Behringer DCX2496 digital X-over

[pinkmouse]

They probably do it in the processor, and the board was never updated due to retooling costs for the production line.

[sfogg]

"You know what, this is funny but the "Mute outs" is still bloody working even if T1 to T6 trannies are out of business !"

They may mute the signal out of the DSP to the DAC for the Mute button function.

Are the hardware shunt mutes used for turn on/off states to avoid transients?

I had the earlier BFD and it had very bady on/off transients.

Something like that in a line leve crossover where a tweeter is directly connected to an amp would be *bad*.

I have never tried the DCX to see if it has on/off transients as I didn't want to risk my drivers. I just leave them on all the time.

Shawn

[Thmartin]

Actually the AK4393 has a built-in mute function.
I rather think Behringer has been playing on the safe side, i.e. another mute, just in case

Anyway, mute buttons operate perfectly and function is dead silent, not the tiniest transient so won't complain.

[ThorstenL]

Konnichiwa,

Originally posted by mbroyles
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.

Which begs the question: "Why don't you know?"

You could have consulted the DAC IC's Datasheet which is readily available on the net and you would have found out that the DAC will drive an 600R AC per phase and a 1K DC per Phase load, has 3.5mA rated output current or theoretically 8.2V Peak into 1K2 Balanced AC Load.

Stoppband attenuation (including the noisefloor from the Noiseshaper whic actually dominates) is better than -75db, a little lowpass filtering (best 2nd order LC LPF) seems indicated.

My final recommended output confioguration would be to use a 1.2K Load Resistor per phase in parallel with 1.5nF capacitance and a 4.7mH Inductor in series to the DAC Chip, per phase and DC coupled, take from each junction point a 2.2uF (or larger) coupling capacitor.

This circuit allows loads down to 10K per phase or 20k balanced, retains the balanced output and allows 1nF per phase or 500pF balanced cable capacitance, in other words more than enough to accomodate most sensible cables and amplifiers, but prohibiting extreme loads (which a 600R:600R line output transformer may actually allow).

FWIW, at the stopband return of the Digital filter (assuming 96KHz operation) the passive filter will be > 40db down, for 44.1KHz operation there will still be 30db attenuation, sufficient for most competently designed amplifiers to handle okay.

Sayonara

[Davey]

Originally posted by Thmartin
Yup, pretty sure. Let's say 99%.
Anyway I'll measure it before I turn the amps on.

Thierry,

Yes, please measure it. You'll find approximately 2.6 VDC on each of the four output pins on each chip. A simple low-pass filter like you described (I hope my reading of your description didn't get lost in the translation...I don't think so because your English seems fine) will not do anything to block or balance this DC level and it will be passed on to the following stage.....whatever that may be.

Please be careful because this is a 1% situation.

Cheers,

Davey.

[sfogg]

"Anyway, mute buttons operate perfectly and function is dead silent, not the tiniest transient so won't complain"

I didn't mean a transient from turning on/off the mute but from turning on and off the Behringer itself.

Shawn

[Thmartin]

Oh yes, I'm with you now. Anyway, as I have been using monster amps for 30 years I always turn them on AFTER the Xover and turn them off AFTER.

I reckon the DCX is noisy in it's standard form.

[AndrewT]

Hi,
re post505 & my post474;-

AN003 fig 2.4 http://www.jensentransformers.com/an/an003.pdf

This circuit blocks DC as well as providing balanced source impedances.

Surprising no one has picked up on this yet.

I would add 1uF polypropylene // bipolar to help reduce distortion in the audio band.

[ergo]

Kuei Yang Wang,

I just made an experiment a few weeks ago with LC filtering on DCX.

My conclusion was that it looks good on paper but was bad in reality. The problem is that there seems to be a lot of RF floating around inside DCX and the L part of the filter will pick it up like antenna.

I can post measurements later when I get home. I looked at the spectrum of 10Hz-40kHz and the LC type filter created a lot more dirt that raised above the noisefloor.

Thus I burried the idea of LC and now my setup will be

RC -> 6ch volume pot -> RC -> OPA2134 in simple gain mode -> output.

I ran the DCX with simple RC filtering and no analog stage a long time, some 6 months or so, but in the end I did not like the sound especially in top octaves. There is something wrong there. At first it gives a very positive impression but gets tiring and irritating after a month or two.

Also the gain is too small for my liking.


Regards,
Ergo

[ThorstenL]

Konnichiwa,

Originally posted by AndrewT
This circuit blocks DC as well as providing balanced source impedances.

Surprising no one has picked up on this yet.

It is not surprising AT ALL. The circuit you cite is aimed at the exact opposite problem to the one faced wwhen directly interfacing the DCX2496 DAC's, that is it is meant to give a Pseudo-balanced output from a single ended source, while we need one that gives a SE source from a true balanced output.

Sayonara