tl,dr: Just looking for apro_sys_v115.zip? Here it is.
Update: On 2023-06-20, three years after I wrote this, the same error popped up again, but this time the recovery procedure doesn't seem to help. I've tried multiple times, but I think it's done for now.
My trusty old Roland A-500 PRO, which has been the center of my workstation, started acting up recently. It booted into reset mode and would say
When I pressed [Enter], it would force me into a sequence that reset the controller (Buttons [L9] and [L3]). After that it would sit there doe-eyed and say it's waiting for SMF.
Problem was that the SysEx was hard to find, and is no longer available on the Roland website. I tracked it down, and am saving it here for posterity. This should hopefully work on the A800 PRO as well since it must be the same firmware, although I haven't tried it.
Here's the solution, copied from a KVR thread:
First, download apro_sys_v115.zip (I've rehosted the file, contact me if the download isn't available).
- Switch off the power to the unit.
- At the underside of the keyboard, change dip-switch EXT A to blank and EXT B to an asterisk *
- Using the USB cable, connect the unit and the computer.
- Switch on the power to the unit.
- Fire up your DAW and load the downloaded p000001.mid SMF file into a midi track. In Cubase, it actually came in on 2 tracks.
- Set the midi track's MIDI Out Device to A-PRO (for both midi tracks in my case)
- If you A-Pro says [PRESS ENTER], Press ENTER!!
- If it says Erase? , Press [L9]. Sure? press [L3]
- Now it should say "Send SMF"
- Now you can press PLAY in your DAW and watch the update arrows and wait for it to Say Completed! It should say this once it gets past the end of the midi file.
- Save DAW session (in case you need to repeat this step), and Switch off the power to the unit.
- At the underside of the unit, change both the EXT A back to asterisk * .
- Power Up the Roland and it should boot all the way in now, press any of the function buttons (i had to do this step twice). Now it should work! Hopefully it's fixed!
Maybe this is a sign of failing EEPROM? Anyway, it works for now, I'll see how it plays out.