Hi all,
Behringer placed the low-pass filter right after the DACs and before the IC1A.
Now, if you don't need balance output and you think that the crazy amount of opamps, caps and resistors Behringer stuffed there is bad (which by the way I think too) you can take an unbalance output directly after C43, you can even replace the cheap electrolytic they placed there with a Black Gate N type.
If you need the Mute function (I don't) the unbalance output can be taken right after R10.
Something I'll do in the future is to replace IC1A with a high quality audio op-amp, BB OPA627BP biased in Class A first comes to mind.
Regards
Padel
Behringer placed the low-pass filter right after the DACs and before the IC1A.
Now, if you don't need balance output and you think that the crazy amount of opamps, caps and resistors Behringer stuffed there is bad (which by the way I think too) you can take an unbalance output directly after C43, you can even replace the cheap electrolytic they placed there with a Black Gate N type.
If you need the Mute function (I don't) the unbalance output can be taken right after R10.
Something I'll do in the future is to replace IC1A with a high quality audio op-amp, BB OPA627BP biased in Class A first comes to mind.
Regards
Padel
Most interesting, thanks everyone.
"The best solution might be a suitable 1:1 Transformer, wound to have a defined and well controlled HF Rolloff above 48KHz, which would give a 2nd order rolloff, DC blocking "
Aren't those xformers like $40?
Why not just use a series inductor? Doesn't the 6.8 cap give DC blocking?
"The best solution might be a suitable 1:1 Transformer, wound to have a defined and well controlled HF Rolloff above 48KHz, which would give a 2nd order rolloff, DC blocking "
Aren't those xformers like $40?
Why not just use a series inductor? Doesn't the 6.8 cap give DC blocking?
Padel said:
Funnily enough even after removing all the **** by cutting off circuitry between X13 connector and the XLRs, the Mute function is still working ! (transistors T1 to T6)
That proves if needed how un-necessary was that stuff for our use.
Did I say each on his own ? Yes I did
Thierry
.
Kevin,
I don't know about a garden variety.....but yes, I think Thorsten is exactly correct and all things considered a transformer is the best option and it will give you a balanced or unbalanced output, whatever you want. However, it is the most expensive. Darn.
Behringer has tied the VREFH pin to 5 volts and VREFL pin to ground. These are the recommended voltage to scale and "window" the four output pins to the proper levels. (This information is on page 4 of the data sheet.) I haven't measured these myself but this should position the AOUT pins at approximately 2.5 volts DC and thus the requirement for a DC blocking capacitor somewhere in the downstream circuitry. (C43-C48 in the output sections.)
I'm sorry I haven't been following this thread for the past few months, but I've been reviewing it a little bit today. Unfortunately there is quite a bit of mis-information in many of the posts.
Cheers,
Davey.
I don't know about a garden variety.....but yes, I think Thorsten is exactly correct and all things considered a transformer is the best option and it will give you a balanced or unbalanced output, whatever you want. However, it is the most expensive. Darn.
Behringer has tied the VREFH pin to 5 volts and VREFL pin to ground. These are the recommended voltage to scale and "window" the four output pins to the proper levels. (This information is on page 4 of the data sheet.) I haven't measured these myself but this should position the AOUT pins at approximately 2.5 volts DC and thus the requirement for a DC blocking capacitor somewhere in the downstream circuitry. (C43-C48 in the output sections.)
I'm sorry I haven't been following this thread for the past few months, but I've been reviewing it a little bit today. Unfortunately there is quite a bit of mis-information in many of the posts.
Cheers,
Davey.
Hi,
now that we are getting closer to the real output impedance of the modified DCX output stage.
Can we estimate the true Zout?
If 10R +33R then 0.1uF is a bit large. 68nF or 56nF might be an option to try.
Thmartin and anyone else,
did you try a range of caps on the output?
How did it sound?
Would moving the RC time constant around help?
What roll off characteristic is ideal following a DAC?
Arcam have used a 3pole Bessel (active) in the past. Also set to the high 40kHz.
now that we are getting closer to the real output impedance of the modified DCX output stage.
Can we estimate the true Zout?
If 10R +33R then 0.1uF is a bit large. 68nF or 56nF might be an option to try.
Thmartin and anyone else,
did you try a range of caps on the output?
How did it sound?
Would moving the RC time constant around help?
What roll off characteristic is ideal following a DAC?
Arcam have used a 3pole Bessel (active) in the past. Also set to the high 40kHz.
Hi AndrewT
I am VERY happy with what gets out from the DCX now.
However, just to try a more "universal" solution I'm planning to remove completely the DC blocking caps by using the DACs in balanced mode.
That'll include a 1KR resistor on each leg AOUT+ and AOUT- and a 3,3nF polyester capacitor between them.
This will avoid having to block DC (DC shall be cancelled in balanced mode) and 1KR probably provides a better protection for the DACs.
This solution will be tested in the next weeks, I'll let you know what comes out.
But I insist on the fact that the DCX modified as it is is a pure wonder.
Thierry
I am VERY happy with what gets out from the DCX now.
However, just to try a more "universal" solution I'm planning to remove completely the DC blocking caps by using the DACs in balanced mode.
That'll include a 1KR resistor on each leg AOUT+ and AOUT- and a 3,3nF polyester capacitor between them.
This will avoid having to block DC (DC shall be cancelled in balanced mode) and 1KR probably provides a better protection for the DACs.
This solution will be tested in the next weeks, I'll let you know what comes out.
But I insist on the fact that the DCX modified as it is is a pure wonder.
Thierry
Thanks, Thierry.
Simple solution, I am going to try it myself.
I will try it into a valve output buffer,
so I'll be able to run longish cables,
we'll see how that turns out.
Plenty of other stuff to do first, tho,
so probably won't get it done until late spring ,
by that time others may have tried that allready?
cheers
Simple solution, I am going to try it myself.
I will try it into a valve output buffer,
so I'll be able to run longish cables,
we'll see how that turns out.
Plenty of other stuff to do first, tho,
so probably won't get it done until late spring ,
by that time others may have tried that allready?
cheers
As best as I can tell, the opamp circuitry is providing:
- current buffering for the DAC
- a high frequency (above audio range), low pass filtering
(both I presume help stabilize the DAC's audio output)
- a shunt circuit for muting signal
- a differential drive circuit for balanced output
Not knowing what the DAC's current drive capability is, how stable and high freq trash free it is, I would be reluctant to chop the first 2 opamps out of the circuit. If you're always going to run an unbalanced output, the last 2 can more easily be chopped without worries.
If the DAC output checks out clean, the mods as shown will work just fine. The downsides are losing the output mute and balanced drive permanently.
MEB
- current buffering for the DAC
- a high frequency (above audio range), low pass filtering
(both I presume help stabilize the DAC's audio output)
- a shunt circuit for muting signal
- a differential drive circuit for balanced output
Not knowing what the DAC's current drive capability is, how stable and high freq trash free it is, I would be reluctant to chop the first 2 opamps out of the circuit. If you're always going to run an unbalanced output, the last 2 can more easily be chopped without worries.
If the DAC output checks out clean, the mods as shown will work just fine. The downsides are losing the output mute and balanced drive permanently.
MEB