Pot/RCtime help

sumdawgy

I built a controller for a dishawasher and had it running....
But, it had an issue with random resets during operation.

Finally had some time to go back and look into it.
-It worked fine on test bench, but failed in unit.
-Disconnected power to A/C motors, worked fine.
-Capped all pins except pot pin to ground & vdd. (As described in harding your stamp PDF.)
<EDIT>

---

unit still failed randomly
</EDIT>
-I dropped the pot pin (only one not hardened.) low to ground it. No problems for a month.

-The wires to the built in thermistor run 18ga unshielded alonside the other A/C wires in the dishwasher's primary harness. This was allowing random voltages to build (thru inductance & static). Affecting the inpt pin and being shunted by the stamps protective diodes. Causing the resets...

Since, I cannot use this in the standard pot/rctime config.
I'm thinking I'll have to use it in an active voltage divider circuit. To asorb the spurious volatges......

1> Does anyone have/know the circuit that was posted a few months back RE: using rctime to read voltages from 0-12 (I tried to search but...)

2> Does anyone have/know of a better solution?

schematics and layout .... http://forums.parallax.com/showthread.php?t=123064&highlight=dishwasher

Mike Green

The technique to read voltages (typically battery levels) using RCTIME is described here: EME Systems. Click on the "app-notes" link at the bottom of the page.

It's bad practice to run low level signals alongside noisy AC power lines. You get exactly what you're seeing ... high induced noise levels. You need shielding or at least twisted pairs to carry the low level signals.

Can you replace the thermistor with a digital temperature sensor? Maybe you could use a 1-wire temperature sensor (like the DS18B22-PAR)? That would need a BS2p/pe/px Stamp though.

sumdawgy

Mike Green wrote: »

The technique to read voltages (typically battery levels) using RCTIME is described here: EME Systems. Click on the "app-notes" link at the bottom of the page.

It's bad practice to run low level signals alongside noisy AC power lines. You get exactly what you're seeing ... high induced noise levels. You need shielding or at least twisted pairs to carry the low level signals.

Agreed on the unshielded low voltage signal issue. But bear in mind this was the layout in a commercial product..(heh....that failed..so....).

I'm trying to stay in the spirit of using existing hardware. I went back thru the origional tech sheet supplied with the unit.... it mentions that heated cycles would run but NOT COUNT TIME untill water temp would rise to a certain level. They let the internal thermosat handle actual temp control (As I've ended up doing) and using the thermistor to indicate when a certain temp is reached.

Unfortunately, I don't have access to a personal O-scope (logging or otherwise)....
Or, I could see exactly what I'm dealing with noise-wise.....

It occurs to me that if I were to use the thermistor in an "powered" circuit I could use a small cap connected across the source wires to "short" the unwanated A/C signal.

I'd still need to either trigger a switching transistor or "read" the resultant voltage from a voltage divider circuit.

Personally, I want to employ the divider circuit using "pot". Somone posted a circuit&code snippet that did just that...I thought I'd saved it, but it's not in my "tooldirectory".

erco

Mike's link worked for me, here's a direct to the app notes page:

http://www.emesystems.com/BS2index.htm#misc

My kitchen electric story: (marginally related, but memorable)

When I installed flourescent lighting in my kitchen, (basically shop lights above a drop ceiling) I tested each light individually on the floor before installing it overhead. It went like this :

1) test on ground, works great
2) install on ceiling, no work
3) remove, test on ground, works great
4) install on ceiling, no work
5) repeat

Turns out that the lamps needed a ground wire to function. My 1957 house was originally ungrounded, and I upgrade as I remodel for GFI's etc. At that point my final ground wire wasn't hooked up to anything yet. But on my floor (raised foundation), they were close enough to the ground where they worked through some mysterious AC ground field. Quite maddening until I figured it out!

sumdawgy

erco wrote: »

Mike's link worked for me, here's a direct to the app notes page:

http://www.emesystems.com/BS2index.htm#misc

My kitchen electric story: (marginally related, but memorable)

Turns out that the lamps needed a ground wire to function. My 1957 house was originally ungrounded, and I upgrade as I remodel for GFI's etc. At that point my final ground wire wasn't hooked up to anything yet. But on my floor (raised foundation), they were close enough to the ground where they worked through some mysterious AC ground field. Quite maddening until I figured it out!

yea I read his comment but SKIMMED the instructions bvelow it...(reflex I guess)
Saw it 2 sec after I posted ...so I edited and deleted that line....you evilly fast man!

I was gonna say we had something similar with shipboard lighting we tried to re-use in a wharehouse in shipyards....silly ground wire...onboard ship....the whole thing is a ground so......

PJAllen

Have you considered using co-ax or a twisted pair (approx 1 turn per inch) for the RCTIME interconnect/run?

sumdawgy

PJ Allen wrote: »

Have you considered using co-ax or a twisted pair (approx 1 turn per inch) for the RCTIME interconnect/run?

Yeah...but...it ties into the main A/C harness. Would require I run a new run....
However......I have "tons" of rg6 cable coax lying around....

I was thinking......I'd mount (in the front panel circuit) an opto circuit controlled by the thermistor & just go for a fixed current/temp setting. Knowing exact temp was just a bonus anyway.