MB-6582 debug

[Berg]

Gents

Well i'm switching to a debug phase on my MB-6582 build. This is a kit from smash Rev2 pcb. All components are pluged in and soldered. CS assembled. PSU for 6582 chips. Ran a 5vdc wallwart and a 9vac wallwart into a 7pin din plug. PSU tests out. J3=5.12vdc, J25=10.34vdc, J4(9v peg)=8.97vdc, J1_sid 9v=8.97vdc, pin 28 on sids also =8.97 when jumper is in place. Pin 9 on lcd equals 5.12vdc. Pics from smash with Mios V1.9F. J11 jumper not installed yet, R80 is installed. I am a bit stumped.

When turned on with all ic chips the following occurs: Inital all led's light for split second, lcd redbacklight lights for a moment, crystals on lcd do a funky dance.

I am a bit lost to what may be next for testing.

Is this normal in a J4 & J3 5vdc - grd continuity check: When i initially touch the testing leads to the 5vdc and grd there is a brief beep then nothing, on all other linked tracks the beep continues. Is this normal?

Some of the ribon cables between the Main PCB and the CS look ugly. I will check these better.

Any ideas on what to check next would be greatly appreciated as I am a newbie noober boob.

[nILS]

What exactly is and isn't working? I am kinda missing that info in the post ;)

When i initially touch the testing leads to the 5vdc and grd there is a brief beep then nothing, on all other linked tracks the beep continues. Is this normal?

Yes. Your multitmeter charges the caps for a moment, once they're charged the beep ends.

[Berg]

I am not sure what is supposed to happen on start-up. In the build guide it says to upload the test tone. Is there supposed to be a different reaction at start-up that says ready or something? When the lcd did a little random dance and then nothing I got confused "lack of knowledge" as to what next. Is it time to start the testone then MB-sid upload? I have not done that before or researched how on the forum yet. The pics already have MIOS v1.9 as they came as so from smash.

In checking some of the ribbon cables i did find a couple of linked ribbons that were side by side. I will fix. Thanks for the confirmation on the 5v to grd check figured it was a capacitor thing.

[smashtv]

lcd redbacklight lights for a moment

This sounds like a weak power supply.

Is there supposed to be a different reaction at start-up that says ready or something?

Yes with the bootloader + MIOS loaded on the PICs (as they shipped from me) the display should show a copyright message then "Ready."

I suggest starting it up with your meter connected and looking for a drop on the 5v power rail as it lights the LEDs.....a working PSU will show little change, but a weak one will often drop below the operational power threshold of the PIC (and other components).

Best regards

Tim aka SmashTV

[Berg]

Much thanks for guidance I will play with the PSU. I bought two wallwarts as the c64 powersupply I bought off ebay did not work. the wall warts may be the issue

[Berg]

On the half broken ebay C64 powersource, the 5V side works and the 9vac side doesn't. I tried it again with soldering in the other 5vdc c64 source (instead of the 5vdc wallwart) and kept the 9vac wallwart for the 9vac pins the same thing occurs (quick flash of all led's backlight flashes on then only fan noise remains). Through elimination it says that if it's the PSu its the 9vac side as both 5v sources yield the same result. I tried putting a meter on when i flick the switch on but the meter is slow the led's flash on and off before the meter gives a good reading. In another post I read that it may be the P2_core trim pots so i tried cranking them and no change. I rechecked the ribbon cable to the lcd and pins 1-16 all have connection. It's a crystalfontz lcd I matched the pins 1-16 at J15_core1 to the 1-16 pins on the lcd.

Any ideas?

[Berg]

I think i may have soldered the LCD up wrong. I soldered number to number not following

http://www.midibox.org/users/jim_henry/building_a_midibox_lcd_cable.pdf

i will change and report back. I essentially matched J15 pin 1 to lcd pin1 then j15 pin2 to lcd pin2 etc. Silly me for thinking it was as simple as connect the matching numbers.

[Berg]

New debug log:

Item 1: lcd wiring now correct

Item 2: loopback chec on midi with mios. all cores check out

Item 3: only one pic of the 4 will link to mios,

Item 4: Uploaded MB-6582 setup hex, to the one working pic

Item 4: IT PLAYS NOTES SENT FROM MIOS AND JOMOX WAHOOOOOOOOOOOO!!! but only in core places 2 thru 4

Item 5: When the pic is in core place 1 it just beeps continuously, not notes just beep beep beep beep beep beep,

Item 6: the lcd does not say anything and the buttons and nobs have no affect, when the working pic is in core places 2 thru 4 the lcd does not light up, when it is in place one it lights. all display gibberish

[Berg]

can any one tell me wha the trim pot beside j11 is for? not the 4 P2_core1,2,3,4's. it is right beside T1.

another queston: how do you know if a lcd needs 4-bit or 8bit mode. I got th crystalfontz 2004A-RGH-JP

Edited December 17, 2010 by Berg

[Berg]

New Report:

1:all pics now have proper ID's 0,1,2,3

2:all pics linking fine with mios studio

3:all pics have mb6582 setup hex uploaded

At start up the following happens:LCD is not displaying anything, matrix lights up then has a pattern that flutters while it looks like it is thinking...booting up?????? some sid noise are present from each sid (changed headphones to each jack), the buttons & knobs are non responsive and

Again no help from the LCD. not seeing "ready"

Anybody out there? It feels so close now that some sounds are present. It seems like the control surface is not enguaged or configured.

[nILS]

can any one tell me wha the trim pot beside j11 is for? not the 4 P2_core1,2,3,4's. it is right beside T1.

That's the backlight trimpot, which determines the amount of current for the LCD backglight.

another queston: how do you know if a lcd needs 4-bit or 8bit mode. I got th crystalfontz 2004A-RGH-JP

By checking out the manufacturers website? http://www.crystalfontz.com/product/CFAH2004ARGHJP. General specs, 3rd line.

I'd suggest removing the CS for now. Tackle one problem at a time. First of all, I'd fix the LCD. Rererererecheck the wiring. You might wanna check out for that.

[Berg]

Back at the debug after a dead hard drive BLUE SCREEN OF DEATH!!!!!!

New report:

I REREREREREREREREREREREREREREChecked the LCD wiring, with confidence its good, so it's being fed garbage is the new theory. Have loaded the lcd interconnection test Hex. and there appears to be a problem. I did also fill in the pads on the bottom of the PCB for 8 bit mode.

When running the test for 5V on the data pins (any of them) D2 & D3 always show up with ~1.6v when they should be 5 or 0 pending the test. All other data pins react as they should.

[nILS]

That's because you shouldn't use 8-bit mode. Two of the pins are used for the CAN bus.

[Berg]

The solder was removed from the pads on bottom. I also ran the testtone app and it seems to work,quite high pitched. It will also play a midi note when the mb-6582setup hex is installed. My inexperiance is shinning at the moment. When i load up the setupmb6582 hex am i supposed to be changing anything in the code or just a simple upload? When i initially had the LCD hooked up wrong way back earlier in this post could it have buggered up the lcd? It's lighting up the back light but none of the crystals are doing anything.

[nILS]

I also ran the testtone app and it seems to work,quite high pitched

1000Hz (correct) or sth else?

When i load up the setupmb6582 hex am i supposed to be changing anything in the code or just a simple upload?

Just upload the mb6582 hex file.

could it have buggered up the lcd?

That's possible, yes.

[Berg]

Testing Sids: pin #8 & #27 reads .999khz.

setup_mb6582.hex was uploaded is there a different mb6582.hex?

[nILS]

That's the correct one and .999kHz is definitely close enough to 1000Hz :)

[Snoozr]

If your LCD is still not working, consider the possibility that you cooked it by wiring it up the wrong way. I have cooked multiple LCDs from wiring errors. They seem to be very sensitive. I lost one on the 6582 from wiring and spent a lot of time trying to "troubleshoot" the LCD when there was nothing that could have been done.

I recently had a Sammich LCD die after about, oh, a few hours use. There is no chance of a wiring error with the Sammich due to Wilba's cool design but the LCD died anyway. This could have been from static electricity as it is rather cold around here at the moment.

The only way to really check these sorts of things is to have a known good LCD that is wired up correctly, and then temporarily transplanting it on to the new unit to see if it works there. I happen to have a 3rd Midibox with a working LCD (but with some horrible noise somewhere that I could not figure out and hence the box was abandoned). This one is my test LCD. It got me through the two troubleshoots above.

[Berg]

i tried another lcd and similar result.

new thought... when i pulled the cs ribbon wires earlier in this build off of the cs panel some of the pads came off of the top of the soldering hole this may have goofed up the plane where jd09 5v attaches. if this 5v plane was affected would it bugger th lcd?

[Berg]

I believe that #4 connects to a 5v plane (light purple in this file=> MB-6582_CS_PCB_R2_Color.pdf ). Can someone confirm? Some of the other solder hole pads i am not as concerned about,leds & encoder pts. most can be connected with a jumper as a fix.

[nILS]

I believe that #4 connects to a 5v plane

#4 is connected to the ground plane.

[Berg]

if the cs ground plane was not connected correctly could it affect the lcd?

[nILS]

No. The LCD is completely wired with... wires :)

[Berg]

New Report

The LCD is working Yipee!! :drool:

So new problems emerge. I made sure that each core had the right .hex and the following happens on power-up.

1) says mios is there by tklose

2) launches CS

3) E002 Pint Lead Chn.1 (it pauses here for a while approx 20-30 seconds)

4)lights flicker and the menus seem to scrool and change values by themselves and always seem to stop again on either the envelope page or the LFO pg. When i press the buttons (any and all) it's like they arent doing anything. If I turn the data knob above in (2) while its pausing it seems to do some scrolling madness andends up freezing.

To me this all screems short in CS. I will keep on with the perserverance game.

[Berg]

I have looked at all of the shift reigisters to look for bad links and re touched all of the solder pointes associated with the registers, still the same symptoms. Is there somthing specific that would point to the buttons not working? any ideas?