Balanced AES/EBU (SPDIF) input circuit (to DIR9001)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm building a DAC with a DIR9001 receiver, which will be fed balanced AES/EBU from a Lynx AES16 sound card. Problem is, I haven't been able to find a balanced input circuit anywhere. And believe me, I've searched...

Anyone have a balanced input circuit? No transformers, it's already transformer coupled at the sound card.
 
Thanks a lot, but I'm afraid it doesn't help me much with the DIR9001, as it only has single ended input. I guess it's possible to do an AES -> S/P-DIF conversion first, and then do a single ended input circuit, but I'm sorta hoping there's a better way...
 
@novec
If output od your soundcard is transformer coupled, signal on secondary is already balanced :D
Better to think about right transformer, how to make correct impedance, signall attenuation, right connector and reflections from DIR9001 input.
Schmitt trigers are nasty :mad:

Nice reading here:
LINK 1
LINK 2
 
Last edited:
Thanks a lot, but I'm afraid it doesn't help me much with the DIR9001, as it only has single ended input. I guess it's possible to do an AES -> S/P-DIF conversion first, and then do a single ended input circuit, but I'm sorta hoping there's a better way...

AES/EBU and SPDIF are virtually the same thing. Use a proper pulse traf and terminate it correctly, and ground one side of the secondary to digital ground, there is no other way AFAIK, except to ground one conductor through a cap.
 
The DIR9001 accepts both S/PDIF and AES/EBU, so AFAIK the only thing I need is level matching. I imagine it would be a pretty simple thing for an experienced electronics dude to cook up, but I'm not one of them...

The cables will go straight from the soundcard to the DIR9001 without any connectors, so off-the-shelf solutions won't do much good.
 
Your Lynx manual says it outputs AES3 on the XLRs. Terminate it into a 110 ohm resistor, couple one end to ground through a .01 cap and couple the other end through another .01 cap to the input pin. Somewhere one line has to terminate into ground. The signal is differentially balanced and the DIR is not. A pulse trafo termination is still the best way though.
 
At the risk of sounding like a broken record, i'd look at the Crystal receiver data sheet. The problem with that data sheet is that the receiver portion assumes a single input into the 8416 input MUX. When you have multiple inputs, chances are that you can't use a transformer for each as the data sheet shows. For example I've got both TOSLink and SPDIF going into the receiver chip. So this chip is also effectively single ended because I can't take the receiver negative only for the SPDIF input.

My solution was to capacitively couple all inputs and the return to ground. That works as the Crystal datasheet says for the TOSLink input, but that can't work for a transformer in the SPDIF line. If I wanted a direct connection with the SPDIF, I'd drop a 75ohm resistor to ground, and couple to the chip through a 0.01uf cap and I'd be done. But I wanted to use an input transformer so I did something like this. The transformer primary is floating on both sides and goes to the BNC connector (you would use an XLR). One leg of the secondary is attached to digital ground and the other leg goes to the receiver chip through a 0.01uF cap and also to ground through a 75ohm resistor. I have the benefit of isolation and also input switching with this scheme. I'm not sure if it's ideal, but it has been working well for me.

here's a schematic from anther thread that shows what I did, and how when using the 8416 input MUX, that the input is really also single ended.

An externally hosted image should be here but it was not working when we last tested it.



Maybe I'm missing something, but I do't see why that wouldn't work for the DIR9001, which is also single ended.
Sheldon
 
Last edited:
Is there any reason I can't just terminate negative to ground with a 56 Ohm resistor and put a pad on the positive? I'm not entirely sure why, but I'd prefer to avoid the transformer, unless that's the only passive way to do it. And any kind of chip requires yet another power supply, making it more complicated again. The life of a DIY-er is never easy ;-)

If transformer is the best/only option, do you have any relatively affordable suggestions? Preferrably in the Newark/Farnell inventory.
 
Properly terminated into 110 ohms you will have a 3.5VP-P signal at the input if the AES16 output is 7VP-P. You don't need a pad, the termination pads it 6db.

Again, use a 110 ohm resistor across the lines and couple the lines to the input pin and ground with .01 caps. It's simple
 
Is there any reason I can't just terminate negative to ground with a 56 Ohm resistor and put a pad on the positive? I'm not entirely sure why, but I'd prefer to avoid the transformer, unless that's the only passive way to do it. And any kind of chip requires yet another power supply, making it more complicated again. The life of a DIY-er is never easy ;-)

If transformer is the best/only option, do you have any relatively affordable suggestions? Preferrably in the Newark/Farnell inventory.

Will 4 resistors do? Fig 1.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.