828 MKII USB,pops,squeals WORKAROUND-READ ENTIRE THREAD!

[sampsa]

I bought a second hand 828 MK2 USB2 and while it "worked" as in sound came out of it, it sounded really distorted. I think the distortion was due to computer/usb not being able to feed enough data to the bus so buffer updates were missed. It sounded kind of grainy as if it was reusing previous buffer. YOu might know the sound I'm talking about. I'm running on a ASUS M51Sn laptop with Vista and XP.

I was running on Vista, so first suspect was the OS. I tried it all. DPC checker, RATTV3 analyzing, KernView analyzing but I couldn't get to bottom of it. When ever 828 was active (audio being fed to it) DPC latency was high as 4000us. Without the interface present it was hovering around 250us (which is pretty high as it is). The sound came out distorted.

Then I tried XP (I have bootable USB HDD with XP installed on it, which is _really_ handy for analyzing the computer), but no avail. It might be psychological but the distortion sounded a bit "milder" but still very present. DPC latency was a lot lower, around 20-30us. With 828 active it was at 400us, still a lot lower than with Vista.

I had gave up hope when I found this thread. Any tricks in it didn't help, and I skipped the old drivers as I thought "that's not going to help me". New drivers should be better, right? WRONG. After removing the drivers and installing the one thehowler gave, everything worked perfectly. No distortion whatsoever. I still need to do some stress testing but I'm really happy now. I got the confirmation that it's not my rig, not 828 itself but the drivers! Thanks very much!

So it seems that MOTU f*cked up something in the process of making new drivers.

Where could I find all the different driver versions? 3.6.7.0 doesn't have vista support, does it?

Also where could I get the latest firmware? I have 1.01 but there is certainly new one? I can't register the 828 on MOTU's site as the serial number is unreadable on the bottom with a lot of scratches on it.

I am going to stress test it over the weekend, but feel free to try this for yourselves.

The driver I used it the FireWire/USB driver version 3.6.7.0.

You can get this from http://downloads.motu.com/firewire/MOTU ... .6.7.0.zip

Try it out and see if you have any success.

[sampsa]

Where could I find all the different driver versions? 3.6.7.0 doesn't have vista support, does it?

Nevermind, I already found the sticky post for the drivers Will test them out later.

[abramco]

I had same problems with 828Mk2 on a laptop ASUS X5DIJ.

It come with Windows 7 64 bit factory preinstaled.
So, I needed to create a bootable instalation CD for XP 32bit including SP3 and SATA driver on it.

After XP SP3 was running perfectly, I instaled last MOTU drivers version 3.7.37320, and all headaches (pops,squeals,crackle,sound loose...)

I have tryed all posible solutions appeared in this thread, but nothing work. Also MOTU tech support recomendations, and nothing work.

So, at last I tryed to install older drivers 3.6.8.1410 (I have these working fine on my desktop PC), but instalation program jump me out with message "you need SP2 or older servipack version" even when I have SP3 instaled.

So, I have copied files "motubus.sys" and "motufwa.sys" from my desktop windows/system32/drivers folder(drivers version 3.6.8.1410) to my laptop windows/system32/drivers folder (overwriting files of drivers version 3.7.37320), and...
OK...now,every thing working perfectly on laptop...NO NOISE,NO DISTORTION, NO CRACKLES.
Even reactivating onboard soundcard (VIA chip), network adapters, usb camera and ADAT optical I/O (previously, MOTU tech support recomended me deactivate them).... every thing keep working fine.

So, I think there is a clear drivers problem in last version 3.7.37320

SO HAPPY!!! I can't beleavit!

[xbass247]

just like to say sorry ,motu ,i still love you ,and it was my flaut why the 64bit drivers would not work ,as it says on motu page look for the right drivers for u , so if you haveing promblems wiv your 828mk2 usb or firewire and ya just moved too 64bit the right drivers are here on this link http://www.motu.com/download/download_m ... duct_id=14" onclick="window.open(this.href);return false;

[sampsa]

Just wanted to give a heads up that the latest drivers, 3.7.39083, seem to work flawlessy on my setup. I'm using Windows Vista 32-bit. With previous drivers I had the distorted signal which is now gone. Also the spurious crashes are now gone. The device still loses connection sometimes when working with sequencer. Might be some issue with iLok.

I still need to test if the drivers can survive sleep mode and hibernation which the older drivers couldn't do, effectively crashing the machine. Also I need to try out the latencies if it's possible to go lower than 512 samples.

I really hope these drivers are now stable. I don't want to see motufwa.sys ever again in the BSOD! I got one hard drive messed up due to the constant crashing caused by the previous drivers. Probably the OS was updating the NTFS lookup table while the thing crashed, leaving nothing but mess behind.

EDIT: Since yesterday I've had one BSOD with motufwa.sys and multiple audio drop-outs. It feels a bit better than the previous one but a crash is a crash. Thinking about getting a new interface...

[jalderwood]

QUESTION. Does anyone know if there are PC systems running MOTU 828MKii which do not exhibit these bugs?

This forum has some real life to it. May those trying to decide what's better for audio, Mac or PC, and many future owners of MOTU gear happen upon this place in their search for reviews. Here's my .02

I do not ever get BSODs. I do not observe any relationship between graphics activity and audio glitches, though I have thought many times that there's a correlation with disk activity. I'm sold on the theory that HAL is part of the problem but I'm not willing to reinstall at the moment. Indeed it seems the 828MKii is more susceptible to Microsoft's drivers than most other hardware. However I also have an M-Audio Ozonic on hand. It doesn't make the same glitches but suffice it to say that I have encountered similar symptoms with that piece while seeking refuge from the 828. In fact those anomalies are what made me realize the problem was not just with the 828.

When used as a standalone mixer my MOTU 828 MK2 works fine next to my custom built PC, which for the record is:

EVGA 256-P2-N765-AR GeForce 8600GTS 256MB 128-bit GDDR3 PCI Express x16 HDCP Ready SLI Supported Video Card
SAMSUNG SpinPoint T Series HD403LJ 400GB 7200 RPM SATA 3.0Gb/s Hard Drive
GIGABYTE GA-P35-DQ6 LGA 775 Intel P35 ATX Intel Motherboard
Intel Core 2 Duo E6750 Conroe 2.66GHz LGA 775 Processor Model BX80557E6750
OCZ Reaper HPC Edition 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory

But when I run the 828 as an audio interface with Ableton I get intermittent noise and glitch with matching front panel chaos, and the whole IO bus on the PC seems to get interrupted. Even the lights on my midiexpress blank and MIDI halts... all with this cacophany of squeals, crackling, ring mod sounding distortion, and halting audio. It is maddening, even more so because it doesn't happen all the time. Sometimes I can get a weekend of work out of it, other times I have to find something else to do.


Troubleshooting (FAIL)

I've been on the phone with MOTU support. Verified latest firmware 1.0 and driver (3.7.39083), Factory reset. Have tried the Optical In/Out on/off voodoo they suggested, which sometimes seems to make things work for a bit. An observation: When the interface is acting up you can see the optical in/out settings greying out in sync with the glitch.

Have tried thehowler's suggestion to roll the driver back to 3.6.7.0.

Did not attempt thehowler's ACPI mod.

Thinking it might be an electrical issue because the problem seems to get worse in the summertime, I added a Furman PL-8 Series II power conditioner to my rack. Initially encouraging but turned out not to help.

While troubleshooting a likely unrelated USB device conflict (system would not recognize AKAI MPK25), I removed Logitech's buggy USB bandwidth regulator LVUSBSTA.SYS. (Apparently gets installed with most of their stuff.)

MOTU recommends that your firewire controller use a TI or LSI chip. Though my Gigabyte GA-P35-DQ6 motherboard has a TI chip, I went ahead and tried adding a Lacie firewire PCI card (TI chip). More glitch ensued. The fact that MOTU recommends this, yet we know the problem exists with firewire as well as USB, suggests there are other firewire issues with the 828MKii.


Sometimes I think this thing has moody gnomes in it. When I need to work on a project it usually shows up and doesn't give me any trouble for a day or two. Other times it just will not cooperate, perhaps for days. As you try to get your studio working, you go back and forth with ideas as to what's wrong and how to fix it. I blame MOTU and M$ for hours and hours of time spent like a nerd trying to catch this ghost, and countless nights of frustration when all I want to do is work. But neither company will take responsibility and just say it's broken and we're not fixing it.

Now, you can either buy a Mac and have this thing work with it, or buy a new soundcard. I'm leaning toward option #1.



p.s. For those 828MKii owners who find themselves in the unenviable position of not being ready to make your next purchase, I offer this small and hopefully obvious bit of solace. For regular glitch free listening you can use your motherboard's built-in sound card, connecting it to one of the 828's many inputs. Glitching only happens when the firewire (or USB) bus is open, like when you're using it in your DAW or if it's set as your primary Windows soundcard. Other than that the 828 stands alone as a mixer. Who knows, you may just luck out when you get that creative spark and try to run your sequencer.

[mcoyles]

QUESTION. Does anyone know if there are PC systems running MOTU 828MKii which do not exhibit these bugs?

*waves*

I run (frequently) a variety of combinations of 828MkIIUSB, 828MkIIFirewire (with a dead backlight), 2x 8PREs and an old 896MkII simultaneously on a Dell Inspiron 6400 laptop, on Windows 7 Pro (32bit), using latest firmware on all devices, and latest drivers. Since moving to Windows 7 I haven't had any issues at all. Same setup also works on 2x Sony Vaio laptops (from radically different eras, one running XPSP3, the other running Vista SP1). To summarise, the gear works fine on more or less every Intel-based laptop I sling it at. Never tried with an AMD laptop.

I have a Quadcore rig at work, self built, Asus P5Q Pro mainboard, low end nVidia card, stacks of ram, stacks of drives, huge psu... running Windows XPSP3. This rig gives me all the grief you guys are experiencing... as does my iMac (Intel Core2Duo), and my PowerMac (Dual 1.8Ghz PPC) - both on OSX 10.5.7, allbeit no BSODs on the Macs obviously, but still all the glitches and dropouts. Not as MANY as the PC, but they still happen. Note, the issues happen with ANY of the devices hooked up on their own... not just the 828MkII-USB... but admittedly the 828MkII-USB does do it more frequently than any of the firewire devices.

I only ever run this lot at 48Khz... never tried any of them higher than that...

[rhythm_kitchen]

Been following this thread. I too have been using built in drivers for my ASUS quadcore...ATI drivers did work. I installed an Intel Chip update and suddenly my dual graphics card became a Radeon... still ATI drivers... weird but not to the point.

USB is a hub. You will notice dual hubs, the Faraday Shielded hubs on the rear and for me, 6 empty sockets (antennas) on the MOBO. EMI poisons the whole hub, USB is like ethernet. Voltages change the link from 1.1 to 2.0 etc (Isosynchronous Mode non CRC USB-2 is the fastest).

Disconnect all those crappy USB-2 cables from the front of your Chinese made cases. I have yet to turn off the unused 6 port hub on MOBO in device mgr as I have other problems and use a PCI audio card. But my convenience port cable is unshielded at the last 3/4 inch and the data pairs untwisted.

In 2000 I informed Digi (having tested their new USB 1.1 box and the live acoustic quartet recording failed- RF spikes -- ruined many takes... this on an icebook - their new (original) mBox. Frustrated, being a Cisco trained IT guy, I showed it to a Telecomm engineer who took one look at the cable and laughed... you need dual ferrite beads on this cable, he told me, on either end. We have been required, a QTL 9000 rated supplier, had been for over a year making these cables, again... REQUIRED ON ALL PURCHASE ORDERS from the big Telecomms. So a word to the wise. USB-2 is ok... but the Chinese case makers have yet to have pressure put on them by the manufacturers in Audio... Telecomm yes. Audio no. http://www.kimber.com" onclick="window.open(this.href);return false; (USB) http://www.granite.com" onclick="window.open(this.href);return false; (1394) options for firewire.

You will see in device manager the two 6 port hubs (at least I do). The good ones, Faraday Caged on the MOBO rear, and those with antennas sticking up unshielded sitting on the MOBO, some have less. My board has 6. EMI can not only poison the data aerials but power as well.

Also beware water pipes may now have RF on them, or ultrasound, we have no idea how these offsite read meters work. Many of us increased our ground years ago by using the copper to add a ground clip or two. Disconnect those. Hire an electrician for a panel as close to your gear as possible with its own ground stake, water it, and keep the ground pigtail (in plastic boxes) as short as humanly possible (this is from Cisco Research, the data on their VPN and I no longer have access). I scoped out my water meter with a good snifer but only in the lower mghz range. We didn't get a signal. Now 'they' claim a truck drives by once every two months and sends an IR pulse to activate the chip, however I sprinkled some trees for over 30 hours and the company arrived immediately thinking the meter was leaking. So this is what the installer told me, the meters could be read by satellite (I take this to mean offsite as he was just a plumber) to be more exact but who knows? I wish I had a sniffer with a higher frequency range... these are extremely hard to come by these days. Who knows could be ultrasonics?

Oh... one other thing. Grounds are now commonly shared with cable and gas pipes. If the neutral goes out on the pole, or your home, the sheild on the tv cable melts, happening all over. I saw this with my own eyes. Insane... now with homeplug alliance or a neighbor with ethernet over AC who knows what is on your ground or neutral? So with all of this, consider good PSU units, try an isolated UPS backup -- i.e. the battery type isolated from AC, synthetic AC but at least isolated from Home Plug Alliance, which so many radio engineers attempted to veto but Bush II and Colin Powell's son (both experts in the field I note with some apprehension) made the decision... not one engineer consulted, they all petitioned against it. So have a little faith in the drivers and look elsewhere, this is my advise. Yes I have advised Motu as Avid 'pleased' with the fact their mBox was unprofessional. Heck they could sell more TDM... I do not blame this on the old team in Menlo Park, but the suits who overule engineering. I will not say more as I deplore this type of marketing strategy and believe it brings only bad karma.

[rhythm_kitchen]

Edit: Notice that Native-Instruments and RME now supply a USB type A to B with a ferrite bead on each end. This is closer to Telecom specs (RBOCS)... Telecoms required back in 2000. So the HD audio industry is 12 years behind.

Telecoms also demand more EMF leak testing on the connector. These are done in Faraday Tents on a lab bench.

Also now most hi-def audio hardware designs are moving the power supply to a separate Faraday Shielded unit. Like Apple's original USB or Firewire connections at the rear of New World G-3s, G-4s, G-5s & MacPro. All are Faraday Shielded. The front convenience ports are the weak link here particular with non-Apple hardware. This may have caused the peculiar noise in the first G-5 Tower with Firewire interfaces.

Granitedigital also sells 1394 cables with dual ferrite beads. Kimber makes & sells USB type A or B with similar ferrite beads on either end.

---> Try plugging into rear of PC. Removed USB cable at MOBO to front of PC case. Or some have been forced to buy a USB PCI or PCIe card. That is on an isolated USB hub. So if EMF poisons the MOBO hub, try this fix first. And if still a no go, a USB PCI card.

Hire an electrician to test for proper grounds, particularly at panels and even if paint or corrosion on the copper stake is preventing conductivity.

Studios in San Diego often had a designated beer drinking spot (or rental return area at the ground stake... ). But there is actually an electrolyte liquid used by radio broadcast engineers but I am not sure of the name. It most likely is banned by the EPA now.

[Th3_uN1Qu3]

Let's make something clear - the OP's issue wasn't EMI or RFI or anything like that. It was DPC latency spikes caused by the power management features of his nVidia card. On laptops for example you can't just change it... The real solution is to just disable PowerMizer and you are good to go.

This issue has been around for years yet nVidia still refuses to address it or tell users what to do when it happens.

I have the same issue (pops and squeals) with a Traveler Mk2 i have in for repair, but it's not due to the nVidia card.

This unit had the classic problem of firewire controller failure after cable plugged in reverse. After replacing the firewire controller IC it works fine standalone, but will sometimes choke on playback from my laptop. Fortunately, if you don't send any audio to it, controlling it and recording from it via firewire works just fine. There is another chip that i believe is responsible for this failure - it is the PDI1394L40BE which is a "1394 enhanced link layer AV controller". Why they would use a video capable chip in an audio interface, beats me. Touching that chip in a certain area (the package, not the pins! touching the pins actually doesn't do anything) triggers the dropouts, and on occasion it will totally die and not be recognized by the computer anymore until it is power cycled.

This leads me to believe that there is a non-critical internal failure in that chip, maybe damaged bond wire. Unfortunately, i cannot find that chip for sale anywhere.

[jalderwood]

Wow this thread just took a slightly conspiratorial turn into issues of EM/RF interference via the Homeplug Alliance. That certainly is a heavy consortium. Given our utter lack of success in solving this problem, I was inclined to go get some Gatorade and pour it in the ground around my house.

So I got a virus last weekend and set about rebuilding my system. I came back here to see what was new. Th3_uN1Qu3's mention of DPC latency got me over to another forum where there was a long discussion of it.

I went ahead and reinstalled XP, making sure not to install the onboard HD audio device. I didn't have a single dropout for a couple of days, then I let Windows update to SP3. Just as I was telling my friend how I fixed the damn thing the first squeal came out. That really got me going!

It's safe to say that Gigabyte mobos have had, and perhaps do have, DPC issues but it's not that simple. It happens on Macs and PCs and cannot be pinned to specific devices, drivers, brands or anything. I'm sure with enough research you could find some consistency but I have not.

The best I can figure is that lax quality control in the whole PC ecosystem led to these problems with the particularly demanding application of realtime audio. Manufacturers have been petitioned for years on this issue. Blame has been well spread and it looks like the companies have been listening.

Once again I'm down to super low DPC latency and have not seen a spike for a couple of hours. (They happen frequently or not at all.) All I did was install the latest motherboard drivers, update to the latest bios, and update the video card driver. I'm also not running Azalia but there is a driver update for that too and I might try it. Will update if anything changes.

[Vincentcnbres]

Just for kicks, I have an open TechLink with MOTU to see their advice - I'll post their response(s) in due course...

[vrangsinn]

questions to see what solution is possible

1. Do you have a computer with AMD processor with win Vista or Win7?
Amd is •••• - I tried using USB card on several computers with amd processors with failure every single time- Different brands of cards - for several years
On Pentium I never had such problems

2. Do you use adat in and out and if so - did you remember to put the sync settings right and to put your external gear as master?

[vrangsinn]

If you do some research you will find that line-6 cards, m-audio cards and just about any brand of USB card will have these "craccles and popples" when running on an AMD machine - And if you got ATI you are even in more trouble

I think maybe it is time hardware manufacturers start stating on their products:

WARNING - NOT COMPATIBLE WITH AMD

[torrentg]

If you do some research you will find that line-6 cards, m-audio cards and just about any brand of USB card will have these "craccles and popples" when running on an AMD machine - And if you got ATI you are even in more trouble

I think maybe it is time hardware manufacturers start stating on their products:

WARNING - NOT COMPATIBLE WITH AMD

This is complete nonsense. It all comes down to user-induced error in almost all cases of things not working well. This includes old outdated drivers being present on the system or even poor choice of antivirus, especially the suite type packages.