BTW anything else need changing for using SE input with original 2210/1678 active filter ? 390R/470R inverted input to ground , 200R inverted output of dac to ground ? Coupling cap on output if not using last stage of original active filter
You should be able to ground the inverting input through a resistor same as with the OPA1632 board. The resistor, and 2.2nf NPO/C0G cap in parallel if desired, would go from the inverting input to ground in order to approximate what the dac output would look like to the input of the output stage board. Whether or not to load the inverting output of the dac itself is IMHO optional.
Another thing you could try is slowing down the differential summing opamp on the output stage board, as I did. At least I thought is sounded a little better that way if used in balanced output mode. Don't know if any difference for SE mode.
Also, it should be possible to get rid of the DC blocking cap by injecting an offset into the output stage boards at some point. For the OPA1632 board, that might be at the COM pin of the fully differential amplifier IC.
Another thing you could try is slowing down the differential summing opamp on the output stage board, as I did. At least I thought is sounded a little better that way if used in balanced output mode. Don't know if any difference for SE mode.
Also, it should be possible to get rid of the DC blocking cap by injecting an offset into the output stage boards at some point. For the OPA1632 board, that might be at the COM pin of the fully differential amplifier IC.
Thanks Mark ,
I need to try comparing with and without the dacs inverting output loading resistor .
I remember that slowing down the differential summing op-amp mod you tried, I'll have a search back to find the details and give it a go .
It would be nice to get rid of the coupling cap
I need to try comparing with and without the dacs inverting output loading resistor .
I remember that slowing down the differential summing op-amp mod you tried, I'll have a search back to find the details and give it a go .
It would be nice to get rid of the coupling cap
That should be 8.2 nF, assuming the capacitors C15 and C42 on the DAC board still have a value of 8.2 nF.
That COM pin on OPA1632 is actually VOCM (output common-mode control voltage) which directly sets output DC so that cannot be used. One way to remove DC from output would be to connect the resistor on negative input to same DC level that is on positive input (i.e. DAC output) instead of ground.
Okay. Sometimes people use current sources for that so that the offset injection impedance is very high and doesn't much affect existing time constants.
Why not do as I did: put a capacitor from - dac output to ground? That way you get the correct dc value from the dac an no need for output capacitor
As I remember, Gene Pope used an Apogee ADC, but I don't know if it used dither. The Nagra D he used in this session was unmodified, and only accepted SPdif input. The mod I made was making it accept Sdif , a professional standard, with three wires Ldata, Rdata and Wclk in order to use a distributed clock for ADC and Nagra.
The distance from the orchestra is kind of a personal taste. Where do you want to sit in the concert hall? By the way I forgot to mention, that the recording was made with pretty low average level, so that could also explain the low level fuzziness (I haven't noticed that), because you can run out of bits that way?
No output cap, but since the cap is connected to the inverting input of the output stage there will be some current flowing through the cap, so it may have some influence on the sound. If the cap has enough capacitance though, it may not generate any harmonic distortion. However, it could still potentially generate some 1/f noise due to leakage current. It could also affect phase of low audio frequencies. IOW, maybe still best not to have the cap if possible.
Usually its the most expensive seats in the house (which often sell out first). Maybe some lucky person has a season ticket for those seats. They are usually around the 5th or 6th row, or maybe at the front of the first balcony. Somewhere around the center of those rows is usually best. However if the seats can't be used to place mics, it may be possible to hang mics from the ceiling. Takes long cables though and possible help from the stagehands.
The ADC may have had a switch to use dither or not. Some people might think adding dither is a source of undesirable noise, but if used properly it can prevent fuzziness (quantizing distortion) and extend bit-depth an extra few bits below the 16th bit. The catch is the lower bits are then below the noise floor, but they can still be audible to some extent.
Last edited:
Sounded really nice - why aren't all recording like this? Would you know how the mics was placed? I'll check for the fuzz...
//
I prefer the balcony instead of the front seats on the floor, and therefore I like the way the Schnittke track sounds. Far too many recordings are too close to the orchestra, so you almost think: this is how the conductor hears it
I am not sure I agree here Markw4.
I think inducing the DC from the DAC at the -in is better than placing a DC blocking cap anywhere.
And if you have a big resistor between the cap and the - out from DAC and another big resistor from the cap to the - in of the filter, that cap is isolated AC wise from the circuit. You could even take the input form the + out of the dac that way. You only need a small value cap = very small current through the cap and no significant leakage current and no phase shift in the low freq.
I think its fine for people to experiment around with the output stage and see what they prefer.
Where I think we can agree is that in SE mode all of the music is there and there is a deeper sound stage. Whereas in balanced mode part of the low level musical details are lost. Moreover it has nothing whatsoever to do with liking distortion. All people have to do is listen to know that is the case.
I would just add that when we get around to trying better clocks with this dac, once again there may be effects not very well predicted by theory.
Where I think we can agree is that in SE mode all of the music is there and there is a deeper sound stage. Whereas in balanced mode part of the low level musical details are lost. Moreover it has nothing whatsoever to do with liking distortion. All people have to do is listen to know that is the case.
I would just add that when we get around to trying better clocks with this dac, once again there may be effects not very well predicted by theory.
How nice of you for them to have your blessing and permission to do so... ;-) and hey, they might come up with something revolutionary👍
//
It was just my way of saying I don't necessarily disagree with koldby. If people like his solution better then maybe it is better.
No. We can't agree on that. What can be stated is that there is potentially higher common mode noise and distortion in SE mode. That is a fact so no agreement is necessary.
I would only expect agreement from the vast majority of those who have actually listened to Marcel's RTZ dac in SE mode. That applies only to the board layout and BOM as designed by Marcel, usually along with power supply pre-regulation, ferrite-free and linear cap bypassing for clocking, etc. For other designs its likely apples to oranges.
Last edited:
Yes, it applies only to Marcel's layout and BOM. Not to my version which has superior performance.