I thought it would be a great idea to showcase the new board design that I have been developing for the past few weeks to show how similar but different it will be compared to the Marantz DP870 version that I posted two months ago.


So here is the schematic for the DP870……




….. And now the one for the ED-901….




Overall they are similar with the big exception being that the SN74HCU04N IC is now playing a bigger part in that it not only provides a buffer circuit (once again) for the coaxial digital input but it now has additional usage as a “fanout” buffer for the AC-3 signal from the Demodulator IC. And upon closer inspection, one may notice that the fanout buffer is being deployed this time to provide an amplified Y split for the purpose of running that AC-3 signal “twice” to the SN74HC151.

So one might be wondering, “why run that signal twice?”. Well…. It simply comes down to this….




In the process to design this new Mux circuit for the ED-901, the presence of a physical switch that can be left in either position while the processor is on the AC-3 RF Input poses an issue w/ how the select line truth table works. My first thought when designing this circuit was the possibility of using two Mux ICs’ (one as primary to choose between AC-3 RF & Digital Input and a secondary to choose between Optical & Coaxial Digital Inputs). I did not find this a practical approach due to the amount of available space on the board to fit yet a third large IC and also add even more parts to power it up.

Hence, I did eventually realize that a simple solution was to run the AC-3 RF (post demodulation) signal twice to satisfy the truth table requirement but an immediate concern I had was that the signal was going to be split three times now (one to the Sanyo digital interface IC already in the unit and the other two to the new Mux switcher) and I didn’t want to introduce any additional compromises to the signal in the form of possible reflection issues w/ the digital 1’s and 0’s and overall signal degradation. After doing some research, I eventually came up with the “fanout” buffer concept which essentially is a zero loss Y-split and it was a fortunate coincidence that I already had an IC ready to go in the form of the unused portion of the SN74HCU04N (unused in respect to how it wasn’t used in the Marantz version).

Moving on, I had previously used the LED indicators on the Marantz unit to provide the switch line voltage for input selection. For the Onkyo unit, that will not be the case as I can take advantage of the Mux circuit inside the unit that works the video switching circuit along with that physical digital input selector that happens to give off a voltage change as well.

So that is where I am at this point. My goal is to have this mod completed in about three months as I figure out how to allocate some time to do it. I’ll update this posting w/ all the details when the time comes.

That sounds like an awesome project.

Note: Someone already did a simpler mod here:
https://raudonis.eu/heimkino-ac3.html

That way you can switch between PCM(/DTS?) and AC3 and use the original coax connector as output with just a single new wire.

Interesting…

I always assumed that Pin 1 on the Sanyo chip was shutdown since that is the case on the Marantz and HK units (we’ve discussed this before in the ADP-303 thread that this pin would work but needs to be enabled).

I’ll test that out first before proceeding with my design to see what it does. If it works, I will undoubtedly scrap my plan.

Thanks for the heads up. This will definitely make it an easier mod for everyone to know this pin is active.

I had exactly that mod working for a few weeks. Wanted to reposition my components a little and after that the ED901 still decodes AC3 via it's 6-channel output but SPDIF isn't working anymore. I assume one would need at least a few more components to buffer the output and adjust levels.

If it worked before it should continue to work. Maybe moving the unit around loosened something up inside?

I‘m not so sure. Some kind of static charge might have damaged the Output as there is no galvanic protection if you apply the mod as described. Haven’t measured it as i don’t have an oscilloscope right now. Might try to use a multimeter at least to check if any signal is being output ….
That is what ChatGPT is saying if i ask:

That would make sense. Definitely should have that output buffered on some level. Also, not sure why the original modder just didn’t keep the Coax input intact and allow the unit to maintain its three input capability; maybe it’s laziness.

I plan to put together an optical output w/ a 75 ohm resistor in line when I get the chance. This is really the better option to maintain isolation. Curious to see how it works.