what is happening with the pots?

[tomcody]

I finished my midibox with 28 pots and 12 buttons, but happens something strange:

if we grouped the potentiometers thus:

[1 2 3 4][1 2 3 4][1 2 3 4][1 2 3 4][1 2 3 4][1 2 3 4][1 2 3 4]

in the first group, the second pot sends messages to the channels 1,2,3 and 4, and the others sends nothing. In the second group happend the same, but the fourth pot sends messages and the others nothing .....  :-\ , and thus with all the groups ... (second pot, fourth pot, second pot, fourth pot ....) what happens?  ??? ??? help me please.

[illogik]

hi; checked your wires? maybe you swapped some..

[smashtv]

Hi Tom

This failure mode sounds like it could have a very similar cause as your issues with the dinx2.....only reading one input vs. scanning them all.....

If its the same PIC used to test both, I would look there only after you are sure nothing is wired wrong or shorted......

Also you only seem to be using 3/4 of the ain, are the last 8 inputs on the ain grounded -or- core J5 pins A3-A7 grounded -or- the app re-compiled to only read 3 4051s (28 pots) vs. 4 or 8 4051s for the default app?

My gut says the clock line on the data bus is not there (bad soldering or bad PIC) or miswired, or shorted to something else besides ground, making it harder to find.

Best

SmashTV

[tomcody]

thanks for reply, I think the PIC is bad, maybe this happened when the SC pin was in short with ground. But, when I put 8 pots without AINx4, they work fine ...  ??? ??? Maybe the multiplexers are bad be cause are sensitive to electrostatic discharge and I touched their pins (I have long hair and this make me a Van der Graff generator to the multiplexers ;D).... well I will buy others and will check. Thank you again.

[smashtv]

wow the blood/coffee level was way wrong when i said

My gut says the clock line on the data bus is not there (bad soldering or bad PIC) or miswired, or shorted to something else besides ground, making it harder to find.

Since there is no bus clock there in your app, that's a bad answer!  ;)

The A,B,& C lines from the core tell all 4051s what input pin to read, and each 4051's output line has its own wire to the core....So if the core wants to read pot 1, it sends a 3 bit word (via ABC lines) telling -all- 4051s to route input 1 to output, then listens on core J5:A0 for the value to read....

My gut still applies though considering that the ABC lines are used by some non-ain apps for din/dout bus......;)

So since its only reading the same pot(s).....follow the ABC lines to make sure they are wired right, all the way up to the pic pins.

Also be sure your app is expecting an ain and not unmuxed pots, that -might- cause same symptoms.

On the static thing:

1. I hear ya brother!  I can always tell when i'm extra charged by the friz level of me mop..... ;)

2. The odds that this is a static death are very low, considering all the 4051s are acting the same.

Best

Smash

[tomcody]

I have been all the day reviewing the boards, the connectors, but I can't find the problem .... when I compile the main.asm and set it with 1 or 2 pot(s) and Multiplexers, the pot(s) works fine - but when I set it with 3 pots, the second sends to channels 2 and 3, and the third pot doesn't work. Well, I will continue tomorrow reviewing ... thanks for all.

[stryd_one]

Curious... If you set it up for three pots (as per your last post) and then only connect pots 1 and 3, what does pot 3 do then? Still nothing?

Don't forget to ground the unused pins....

[tomcody]

Yes, stryd_one: Still does nothing ..., I have a question : Each multiplexor uses a 3 bit word (the A,B and C lines) to indicate to the PIC which of the eight inputs is changing its value? that means 000 for input 1, 001 for input 2, ... , 111 for input 8? Then I could check these values when I move a pot .... is this right? Thanks for your replies :) .

[smashtv]

I have a question : Each multiplexor uses a 3 bit word (the A,B and C lines) to indicate to the PIC which of the eight inputs is changing its value? that means 000 for input 1, 001 for input 2, ... , 111 for input 8?

close....except the pic is telling the 4051s which input to let through via the ABC lines, scanning through them as fast as possible (almost as fast as a DIN scans buttons, but slow enough to allow for needed settle time).....

The 4051 datasheet has a table showing ABC state vs. selected input....(sorry no link, i'm on site with me Palm)  ;)

Then I could check these values when I move a pot .... is this right?

The ABC lines will be switching way too fast to read without a strip-chart style logic analyzer........but.......

You can set the ABC lines manually for some low speed testing, just tie the lines hi/lo as shown in the chart, and the ain should switch inputs as expected.  (obviously with ABC lose from the pic, in case it is sending something unexpected on ABC)

Luck be with ye!

Smash

[tomcody]

  Thanks for clarify my idea of how it works. Now I can try other ways .... thank you I'll never give up!!  ;D  ;D  ;D