@iancanada,
Looking at the "9038Q2MPi Dual Mono Plus DAC HAT user’s guide," on page 3 it says:
J1: 3.3V DC power input for AVCC
J4: 3.3V DC power input for VCCA
J2: 3.3V DC power input for DVCC
Connect three 3.3V DC / 100mA (minimum) power supplies to those 2-pin 5.0mm terminals
Should I conclude there is still only a single AVCC rail shared by both dac chips, and not separate AVCC power connectors for each dac chip (equivalent to separate AVCC_L and AVCC_R power inputs for a single ES9038Q2M stereo dac)?
If shared AVCC, I would expect stereo separation to still be less than what I would like to have. Also, I prefer not to have ceramic capacitors on any AVCC rail if I can help it.
Looking at the "9038Q2MPi Dual Mono Plus DAC HAT user’s guide," on page 3 it says:
J1: 3.3V DC power input for AVCC
J4: 3.3V DC power input for VCCA
J2: 3.3V DC power input for DVCC
Connect three 3.3V DC / 100mA (minimum) power supplies to those 2-pin 5.0mm terminals
Should I conclude there is still only a single AVCC rail shared by both dac chips, and not separate AVCC power connectors for each dac chip (equivalent to separate AVCC_L and AVCC_R power inputs for a single ES9038Q2M stereo dac)?
If shared AVCC, I would expect stereo separation to still be less than what I would like to have. Also, I prefer not to have ceramic capacitors on any AVCC rail if I can help it.
Last edited:
yes, and? you were talking about Ian's dual mono 9038q2m DAC, not his non existent 9038PRO multi channel design. it is a dual mono design, not a 4 channel design. I think you need to learn what it is you are looking for. Dual Mono ... 2 x 1 channel dacs. just because they can be stereo dac chips, doesnt mean they are in every application.
indeed the old multichannel fifo is .... a multichannel fifo and therefore it would be suitable for reclocking a digital crossover. the mere fact the 9038 dac chip exists, doesnt mean just plugging it into Ians design will make for a good crossover product and it is not that simple. the MC fifo was created for that very purpose AFAIK, but you must pair it with an i2s input (DSD... not so much) multichannel dac and a multichannel i2s source (DSP with MC i2s output, whether that be a PC with a USB->MC i2s board, or a DSP board with multichannel I2S output).
Last edited:
@nicoch58
Sorry for my absence.
MCFifo/MuDualXO solution is the only multi channel FIFO solution so far. It's great for all multi-channel applications. So I think should be no problem for the DSD XO. But if it's possible, please give me some details about it just to confirm.
My ES9038Q2M Dual Mono Plus is a stereo DAC, not a multi-channel DAC. It uses two ES9038Q2M chips in mono block configuration just for better performance. However you and use more pieces of them to set up a multi-channel DAC. McDualXO has enough DATA output signals (7 in total) to drive more DACs.
Regards,
Ian
+1
@iancanada Any plans to update the design to include the separate rails option, or would it all not fit on the given PCB real estate?
@iancanada Any plans to update the design to include the separate rails option, or would it all not fit on the given PCB real estate?
You can try multiple DAC for the multi-channel configuration, but have to share same controller.
Regards,
Ian
He wants to know if you can hack the dual mono dac into 2 stereo dacs ... he wouldnt accept my answer of no. I mean you could maybe do it, with a different i2c controller, or no i2c controller and hacking a whole bunch of traces, but given I suppose you are using the gpio pins to send the i2s from one dac to another, this seems unlikely.
Last edited:
@Markw4 @xaled
Yes, theoretically separated AVCC for both left and right channels would be in better performance. But when I use ultracapacitor power supply or ultracapacitor /LifePO4 hybrid power supply for my DAC, I didn't feel any difference. Because the ESR of those power supplies is very low in just a couple of mohm level. And also, separating AVCC could result in different DC offset in I/V stage, unless the I/V stages are also in dual mono configurations. For two AVCC voltage rails can not be 100% identical. I'll take it into consideration if I got chance to design a AK4499 or ES9038Pro DAC.
Agree with you that using MLCCs along for analog is not a good solution. However, when using organic polymer capacitor as main decoupling capacitor, MLCCs could be a good compensation for the high frequency range. But there are something we have to keep in mind
1. Have to be carefully choosing MLCCs for low piezoelectricity effect.
2. Have to combine with small package NP0/C0G capacitors.
Traditional aluminum electrolytic capacitors are no longer suitable for modern DACs because the MCLK frequency could be extended to 100MHz range.
Please refer to the picture of my ES9038Q2M Dual Mono Plus DAC HAT for more information.
9038Q2MpiDmPlusa by Ian, on Flickr
Good weekend.
Ian
Last edited:
I don't think you can do it because the circuits of the multi-layers PSB are too small to make any modification.
Ian
Last edited:
You just keep on pretending its a good idea mate and post your pics when you are finished. yes, nothing to lift. no pin and if there is, for example, since its a 4 layer board, a qfn, which has no pins, could have an i2s, or other pin that is shared with the second chip and is routed underneath through a via in one of the inner layers. if it had pins, you might be able to carefully lift one and tap in; with a qfn, this is not possible.
Last edited:
Pulsar, CCHD957, AS318-B ....
Pulsar is the best of the best
Ian
Compared with AS318-B and CCHD957, AS318-B is slightly better, but the price of AS318-B is very low.
I also have Pulsar, but it’s a pity that Pulsar is discontinued again, so the only option is AS318-B.
The ES9038Q2M datasheet only mentions the DNR of 129 in stereo mode, and does not mention it in momo mode. May I ask where the 132dB (theoretical) you claimed came from.
http://file2.dzsc.com/product/18/05/25/829029_170233543.pdf
In fact, the manufacturer only mentioned that the DNR of ES9038PRO in MOMO mode is 140. In other words, I am not sure, is a DNR of 140 the theoretical limit value?
