cathode Follower balanced-to-single-ended converter.

Need I mention well filtered DC for all filaments?
 
I'm stumped on this choice for a future design myself too.

417a's used to get a good rap. Obviously it was too good, since prices for NOS went from $10 to $60 pair. This was going to be my choice, but apparently you have to screen 20 of them to find a single quiet pair. And at these prices (and increasing), I think I'll be looking for another type!

So I'm trawling though the other options such as triode-connected pentodes and the like. (D3a, C3g, E280F, etc. And some of these are already not that easy to find and afford now too).

Of course if it's 'current production', the the 6DJ8 seems the way to go. I once read though on a DIYers site that compared to the 417a the 6DJ8 was something like "50 times less sensitive to input signal variations from the phono cartridge" thus making the 417a a better choice for audiophiles. I never bookmarked the site and have no idea what that claim was being based on. If someone can explain (or refute) it to me, I'd be interested to hear.

All the foregoing notwithstanding I think balanced input is the way to go even if it was for just the first stage. Although there are a few advantages, in a tube context the major one is that the input capacitance seen by the cartridge is halved as compared to a single tube; which gives us a little more scope to maneuver.
 
SY said:
I do. It's what's left on the ground after the bull leaves.
a Matador? :D

Seriously though, I remember the page quoting respective tube figures for each type to support the claim. (Whether or not said use of the figures actually validated it of course is another matter). But since I can't seem to re-find the page, I guess the point is moot for the moment anyway.

dsavitsk:
Any comment on comparative sonics of the 6J9P ?
 
Guys,

Another alternative to the 417A is the Russian 6s45p. All of the ultra-high gm triodes require extraordinary measures against parasitic oscillation. Don't forget that these "puppies" were intended for service as grounded grid RF voltage amps.


----------------------------------------------------------------------------


What follows below is "blue sky" conjecture. Make of it, what you can.

An EF86 in the 1st gain block neatly disposes of the Miller capacitance issue. The price is marginal S/N performance. The S/N balance can be tipped in the EF86's favor by using a very large load resistor, but such loading of pentodes restricts HF bandwidth. The bandwidth limitation would have to be factored into the equalization.

Another way to take advantage of the large stage gain of an EF86 is to use a CCS loaded 6922 section in the 1st gain block. Even with the 20 dB. loss from the EQ network, the signal level is high enough to quietly drive the pentode.
 
Deux centimes

The great thing about a forum such as this is that people who don't know much about a particular - or indeed - any topic, can still add their two cents/pence/centimes or even xu to the discussion.

In this case I refer to myself.

I have often read that since noise is a major issue with RIAA stages and low-output MC cartridges, then we should like to see high gm at the first stage. However, as remarked above there is a penalty: the high-gm types have µ of >40 (85 in the case of strapped D3A) thus imposing a Miller capacitance penalty.

So if I am correct, we are seeking a valve with high gm and medium µ.

Once again I present to you one of my little hobby-horses: the E282F (triode-strapped). With its µ of 25 and gm of 30 I would guess that it is a candidate?

7N7
 
Paul, that seems like a really good candidate, but it will not be trivial to find in the US.

For an MM like the Shure, ultra-high gm is not that important- if the cartridge has an internal resistance of 1k or more, the difference between an equivalent noise resistance of 60R and 600R is only about 3dB of S/N. For scale purposes, an ECC88 has an equivalent noise resistance of about 300R at 10mA current, so there's only 1.5dB or so more noise than a "perfect" tube.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.