>>> KIT AVAILABLE <<< All In One v2.8 PCB Kit to sell !

[Synthex]

Ken, I do not understand what you say ...

PIC and Atmega are not working together.
The USB port can not be used if only the MIDI port is used.

[Ken Forgettable]

PIC and Atmega are not working together.

Make both chips reset on a reset condition like this:

(Using v2.8 schematic)

ICSP_Atmega - ICSP_18F2550
pin 4 <-----------> pin 1

???

[MaberY]

Thanx, but after a short while I encountered problems and lost the display. I first thought it was a loose cable connection
but after repairing the connection it acted weird.
MIDI led stayed on for long after booting and after a while it did not work any more.
Strangely the PIC still responds to windows

I think I fried the Mega32 somehow, not sure if the display is still intact. I'll check if I can program with ponyprog trough the ISP connector.

I'm pretty sure it worked OK , because I connected a crystal headphone and when tapping that the VU display reacted and MIDI led lit up.
I'll check If i can order a replacement Mega32 and display soon.

I already have some glassfiber maze cloth for fabricating the pads.

Somehow the display and the Mega got fried
I tested the display with LCDSmartie and got just junk in the display.

I ordered 2 new mega32's and I made a ISP adapter
I took another display (14 pins) i had in my junk drawer and
hooked up the stuff

I could load the boot loader in the Mega32, but I could not load the firmware in the Mega32 trough MIDI (internal USB and using M-Audio UNO) using MTC or MIDI-OX
I flashed the firmware directly trough ISP and that worked.
Soooo. All stuff working again so it seems. I'm still puzzled how both the LCD and the MEGA32 could get fried in the first place. I checked and double checked all 3 times before hooking up things and it worked right after the first powerup

[Synthex]

Be careful to static electricity.

[Ben]

Here is a new version of the PCB (being designed).

- New dimension : 120mm x 100mm
- 56 inputs available
- Use of 18F14K50 for USB

Can you give me your comments and wishes ?

MD56i_v31.gif

Comments on trace routing: Avoid stubs on high speed nets. Having two connectors on USB creates such stubs. Use stuff options with zero ohm resistors to employ one or the other connector. <via to header> R <via to/from chip> R <via to USB connector>

Is there a good ground plane under the USB nets?

[Ken Forgettable]

Avoid stubs on high speed nets.
Is there a good ground plane under the USB nets?

All valid points I'm sure Ben, unfortunately many end up with stubs on high speed nets.
Usually these stick out from fly splattered teeth

Dmitri's 22K-47K (lower the better) pull down on the pgm line works a treat but then - I'd like the vision of pull downs on the mux outputs to vanish too...

[dmitri]

I'd like the vision of pull downs on the mux outputs to vanish too...

These are not pull downs.

[Ken Forgettable]

These are not pull downs.

I'll try; components added as bleed resistors.
They try and limit the voltage swings into the ADCs.

[dmitri]

These are not pull downs.

I'll try; components added as bleed resistors.
They try and limit the voltage swings into the ADCs.

Not even this.

[Ken Forgettable]

Not even this.

Ah, these are the reason you wanted Roland wiring for the positional sensing. I've been here before
Remove the negative portion of the wave - pull ups - tush...