Multiple AOUT_NGs

[Zam]

Hi Thorsten

tks for info, then I will put 5631 in the basket and try them.

Best

Zam

[Highcooley]

Hi Thorsten

Sorry for warming up this thread, but I've got trouble with my fourth AOUT_NG module again. After a break which unfortunately took me a couple of months, I started working on my synth again. I soon noticed, that something is amiss with the fourth board. The outputs don't react at all. To test correctly, I used a config file with a reset command and this single line "AOUT  type=AOUT_NG  cs=1  num_channels=32". Testing aouts with the caliaout command, I was able to test different outputs from 1-24. Everything worked except 25-32 on last board. Since I am pretty sure that they once did work, I switched the fourth board with a spare one, but nothing. So, I started interchanging boards as well as cables to rule out defective hardware. All boards and cables in any position work except for the fourth board in the chain.

I did test with NG V1.033 pre16, V1.033 pre10 and V1.035

Do you have any idea, what could be wrong in the software?

 

Cheers

Andy

[Highcooley]

In case anybody else has a similar problem, here the solution: I figured out that whenever I had mi DSO attached to the FC line, the fourth AOUT_NG module worked fine most of the time. If I detached the probe after the initialisation phase of the midibox, the chip stopped working until I reattached it. This smelled like an AC termination issue, or better to say the lack of it. Since an AINSER64 in series with four AOUT_NG modules results in quite a long total cable length, these problems can occur. Basically, the signals can get reflected at the end of the line causing timing problems between the clock, FC and signal line. My solution was to terminate all three lines with 100ohms and 100pF to ground like the picutre below. Now everything works a treat again.