I have the Multiface with one of the original PCI cards (not PCIe) and it worked well for many years but recently has been playing up.
A while ago, immediately after loading the firmware it would sit in a state with the red “error” light flashing, but at a different speed from the firmware load, along with occasional flashes of the MIDI out light. Attemping to use it caused a “device or resource busy” error.
Now, with the most recent update to kernel 4.16.11-1-ARCH the hdspconf and hdspmixer tools will at least talk to it but it isn’t really usable. The red light still flahses occasionally, though not as fast as before, and co-incident with this hdspmixer shows all the meters peaking in the same was as when first initialising as if, perhaps, the driver is resetting something at that point.
Before this recent kernel upgrade I was wondering it had developed a fault but the fact it has suddenly got close to working suggests this may be a driver issue.
mhartzel is correct. Usually the caps go before anything else. However, as long as they are not surface mounted and you are handy with a solder iron, they are pretty simple to replace. Just make sure to keep the polarities correct.
I run a multiface II on arch for almost a decade with not one single dropout. But i have the same install problems everytime i set up a new distro with hdsploader not working, etc… sometimes it helped compiling the new alsa-firmware from source (on ubuntu) … but as mhartzel suggests this could be as well a hardware problem in your case…
Old hardware does die. It is often due to capacitors failing. They have a limited life span. Your card is not necessarily dead, you might want to test it in another computer. Another option is to try with an old version of AVLinux (= old kernel) if you can find one.
There should never been any reason to use hdsploader. The “firmware” is loaded by the kernel as the module is loaded. Note: it isn’t really “firmware” in the traditional sense, but it is configurable data that is required for the RME to function as intended. hdsploader was written at a time when the kernel couldn’t do this and/or there were often reasons to reload that data into the card.
Multiface II user here, and yes, works with latest 4.16 kernel. Haven’t used the old multiface in at last 8 years, Glad to hear they’re still around, seems to be solid engineering ;-D
You may want to check if your specific hardware requires an older firmware version (Google and the LAC archives may help)… The output of “lspci” might provide useful search strings. Other than that, no idea really.
Thanks for the replies everyone. I do have a second PCI card in another machine but that does seem to have been victim for the old capacitor problem in that there was a pop and a nasty smell shortly after turning it on. Interestingly that didn’t cause the PSU to shut down. I’ll see if I can set up a different machine with the other PCI card and see how it behaves there.
I’ve also just checked and mine is a Multiface II. Firmware-wise I notice there are two, different (as in cmp says they are different, not just different names) firmware files on my system: multiface_firmware.bin and multiface_firmware_rev11.bin and I don’t see anything in the logs to say which one is being loaded. I also note that these are part of a packaged called alsa-firmware 1.0.29-1 which was last updated in 2015. Looking on the ASLA website it looks like this has been dropped in favour of linux-firmware which looks like it is being maintained by the kernel team but doesn’t include the multiface firmware.
I did find these in the output of dmesg, though:
[ 2.516564] snd_hdsp 0000:05:00.0: wait for FIFO status <= 0 failed after 30 iterations
[ 2.516814] snd_hdsp 0000:05:00.0: Multiface found
[ 2.523136] snd_hdsp 0000:05:00.0: loading firmware
[ 6.036921] snd_hdsp 0000:05:00.0: wait for FIFO status <= 0 failed after 100 iterations
[ 6.036922] snd_hdsp 0000:05:00.0: enable_io fifo_wait failed
[ 6.036926] snd_hdsp 0000:05:00.0: couldn’t get firmware from userspace. try using hdsploader
[ 6.036927] snd_hdsp 0000:05:00.0: card initialization pending : waiting for firmware
This doesn’t look right. Apart from the FIFO wait messages why would it have trouble getting the firmware from userspace after having already reporting loading it successfully? Is it trying to load it twice and, if so, why?
So, with the 2nd PCI card in the 2nd machine the card finishes loading the firmware and the red light goes out and stays that way. The machines do obviously both have PCI and they both run Arch Linux but there are many differences. I picked this particular machine because it has been sat in a corner unused for a while and therefore has not had the software upgraded recently. Sure enough, it is running kernel version 4.7.6-1-ARCH #1 SMP PREEMPT compared to 4.16.11-1-ARCH #1 SMP PREEMPT on my current machine, the one with the issue. The firmware is the same as I had to copy the alsa-firmware package across and install it.
So, I am beginning to think this is a software issue, though it could still be that the first PCI card has gone faulty while the I/O box remains good.
Also, for reference the hdsp lines in dmesg on the older, working machine are:
[ 8.659841] snd_hdsp 0000:01:01.0: wait for FIFO status <= 0 failed after 30 iterations
[ 8.659951] snd_hdsp 0000:01:01.0: Multiface found
[ 8.670662] snd_hdsp 0000:01:01.0: loading firmware
[ 12.036741] snd_hdsp 0000:01:01.0: finished firmware loading
I tried reverting to kernel version 4.16.1-1 and that seemed to work briefly until I tried to write the output of dmesg to a file in a directory I didn’t have permission for whereupon the meters in hdspmixer went crazy. After that no other versions seemed to work and even going back to 4.16.1-1 no longer worked after trying versions in between that and 4.16.11-1.
At that point I tried swapping the PCI card and then it started working. Even now I can’t be sure it is actually faulty or whether the connector has become mis-seated or dirty - that would certainly account for things being a bit intermittent.
you could boot from a different live distro to see if the problem remains to rule out a software issue…? I ran AVLINUX for a long time with the multiface without having problems…
I ran AVLINUX for a long time with the multiface without having problems
The recent AVLinux 2018 with kernel 4.9.76 RT runs fine for me with Multiface 2. Just one data point for what it's worth.
Download from https://mirror.bytemark.co.uk/avlinux if you want to try it. It's a live boot from which you can install if you want.
I’m using, for I think 8 years, succesfully the multiface2 with pcie card. But since month or 2-3 I’m getting cracks and pops when I’m using hardware synths. Switching the FireWire cable didn’t help so I presume it’s faulty. It’s going for repair now because it’s still a great interface.