Last Gaussmeter Problem? (solved!)

[RedHouse]

gaussmeter has arrived yesterday.....




....data lines are connected but there are not data....


the problem was a short (but they were 2) in pcb .... you can see in picture.... (note that is picture posted from Jeff..... i only add rectangles

one of that pin is the clock out of MCU.... with short the clock was locked and firmware did not run....and data lines did not have data....

now i don't know what to say.....

Jeff, in a my email i said you to ship your gaussmeter without box because economic shipping from italy is max 5cm width.. now with box redhouse have to pay much more for the re-ship.......

....tomorrow i'll test with sensor....

bye

BTW, Redhouse doesn't have to pay for re-ship.

The picture doesn't show any shorts, was there some additional soldering on those three pins when you recieved the board?

If you use ANSEL=00000011 that will turn pin 10 (RA6/OSC2/CLKOUT) into a digital I/O and in conjuction with INTOSCIO and setting all of Port-C to digital (since you are not using port-C anyway) then a "short" such as you describe wouldn't have made any difference in the PIC's operation.

Another less than optimal design "feature" would be leaving pin-1 (MCLR) floating as you have done. Leaving the programming pin (pin-1) floating can lead to mis-behavour in PIC's during runtime.
(week pull-ups (WPUB) are only available on port-B in the 16F883)

When you do your next PCB for your next version you should think about adding a ~10k resistor to hard wire pin-1 high.

[-Elepro-]

ah you are studing pic programmng.....

If you use ANSEL=00000011 that will turn pin 10 (RA6/OSC2/CLKOUT) into a digital I/O and in conjuction with INTOSCIO and setting all of Port-C to digital (since you are not using port-C anyway) then a "short" such as you describe wouldn't have made any difference in the PIC's operation.

ah ah ok next time i program for all possible bad-pcb-caused shorts

but above all:

Another less than optimal design "feature" would be leaving pin-1 (MCLR) floating as you have done. Leaving the programming pin (pin-1) floating can lead to mis-behavour in PIC's during runtime.
(week pull-ups (WPUB) are only available on port-B in the 16F883)

When you do your next PCB for your next version you should think about adding a ~10k resistor to hard wire pin-1 high.

one for all that says how many you know about PIC

from 16f883 datasheet (config1 register)

......bit 5 MCLRE: RE3/MCLR pin function select bit
1 = RE3/MCLR pin function is MCLR
0 = RE3/MCLR pin function is digital input, MCLR internally tied to VDD

but my bit 5 in cnfig1 is low and i have not any pin MCLR because it is "internally tied to VDD" then i never will need any resistor there.....

bye bye

[RedHouse]

but my bit 5 in cnfig1 is low and i have not any pin MCLR because it is "internally tied to VDD" then i never will need any resistor there.....

bye bye

Ahh good, then no problem there. Most people don't rely on that and hard wire it high when not using ICSP.

But you asume too much, do you think nobody else here has any programming experience?

BTW; programming for all aspects of user related scenario's (newbie's included) is considered good practice, leaving things to asumption is not so good.

[Jeff Callahan]

Thanks Elepro, let me know how much the shipping is and I'll get you the money. Do you do paypal?

[Joe Gwinn]

The picture doesn't show any shorts, was there some additional soldering on those three pins when you recieved the board?

The picture does show the shorts, but magnification is required to see them. It looks like a stray piece of very fine bare wire became involved, but was not noticed.

[Dave Kerr]

Regardless of how those shorts got there, you've got to think that a guy who'd charge a fee to assemble the board would've noticed them.

[David Schwab]

The picture doesn't show any shorts

Yes it does. Look at the photo full size (click on it to open in a new window). There is a thin strand of solder between the adjacent pads.

[RedHouse]

Regardless of how those shorts got there, you've got to think that a guy who'd charge a fee to assemble the board would've noticed them.

Yea, sure would.

The picture does show the shorts, but magnification is required to see them. It looks like a stray piece of very fine bare wire became involved, but was not noticed.

Aha, I didn't see that when I looked , for some reason I thought he was describing a solder bridge type short.

Reminds me of a repair job I did recently on one of those Floyd Rose guitars, customer brought it in for having "intermittent cutting out" and "crackeling" noises which turned out to be some bad wire connections one of which was a loose thread grounding-out the Jack/Plug, the other is with a mis-connected treble bleed, here's the pics:
(the treble-bleed parts wer not actually soldered, just touching each other and making periodic contact)

The FR guitar was one of the last ones assembled in Redmond before they too the operation off-shore.

[David King]

I see shit like that all the time. Guys will bring in guitars that supposedly worked everyday for years and I'll find that no one bothered to solder the PU leads to the center lugs of the pots. The just stuff the wire through the hole and call it good and that's from the factory. Don't get me started on diy soldering jobs..

[LtKojak]

customer brought it in for having "intermittent cutting out" and "crackeling" noises which turned out to be some bad wire connections one of which was a loose thread grounding-out the Jack/Plug, the other is with a mis-connected treble bleed, here's the pics:
(the treble-bleed parts wer not actually soldered, just touching each other and making periodic contact)

The FR guitar was one of the last ones assembled in Redmond before they too the operation off-shore.

Man, what a HACK! I have no words...

[Jeff Callahan]

Thanks for all the help

Just thought I'd post one last time regarding the gaussmeter. Just got back from Elepro, and is working, he refused to charge me for his time, thank you. And thanks to Redhouse, for putting this thing together, and all that helped me figure this out. I've learned so much from all of you and it is really nice to have guys like you out there.

Jeff

[-Elepro-]

perfect....

bye

[Joe Gwinn]

Elpro, what did the problem turn out to be?

[Jeff Callahan]

It was a couple sloppy solder joints that my guy did, a couple of the pads were bridged together with solder.