Phono stage design considerations part 1: choosing 1st stage tube

[Merlinb]

ms, a cascoded DN2540 CCS has a current noise of about 14pA/rt Hz.

How did you calculate that? The 120R resistor alone will contribute about 12pA/rtHz!

 

[PMA]

What is the full scale zero in dBV? Looks like a logarithmically averaged chart, correct?

0dB = 2V. Exponential averaging.

 

[SY]

How did you calculate that? The 120R resistor alone will contribute about 12pA/rtHz!

Measured, not calculated. Walt Jung got about the same number in his published AX article using nicer equipment (AP2).

 

[Salas]

0dB = 2V. Exponential averaging.

OK. At what gain is the DUT set? X100? X1000?

 

[PMA]

+40dB/1kHz.

 

[Salas]

Thanks. Was just trying to roughly approximate the equivalent input noise in SQrtHz by that chart's noise floor & extra info. So it roughly seems to me about 6.5nVrtHZ with negligible 1/f hump. Some BJT input op amp.

 

[PMA]

6.5nV/rtHz might be quite right, input opamp is OPA627.

 

[Salas]

Very nice 1/f for a FET input opamp. Should be dominated by the source also. Does not look like 0Rs and FET. That is why I though BJT input instrumentation type. Nice result, congrats.

 

[PMA]

Thank you

 

[Joshua_G]

After looking at your posts I thought you might get something out of these pages. First is from the Sound Practices reading club list and includes the article on the Siren Song, an old design but one that others have used as a starting place for one of their own designs, and the Second is J.C.'s Blog page on a circuit I thought might interest you.

Hi,
Thank you.
I'll read the linked articles later on.

 

[Joshua_G]

Hi Thomas,
Thank you.

It takes much more to achieve good sound than just selecting the right tube.

Indeed, however choosing the 1^st stage tube may be an appropriate start.

Since you already ruled out so many suggestions, maybe you should give some more specifics like range of current, rp and mu you are looking for.

I don't have a list of preferred like range of current, rp and mu.
I've listed the parameters I'm looking for in choosing the 1^st stage tube:
1. Very low noise and microphonics.
2. Reasonably low miller capacitance.
3. Good linearity is preferred, especially low 3rd and higher odd harmonics distortion levels.
4. Reasonably low plate resistance.
5. At least 5mA operating plate current.
6. Availability and reasonable cost.
7. Long life.

The above list is in order of priorities, that is, very low noise and microphonics is the first priority, long life is the last priority.

I would have suggested the 6SF5 for a restively loaded first stage. I have been using it in my lower cost phono preamps and it works beautifully despite it's low plate current in the region of 1mA. It is available cheaply too:

VinylSavor: Tube of the Month : The 6SF5

I don't see how it meets my requirements.

I also used the D3a successfully but only in transformer coupled and in LCR EQed phono stages.

CCS is ruled out for me, as well as cathode follower and GNFB.
Loading/coupling can be either RC or transformer.

However I don't see how choosing the type of RIAA equalization (passive or LCR) affects the choice of the 1^st stage tube.

What plate current did you operate the D3a in and what transformer did you use?
Also, how do D3a is compared to other contenders in the above application?

Another cheap tube is the 6AM4:

VinylSavor: Tube of the Month : The 6AM4

How does it compare to other tubes concerning noise and microphonics.

If you want best sound you should also think about circuit and EQ alternatives as they have a huge impact. In my experience 600 Ohm LCR EQ is giving the best sound.

Thank you.
Choosing the type of RIAA equalization was going to be my second consideration.
Again, how choosing the type of RIAA equalization (passive or LCR) affects the choice of the 1^st stage tube?

 

[Joshua_G]

Hi,
Thank you.

You can start with a ec8010 as Morgan Jones dos.

How does it compare to other contenders, regarding the list of my priorities?

What transformer you have?

None yet, I'm considering S&B TX 103.

 

[Joshua_G]

Hi,
Thank you.

Try decent (I use Sylvania 3 mica) 5751 instead of 12AX7. Slightly less gain, but much better sound.

I did many tubes rolling in my previous phono stage, including various brands of 12AX7 and 5751. As I wrote above, I'm looking for something better.

My phono (used every day): 5751 (green LED bias, 1mA), passive RIAA, 5751 (green LED bias, 1mA), ECC82 (3.5mA) cathode follower.

CCS:
As you can see, this is CCS loaded phono (all tubes!), and almost invisible harmonics.

http://www.diyaudio.com/forums/tubes-valves/246392-c3g-triode-strapped-linearity-question-9.html #85

When the measurements of D3a and C3g were so good, why do you use 5751 based phono stage?

Anyhow, once again, 1 mA plate current tubes like 12AX7/5751 are ruled out for me.
Also, CCS loading, cathode follower and GNFB are ruled out for me.
As I said above, it isn't business, it's only personal.

 

[Vinylsavor]

Hi!

However I don't see how choosing the type of RIAA equalization (passive or LCR) affects the choice of the 1^st stage tube.

To drive a 600 Ohm LCR RIAA puts very different requirements on the 1st tube compared to a RC RIAA with medium or high impedance

What plate current did you operate the D3a in and what transformer did you use?
Also, how do D3a is compared to other contenders in the above application?

18ma plate current. It is very high on the list of my favourite tubes for this application, only topped by EC8020 or WE437. I prefer it over E810F, E55L.
I used it with Lundahl LL1692A

How does it compare to other tubes concerning noise and microphonics.

I prefer the tubes mentioned above over the 6AM4. I just listed it since it would full fill the requirements for availability and low cost

Best regards

Thomas

 

[Joshua_G]

Hi,
Thank you.

Again, you're letting your preconceptions prevent you from arriving at good solutions.

Possibly it is so, possibly it isn't so.

Perhaps you could take a look at the distortion spectra of my preamp, where everything is loaded by CCS and point out the 3rd, 5th, or higher harmonics. And there's dreaded cathode followers.

Harmonic distortion, including the spread of the various harmonics, in itself, is but one parameter I'm looking for – not the only parameter.

As I wrote above, I'm looking also for comparative listening evaluations, besides measurements data.

 

[Salas]

Thomas, do you use the EC8020 for direct input MC?

 

[Joshua_G]

Hi,
Thank you.

I would also like to provide you with one.

What amp is it?

And my suggestion would be - get away from tubes as far as possible

Thank you for your advise.
Yet, I intend to build a tube based phono stage.

 

[Joshua_G]

Very nice 1/f for a FET input opamp. Should be dominated by the source also. Does not look like 0Rs and FET. That is why I though BJT input instrumentation type. Nice result, congrats.

As I stated above few times, noise is but one parameter I'm looking for – it isn't the only parameter.

 

[Vinylsavor]

Hi!

Thomas, do you use the EC8020 for direct input MC?

I generally use MC step up transformers.

I did try a D3a phono stage with a MC direct in once and it was well usable.
But I prefer to have the MC step up. No point in trying to avoid the transformer at the inout when the entire phono stage is transformer coupled

Best regards

Thomas

 

[Salas]

As I stated above few times, noise is but one parameter I'm looking for – it isn't the only parameter.

Noise in high gain MC like 60dB-63dB (X1000-1400) is THE consideration priority. Not just a parameter. If you can't fetch the signal across strongly over the noise floor then its so contaminated that you lose much original information. That alone renders the tonality of what portion remains intact a purely cosmetic exercise.