So you are saying within certain range all a DC offset on the output is going to do is add 2H?
I did not mean to generalize just pointing to the only known puplished measurements taken on DC nulled passive I/V and publicly shared, not worth getting defensive about it, for it is only discussion.
My personal experience is with passive I/V using a 6n6pi which is a tube known for 2H and minimal higher harmonics which gave a 2H at 0dbs -70db and that level of 2H gives a loss of definition when followed by a SET amp, followed by a good SS amp it was not an issue but in combo with the 2h from the SET it was too much. Again just sharing experience, no one is trying to generalize.
The problem as you know is distortion from a DAC is attenuated and amplified, sometimes twice, so distortion not considered an issue with an amp can be with a DAC.
Agree about the ECC88, not a great choice. And I also think his final result with distorion below -90db by "balancing" the load with a heavy loading resistor on the output, will probably give a different (worse) distortion depending on frequency. I think it is a mistake he made, I did the same with an SRPP. Their performance gets better with a higher load up to a point but it is frequency dependant and probably not a great choice for a line level output.
I also tried the CCDA with a CCS on the anode of the first tube and a ccs on the cathode of the follower and it didn't help much.
A differential set-up with a tube push-pull with a little shielded OPT nickle transformer is said to perform nicely, as does the Lazlo stage, but as interesting as these aproaches are they complexity of NOS differential is too much at this time for me.
So thi I am definately leaning toward the D3a passive I/V common cathode, But before sinking the money in the tube components it is tempting to revisit one of the newer opamps if there is proof that there is none of this skyrocketing input impedance with frequency. There are fairly old published studies with the 627 IV in conjuction with the tda1541 (on the other site) that look promising.
Thanks again.
Last edited:
Just thinking out loud here - pure 2H generation is a waveform asymmetry. So say in circuit A (a DAC) the positive side of the waveform is compressed more than the negative. So then if circuit B (an amp) is placed after A you have two circuits which generate asymmetry. Can't you reduce the composite effect by arranging them so that each compresses the opposite half of the waveform? If both happen to compress positive excursions more, then you'd need to invert circuit A's output before feeding it to circuit B.
This is similar to how balanced circuits correct 2H.
Hi,
No.
I am pointing out that the ECC88 has a significant amount of 2nd HD and what is shown may be the result of adding 2nd HD of the correct amount and polarity to the output of the DAC, especially given that extra current source is a relatively non-linear device.
We simply cannot extrapolate from this one set of examples without testing more (e.g. using resistors to null the offset etc.).
I do not think you identified the problem correctly, I accept the part about the 6N6 sounding in a certain way in a certain certain circuit (certainly I would not use it), however I think you are misidentifying the cause.
Have fun trying more things.
Ciao T
No.
I am pointing out that the ECC88 has a significant amount of 2nd HD and what is shown may be the result of adding 2nd HD of the correct amount and polarity to the output of the DAC, especially given that extra current source is a relatively non-linear device.
We simply cannot extrapolate from this one set of examples without testing more (e.g. using resistors to null the offset etc.).
I do not think you identified the problem correctly, I accept the part about the 6N6 sounding in a certain way in a certain certain circuit (certainly I would not use it), however I think you are misidentifying the cause.
Have fun trying more things.
Ciao T
Once again, datasheets and sound quality as being perceived by the ears may be two completely different things.
Did you looked at the the THA4032 specs? It's THD stays below -90dB up to 1MHz...
Thanks for trying to help but I don't see this has any bearing on this discussion to the discussion at all, Do you know what an opamp I/V input impedance vs rf curve is or its revalance to pre sigma-delta switcher "DAC's?
Hi,
Yes, suitably scaled.
I have abandoned this style in my commercial designs, but there I have extra complexity to exploit and other requirements, for DIY I'd stick to LCR EQ...
Ciao T
Yes, suitably scaled.
I have abandoned this style in my commercial designs, but there I have extra complexity to exploit and other requirements, for DIY I'd stick to LCR EQ...
Ciao T
It's lower than the resistors that are needed in order to raise the voltage level enough to get some decent SNR.
Don't like the output dynamic impedance of naked OpAmps, add in the loop a discrete stage with MOS-FET. Would be less cucumbersome than any tube concoction.
HelloThorsten,
I have been following this thread with interest as i have read about the lifelike and musical sound of NOS DAC,s and have also read rave reviews of your AMR CD77 based on TDA1541 DAC,Which i cannot afford, I am asking you a commercial question,as being from a country where i will have to import every component from abroad and being not a DIYer myself, I want to buy a NOS DAC with tube outputs,say if i buy a DAc from source like EC designs and the rest of the other components from other sources to complete the DAC,can you recommend how and where i can source and build parts for such a DAC.
Thanks
Hi,
I do like to help, however I really cannot help with this.
There are a number of DAC's and CD-Players that use the TDA1541A and India has many DIY'ers and capable technicians.
I suggest you try to find a suitable "donor" Machine (may be even in india) and then have someone talented and capable do the modification work.
Ciao T
I do like to help, however I really cannot help with this.
There are a number of DAC's and CD-Players that use the TDA1541A and India has many DIY'ers and capable technicians.
I suggest you try to find a suitable "donor" Machine (may be even in india) and then have someone talented and capable do the modification work.
Ciao T
Thanks thorsten,I have a friend who is a master technician,but difficulty is in sourcing different components and he does not use net,so that otherwise i may have asked him to join this forum to ask for different inputs from forum members to DIY a DAC,now he is repairing my CD player which since last 6 months is lying with him for repirs,as i have to source components from the manufacturer in US ,first i imported lens assembly ,then full drive and when i bought the player home after these fitments sometime back,i found one channel not working ,now i am waiting for the DAC from manufacturer
mekr,
Suggest you look for 2nd hand older machines - I do my "playing about" with the old Rotel RCD855 and 955 units as they are cheap and a good simple build with enough room for mods and less SMD components like many of the same age Marantz, Phillips, Grundig machines (that are still good source of the player mech (CDM4/19) common to nearly all) - Arcams also good to upgrade.
If you do get one - 2 suggestions ....
When shipping, insist on having the red plastic transport "catches" fitted (underneath) or have the lid removed and a foam cushion to hold down the mechanism - the top part (on spring suspension) of the mech (CDM4/19) easily falls out of its spring mounts when turned on it's side or upside down and it's a PIA to refit.
Also, suggest replacing the standard 2 wire power cord with a 3 wire power cable that does include a chassis earth and add a "100 ohm floating resistor" between 0volt point on the pcb and the chassis earth - back then, they just left off the chassis earth for various reasons - very common, but quite unsafe!
... 2 cents
Suggest you look for 2nd hand older machines - I do my "playing about" with the old Rotel RCD855 and 955 units as they are cheap and a good simple build with enough room for mods and less SMD components like many of the same age Marantz, Phillips, Grundig machines (that are still good source of the player mech (CDM4/19) common to nearly all) - Arcams also good to upgrade.
If you do get one - 2 suggestions ....
When shipping, insist on having the red plastic transport "catches" fitted (underneath) or have the lid removed and a foam cushion to hold down the mechanism - the top part (on spring suspension) of the mech (CDM4/19) easily falls out of its spring mounts when turned on it's side or upside down and it's a PIA to refit.
Also, suggest replacing the standard 2 wire power cord with a 3 wire power cable that does include a chassis earth and add a "100 ohm floating resistor" between 0volt point on the pcb and the chassis earth - back then, they just left off the chassis earth for various reasons - very common, but quite unsafe!
... 2 cents
Is this mod designed to make the unit more safe? There's nothing particularly unsafe about the original configuration - its typical for Class II (double insulated) appliances. Having a ground via 100R might turn out to be more unsafe - not to mention the 3 wire mains cable might be only Class I rated.
Sorry abraxilito,
Dunno about the double insulated rated safety standards gear, but if you have 240 volt and just the usual 2 wire mains supply wire into a metal cased cd player of 15 - 20 years ago, the metal case will rarely have a chassis earth connection at all, but relies on the signal's sheild wire of the interrconnect for safety earth - with the insulated plastic cases like some of the Marantz, etc, players, this won't apply provided the lid is kept on and so on ...
However, if you are into DIY and maybe using older metal chassis players (without lid, bare mains fuse holders, bare wires at the switch, etc, etc) and it doesn't have a chassis earth connected via a 3 wire cable, I would stongly suggest you add one (or an IEC socket) and tie the metal chassis to the earth pin for safety and then isolate the phono plugs from the metal case - and maybe seperate/float the original central 0 volt point on the pcb from the new chassis earth via a resistor, thermistor, etc
I see a ridiculously large percentage of diy gear that's dangerous, mainly thru ignorance, and a strange belief that "they won't get bitten" or that the earth leakage safety switches fitted to the house switch board will save them from a lethal jolt.
Dunno about the double insulated rated safety standards gear, but if you have 240 volt and just the usual 2 wire mains supply wire into a metal cased cd player of 15 - 20 years ago, the metal case will rarely have a chassis earth connection at all, but relies on the signal's sheild wire of the interrconnect for safety earth - with the insulated plastic cases like some of the Marantz, etc, players, this won't apply provided the lid is kept on and so on ...
However, if you are into DIY and maybe using older metal chassis players (without lid, bare mains fuse holders, bare wires at the switch, etc, etc) and it doesn't have a chassis earth connected via a 3 wire cable, I would stongly suggest you add one (or an IEC socket) and tie the metal chassis to the earth pin for safety and then isolate the phono plugs from the metal case - and maybe seperate/float the original central 0 volt point on the pcb from the new chassis earth via a resistor, thermistor, etc
I see a ridiculously large percentage of diy gear that's dangerous, mainly thru ignorance, and a strange belief that "they won't get bitten" or that the earth leakage safety switches fitted to the house switch board will save them from a lethal jolt.
hmmmmm.... metal case should be DC coupled to PCB ground at single point, via low impedance connection (large contact area & low contact resistance between PCB ground and metal case) - best place is PCB ground area with lowest noise, i.e where the RCA's are located - at inputs / outputs.
The rest of the PCB ground should be AC coupled to metal case at as many points as possible, using parallel combination of 0.1 and 0.01uF capacitors with very short leads - surface mount capacitors are perfect -> this requires some planning -> proper PCB layout at design phase.
Mains earth should not be connected to metal case -> this applies to all HiFi components in a system.
Mains earth is VERY noisy...... AND it contains induced 50Hz artifacts caused by close proximity and long parallel run of 240V phase lead, and earth lead -> which includes power cables as well as the house wiring.
Boky
The rest of the PCB ground should be AC coupled to metal case at as many points as possible, using parallel combination of 0.1 and 0.01uF capacitors with very short leads - surface mount capacitors are perfect -> this requires some planning -> proper PCB layout at design phase.
Mains earth should not be connected to metal case -> this applies to all HiFi components in a system.
Mains earth is VERY noisy...... AND it contains induced 50Hz artifacts caused by close proximity and long parallel run of 240V phase lead, and earth lead -> which includes power cables as well as the house wiring.
Boky
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.