OPL3 chip damaged?

[tomcody]

Hello, I'm making the MIDIBox FM and at this momment my system is like the minimal: CORE with PIC18f4685 + MBHP_OPL3 + LCD + bankstick.

In this I posted that some pin of the YAC512 was unconnected to ground. Finally I finnished the MBHP_OPL3 without ground plane, so the ground of 5+ was unconnected to the ground of +12-12. When I tested the module for the first time the OPL3 chip got a little warm, I did the connection test (mbfm_interconnection_test_v2) and the D0, D1, D2 & D3 pins always reads 5V. The remaining pins works OK (the LCD worked OK too).

After this I loaded setup_pic18f4685_mbfm_tk.hex with all my system interconnected hoping the LCD will shows me something but nothing happened (just the upper LCD line in black). Then I loaded the 1kHz triangle test and no sound :sad:.

After 4 or 5 times the OPL3 chip doesn't got warm anymore :huh:.

Then I realized that the oscillator's GND pin was connected to the +12-12's ground and not to the +5's ground so the oscillator never got power and the OPL3 chip never got the clock signal, now it's fixed. I did triple check both circuits.

summarizing:

CORE are working OK except D0, D1, D2 & D3 pins (I think).

LCD working OK.

MBHP_OPL3 PCB OK and all components correctly mounted, green led lighting.

CORE & MBHP_OPL3 connection OK. Not soldered, but through ribbon cable connection (see pic).

I'm afraid my attempt to save toner was the error (no ground plane).

Could this damage the OPL3 chip :bye:?. Maybe the clock pin was floating because of that and that's why got warm I dunno...

Thank you in advance for any reply.

[Imp]

Your D0-D3 issue sounds familiar.

I think the reason for this behavior is that the interconnection test was written for pic18f452, which handles the LCD pins differently, but don't take my word on it.

Try switching between 4bit/8bit mode (via pic-header), to make the LCD work in mbfm too.

Your attempt to save toner was a nice idea, but what about the wasted etchant? :tongue:

I can't say if this damaged anything, but i guess it would kill the Yacs first.

Maybe the oscillator got a little too much voltage bedause of the wrong ground.

If you have a multimeter that can measure frequency, check if the osc still does its job.

[tomcody]

Thanks for your reply Imp, fortunately I have a 18f452 and did the connection test with it and all OK ... :thumbsup:.

Then I loaded setup_pic18f452_mbfm_tk.hex and the LCD showed me the mbfm menu ... :thumbsup:.

Then I loaded the 1kHz test sound and it sounded in all the channels ... :thumbsup:.

:laugh: :laugh: :laugh: :laugh: !!!!

I don't understand why the LCD doesn't showed anything with the 18f4685 ...

I used the latest .hex files:

bootloader_v1_2b_pic18f4685.hex

mios_v1_9g_pic18f4685.hex

setup_pic18f4685_mbfm_tk.hex

Maybe because of using the minimal requirements?

Your attempt to save toner was a nice idea, but what about the wasted etchant? :tongue:

Well I don't use ferric chloride, instead I use nitric acid and takes less than a minute and goes a long way ... in fact I have to water down a lot this thing or the PCB will end with no copper and all the toner floating in 3 seconds :rofl:

Well I think for the momment is solve my problem but the 18f4685 thing. Thank you again for taking the time.

Tom.