Clocking seems to be set wrong. What are your settings in the HDSP settings app and the front panel of the ADI-2? Do you have connections for Out and In between the ADI and HDSP?

Your reply does not answer the question completely, but it seems like you have both devices running as Master clock. Set the ADI-2 to sync to input instead. The flashing LED indicates clocking setup is wrong.

Hold on - I've made it, I guess!
Set ADI-2 to let the right upper lamp of the clock settings shine alone. It doesn't blink.
But I don't really understand it. I intended to use the ADI-2 as a master.
Doesn't that mean right now that the 9632 is the master (if synced to input it is in my emagination synced to 9632's clock)?

Yes Vincent, that's the case: I've connected both ins and outs adequately (if 9632 -> ADI-2 were connected wrong there would be no sound, so ADI-2 -> 9632 must be connected right, and both directions ARE connected).

I'm using coax at the moment, so it should be AES/EBU automatically, right?! There's no switch to choose from in that regard.
And with AES/EBU using the ADI-2 as a master should be advantageous (at least in theory).
But is ADI-2 the master right now? What would tell me it REALLY is if 9632 is set to AutoSync, but ADI-2 is set to sync to input?

Yes Jeff, did that.
Looks like it works now.

Why has ticking Professional in SPDIF out (9632) no noticable effect?
I thought it should be the same as in ADI-2.

And what about the AES/EBU function?
Is it automatically used by ADI-2?
If I tick it in 9632 the 9632 becomes master.

Okay, reread it.

How can I use AES/EBU with ADI-2 as a master then?
Can't find anything concerning that in the manual.

Only 9632 as a master seems to provide AES/EBU, as far as I can extract it.

I understand that AES/EBU is kind of a subcategory of SPDIF, providing additional information or capabilities.
In theory there should be an advantage using AES/EBU - in real world I don't know yet.
But I wanted to try it.

In the manuals (e.g. ADI-2: page 9) there is explained, as far as I understand it, how to use the cinch connections of the ADI-2 together with balanced output of the 9632. I conclude that even though the inlets of the ADI-2 look like being unbalanced cinch connections they can accept balanced AES/EBU from the 9632 if I use RME's digital brakeout cables (e.g. RO968), even though on the RO968 its cinch-outs are used: I expect the pin-connections of the 9-pole DIN plug of the RO968 to be set in such a way that a balanced AES/EBU signal can be sent from 9632 to ADI-2 using cinch out on RO968 (at least thats my understanding of the pin-usage explanation on page 9 of the ADI-2 manual).
That had been confirmed by Matthias in this forum (see http://www.rme-audio.de/forum/viewtopic.php?id=10925).

Otherwise I'd need an adapter XLR->cinch from 9632 to ADI-2 which I would have to build myself paying attention to the pin-usage explanations on page 9 of ADI-2's manual for sending AES/EBU from 9632 via RO968 (using its XLR line) to ADI-2 (using its cinch input).

I'm trying to understand it at the moment. If all that will be beneficial is another discussion that I don't mean to have here.

Thanks