MidiSense Board Not recognised by midisense software

[humanpyramid]

Dear Ladyada,

I recently ordered two boards - an ADIO and the Sensor board. The ADIO works a treat and i've just set up an infrared sensor (one with a wider range up to 120cm) and i'm over the moon with it - ansolutley amazing/!!!!! But! the Sensor board for resistive sensors won't work...
The green power light comes on. The red/green led flashes green - it's sending midi. My UM EX 1 edirol midi to usb converter flashes to say it recieving midi. The audio setup on my mac book pro sends a test signal to say midi is being sent and received. But, the midisense software (which works great with the ADIO board states that it has failed to find any midisense boards when i try to connect to the sensor board. I've checked all my solder points - no cold joints etc. What could the problem be? Hope you can help. Ben

[adafruit]

hmm, i will check this out. odd that the adio board works but not the resistive one

[humanpyramid]

Strange isn't it that one works (ADIO) and one doesn't....
Could it be the R3 and R6 resistors? I found the build instructions for these are a little bit confusing.... Here's the resistors i thought the instructions were telling me to use:

R3 220 ohm
R6 220 ohm
R7 10k ( the instructions don't seem to say which one to use - i used the one i had left...)
R13 220 ohm

If these are wrong it should work if i put the right ones in!

Thanks for your help with this,

Benjamin Freeth

[humanpyramid]

Can you clarify which resistors should be used in the following positions:

R3, R6, R7, R13?

Thanks,

B

[michu]

Can you clarify which resistors should be used in the following positions:

R3, R6, R7, R13?

Thanks,

B

same error here! my debug log shows the same error as in this thread: http://forums.ladyada.net/viewtopic.php?t=5116, but the oszi is right, 16Mhz.

R13=1k
R10=10k

this information is from the build instruction side. Also noted there:

>Note: there was a PCB error in some very early MIDIsense resistive
>boards, R3 and R6 may be swapped! Check that the 220 ohm resistor
>connects to the MIDI jack and the ~1.5K resistor connects to the bi-color
>LED.

What early boards? I just mounted as instructed (i did not switch r3 and r6)
R3=220Ohm
I also noticed, R3 and R6 are swapped in those 2 pics!

[michu]

ok i swapped R3 and R6 - still the same error. I looks like midisense dont receieve data (red blink), even if th midi controller sends data, according to its led. Ill guess i will have a look tomorrow with my voltmeter...

some quick information:
-usb drivers are ok, tested with midiox
-midi cables are not reversed
-soldering should be fine, as i double checked it
-green led is blinking, red led is only blinking when using midiox

[michu]

Ok i read the instructions again:

Check that the 220 ohm resistor connects to the MIDI jack and the ~1.5K
resistor connects to the bi-color LED.

so - in other words
red-red-brown 220Ohm - connect to midi jack - r3
brown-black-red 1kOhm - connect to bi-color led - r6 connects to led

so i switched back the resistors...

then
-Make sure the amtel sits tight on the socket (mine wasn't really tight)
-i set dip-switch 1 and 2 to on
-there are no drivers from the manufactor (midi sonic 2x2, infrasonic)

Using Win XP SP2, I could connect to the device! I was unable to connect using Windows Vista SP1 (I also tried to "runas administrator")!

There were some errors in the connection:

Code: Select all

00000043   19.19594955   [6476] 22:48:09: adding checkboxes    
00000044   19.19656754   [6476] 22:48:09: adding names    
00000045   19.21066666   [6476] 22:48:09: adding typechoice    
00000046   19.25665092   [6476] 22:48:09: adding new panel: 13    
00000047   19.37583733   [6476] 22:48:09: MIDI error: PortMidi: 'Buffer overflow'    
00000048   20.18521881   [6476] 22:48:10: Done    
00000049   20.18528557   [6476] 22:48:10: Found 1 boards    
00000050   20.18531990   [6476] 22:48:10: Configuring board #13    
00000051   20.21488380   [6476] 22:48:10: Sending Packet: 0xf0 0x7d 0x20 0xd 0x0 0xf7     
00000052   21.23618507   [6476] 22:48:11: Sending Packet: 0xf0 0x7d 0x20 0xd 0x0 0xf7     
00000053   21.66623878   [6476] 22:48:11: ***************Recv'd bad checksum: 16 sum = 86    
00000054   21.66629791   [6476] 22:48:11: Recv'd Sysex: 0xf0 0x69 0x21 0xd 0x3f 0x10 0xf7     
00000055   21.66631126   [6476] 22:48:11: enabling sensors 0x3f    
00000056   21.67483521   [6476] 22:48:11: stop    
00000057   21.67488861   [6476] 22:48:11: Started Listening    
00000058   21.67504120   [6476] 22:48:11: Sensor 0    
00000059   21.68149757   [6476] 22:48:11: Sending Packet: 0xf0 0x7d 0x32 0xd 0x0 0x0 0xf7     
00000060   22.01914406   [6476] 22:48:12: Sending Packet: 0xf0 0x7d 0x10 0xd 0x1 0x0 0xf7     
00000061   22.44915390   [6476] 22:48:12: MIDI error: PortMidi: 'Buffer overflow'    
00000062   23.35701180   [6476] 22:48:13: Sending Packet: 0xf0 0x7d 0x10 0xd 0x1 0x0 0xf7     


I started to verify the firmware... and canceled this procedure... NOW my midisense board act's "funny", when i power it on, the led flashes once red (guess this means that the bootloader should be fine), but the led does not flash green anymore! I guess when the bootloader works, it should be possible to upload a new firmware... or not?

[michu]

OK it looks like I was right about updating the firmware...

But after erasing the eeprom, midisense crashed:

Code: Select all

...
[7016] 23:24:28: Sending Packet: 0xf0 0x7d 0x44 0x1b 0xf7 
[7016] 23:24:28: sent
[7016] 23:24:28: bootloader, skip checksum
[7016] 23:24:28: Recv'd Sysex: 0xf0 0x7d 0x33 0x9f 
[7016] 23:24:28: Got reply, 0 bytes
[7016] 23:24:28: Sending Packet: 0xf0 0x7d 0x44 0x1b 0xf7 
[7016] 23:24:28: sent
[7016] 23:24:30: timeout
[7016] 23:24:30: Sending Packet: 0xf0 0x7d 0x44 0x1b 0xf7 
[7016] 23:24:30: sent
[7016] 23:24:30: bootloader, skip checksum
[7016] 23:24:30: Recv'd Sysex: 0xf0 0x7d 0x33 0x9f 
[7016] 23:24:30: Got reply, 0 bytes
[7016] 23:24:30: look for reply
[7016] 23:24:30: eek!


ARGH.. any hints?


EDIT:
after erasing flash, midisense lost connection to the board. after i confirmed this message, midisense tries to program addres #0 - and crashes. the same error occurs trying to verify the fw. this looks like a bug of midisense...

I also noticed that the flash process needs allmost 100% of my cpu!

Ahh, I guess a warning message should be displayed, that verify firmware should not be interrupted, or the fw could get damaged!

[michu]

yes, another message, ill talk to myself ... i guess the problem is this f3$!@g usb-to-midi adapter, which comes without driver and manual, thats why the performance is that bad (>90% cpu use of midisense).. just and idea. anyone tried allready this adapter: http://www.infra-sonic.com/page/midisonic.html ?

[michu]

ok, i stopped all cpu intensive tasks (wlan services, virus scanners etc..) and then I retried it. Uploading the firmware is quite timing relevant, if i started midisense too early, i received alot timeouts, after some tries, the firmware upload went fine. after uploading the firmware successfully, midisense tries to verify the firmware. at verifying adress 64 (or 128) midisense just freeze, no error is displayed in gui or in debugview... any hints here?

[michu]

Ok I just realized, that vista is not supported due the portmidi lib. we use quite an old version here (15 nov 04), check the changelog:
http://portmedia.svn.sourceforge.net/viewvc/portmedia/portmidi/trunk/CHANGELOG.txt?view=markup

So i created an up-to date version of midisense. I also created a version, which does not verify the firmware and a version with an larger timeout (6s).

Ill test this this evening... cheers