[SOLVED] bankstick(s) not working

[mescalinum]

hi

I'm finishing a modular MBSID. actually I've done 1 CORE_V3, 1 SID_V2, 1 LCD 16x2 and works like a charm.

I wanted to add banksticks.

I made a little two-socket on veroboard (see attached image) (I used eagle cause it helped me to cross the less wires while working on a single layer)

anyway, the bankstick appears dead, I double-double-checked solderings, wiring, checked for shorts. I also checked that I have the two air-wires on the core (J4:SC->pin28 and J10:PWM->pin17).

the symptom is: when I send program change, it says <No Bankstick>.

I'm stuck, I don't know what to troubleshoot further

any help would be appreciated, thank you

EDIT: oops, it seems the attached image never got here.

here it is:

[nILS]

The schem is correct. A picture of the veroboard might help :) Some shots in the dark:

- R2/R12 on the core not mounted/wrong value

- Cable between core and bankstick is the wrong way around or broken

- You made a mistake in the eagle->veroboard process

[mescalinum]

thanks for your interest

The schem is correct. A picture of the veroboard might help :) Some shots in the dark:

sorry, I do not have a camera that can take decent photos, so I toke a couple, and added visual hints to one, in case you can't see every spot

(click to enlarge)

by the way, here are some test I made: (I'm repeating test now, just before hitting SEND)

with the bankstick circuit unplugged and ICs unmounted(looking at the schematic I made in eagle), using OHM-Meter:

- every IC pin that should be connected to GND, it is in connection with GND pin. every other: not.

- every IC pin that should be connected to +5V, it is in connection with +5V pin. every other: not.

- the 2 wire data bus does not touch +5V neither GND

- the 2 wire data bus is consistent (J4:2 touches only IC1:6 and IC2:6, J4:3 touches only IC1:5 and IC2:5)

with the bankstick attached to core, and core powered on, I can measure 5V between GND and 5v on every IC

that should exclude problems in the bankstick circuit itself

- R2/R12 on the core not mounted/wrong value

R2: ok (brown,black,red,gold = 10*100 = 1k)

R12: ok 1k too

- Cable between core and bankstick is the wrong way around or broken

as I said (text in the photo), I refer to red wire on ribbon cable as GND, and I measured +5V between GND and +5V on the bankstick circuit

- You made a mistake in the eagle->veroboard process

given the tests I made above, it should be very unlikely.

unfortunately I don't have access to a logic probe or an oscilloscope... I don't know what to do then.

every other part is functional (sid functions, audio output, midi io, etc..)

[mescalinum]

I tried also 24LC256 ICs (always from Microchip) but same result <No BankStick>

this should exclude a fried IC.

the only thing I couldn't check now is the 2 wire IIC bus... later I'll dismount the core module again and do some more tests...

btw, I remember having tested with OHMmeter every single wire after etching CORE, prior to begin soldering

[mescalinum]

damn! :D

i got some time to unmount and check CORE module: air-wire from IIC pin to R12 was soldered to the wrong node

so I fixed it, and the midibox started to see external banks :)

sorry for the fuss

(I don't know how to put [sOLVED] in the subject so that this can be helpful to someone...)

cheers!

[mescalinum]

p.s. here a couple of shots of the unfinished boy - I steel need to make the panel, I can't find detended encoders through local shop (they have gray code rotary encoders at 5 bits only), I'll have to search on internet shops

[nILS]

Looking good!

I changed the subject for you. You can do that by editing the 1st post and clicking "Use Full Editor".

[jbartee]

This looks great! What did you use to cut the case like that? That's a very precise, tight seam.

[mescalinum]

This looks great! What did you use to cut the case like that? That's a very precise, tight seam.

I first used a manual iron saw (30-40cm blade).

Then I smoothed the cut with a belt sander, and then used fine-grain sandpaper - sticked on a very straight plane - to refinish by hand.

If you do not have belt sander you can just do papersand by hand... it's just a bit more work.

[jbartee]

I first used a manual iron saw (30-40cm blade).

Then I smoothed the cut with a belt sander, and then used fine-grain sandpaper - sticked on a very straight plane - to refinish by hand.

If you do not have belt sander you can just do papersand by hand... it's just a bit more work.

Well, that sure beats the hell out of the dremel I've been using for my plastic cuts. A saw like that is a little out of my price range, but it's hard to argue with your results. Beautiful work!

[nILS]

That's what I use for stuff like that:

A bit of a pain with bigger stuff but definitely affordable @ ~15eur