Radio interference with SPDIF

Status
Not open for further replies.

Frank B

Senior Member
Today I worked on a little bug in SPDIF-in, and I noticed a problem with radio interference:

I had a Teensy 4 with a short program which uses SPDIF-OUTPUT (and a second with SPDI-IN - but that does not matter here)

If I connect a 10cm patch cable to any of the three SPDIF-Outputs, my DAB radio - 2 meters away, with a telescope antenna - stops receiving! This is reproducable.
I never noticed this with a Teensy 3 (but quite possible, that it happens with the T3, too)

How can I fix that?
 
I guess the patch cable act as antenna radiation some MHz clocks, that may interfere with DAB antenna frontend (most likely an active antenna).
(I have no idea on SPDIF clock rate, but is the effect sampling rate dependent?)
 
The clockrate is ~11.5 Mhz (with 44.1kHz) with a Manchester encoding. Isn't DAB around 120MHz?
I'll try other samplerates.

Yes, it is an antenna - and it is not needed to connect anything to the other end of the cable.

Maybe I discovered one of the reasons for the usual optical connections..
 
The clockrate is ~11.5 Mhz (with 44.1kHz) with a Manchester encoding. Isn't DAB around 120MHz?
I'll try other samplerates.

Yes, it is an antenna - and it is not needed to connect anything to the other end of the cable.

Maybe I discovered one of the reasons for the usual optical connections..

Now modulate 120 MHz with 11.5 MHz.
Also not sure if there is a IF (typically 10.7 MHz in double conversion radios) also in digital radios.
 
Ah, ok. Thanks! I thought it could have been the very sharp signal edges the T4 can produce.

I'll try on a Teensy 3, too.
 
A 10 Cm patch cable is probably going to look like like a pretty good antenna for the sharp edges a T4 can produce. The receiver front end may be being "desensitised" by strong out of band signal.

For some food for thought, here's a couple of links discussing SPDIF circuitry.

https://xse.com/leres/circuits/digitalaudio/
https://circuit-diagramz.com/coaxial-spdif-output/

Which highlight the need for some kind of buffer to drive the *theoretically* 75 Ohm load seen at the other end. Any of the choices of silicon is also gonna act as a slew limiter and blunt the T4's sharp edges as a bonus.

To date I've not seen any off-the-shelf gear that uses the pulse transformer providing isolation. Bean Counter Syndrome..
 
Ok, i've tested various things:
- My radio still works if I use SPDIF-Out on a Teensy 3.2
- Varying the PAD-Config (PAD_CTL) does not help.
 
I got it working by adding a 560pF ceramic capacitor (hat no other small values at hand) between SPDIF-OUT and GND. Maybe other values are better.

Correction: - nope, that' s unreliable.

I'll just try a shielded cable.
 
Last edited:
More food for thought.

Does using different T4 clock freqs change the radio's behaviour?


Here's another. http://www.hardwarebook.info/S/PDIF_output


WRT the Simple S/PDIF output without galvanic isolation example:-


Note the paralleled buffers. I don't think a T4 output has the current capability. If a 3.3v powered buffer is used, the 394 Ohm needs adjusting.


Maybe borrow the voltage divider. Let's call it R1, decrease 394 Ohms to suit 3.3v and play with some bypass C across the 93 Ohm (R2). Shortest path possible, R1 straight to the T4 pin and R2 to a solid ground and any bypass C to same place. Then think of coax.

In using a buffer, I'd think it wise to allow for some series decoupling R between the T4 SPDIF out pin and the buffer input.

Hope this is helpful.
 
Status
Not open for further replies.
Back
Top