Yes, I'm sure this has been answered before, but I can't find any good answers by searching. Sorry...
I've got a Lynx AES16 digital sound card with balanced AES/EBU inputs and outputs, and need to convert them to and from optical and coax S/PDIF. Circuits for balanced AES/EBU to coax S/PDIF are abundant, but I haven't found anything for optical. Anyone have anything up their sleeves?
I've got a Lynx AES16 digital sound card with balanced AES/EBU inputs and outputs, and need to convert them to and from optical and coax S/PDIF. Circuits for balanced AES/EBU to coax S/PDIF are abundant, but I haven't found anything for optical. Anyone have anything up their sleeves?
Maybe not so simple. Putting aside the issue of professional and consumer flags and that some consumer gear flatly refuse to work with professional flags and the sample rate indicators on my old school Dacmagic won't light up with them, the output levels are different with AES/EBU at 2-7 Vpp and SPDIF at 0.5 - 0.6 Vpp. Not only is the output of an optical receiver ill-suited to driving such levels, it is also single ended. In your shoes, I would seriously consider an actively driven transformer based solution.
Input is fortunately very simple, and since the Lynx has output transformers I'm just adding a voltage divider on the coax S/PDIF outputs. The optical outputs can take the AES/EBU voltage directly.
Flags on the outputs may undoubtedly cause problems in certain situations, but I'll tackle that problem if it arises. Their main functions will be to interface with home made DACs if I ever need more channels.
Flags on the outputs may undoubtedly cause problems in certain situations, but I'll tackle that problem if it arises. Their main functions will be to interface with home made DACs if I ever need more channels.
Bill: In that case I might get away with terminating the return wire and run the live directly into coax for a 3.5V output. Would make for a cleaner signal path when running XLR, too. And I'd need to get the actual output voltage from Lynx to get correct resistor values for the voltage divider to keep S/PDIF coax voltage within specs.
Bottom line: Do as little as possible, and fix it if things blow up
rfbrw: How so? The Lynx can take both AES3 and S/PDIF code, and AES has an input voltage range of 0.2-7V to cater for both S/PDIF coax, optical and AES/EBU.
Bottom line: Do as little as possible, and fix it if things blow up
rfbrw: How so? The Lynx can take both AES3 and S/PDIF code, and AES has an input voltage range of 0.2-7V to cater for both S/PDIF coax, optical and AES/EBU.
Last edited:
To allay any fears, look up the datasheet for the CS8416 receiver chip, and check out Rane's website. They have quite a bit of info about terminations and other useful stuff. Any time you match impedances you attenuate the signal by 50%.
If you ever do need to buffer or raise the level of an SPDIF output you can do it easily with a flip-flop chip like the 74hc74. Easy as pie.
Best, Bill
If you ever do need to buffer or raise the level of an SPDIF output you can do it easily with a flip-flop chip like the 74hc74. Easy as pie.
Best, Bill
I've built plenty of these cards. I wouldn't call these "audiophile", but neither is AES/EBU (who the hell decided XLR was a connector worth using for >1mbps data, anyway?)
For SPDIF-AES, use a TORX177 receiver and feed a 74HCT86 XOR gate with it - usually I'll use all four gates to buffer the '177 output, two configured as inverting and two as non-inverting. Put a 100 ohm resistor on every gate output, tie the inverting/non-inverting lines together after the resistors, put a 1uF ceramic in series with one of them and feed an AES/EBU transformer with it. Output side of the transformer is your AES/EBU output.
For AES-SPDIF, feed the AES/EBU input into a transformer, terminate the secondary of the transformer with 110 ohms (I use a pair of 1% 221's) then feed the input into a fast (>=10mbps) RS422/RS485 receiver chip. Connect the receiver output to a TOTX177 input.
In each case power the thing off a 5V wall wart, or if you're lazy, put a USB connector on the card and power it from a USB port on your PC. And make sure you decouple everything well, especially the HC86 in the first design.
For SPDIF-AES, use a TORX177 receiver and feed a 74HCT86 XOR gate with it - usually I'll use all four gates to buffer the '177 output, two configured as inverting and two as non-inverting. Put a 100 ohm resistor on every gate output, tie the inverting/non-inverting lines together after the resistors, put a 1uF ceramic in series with one of them and feed an AES/EBU transformer with it. Output side of the transformer is your AES/EBU output.
For AES-SPDIF, feed the AES/EBU input into a transformer, terminate the secondary of the transformer with 110 ohms (I use a pair of 1% 221's) then feed the input into a fast (>=10mbps) RS422/RS485 receiver chip. Connect the receiver output to a TOTX177 input.
In each case power the thing off a 5V wall wart, or if you're lazy, put a USB connector on the card and power it from a USB port on your PC. And make sure you decouple everything well, especially the HC86 in the first design.
Behringer SRC2496
Try a Behringer SRC2496. Must have product for a cheap price.
.
Behringer SRC2496 Ultramatch Pro at zZounds
.
Try a Behringer SRC2496. Must have product for a cheap price.
.
Behringer SRC2496 Ultramatch Pro at zZounds
.
sendler: No way, I'm going all out on the DIY My main goal is to build a good 10-channel DAC, inputs and outputs are a less important afterthought that I'm not too quality concious about.
Bill: Do you have the schematics handy? A quick google search didn't give me much.
Just to clarify my situation at this point: I've already had the DAC boards made, which is simply a DIR9001 S/PDIF receiver feeding dual mono PCM1794A DACs with DATA, BCK, LRCK and master clock. That means any kind of reclocking has to happen in the S/PDIF domain, and that'll simply have to wait. The HT receiver I'm using at the moment just burned its line level outputs, so I gotta get this thing moving... And the only thing I'm missing to get that done at the moment are the S/PDIF inputs and outputs.
My main goal is to build a good 10-channel DAC, inputs and outputs are a less important afterthought that I'm not too quality concious about.Bill: Do you have the schematics handy? A quick google search didn't give me much.
Just to clarify my situation at this point: I've already had the DAC boards made, which is simply a DIR9001 S/PDIF receiver feeding dual mono PCM1794A DACs with DATA, BCK, LRCK and master clock. That means any kind of reclocking has to happen in the S/PDIF domain, and that'll simply have to wait. The HT receiver I'm using at the moment just burned its line level outputs, so I gotta get this thing moving... And the only thing I'm missing to get that done at the moment are the S/PDIF inputs and outputs.
who says you have to use XLR for AES? I agree any number of other connectors are more suitable (lemo, hirose, RJ45, HDMI, some molex etc), but that doesnt make AES unworthy IMO, sure i2s is superior for audio, but once its out of the case it loses ground (but you know that)
10 channel dac woah!! i'm going for 6-8 channels for an XO, you? whats the end goal for all those channels an XO as wel?
10 channel dac woah!! i'm going for 6-8 channels for an XO, you? whats the end goal for all those channels an XO as wel?
Last edited:
For the DAC I'm using a DB25 breakout cable, but I want to use standard connectors for other I/O.
I2S in its common application is only good for short in-case connections, as you point out, but PWatts did some very promising work on LVDS buffers over CAT5 cable. He did some measurements on an Audio Precision and had almost perfect transmission over 10-20 meters. But like I said, that'll be later, if ever. He was working on some FPGA-based power DACs (digital amps) too, but that sorta died out.
The 10 channels will be for three-way fronts (dual mono DACs, LC Audio ZAPfilter, balanced JoshuaTree relay attenuator), two-way center and two surround speakers (stereo DACs, simple opamp I/V, single ended JoshuaTree). AES outputs are mainly there for future upgrades like active XO surrounds, 7.1 speakers and the like.
I2S in its common application is only good for short in-case connections, as you point out, but PWatts did some very promising work on LVDS buffers over CAT5 cable. He did some measurements on an Audio Precision and had almost perfect transmission over 10-20 meters. But like I said, that'll be later, if ever. He was working on some FPGA-based power DACs (digital amps) too, but that sorta died out.
The 10 channels will be for three-way fronts (dual mono DACs, LC Audio ZAPfilter, balanced JoshuaTree relay attenuator), two-way center and two surround speakers (stereo DACs, simple opamp I/V, single ended JoshuaTree). AES outputs are mainly there for future upgrades like active XO surrounds, 7.1 speakers and the like.
Attachments
If all you need to do is feed transformer isolated AES3 to a DIR9001, all you need to do is terminate the line with a 110 ohm resistor, the DIR9001 can handle it.
If you haven't yet, go to rane.com and search for spdif, all the info you could ever need is there.
I'll see if I can find a link to using a D flip-flop for buffering.
Best, Bill
If you haven't yet, go to rane.com and search for spdif, all the info you could ever need is there.
I'll see if I can find a link to using a D flip-flop for buffering.
Best, Bill
Thanks, Bill. I've read the "Interfacing AES3 & S/PDIF" note on Rane, but your link is far more useful when it comes to optical interfacing. My brain must have had a case of bad input voltage itself when it thought I could feed a TORX177 straight into AES...
rfbrw: AES output voltage is 2.0 to 7.0 Vpp, but the minimum input voltage is 0.2 Vpp, as stated in the Rane note above. So AES inputs can take S/PDIF, but not necessarily the other way around.
rfbrw: AES output voltage is 2.0 to 7.0 Vpp, but the minimum input voltage is 0.2 Vpp, as stated in the Rane note above. So AES inputs can take S/PDIF, but not necessarily the other way around.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.