PRO 2 dead after Sysex Lockup

Techbuzz · « **on:** August 06, 2018, 04:14:20 AM »

Has anyone experienced there Pro 2 lock up on a sysex file? Last night I sent a file to my Pro 2 and it just locked up. I power cycled the synth and now the display doesn't work and nothing works. Most of the buttons light up but if you press them nothing changes. Pressing any key does not output any midi information or sound. So I think the OS got corrupted.

I tried reloading the OS but that didn't seem to work at all since the screen isn't working. Does anyone know of a hardware reset button combination? I looked in the manual and couldn't find one. I have a request into the support dept. I just hope I don't have to send it back to get the FW flashed or something like that.

Thank you.

dslsynth · « **Reply #1 on:** August 06, 2018, 07:14:35 AM »

Send an email to DSI support immediately. They will help you out. There may or may not be a bootloader on Pro 2 and if there is one it may be possible to recover from such an update failure without sending the instrument in.

Techbuzz · « **Reply #2 on:** August 06, 2018, 07:58:47 AM »

Thank you dslsynth. I did. Will see what they have to say. A failed syex file for a patch shouldn't corrupt the software. Kind of disappointing.

Techbuzz · « **Reply #3 on:** August 06, 2018, 03:44:35 PM »

According to tech support, there is no way that sending a sysex patch file can do what I am claiming happened. Load sysex file, synth freezes up, power cycle to recover, then dead. They explained that the only way this can happen is a power cycle in the middle of a short period of time when the OS is being written that will brick the synth. Well, I didn't send an OS file, it was a patch file, the synth locked up and I power cycled it when it froze. I tried hitting global, write, compare or any button to escape from the freeze up. Didn't work. So I power cycled.

So now it has to be sent in to be repaired. Not too impressed.

dslsynth · « **Reply #4 on:** August 06, 2018, 04:53:52 PM »

Quote from: Techbuzz on August 06, 2018, 03:44:35 PM

So now it has to be sent in to be repaired. Not too impressed.

Ask them if they can arrange a board swap instead of sending in the instrument. That of cause only if you feel fine with opening the instrument and replacing a board.

And yes, its sounds a bit surprising. Do you happen to still have that file around that caused the problem? Could be interesting to study its contents.

Techbuzz · « **Reply #5 on:** August 06, 2018, 05:35:09 PM »

I am going to just send in the boards. I am fine with taking the thing apart. I do this kind of work for a living all day in an industrial environment. I was hoping they could point me to the FW tool or whatever is needed to restore the basic software on the synth to get back up and running. There is an Analog devices chip and a PIC controller inside. I am sure there is specific software for those two chips to restore them to factory and reload whatever FW it is for the instrument to work.

Opening it up there is a header that looks like a connection for a serial port to the PIC controller. My guess is it is for the PIC microcontroller as the traces lead to that chip. Looking up the datasheet it has a USB, Ethernet and CAN interface. So my guess is they will jack into this port to restore the thing. I doubt anything is physically damaged.

Attached is the exact file it choked on. While any patch file should not lock the synth up or cause it to fail, this one did. It is the correct length of bytes for a PRO 2 sysex file. The header and end are correct. If the data is wrong or a parameter value is wrong, it should dump the file and not freeze up. Maybe give a parameter value error or something like that. But it shouldn't lock up that you have to power cycle it.

Thank you for your time. It is appreciated. I sent tech support the same file and asked them to load it on a PRO 2 there to see if they get the same results. I hope they do. If so maybe they will find a fix. I wasn't doing anything other than loading this file. No other MIDI data was being sent to the synth at the time either. No clock information nothing. Just a patch. The previous sysex file worked like a champ. This one killed it.

dslsynth · « **Reply #6 on:** August 07, 2018, 04:25:36 AM »

Very happy you are so technically skilled. Very impressed that you could decipher the hardware so easily and check its features.

I have said this above but as it could be an advantage for you: turnaround time may be faster if they can offer you a board swap in which they send you a new working board and you send your old board back. Not sure if they always offer that but it have happened in the past.

Took a quick look at the sysex file. It does indeed look fine.

Techbuzz · « **Reply #7 on:** August 07, 2018, 04:34:01 AM »

Thank you for the look at the file. Tech support confirmed that the file will lock up the instrument. But they were able to power cycle and recover. The instrument is out of warranty so I have to send the boards in. Bummed but what I am I going to do.

Thank you again for your time.

dslsynth · « **Reply #8 on:** August 07, 2018, 04:41:34 AM »

Sounds like a plan. Its good to know that the patch file is part of the problem. Maybe there are timing or other tolerance type issues with the hardware? Also, DSI offers very reasonable repair prices so I would be surprised if this repair will cost you a fortune. Anyway, good luck with it!

ARNK · « **Reply #9 on:** August 09, 2018, 09:19:08 PM »

Quote from: Techbuzz on August 06, 2018, 05:35:09 PM

I am going to just send in the boards. I am fine with taking the thing apart. I do this kind of work for a living all day in an industrial environment. I was hoping they could point me to the FW tool or whatever is needed to restore the basic software on the synth to get back up and running. There is an Analog devices chip and a PIC controller inside. I am sure there is specific software for those two chips to restore them to factory and reload whatever FW it is for the instrument to work.

Opening it up there is a header that looks like a connection for a serial port to the PIC controller. My guess is it is for the PIC microcontroller as the traces lead to that chip. Looking up the datasheet it has a USB, Ethernet and CAN interface. So my guess is they will jack into this port to restore the thing. I doubt anything is physically damaged.

Attached is the exact file it choked on. While any patch file should not lock the synth up or cause it to fail, this one did. It is the correct length of bytes for a PRO 2 sysex file. The header and end are correct. If the data is wrong or a parameter value is wrong, it should dump the file and not freeze up. Maybe give a parameter value error or something like that. But it shouldn't lock up that you have to power cycle it.

Thank you for your time. It is appreciated. I sent tech support the same file and asked them to load it on a PRO 2 there to see if they get the same results. I hope they do. If so maybe they will find a fix. I wasn't doing anything other than loading this file. No other MIDI data was being sent to the synth at the time either. No clock information nothing. Just a patch. The previous sysex file worked like a champ. This one killed it.

That is impressive, do you work with electronic equipment such as synthesizers or can your work easily be applied to synthesizers?

Techbuzz · « **Reply #10 on:** August 10, 2018, 05:02:58 AM »

Quote from: ARNK on August 09, 2018, 09:19:08 PM

That is impressive, do you work with electronic equipment such as synthesizers or can your work easily be applied to synthesizers?

I work with industrial electronics for machines. I build HMI's and control programs for these machines. So for the PRO2, the interface screen, and all the digital controls, I think I could handle no problem. Thinking about what I was asking support means they would have had to expose the base code of the machine to use those software tools. They simply are not going to do that. So I have to wait until the boards return and try again. I just hope that trying again doesn't produce the same results.

I am trying to build a interface for the synth in m4l to integrate into Ableton. With version 10 you can output sysex. I was trying to take my parameters and when a version of the patch is put into a track, it sends a sysex file to the sysnth with the parameter values so that I am essentially storing the patches in Ableton. I do this with a Voyager, a Minitar and an X4 now. The difference is they can receive NRPN and CC in a batch. I did the same thing with the PRO2 and it didn't work. I sent all parameters minus the seq via NRPN and it just wouldn't work. I verified all of them one by one in MIDI monitor to be sure. My guess is it is too many parameters at one time and barfs. So, I was just going to wrap it up into a sysex message and send that on instantiation of the patch. Which should work and is basically what soundtower does. I was really close and then the PRO2 died.

I could, of course, buy the sound tower VST but I don't like the configuration to integrate into Ableton. So, since I make my own interfaces at work all day, and have done it for other synths, I made my own. I clearly screwed something up. If it is just a software issue I will be bummed. Meaning if they simply reload base FW and reload the OS I am gonna be bummed. If a capacitor or other hardware component died then I will understand and it just happened at the exact same time I was testing my sysex message. Which I happened to lock this up multiple times with NRPN messages and power cycled without a problem. First two sysex messages and it died.

Attached is a pic of what I made which works to control the synth and receive the control in return. I just needed to finish the send patch when the instance of it is put into a track or a project opens to send the information to the PRO2.

BobTheDog · « **Reply #11 on:** August 10, 2018, 12:38:38 PM »

I'm guessing that a particular malformed sysex message could put it into firmware update mode but also I am guessing it wouldn't flash the firmware unless it got valid data with a valid checksum.

I'm interested in hearing what the problem is when you find out.

Techbuzz · « **Reply #12 on:** August 10, 2018, 03:55:51 PM »

Quote from: BobTheDog on August 10, 2018, 12:38:38 PM

I'm guessing that a particular malformed sysex message could put it into firmware update mode but also I am guessing it wouldn't flash the firmware unless it got valid data with a valid checksum.

I'm interested in hearing what the problem is when you find out.

That could be possible. Looking at the headers of the sysex files I don't think I did that.

Headers from various sysex files

F0 01 2C 02 00 00 00 7F    The file that locked up my PRO2
F0 01 2C 02 00 00 00 18    The file I used to start from
F0 01 2C 79 03 51 18 40    The start of the FW update file
F0 01 29 75 01 12 16 40    The Mophox4 FW file

The 4th byte I think indicates that it is a FW file or not. This value isn't documented in the manual as I wouldn't expect it to be.

BobTheDog · « **Reply #13 on:** August 13, 2018, 07:35:04 AM »

Yep, they are pretty different

BobTheDog · « **Reply #14 on:** August 25, 2018, 11:15:33 PM »

Did you hear back about the fix yet?

Techbuzz · « **Reply #15 on:** August 26, 2018, 06:27:07 AM »

Quote from: BobTheDog on August 25, 2018, 11:15:33 PM

Did you hear back about the fix yet?

I heard back that reloading the software didn't work. I was told it is a hardware failure of some kind on the main board. I still have not received the boards back. When I get the full story I will post the outcome.

My guess right now for the corrupted sysex I generated is the last 3 bytes of data. If you look at a bank dump, all patches, the last 3 bytes are always 03 7F 7F. Why I am not sure but in my file, they were not. When I get the boards back I am going to try again. I will post more information when I can test again.

Techbuzz · « **Reply #16 on:** September 12, 2018, 09:55:41 AM »

Quote from: BobTheDog on August 25, 2018, 11:15:33 PM

Did you hear back about the fix yet?

Finally heard back. The PIC microcontroller died. The main board had to be sent back to the factory to be reworked to have it replaced. According to tech support the problem with my sysex file was an out of bounds error. Basically, I have a value that is larger than it is supposed to be for that parameter. They were kind enough to send me the sysex map.

dslsynth · « **Reply #17 on:** September 12, 2018, 03:43:08 PM »

Quote from: Techbuzz on September 12, 2018, 09:55:41 AM

Finally heard back. The PIC microcontroller died. The main board had to be sent back to the factory to be reworked to have it replaced.

Did you figure out how much you have to pay for the repair?

Techbuzz · « **Reply #18 on:** September 12, 2018, 03:52:59 PM »

Quote from: dslsynth on September 12, 2018, 03:43:08 PM

Quote from: Techbuzz on September 12, 2018, 09:55:41 AM
Finally heard back. The PIC microcontroller died. The main board had to be sent back to the factory to be reworked to have it replaced.

Did you figure out how much you have to pay for the repair?

The repair fee is a flat $25 for an instrument out of warranty. Not bad. The owner is responsible to pay for shipping and DSI pays return. It cost me $60 or so to ship just the boards. Overall not a bad cost to get fixed compared to a new instrument. I just hope it doesn’t happen again.

dslsynth · « **Reply #19 on:** September 12, 2018, 04:10:12 PM »

Quote from: Techbuzz on September 12, 2018, 03:52:59 PM

The repair fee is a flat $25 for an instrument out of warranty. Not bad. The owner is responsible to pay for shipping and DSI pays return. It cost me $60 or so to ship just the boards. Overall not a bad cost to get fixed compared to a new instrument. I just hope it doesn’t happen again.

Definitely not bad for an out of warranty repair. Fingers crossed for hardware failures not happening again.