This is my revised synthesis for olo111' PCB. Post#52 didn't make DSDON active. DSM is 5th order 160OSR or 80OSR(with EEPROM-5pin 1kohm pulldown). I don't think high-order DSM has the advantage because the lowest order, which can have adequate noise-shaping, is enough. Your high performance in the digital domain has nothing to do with those in the analog domain.
Attachments
I'll reserver further investigation... but from the surface seems to be the real deal.
Keep you posted.
Its delayed due to fire but, can we use akm chip as pcm2dsd conversation?
AK4191EQ | Audio D/A Converters | PRODUCTS | Asahi Kasei Microdevices (AKM)
AK4191EQ | Audio D/A Converters | PRODUCTS | Asahi Kasei Microdevices (AKM)
Last edited:
The shipment status on my order(xc6slx9-3TQG 10pcs)from Mouser is Aug. 23. Their estimation is almost correct in my experience. Three months isn't good but not terrible.
Hello xx3stksm san,
You do not have to wait mouser. Trustable source winsource electronics has it in stock.
XC6SLX9-3TQG144C Xilinx Inc. | WIN SOURCE
If anyone interested verilog code of pcm2dsd board from Mr. Marcel is here below.
https://linearaudio.net/sites/linearaudio.net/files/additionaldatavalveDACversion2p1.zip
https://linearaudio.net/sites/linearaudio.net/files/additionaldatavalveDACversion2p1.zip
AK4191 is designed to interface with AK4498 only. IIRC, AKM said the chips would only be sold in pairs, and that there would be no support for any other uses. It was said that AK4191 existed in FPGA form, but a final production version was never available for sale.
AKM proposes D/A Separation Solution for High-end Audio DAC achieving Real Live Sound | News | ABOUT AKM | Asahi Kasei Microdevices (AKM)
You'll find a more compact version here:
https://www.diyaudio.com/forums/dig...linear-audio-volume-13-a-152.html#post6569767
https://www.diyaudio.com/forums/dig...linear-audio-volume-13-a-153.html#post6570538
but that still needs an XC6SLX45 FPGA (as opposed to an XC6SLX75 for the version on the Linear Audio site) and an SRC4392, so it's nowhere as compact as the XC6SLX9 designs in this thread.
I have repeated the test with HQPlayer as player also for the Simple DSD modulator with PCM output and the result is quite the same.
It is obvious that the conversion made by HQPlayer is of higher level but in any case the result obtain with this module is very good.
Did you do this test in the same hardware chain (PCM and DSD playback)?
I measured the DSD performance of D90 because I have an ADC board(pic.1) with xc6slx9 and HDMI-I2S. It's straightforward to reassign the pin assignment for this board to feed D90 256DSD via HDMI-I2S. MCLK is 22.5792MHz because D90 doesn't accept a 48k family. The result is better than expected, though SNR is slightly worse due to the DSM amplitude penalty(-6dBFS is usually max). I guess DSD256 in 4th order is probably better than DSD128 in 5th order. 0dBFS performance is pic.2. THD is almost the same as USB, but SNR is 6dB worse. Since D90 doesn't have a dedicated 1bit converter, -6dBFS max means 6dB loss in SNR. But ES9038 doesn't have such excellent performance in DSD256 as long as I know. I would say AK4499 is the best monolithic chip in DSD mode. The attached Zip file is 4th order 256OSR for olo111's board, which is the same one used for D90.
Attachments
Thank you very much xx3stksm for sharing your work again. Tested, works fine. Good job!
Tested PCM and DSD sound (pass-through works ok)
These differences are too small to hear and remember. That would have to be compared one to one. But you can see e.g. -10dBfs (first photo DSM5, second curent DSM4):
