My new tube DAC output

[Algar_emi]

Here my new tube voltage gain stage for my new DIY DAC. It is using a HV 6X4 Tube rectifier and a double triode Svotek 6922. The 6922 is using DC filement voltage using schottky rectifer and LM317T regulator. The HV is filtered using a CLC filter with two 47uF cap. an a Hammond 9H, 40ma choke.

 

[Algar_emi]

Coupling caps

And I just forgot. I'm also using in the signal path all copper Riken resistors and the very good Jensen copper foil and silver lead.

I will listen to this beauty soon...

Comments are welcome. Any suggestion on my tube choice will also be appreciated.

Don't be too hard on me, it is my first tube circuit. Better late than never.

Bye...

 

[Algar_emi]

And a few measurements

Since the DAC is not completed, here a few measurements to test everything:

Bandwidth (0.767V, 1Khz reference): 20Hz-300Khz in 50 Kohms

Voltage Gain (in 50Kohms)
0.5Vin = 22.5
31.6mVin (Vout=0dB) = 24.2

Distorsion (in 50Kohms, Vout 0.767V)
100Hz: 0.5%
1Khz: 1.2%
10Khz: 1.4%
20Khz: 1.5%

Are these values of distorsion typical or an other tube brand can give me better values?

 

[Peter Daniel]

Re: Coupling caps

And I just forgot. I'm also using in the signal path all copper Riken resistors and the very good Jensen copper foil and silver lead.

You know that those caps have outer foil marked and I found that when outer foil is directed towards the load, it sounds better.

 

[Algar_emi]

Cap direction

I was wondering I to connect them. Thanks Peter for the tip...

SB

 

[Peter Daniel]

The outer foil is marked with a line.

 

[mig-ru]

Algar_emi,

That looks awesome! I think I might just try tackling a DIY tube DAC. Where did you get the schematic for it, would be interesting to look at.

 

[JOE DIRT®]

I dont know the exact Engineering on those caps but its almost like they are polarised like an electrolytic cap and what peter has said seems to be true....and to comment on your distortion levels....isnt that what your looking for in a tube design?????....that does make the nuiance warm and proprietary with a tube design.....Live music is never perfect and audiofiles are ultimatley trying to re create a live performance not a mixed studio version


Cheers!!The DIRT®

 

[colwaiting]

tube dac

Excellent. All done on a R.S. project board!
I' d like to know about that ceramic 6dj8 tube socket.
Also evrything you know about the 1/v conversion at the output of the dac. (impedances, levels) this area still confuses me a bit.>>Doug

 

[Algar_emi]

Tube socket

I got them from www.partsconnexion.com in Canada. They have gold plated pins also.

 

[Algar_emi]

Schematics

See

http://www.diyaudio.com/forums/showthread.php?s=&threadid=12464

for details and other photos.

 

[Gunders]

Hmm.. i think i know those sockets(maybe). I have a couple of them myself. Both the ceramic and the type used for the 6DJ8.
I bought mine at www.demostenes.no, the price was about 4 Euro.
The socket used for the 6X4 is a good ceramic socket but the socket used for the 6DJ8 tube is in fact not ceramic, it's some kind of plastic material. They are not very good
for tubes producing much heat(small output tubes). I used some of those sockets in an EL84 amp and it did actually melt a little bit because of the heat

 

[Algar_emi]

Found my problem with 6922

I found the problem with the high distorsion. It was the tube. Hey guys, I begin to understand your love of tubes. It is so easy to unplug the bad one and try an other ones!

Anyway, all my others 6922 tubes test like this (pretty consistent also, incredible for a transistor guy like me :

Distorsion (small signal 0dB, 0.767V in 50Kohms, 1 kHz):
0.3 to 0.5%. I will use the 0.3%

Bandwidth: 200 Khz
Noise: <0.5mv

Tubes rules!

 

[Algar_emi]

Soft Clipping and Linearity

I veryfied indeed that the tube go softly into clipping and can drive a lot (in high impedance at least):

Check also the good linearity indicated by the slight variation in the gain.

Distorsion (indicating clipping coming into effect):

Vin(V) Vout(V) Dist(%) Gain
0.03 0.767 0.3 25.6
0.22 5 0.72 22.7
0.43 10 1.05 23.5
0.87 20 1.9 22.9
1.38 33.8 3.0 24.5 (no visible clipping)
2.00 44.6 3.6 22.3 (begin to see it)
2.78 57.7 5.5 20.7 (clipping, see gain is dropping)

This picture is showing the clipping just beginning at 57.7V

 

[Algar_emi]

Very good response too

Here the three waveforms, small signal 1V but at 20 Khz. Very nice for a small preamp...

 

[fdegrove]

TRY THAT WITH SAND.

Hi,

Congrats on the project. This afternoon I went through your posts in the digital section and I feel you did a great job.

Now, if you could tell as how it sounds compared to whatever you feel it can beat?

Thks,

 

[Algar_emi]

Microphony and Vibration

The svotek 6922 is not that sensitive to shock (microphony). Here a small suspension that I want to implement for the tube PCB. They are small rubber post that I found in a surplus store.

I can see small jump in the noise level when I knock on the table where the PCB rest. Vibration effect dissapear when I put the PCB on the rubber post, Great.

 

[Algar_emi]

I'm almost done before I can hear it.

Yes me too I want to hear it. I have just to connect all three PCB and try it. However I don't want to rush it. It is late here in Quebec and everytime I tried to complete a project late at night I did some stupid mistake and burn something. It happened with my ALeph30. So I do it slowly but correctly.

I planned to install everything on a piece of wood, use nice interconnect wires and try it with my Copland 277 used as a transport. I also ordered the XO low jitter clock for my two CD players including the Copland. I still have to installed this too..

Bye.

 

[stokessd]

What's the quiescent current through the tube stage? From personal experience, I've found that less than about 4 mA through a 6DJ8 tube can crank up the distortion figures and lead to less than ideal sound. If you crank your bias up, your distortion figures may drop even more.

I use a pair of 6DJ8 type tubes in an SRPP configuration with cathode degeneration in my tube DAC. I like it and lots of others seem to enjoy it as well. The SRPP stage also has a low output impedance which can be a big deal depending on your cable lengths and cable choices.

Sheldon

 

[Gunders]

Have you considered to remove the plastic from the capacitors and oil the PCB and the components with linseed oil?