I don't think it would work well, because the outputs will strongly affect each other through the coupling between the primary windings.
yes, that's what I'm afraid of too. But maybe in push pull configuration because of high side low side cancelation it might work? 32 primaries in comparation to ddpd which is true 1bit this one will form 32bit moving average coupling, similar to what dsc2 dac does with resistors, I would like to try it, the problem is the price of iso gan fets, they are not very cheap for a test like this, the whole project would not be cheap : )
Experimented single end passive output with 3rd order Butterworth filter,
Leveraged same MUTEREL pin as the filter board to shunt the output after dc blocking capacitor,
Sorted out a quick solution for the mute function otherwise the pop during tracks esp sample rate change is pretty loud,
Cut the trace for mute-Pin7 of 1G74, and tied to Vcc.
Pop noise is still hearable while much weaker
Will try analog switch like TS5A3167DBVR for quicker response.
Thanks @Markw4 for your insights.
Leveraged same MUTEREL pin as the filter board to shunt the output after dc blocking capacitor,
Sorted out a quick solution for the mute function otherwise the pop during tracks esp sample rate change is pretty loud,
Cut the trace for mute-Pin7 of 1G74, and tied to Vcc.
Pop noise is still hearable while much weaker
Will try analog switch like TS5A3167DBVR for quicker response.
Thanks @Markw4 for your insights.
Hi 2A3SET,
I am not sure this may be relevant to your design but is there any chance that you are shorting a DC voltage to GND every time the relay "disconnects" the output? If this is the case I reckon there would be a relatively loud pop ... If there is a DC level might it then be an option to maybe short the DAC output to a capactor-to-GND connection (through a relay) where the capacitor is adjusted to have a DC level similar to the DAC output DC level? That is: DAC output w. DC level -> relay -> capacitor at same DC level connected to GND. The capacitor should have a capacitance value allowing for shorting even low frequencies, and the currents through the relay should be within the specs for the relay ...
You may of course already have thought about this ...
Cheers, Jesper
I am not sure this may be relevant to your design but is there any chance that you are shorting a DC voltage to GND every time the relay "disconnects" the output? If this is the case I reckon there would be a relatively loud pop ... If there is a DC level might it then be an option to maybe short the DAC output to a capactor-to-GND connection (through a relay) where the capacitor is adjusted to have a DC level similar to the DAC output DC level? That is: DAC output w. DC level -> relay -> capacitor at same DC level connected to GND. The capacitor should have a capacitance value allowing for shorting even low frequencies, and the currents through the relay should be within the specs for the relay ...
You may of course already have thought about this ...
Cheers, Jesper
Loud pops with DSD can be caused by player using PCM silence (i.e. "0" data) instead of DSD silence e.g. when sample rate is changed or even when new track is started. The USB-I2S board should take care of this with mute pin. But if mute pin does not do its work properly correcting this downstream may be quite difficult.
... just a brief comment to @bohrok2610's post above: I have been using the JLSounds' board for quite some time now and I have never had any mute issues with it. Even when the playback software - likely due to my laptop not being fast enough - generates complex drop-outs (like many opposite signals in a short while) the mute function is efficient and never causes any ticks or pops. Quite impressive IMHO.
Cheers, Jesper
Cheers, Jesper
I have implemented such mute function on my STM32 USB-I2S boards and it can be quite complex since players work differently. Another similar issue is UAC asynchronous feedback. A "torture" test for both is to switch between extremes. E.g. 44k1 PCM to DSD512, 64DoP to 768k PCM, 48k PCM to 705k6 PCM and so forth.
Hi Jesper,
The relay is after the DC blocking capacitor.
When mute is engaged, it resets the 1G74 so output data are all 0, this introduces the pop for single end output, while has no impact for balance since +/- sides are same.
The silly thing is that in a very early version, you could simply change some 0 ohm resistors to have the mute function disable the outputs of the 74LV574A's instead (drive their "output enable not" inputs high, pin 1). As the present mute worked fine in differential mode and I hadn't anticipated that people would want to experiment with single-ended versions, I removed those 0 ohm resistors from the layout.
Hi 2A3SET ... Long story very short: Do you think that what Marcel initially designed for, i.e. disabling the LV574 outputs, as described above, would be an option?
Is the relay in series with the output from the DAC / input to the analog amplifier, or does it short to GND?
Cheers, Jesper
Is the relay in series with the output from the DAC / input to the analog amplifier, or does it short to GND?
Cheers, Jesper
Hi Jesper,
I believe Marcel's design is perfect for full balanced output, but not considering for single end output which uses only one side of the balance.
Same as filter board the relay shunts the output after the DC blocking capacitor.
I believe Marcel's design is perfect for full balanced output, but not considering for single end output which uses only one side of the balance.
Same as filter board the relay shunts the output after the DC blocking capacitor.
I remembered it incorrectly. On the prototype that I made, you could remove some 0 ohm resistors to get access to pin 1 of each 74LV574A, but then you still needed to add an inverter and wires to actually drive them from the muting circuitry.
In any case, you could still use pin 1 for muting if you manage to lift it and add an inverter and some wires. I know @Hans Polak lifts pins sometimes, but not how.
@2A3SET : I think then that the issue is that the relay switching speed and the "pulling to zero" speed of the combination of the 1G74 & the LV574 does not happen simultaneously, thus creating DC offsets - which when the relay releases or shorts - causes noise to appear on the output.
As I mentioned in my first post above I have solved this myself by changing the relay short-to-GND with a relay short-to-same-voltage through a large capacitor (low cut-off frequency) to GND. In this way the DAC's single-ended DC output voltage "shorts" to the same voltage and thus pops etc. does not appear. The disadvantage of this arrangement is that there will be a DC offset from the DAC, which, depending on the surrounding circuitry, may or may not be an issue.
Otherwise, apart from Marcel's suggestion, I do not have any other ideas ....
Cheers, Jesper
As I mentioned in my first post above I have solved this myself by changing the relay short-to-GND with a relay short-to-same-voltage through a large capacitor (low cut-off frequency) to GND. In this way the DAC's single-ended DC output voltage "shorts" to the same voltage and thus pops etc. does not appear. The disadvantage of this arrangement is that there will be a DC offset from the DAC, which, depending on the surrounding circuitry, may or may not be an issue.
Otherwise, apart from Marcel's suggestion, I do not have any other ideas ....
Cheers, Jesper
That's very easy.
First remove all solder from above with solder wick.
Heat the pin again and carefully slide a scalpel or lancet slightly between the pin and the PCB, just far enough to get a momentum to push the pin upwards.
Once the pin is separated from the PCB , you can carefully bend it further upwards.
Now a small diameter wire can be soldered to the pin to connect it to elsewhere.
Hans
Happy 2025 to you and to everyone too!