ESS Sabre Reference DAC (8-channel)

[rossl]

Originally posted by Dougie085
But then that bypasses the DSD support of the DAC. And if you want to do that you can just feed the DAC SPDIF or something.

absolutely that is what I want to do.

[sidiy]

Question for Russ: how does the S/PDIF buffer achieve 75 ohm of input Z? As is R13 can not determine the input Z of the circuit, but perhaps I'm wrong.

I hope it's not too late to correct this.

[Russ White]

Originally posted by sidiy
Question for Russ: how does the S/PDIF buffer achieve 75 ohm of input Z? As is R13 can not determine the input Z of the circuit, but perhaps I'm wrong.

I hope it's not too late to correct this.

Hi,

I have never implemented that circuit before, but I saw it in the demo board schematic, so unless there is an error in the schematic, I assume it works ok.

I Actually do think the impedance of the circuit is 75 ohms, at least at the working frequency.

Maybe someone more familiar with this input buffer can chime in.

Cheers!
Russ

[rossl]

Originally posted by sidiy
Question for Russ: how does the S/PDIF buffer achieve 75 ohm of input Z? As is R13 can not determine the input Z of the circuit, but perhaps I'm wrong.

I hope it's not too late to correct this.

It's OK. The incoming signal 'sees' a 75 ohm impedance into the high impedance node of the comparator. It will correctly terminate and will not reflect any energy back into the 75 ohm cable impedance.

[sidiy]

Please correct me if I'm wrong.

Originally posted by rossl


It's OK. The incoming signal 'sees' a 75 ohm impedance into the high impedance node of the comparator.

Then the source sees 75Ohm + something (high) to ground. Which means a lot of reflections back into the source.

[dusfor99]

Originally posted by sidiy
Please correct me if I'm wrong.



Then the source sees 75Ohm + something (high) to ground. Which means a lot of reflections back into the source.

Hi Sidly,

Yes you are right this circuit will not do a true 75 Ohm load to terminate the SPDIF line. It was used since it worked (All the time) . Now that you have brought it up, here is a properly done circuit.

[Russ White]

Originally posted by dusfor99



Hi Sidly,

Yes you are right this circuit will not do a true 75 Ohm load to terminate the SPDIF line. It was used since it worked (All the time) . Now that you have brought it up, here is a properly done circuit.

Thanks Dustin, that looks much more conventional.

I just simulated the previous circuit, and was about to post.

Fortunately there is plenty of time to make the change.

Cheers!
Russ

[sidiy]

Hi Dustin,
Indeed we'd be in deep trouble if that type of circuit would not work. Every CDP needs a data slicer afterall (this because despite using NRZI coding, targeted to eliminate the DC content, the RF signal from the laser diode still 'rides' on a varying level of DC)

I drawed quickly two alternatives. Just a note on the bias through R2 or R7: the voltage drop across them, mainly determined by R3 and respectively R8, must be sensibly higher than the hysterezis of the comparator.

I'm using the second one in one of my DACs, inspired from the demo-board of the PMD100 digital filter.

Hope it helps, cheers!

[rossl]

Originally posted by sidiy
Please correct me if I'm wrong.

Then the source sees 75Ohm + something (high) to ground. Which means a lot of reflections back into the source.

No, I'm sorry, I disagree with your statement. (grin)

The circuit didn't reflect back to the source. That is why it worked.

The term (+ something (high) to ground) is a high impedance in parallel with the 75 ohm.

The incoming AC coupled RF signal doesn't really care about it.

RF signals don't care about 'ground' in general. That is why they are so hard to conceptualize. They will leave the circuit board and travel through the air at a whim.

The terminating impedance was close enough. +/- 10% is usually close enough.

I wouldn't have done it that way, but it was good enough and it worked.

You can't argue with that.

[Dougie085]

Meh as long as it works and sounds awesome I really can't wait to see how you think this DAC compares to the WM8740 that the Opus uses.