5687 fully balanced line stage

[cedus]

Here is something I drew yesterday.
http://ccedus.googlepages.com/full_balanced_5687_preamp.png
This is my first line stage preamp design and thus I need the help of the ones "who know" to tell me if there is a problem with my design before I start designing the PCB.

My constraints for the project were the following.

- Use the 5687 because I have lot of them
- a fully balanced line stage which could be also used with SE sources.
- use CCS (because I never used them on any of my previous tube amp or preamp).

For info the PS is tube shunt regulated. The cascode CCS with LM317 and DN3545 comes from Walt Jung great article about CCS.

Part 1 :
http://waltjung.org/PDFs/Sources_101_P2.pdf
Part 2 :
http://waltjung.org/PDFs/Sources_101_P2.pdf
Addendum:
http://www.waltjung.org/PDFs/AX_Letters_0907.pdf


Could someone help me to calculate the gain of the whole thing. I don't know how to handle the 500ohms degeneration resistors in my gain calculation as well as how a ccs can affect this gain.

Thanks in advance for your help

 

[Cauhtemoc]

Could someone help me to calculate the gain of the whole thing. I don't know how to handle the 500ohms degeneration resistors in my gain calculation as well as how a ccs can affect this gain.

Thanks in advance for your help

Spice can be a great help.

 

[cedus]

...especially when everything is drawn with gEDA

But what I was waiting for are the formulas so I can calculate it by myself next time.

 

[cedus]

Here is a new updated schematic.
http://ccedus.googlepages.com/preampv2.png

 

[dsavitsk]

...especially when everything is drawn with gEDA

But what I was waiting for are the formulas so I can calculate it by myself next time.

Try here http://www.aikenamps.com/CommonCathode.htm

 

[jmbee]

hello,
I think you can use common cathode formulas then half the gain for a differential stage. For 5687, considering Ri= 3,5 K and mu=17, that gives a gain near 14dB
(5) for assymetrical in /out.

 

[jane]

This spreadsheet will calculate common cathode gain and output resistance. --> http://www.veiset.net/tech/rk.xls

Jan E.

 

[46davis]

5687

I've been working with 5687s for a while and I think what you'll run into is a noticeable amount of distortion from the first (plate follower) stage. Even if the tube is run in a relatively straight part of the transfer curve, drawing at least 10 ma, it generates a measurable amount of second order harmonic distortion. I tried about 5% or about 4dB of local negative feedback to get it under control. Also, this tube is a real current hog making PS design expensive and it runs hotter'n hell. Other than that, great tube.

 

[sajti]

I plan to build balanced preamplifier. I have 2 topology, and I'm not able to choose, which would be better:

The first:

An externally hosted image should be here but it was not working when we last tested it.

The second:

An externally hosted image should be here but it was not working when we last tested it.

Sajti

 

[jmbee]

Hello Sajti,
In my humble opinion , the second topology has no decisive advantage
vs the first; it makes me think to a Audioresearch LS first 2 stages:
http://www.arcdb.ws/LS22/ARC_LS22_schematic1.gif
Unles you need a lot of gain, you might increase R1 and R4 for better linearity.
cordialement
jmb

 

[sajti]

The advantages of the second solution:
- higher load impedance for the cathode follower, makes less distortion
- no DC current, and less AC current flow into the ground point

But maybe I'm wrong.

Sajti

 

[jmbee]

you are not wrong but distorsion mainly comes from the first stage with medium high
impédance loads.
If each stage mainly generates second harmonic distorsion, say first satge DH1, second stages DH2 , total distorsion is ( DH1 - DH2) because phase turns.

crd
jmb

 

[hollow_man]

Hi Cedus,
you might want to stay clear form tail current sources, as me experiments have shown, see
http://www.diyaudio.com/forums/showthread.php?s=&threadid=135356
I think Satji has got the idea.

 

[sajti]

I want to avoid to use any ccs. The psu will be unregulated.

Sajti

 

[hollow_man]

...and it will sound better!