PC Flyer : Port A

- Brain2000 3/14/2002 11:00pm- Revised 3/28/2002 6:00pm- Revised 10/15/2002 1:18pm
NOTICE: MOTU has released new PC 32 bit drivers on 8/15/2002 for WinME/2000/XP version 1.0b1, and so far they seem to overcome most of the problems listed below. To download these drivers, goto http://www.motu.com, click DOWNLOADS, then select "USB MIDI Interface Software". The driver covers both USB and Parallel versions, however, the driver is not listed under the "Parallel MIDI Interface Software", so you will have to choose the USB link. Why they did not put the driver in both places is an oversight on Motu's part. After doing more extensive research, I will update this link with any major bugs that I have found. But this is a step in the right direction for MOTU!! 
Does your MOTU MIDI Parallel device keep coming up as a PC Flyer Device?Then this page may just be your fix for MOTU's new WDM drivers for Windows 2000/XP.
So, you're MOTU is sucking? Well, that shouldn't surprise you too much. However, since you're in the same boat as many people are already, here's the article that will hopefully get you up and running.
The first thing you should know is there are FOUR layers of bugs that surround this unit. Getting past these bugs is not that difficult, just a real pain in the ass. Let's begin.

BUG LAYER 1
The first bug deals with the MOTU MIDI unit itself and the parallel port. If you change your BIOS's parallel port settings, you must power cycle the MOTU or else it will not talk properly. In fact, after switching between SPP and EPP several times, I've gotten my MOTU into a state where it would not acknowledge anything coming down the parallel port.It's a good idea to run through the following steps to make sure that your MOTU unit is even going to talk in the first place.
Step 1: Unplug ALL MIDI cables going to your MOTU MIDI unit. This step is very important and will be discussed in bug layer 3.
Step 2: From the device manager, remove "MOTU Midi Device", "Motu Parallel MIDI Interface", and "Printer Port (LPT1)" IN THAT ORDER.
Step 3: Reboot and go into your computer's BIOS settings. Under "Integrated Peripherals", set the type to 378/IRQ7 EPP 1.9 (not ECP+EPP). If that won't work, 378/IRQ7 SPP (standard) is acceptable. Save the BIOS settings.
Step 4: While the computer is posting after the BIOS has been saved, power cycle your MOTU.
Step 5: As your computer goes from the BIOS Post screen to the first Windows 2000 boot screen, watch the LED's on the front of the MOTU, you should see a quick flash on one of them as the parallel port is detected. If you do not see this, chances are it will not work at all once you are booted all the way up. Maybe try power cycling your computer and MOTU together.
Step 6: Follow MOTU's instructions for installing their driver again. Set up LPT1's interrupt, then add/remove hardware, etc.. etc...

BUG LAYER 2
The second bug deals with MOTU's drivers. When you boot your system, your MOTU is actually detected twice. Once when the "MOTU Parallel MIDI Interface" driver loads, and then once when the "MOTU Start LPT" program that was added into your "Startup" group runs. The "MOTU Start LPT" is MOTU's failed attempt to fix the problem.These next steps will change your MOTU driver in such a way that your MOTU unit is only detected once, and the drivers will load, in my opinion, more appropriately.
Step 1: Download MOTUFIX.REG and put it into your registry by double clicking on it.
Step 2: Delete the "MOTU Start LPT" from you Start->Programs->Startup folder. As far as I can tell, this is MOTU's bandaid to attempt to make their drivers work.
Step 3: RIGHT click on "Start->Programs->MOTU->CONSOLE" (note: your CONSOLE will be named according to your MIDI unit, i.e. MTPAVCON = MTP AV Console), then click "Properties" from the menu that appears. Put a checkmark in the box labeled "Run in separate memory space". Click "OK".

BUG LAYER 3
The third bug deals with detecting the MOTU MIDI unit itself through the parallel port. You can force your MOTU to be redected at any time by running "START->Programs->MOTU->LPT Refresh" (make sure you installed this during MOTU's setup). Watch the LED lights blink when the unit is being detected. Notice the output LED rows light up? Do you think this is the unit initializing? Actually, it is not. It is a stream of garbage SYSEX data sent to all MIDI ports. This is very bizarre and I believe it is caused directly by the MOTU driver detection routine. Now, hook your MIDI cables back up, make sure all your MIDI instruments are powered on, and redetect again. This time you will notice the output LED's light up, and also possibly some of the input LED's. What's happening here is the SYSEX data is being echod back by your devices to the MOTU. In my setup, I found the Ensoniq TS12 (if SYSEX is turned ON) and the EMU Proteus 2000 will echo this data back.
When the MOTU unit initializes, this garbage SYSEX data is being echoed back to your MOTU and getting sent back through the parallel port causing the driver's detection routine to fail. When the MOTU driver fails to find anything it recognizes, it creates PC Flyer MIDI devices.
So how do you fix this? There's a couple of ways. Once you have determined which of your MIDI units, if any, are echoing the SYSEX data back, you can either figure out how to turn that unit's SYSEX off (some can't be turned off though). If that's not possible, you can go into MOTU's MIDI Console program and filter out SYSEX data on either the INPUT or the OUTPUT of that particular MIDI port. Once you break the garbage SYSEX echo loop, your MOTU will detect properly.

BUG LAYER 4
This is a nasty bug, and it exists in at least the MTPAV (1.2.0 firmware), and who knows what other MOTU MIDI devices as well.
When sending or receiving SYSEX data to and from the PC, it can and will corrupt the data. However, when routing SYSEX packets through MIDI 1-8, the SYSEX data seems to stay healthy.
It also appears that the MOTU MIDI device is checking each SYSEX data packet. In one of my tests, I was simply routing SYSEX data through it, not using the Parallel cable to the PC, and halted the SYSEX stream right in the middle of the send. The MIDI unit continued to route all data properly, but it would no longer send or receive ANY data (not just SYSEX) to and from the PC Parallel cable. I then routed a complete SYSEX packet through the MIDI unit and the PC Parallel cable was alive again. It is as if the firmware goes into a scan mode at the beginning of each SYSEX packet, thus pausing PC I/O, and continues to do so until the ending SYSEX data is found.
It doesn't bother me so much that it stops responding if you halt during the middle of a SYSEX send, because you normally do not do that. However, it does bother me that the SYSEX data is corrupted. EMagic's SoundDiver has a loopback test, and you'll find it stops almost every time on the SYSEX part of the test. Like I said though, you can ROUTE SYSEX through it as long as you don't use the PC. Currently I'm using my Layla audio card's MIDI ports routing through the MOTU MIDI in order to get data to and from the PC, and that seems to work just fine.
_________________________________________________________________________
That's it, you're done. Hopefully your MOTU unit won't give you much problems after this. If it does, you may also want to experiment with the BIOS Parallel port modes EPP 1.7 / EPP 1.9 / SPP (not ECP) and see which one works for you. Remember to POWER CYCLE the MOTU after you change the parallel port mode.


I hope this page has helped some poor musician out there who just can't take their equipment not working any longer...