Microcontroler/and i2s commuting lines modification Denon DVD-A1

Hi, I have upgraded Denon DVD-A1 (8x pcm1704 , Alpha 24 procesosor, spdi freceiver input) a Mark levinson no360s analog IV stage is built in.
sound is much worse than Mark Levinon No360s DAC using same dac chips and same DAC IV I thought it was power suply (no360s uses discrete) so I put complete no360s DAC/IV boards to DVD-A1. to my surpise is still sounded much worse than No360s. now it was only the digital front end diferencies.

I looked deeper in the schematics of DVD-A1 and see that I2s doesn’t go from DIR to AL24 chip. instead of that it goes trought few DSP chips –DVD processor chip ES6038, and ADSPMEL322. so I connected i2s lines from DIR directly to AL24 chip and bingo- good sound at last.

I make micro small swithches near rj45 jack (denon link jack whch is perfectf or i2s input which I will add, remooved two chips to have this jack free from any conections)

now is problem with comutation. I do not want loose cd-dvd source. dvd-a1 have 3 source selector switch- DVD/Optical/Spdif I need this- WHEN DVD position is set relays are eneergized and i2s line ir connected ( or maybe is more elegant decisions) I am not familiar with micrcontrolers. How to get signal from it to relays to be energized when dvd rotaty position is set? thanks in advance.
 

Attachments

  • i2s .interface.png
    i2s .interface.png
    541.1 KB · Views: 106
  • i2s main.png
    i2s main.png
    842.9 KB · Views: 107
  • i2s rotaty.png
    i2s rotaty.png
    17.5 KB · Views: 107
  • i2s-dvd input.png
    i2s-dvd input.png
    10.6 KB · Views: 95
If the lines you want to switch are I2S bus, its usually done with logic chips such as those used in the following devices: https://www.ebay.com/sch/i.html?_from=R40&_trksid=p2380057.m570.l1313&_nkw=I2S+switch&_sacat=0
Basically, the chips themselves are like those found at: https://www.ti.com/logic-voltage-tr...s/digital-multiplexers-encoders/products.html
You need a mux function that is compatible with the logic families used in your equipment.

The logic chips can be driven/controlled by microcontroller I/O pins or maybe through something like an I2C bus expander chip that can provide some extra I/O pins.
e.g. https://www.ti.com/interface/i2c/general-purpose-ios-gpios/products.html
https://www.ti.com/lit/ds/symlink/pcf8574.pdf

Also, if you want to drive actual relays, you can use relay driver chips and or discrete transistors that are controlled/driven by microcontroller I/O pins.

One thing to consider when switching I2S signals is that jitter can be introduced if proper care is not taken. In some cases reclocking may be used to de-jitter I2S signals.

Not sure if that helps?
 
Last edited:
Thanks Mark,really helpful. During waiting time I have redesigned my comutation to more elegant and simple "raw" way. I will use another 3 way turn button (which turn of screen and some power suplies which is not revelant for me, now it will be real pure direct as will bypass all dsp chips :) )
I will wire +12V suply at this switch and simply redistribute it to relay 1 or relay 2 or both is disabled. I find very small 12v relays (from datashhet its seems normal not solid staterelays) for this purpose -

https://www.elpro.org/gb/goodsky-st-series/30170-st-sh-112l.html

- operation time 5mS maximum
- operating frequency 30Hz
DIMENSIONS:
length 12.6 mm
width 7.8 mm
height 10 mm

now questions is how to implement this -
1)are using smallest relay is best or better use bigger (mabe less interference with coil?)
2) better use 5V or 12V?
2) I can simply feed them via LM7812 (do not limit any current?) just ON/OFF suply of voltage ?
3) beter disconect GND or 12V on relay when not working?


there is 74HCU04AF just before alpha 24 filter processinng chip.
 

Attachments

  • i2s-dvd input.png
    i2s-dvd input.png
    7.9 KB · Views: 55
  • pure direct.jpg
    pure direct.jpg
    74.2 KB · Views: 46
The relays you linked to have a minimum switching current of 2mA. Sometimes the contacts in a relay don't conduct well until enough current is run through them after the contacts are closed. If you want to use relays maybe consider Panasonic or Omron low-signal miniature relays with gold and or palladium contacts. Maybe something like this: https://www.mouser.com/datasheet/2/315/Panasonic_03172022_TX_Relays-2934354.pdf

For 12v relays you could use an integrated driver chip such as NUD3160: https://www.onsemi.com/pdf/datasheet/nud3160-d.pdf
For lower coil voltages maybe a transistor and some associated components would offer less voltage drop.
 
Last edited:
  • Like
Reactions: 1 user
The relays you linked to have a minimum switching current of 2mA. Sometimes the contacts in a relay don't conduct well until enough current is run through them after the contacts are closed. If you want to use relays maybe consider Panasonic or Omron low-signal miniature relays with gold and or palladium contacts. Maybe something like this: https://www.mouser.com/datasheet/2/315/Panasonic_03172022_TX_Relays-2934354.pdf

For 12v relays you could use an integrated driver chip such as NUD3160: https://www.onsemi.com/pdf/datasheet/nud3160-d.pdf
For lower coil voltages maybe a transistor and some associated components would offer less voltage drop.

Thanks Mark, I will look in to these (but i have ordered that small ones) I would like to avoid relays drivers, so to be sure - technicaly speaking nothing prevents from runing relays with "raw" 12v suply diceonecting and conecting them by open /close circuit. I used that way in my car without any issues.
 
Tangential, but one quick way to test an alternate digital filter pathway in this unit is playing an HDCD/streaming HDCD via SPDIF, thus using PMD-100 processing vs. AL24. Nice to see you got the I2S direct working, what are you using as a I2S source?
there is no PMD100 inside. digital filtering is done via AL24 PLus(capable of 192khz) and HDCD decoding done with DSP, I believe(unless al24Plus have integrated HDCD decoder)

but i have DVD-5000 which is similar a it older machine and have selectable PMD100/ AL24 filter HDCD materials goes trougth PMD100 and the rest via AL24. it was easy task to comapre it when fed via usb-spdif playing HDCD file( I have HDCD decoder plugin in foobar 2000): as soon as I decreaeed digital volume from 100% to 99% its switched from PMD100 to AL24 ( HDCD led goes off, Al24 led goes in and relay adjusnting HDCD/non HDCD levels clicks)

so in real time adjunsting swound from 100 to 99% it is posible AB testing.I have found AL24 superior to PMD100. after AL24 PMD 100 sounds rounded, mellow,lumpy and without real sharp edged resolution.

Main reason why I started this I2s swithing is that DVD-A1 are capable of 24 bit 192khz (internal DVD-audio CD) and 24bit 96khz external ( even spdif receiver are capable of 192khz) when using external spdif only 44 and 48khz works, as soon as I switched to 88khz or 96khz sreens shows locked to frequency output relays clicks and very loud white noise with some excerps of music apears. i have tested many usb/spdif converters , usb wires- nothing helped. it seems there is some internal failure of selecting clocks or firmaware upgrade needed (which is impossibel to get after 20 years) so I put straight i2s Righ justified to Al24 PLus and it worked like a charm with 32khz/44.1khz, 48khz,88.2khz, 96khz and 192kHz.
DVD-5000 was impressive build unit (19kg, 4xpcm1704, AL24, PMD100) but very medciore performance) after upgrade (see pics it realy changed)
DVD-A1(also 19kg, 8x pcm1704, Al25 plus) was better sounding as stock but there is hughe performance wall between those units and Mark Levinson and Wadia units I have, considerng they are same build quality (Denons are build even nicer than Wadia)
 

Attachments

  • before.jpg
    before.jpg
    129.7 KB · Views: 124
  • after.jpg
    after.jpg
    195.7 KB · Views: 77
D'oh. I've read too many data-sheets and, as you deduced, confused the 9000 and the 5000. That's a nice looking output stage on the 5000. Is that a current-mirror-fed diamond buffer on the output of the I/V?

I run 16 relays off a 5VDC rail in a line-level stereo attenuator, no problem with electrical noise/interference when switching the coils from the "B+" and still operational after ten-plus years. Your circuit appears to be carrying a control signal, so I'd think any conveniently-interfaced relay would work, providing it doesn't load the line too much. The data sheet will give you a coil-resistance, which will give you the relay's current draw, which should be enough to actuate the relay at its rated voltage. Apologies in advance if this a simplistic answer.
 

Attachments

  • d'oh.png
    d'oh.png
    26.6 KB · Views: 41