Tubed CCDA I/V Amp for TDA1541(A)

regal

Member
2004-01-04 8:41 pm
MD
Here is the schematic, I'V resistor not shown. 6n6p tube wonderful sound measures very good.


compund-2.jpg





You have to "select" the 22mA resistor by using a 9V battery to trial the 10M4S cascode, see the 10M4S thread for details.
 
Last edited:
Hey regal,
As I told you before it is not realistic to use 6N6P as gainstage after a TDA1541.

When properly loaded by 20-30ohm you won´t get more than ca 700mV out.

So if one wants to go this route, it is still a tube like 5842 or 6S45 (with about double the mu of 6N6P) that is suitable.
 

regal

Member
2004-01-04 8:41 pm
MD
Hey regal,
As I told you before it is not realistic to use 6N6P as gainstage after a TDA1541.

When properly loaded by 20-30ohm you won´t get more than ca 700mV out.

So if one wants to go this route, it is still a tube like 5842 or 6S45 (with about double the mu of 6N6P) that is suitable.

5842 would drop right in the circuit with a change to the cathode bias thats the point.


You could also stack the TDA1541's so you could use a 30 ohm I/V.


Think for yourself, but I'm telling you that the cathode output of the CDA kills the sonics vs the single anode follower. I even tried brockies CDA with the CCS's.
 
Last edited:

revintage

Member
2007-05-30 4:12 pm
You could also stack the TDA1541's so you could use a 30 ohm I/V.

A single 1541 should be used with 30ohm.

5842 would drop right in the circuit with a change to the cathode bias thats the point.

So would the 6C45, the difference is higher mu and half Zout, you can look at as very close to a 437 or triode-strapped E810F. A matter taste but I would run the 5842/417 at 10mA.
 

regal

Member
2004-01-04 8:41 pm
MD
Hi regal, could you post a picture of thd+n for your dac ? thats greatly appreciated because I want to build one too :)


I will try to find them, I had it hooked up to a PCM63k with a 47 ohm I/V resistor, sounded better than the D1 stage and any other tube stage I tried. I do believe that Revintage may be correct about stacking the TDA1541's but there are two schools of thought on this issue: Minimize impedance (load) seen by the DAC and the other is minimize voltage (compliance) seen on the DAC's ouput. My theory is that if you are under .4V's or so you are just dealing with minimizing the load, so stacking the TDA1541s driving a single 30 ohm I/V would sound as good as a single TDA into a 30 ohm load. Again 2 schools of thought and only the original chip designers know the answer.
 
How large can the passive I/V resistor be without giving distortion?

You havent read my homepage. This is the latest findings I had to do myself, because of lacking and inconsistant information on the net regarding this issue.
How large can the passive I/V resistor be without giving distortion?

In short. One TDA1541(A) = 56 ohm . Stacked (two) = 27 ohm and so on. You also need a CCS to get rid of DC levels. Should be lower then +-10mV on TDA1541(A) output. The last one show up on THD figures at low levels, where it has a large impact.
 
Last edited:

revintage

Member
2007-05-30 4:12 pm
I have read your homepage before. As far as your page reads you are using 33ohm, think I also saw 27ohm somewhere, which is what I recommend. Using higher values increases distortion and impair sound quality(grain). Going lower will be stiffer. This is the conclusion the mainpart of listeners have drawn.

Going higher in resitance is only to get more output if the gain of the following stage is insufficient, not sound enhancing! Normally a mu of 40-50 is enough as we will then have an output in the ballpark of 1Vrms.

Actually there is a high gm(19mA/V) triode with mu=100 available. That gives us a Ri of 5k. It is the dirtcheap russian 6F12P. Can send you a pair if you want to experiment, just mail me. The pentode can be used as CF if one so wish.
 
It is about loading the 1541 optimal, and that is in the ballpark of 30ohm/DAC.

So if you want to go with two DACs you should go around 15ohm.

And again I say one DAC 56 ohm two Dac's 27-30 ohm.
That is about 80-85 mV over the I/V 0dB.
Thats about the same Pedja Rogic also discovered some time ago.

Thanks for the offer on testing tubes. Maybe I will sometime in the future :)
 

revintage

Member
2007-05-30 4:12 pm
OK Ove,
This shows you have two things two rethink, both I/V-resistorvalue and gainstage-configuration;).

About this Pedja, he can´t be the one who represents the absolute truth in cyberspace-hifi;)?

Why not try ca 30ohm and a single stage GC? It could even sound nice. I´m sure you´ve heard about KISS.
 
Keep it simple?

Why not try ca 30ohm and a single stage GC? It could even sound nice. I´m sure you´ve heard about KISS.

Yes it does sound nice with my CCDA as well. But then I have to use ECC83 to get descent output. The E88CC gives lower amplification but very little THD and uncolored sound. I'm not after sweet colored "tubesound" in particular but as clean and good sound as possible from tubes, to really compare a discrete robust tube circuit with good sounding modern OP amps.

Today the easiest way to get sound with the coloration of your taste, is to use a DSP in the DAC and forget about tubes and OP amps signature.

Using just a GC stage It's hard to get the low output impedance I want, so that I can choose whatever equipment I want as control amp or endstage after this circuit, without altering sound or THD figures.

Even though I have put a lot of work into this project, empirical tests of my own which I later discovered matched up with certain other peoples findings, I am not one of those that says OP amps are bad or even sounds bad.
In fact I now have difficulties to hear difference between my two setups. That's why I also built this A/B test piggybacked PCB.

So... in general tubes are over hyped with their magic glow and OP's have much to much unjustified reputation. At least the good OP's I have at hand, LME49710, AD797 and OPA627.
The only way to find out is to build, listen and measure for yourself. To try to get a good picture of things at this forum is nearly impossible. To follow someone you think has the knowledge and discover a year later that even though the person seems to have booth feets on the ground, maybe a company that sells hitech stuff and so on.... ,there are sometimes flaws and a lot of subjective "findings" everywhere.

So please use my ideas and read my thoughts, but most of all, try for yourself.

Lars I promise, I will use a simple anode follower on my controlamp. :) A 12B4 with CCS as anode load. I will have it hot biased for sure. A project maybe this autumn.


"Everything should be made as simple as possible, ... but not one bit simpler"
Albert Einstein
 
Last edited: