HiFi Haven Group Project - Building The "World's Best Tube Phono Preamp"

[J-ROB]

I was curious if that tape amp used a transconductance amp or whether the reference was to servo-based DC coupling.

I originally rejected the idea but a servo can do a lot to simplify the audio path, at the expense of DC control complexity.

I got that current source circuit from Camille before Valve existed. I loved that guy. He really schooled me on grounding and noise, even beefed up a Tek 7000 scope so I could get rid of gremlins in some radio gear I was working on. That's how I learned that almost all 417A preamps, a popular design of the 90s, are likely putting out a signal on ~150mHz! I had to go to extraordinary lengths to get mine stable. He was a great teacher.

That experience and education is why I worry about D3as going wild. Not all builders have the test gear to know for sure.

I used the Bottlehead C4S a number of times but I was too "vintage crazed" back in the day, wouldn't use anything thought up after 1948. Now I am a lot more open-minded and even have a bag of Ixys current sources on my desk right now! Alas, I am too lazy to crank out builds and compare stuff like I used to.

Putting a new spin on well-worn circuits as Shindo does, and Kondo did in the M7, using tubes in odd operating conditions, is a Japanese artform that appeals to my vintage side. I suppose it can sound good too.

The linked article talks about the Shindo circuit sounding almost too nice. Erring in that direction might not be such a bad thing.

 

[paul_b]

Camille was a very valuable resource, I came along after his death, but PJ and Doc B have been meticulous about keeping his notes around. You are certainly right to worry about what a D3A can do without some efforts applied to tame potential oscillations. We end up using a lot of MLCCs, shielding, and stoppers to keep things happy.

The Tape Preamp has constant current sources between B+ and plate but plate loading resistors from plate to ground (I believe Bench was doing this back in the 90's). There is a servo that permits the direct coupling of the two stages.

How have you found the IXYS parts? My last experience with them left me thinking that you'd want to have a bag of them around to periodically replace them.

 

[J-ROB]

How have you found the IXYS parts?

I have used the IXYS parts a few times myself but jc and Dr. Bae use them by the boatload. They are illegal to sell in/to Korea for some (contractual?) reason, but I smuggled in a couple hundred 10M45s There have not been any failures in the field as far as I know, but I'll ask if I remember to do so.

Back in the old days we used 2 port "preset" Motorola CCS the size of 1N4004s. Those things were expensive, limited in available values, and died like crazy. Discrete jobs like the Camille C4S are definitely a step up from those firecrackers.

I was talking to somebody at ETF who reported big sonic shifts between the different models of INYX CCS. There might be more to it than is obvious from the spec sheet.

 

[mhardy6647]

but I smuggled in a couple hundred 10M45s

umm...

 

[prime minister]

Just to make sure this doesn't go off the rails, it's time for everyone to get back to work.

So, from my understanding, we have two chassis. We are using D3A tubes. Now, what else do we need?

 

[paul_b]

The D3A is a good choice for the first stage, but the tougher choices come with the second (and potentially third) stages.

 

[mhardy6647]

speaking of vacuum tube phono preamps and our friends JE and JROB and Sound Practices...
see this one posted at the Tribute to Sound Practices "group" at Flikr.
https://flic.kr/p/L8DSW3

 

[TubeHiFiNut]

The D3A is a good choice for the first stage, but the tougher choices come with the second (and potentially third) stages.

What candidates should be considered?

 

[TubeHiFiNut]

speaking of vacuum tube phono preamps and our friends JE and JROB and Sound Practices...
see this one posted at the Tribute to Sound Practices "group" at Flikr.
https://flic.kr/p/L8DSW3

Looks interesting.

Please elaborate.

 

[paul_b]

What candidates should be considered?

This totally depends on your preamp!

 

[prime minister]

This totally depends on your preamp!

How so?

 

[mhardy6647]

Looks interesting.

Please elaborate.

I know nothing othere than what's on that page, sorry

 

[TubeHiFiNut]

I am certainly not as technically adept at circuit design as everyone else on this thread so please pardon my questions.

The consensus seems to be that a reasonable path forward would be to follow the D3a/5687 approach as offered by @Salectric ?

With good tube availability, the main challenges revolve around layout and implementation to prevent the D3a from oscillating?

 

[paul_b]

How so?

If you have an active preamp with a 100K input impedance and moderate gain, you can get away with something like the 5687 proposed. If you have a 10K passive attenuator and a solid state power amp, things no longer look so great.

 

[Salectric]

If you have a 10K passive attenuator and a solid state power amp, things no longer look so great.

To drive a 10K load, I would stick with the 5687 and run its output off the CCS in mu-follower mode, which will reduce the output impedance, and increase the size of the output coupling cap to maybe 5uf. That should work fine.

 

[paul_b]

To drive a 10K load, I would stick with the 5687 and run its output off the CCS in mu-follower mode, which will reduce the output impedance, and increase the size of the output coupling cap to maybe 5uf. That should work fine.

And I'd add a 3rd tube stage and a small output transformer (or potentially a cathode follower if space is a consideration).

This is where things kind of diverge a little bit, and subjective performance considerations will start to be very system dependent.

 

[Dan ʕ•ᴥ•ʔ]

It would be nice if the final stage could provide a true line level output voltage of 1.5 to 2.2 volts.
I keep running into phono pre's that don't come anywhere close to that....but instead have outputs of 300-500mV. :frown:

 

[paul_b]

It would be nice if the final stage could provide a true line level output voltage of 1.5 to 2.2 volts.
I keep running into phono pre's that don't come anywhere close to that....but instead have outputs of 300-500mV. :frown:

A lot of them will do 1.5-2V with a 5mV cartridge. The problem comes from those middle of the road HOMC cartridges that put out significantly less than 5mV and significantly more than 0.5mV.

 

[Salectric]

There is much confusion about phono gain requirements which I believe stems from such great differences in gain among other components. Linestages are the most extreme example. Some linestages have gain of 25db or more; and at the other end some use passive attenuators such as an AVC with zero gain. Power amps also come with different sensitivities—-some reach full output with as low as 0.3 volts and others need 2.0 volts or more.

The phono preamp may be used with a low output moving coil (.2mv) or a typical moving magnet (5.0mv). In my opinion, since I like LOMC cartridges feeding stepup transformers, I think it makes sense to design the active phono preamp with enough gain for a moving magnet cartridge (35-42 dB gain) and then rely on the SUT to raise the level of the LOMC cartridge to that of a MM. Using that logic, a 1:10 SUT which provides 20 dB of gain gives a total gain of 55-62 dB, and that total gain has worked well with all the MC cartridges I have tried. With some cartridges, the level may still be a bit lower than the typical CD player but it should be close.

What I don’t understand are active phono preamps with gain of 54 dB or thereabouts (ARC for example). In my opinion that is too low for a LOMC going directly in but too high for a MM or a LOMC with a stepup.

 

[paul_b]

54 dB is indeed a little high for a 5mv catridge, but more useful with HOMC carts.

I also prefer LOMCs with step-up transformers. I still have 50db of gain available on my moving magnet phono preamp, but not a ton of linestage or power amp gain.