1 (edited by Delta87 2017-03-19 03:19:36)

Topic: HDSPe AES not discovered in Windows 10

I've had my AES card for about a year now. It was working fine in my Intel 5820K/Asus-X99A computer setup. I purchased a new AMD Ryzen 1700x setup, and the first motherboard I tried was the Asus X370 Pro on Windows 10 Professional 64 bit. It did not recognize my AES card in Windows *at all*. There was no trace of it in Device manager or the sound control panel. Even after installing the latest driver. I RMA'd the motherboard, thinking it was faulty, and rebuilt the computer today using an ASRock Killer SLI/ac board. Same problem, no trace.

Pretty sure the card still works, it was running just a couple of days ago in my Intel setup. I also have a Babyface Pro that installs just fine (and did for either AMD motherboard).

RME support asked me to check the BIOS to make sure that any PCIe sleep functions were disabled but I didn't see anything like that in either BIOS.

The only strange thing that I noticed is that when I went to install the latest RME drivers, it would install extremely quickly, if not instantaneously, and then afterwards Windows would prompt me that "Maybe that didn't install quite right, would I like to install it using a compatibility mode?". Even if I click yes, it doesn't solve the issue.

I did poke through the menus in Device Manager and scan for hardware changes. No detection. If I choose to install a "legacy device" in the Device Manager I can select an RME HDSPe AIO from the Microsoft menu (which should have the same driver as the AES), but when I select that option, windows immediately goes to a Windows crash screen and tells me it will collect information on the problem and reboot. The file at the bottom of that crash screen seems to indicate a problem with something similar to hdsp.sys

Does anyone have any ideas, why not only one, but two motherboards cannot see my PCIe card? I had literally zero issues with this thing prior to the build.

Re: HDSPe AES not discovered in Windows 10

Have you tried different slots on the mobo?

Some of my 'stuff'. https://www.youtube.com/user/puutappi

3 (edited by ramses 2017-03-19 08:30:52)

Re: HDSPe AES not discovered in Windows 10

Also look at this thread:
https://www.forum.rme-audio.de/viewtopi … 82#p119582

So maybe try to disable fast startup.
https://www.howtogeek.com/243901/the-pr … rtup-mode/

Is it possible to make a parallel installation of Windows 7 whether the cards works there ?
With this test you could eventually find out, that the card works and that BIOS settings are ok.
So to say Win10 related topic.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: HDSPe AES not discovered in Windows 10

I haven't tried a new slot on this ASRock mobo yet. I did move it around on the Asus and it made no difference. I also just turned off Fast Startup and still no dice.

Re: HDSPe AES not discovered in Windows 10

So far I've had the two AM4 X370 motherboards and cannot get the card to work in the short slots. The card *does* work in the x16 long PCI-E slots which typically hold graphics cards.

Re: HDSPe AES not discovered in Windows 10

I tested various PCIe cards with an Asus Crosshair Hero VI, including some RME cards. At the moment x16 slots with PCIe lanes directly connected to the CPU are the most compatible to use on AM4 platforms. Usually this means the upper and middle slots, where you put the GPU in the upper slot and the RME card in the middle slot.

Since this seems to happen with different boards it's likely an issue only AMD can solve. A new AGESA (BIOS) is said to be published by AMD in 1-2 weeks, but it's mostly for better RAM compatibility and then the mainboard manufacturers have to implement it into their own BIOS variants first.

Re: HDSPe AES not discovered in Windows 10

I have tested an ASRock AB250 Pro 4 board privately (XI-MACHINES does not provide AMD systems at the moment), the HDSPe MADI FX works fine in an x1 slot. Actually any PCIe card so far works in any slot aside from a UAD-2 solo which seem to suffer the same issue as on socket 2011/2011-3 boards (and the nMP).
I have not tested an AES card though.