DS2423 going crazy

Website · Post by **simplemind** » Fri Sep 08, 2006 12:22 pm

Robert,

The 741 opamp is a very good amplifier that is best driven by a dual powersupply, +5, 0, -5. However a single supply 0, +5 is also possible. But it depends on the use. Used with a single supply in some cases you have to take precautions to get the output nice in the middle for AC use and need to use a capacitor to stop the DC current flow through the output. But that is not the case here, we use this amplifier for DC switching and amplifying.

I use the LM324 because this opamp is designed for a single powersupply and can switch from 0 to almost 5V when driven by a single 5V supply. There are 4 opamps in this case so when you need a buffer amplifier just use 'm.

Still the signals you get from the 741 looks very good, so maybe it is not as worse as it looks. If you made a serious mistake the 741 would have explode [:D]

Draw the schema again and publish it.

What do you use as busmaster for the DS2423? Check if this busmaster has already the pull-up resitor on the 1-wire bus line. This resistor 1k5 .. 5k is important. The devices which pull down the line must get enough sink current otherwise you get strange results.

Regards,

Marcel

Website · Post by **simplemind** » Sun Sep 10, 2006 7:53 pm

I read your 'older' topic again and saw that you use a 1-Wire serial adapter from Dallas. So the question of the busmaster is solved. I think the pull-up resistor I'm talking about must be already part of this device. If you have a schema of it you can check it.

Regards,

Marcel

Post by **Digit** » Sat Sep 23, 2006 1:06 am

Hi,

It has been a while since my last post, i'm just waiting for errors to occur and trying to log everything that happens while i'm not at home...

(

The counter is stable for quite some time now. I just had 2 errors this day, and both were TMEX Transport related:

This morning i got an error -12 (non-existant actually, or does it mean PORT_NOT_INITIALIZED (-1) + PROGRAM_NOT_POSSIBLE (-11) ???), and later on i got error -7 (NO_DEVICE):

22-9-2006 07:35:17 : 5200000007C4761D: 575749 143969
22-9-2006 07:35:17 : 5200000007C4761D: Time: 210
22-9-2006 07:45:45 : 5200000007C4761D: Error 1 -12
22-9-2006 07:45:45 : 5200000007C4761D: 0 143969
22-9-2006 07:45:45 : 5200000007C4761D: Time: 871
22-9-2006 07:45:47 : 5200000007C4761D: 575749 143969
22-9-2006 07:45:47 : 5200000007C4761D: Time: 200

22-9-2006 13:32:17 : 5200000007C4761D: 575785 143969
22-9-2006 13:32:17 : 5200000007C4761D: Time: 200
22-9-2006 14:20:18 : 5200000007C4761D: Error 1 -7
22-9-2006 14:20:18 : 5200000007C4761D: 0 143969
22-9-2006 14:20:18 : 5200000007C4761D: Time: 902
22-9-2006 14:20:20 : 5200000007C4761D: 575785 143969
22-9-2006 14:20:20 : 5200000007C4761D: Time: 190

But both times the counter did return to it's original values as you can see, so in fact this kind of problems can be (and are) handled.
I don't seem to need any pull-up resistors, since the serial adapter provides it.
Power fluctuations don't seem to have been an issue in the errors mentioned above either, by the way.

But still, waiting for "it" to happen again....

Regards,
Robert.

Website · Post by **simplemind** » Sat Sep 23, 2006 10:29 pm

Ok, just waiting.

Regards,

Marcel

Post by **Digit** » Wed Nov 01, 2006 5:57 pm

Yes!!!

I found the cause for the crazy counts...it was the power supply...
Read on...

I wrote a routine some time ago that automatically makes a screendump whenever strange countervalues are read. It finally paid off: something wrong with the voltages during a period of 2 seconds on October 22nd:

But, everything went back to normal after those 2 seconds...so, by the time it was discovered, there was nothing to be 'seen' anymore. But something told me it had something to do with power failure or short-circuit somewhere. A physical inspection of all wires led to nothing...no loose contacts, blank wires or anyhing else.
Still no luck.

But yesterday, the problem got worse: i suddenly had thousands of screendumps like this:

And the problem was still there when i discovered it!
So now at last i could inspect everything again and this time i found the cause very rapidly: It was the 5V power supply! I immediately saw the LED on the power supply was off!
Apparently the power supply didn't make good contact to the wall outlet, because a gentle push made the problem disappear; the LED went back on and everything went back to normal.

Aarrgghh! Now i'm the one that was getting crazy...

But, OTOH, finally the mystery has been solved, and i can move on!

Thanks everybody for your suggestions, but who would have thought the solution to be this "easy" ..??

Regards,
Digit.

Post by **Bwired** » Wed Nov 01, 2006 9:11 pm

Ook mooi [:D]

Website · Post by **simplemind** » Wed Nov 01, 2006 9:38 pm

Nice done Digit! I appreciate your perseverance.

I myself would never have expected that the ac-outlet caused this. Though lesson.

Regards,

Marcel

Post by **pszczerba** » Mon Nov 20, 2006 4:43 am

How are you triggering Channel A

Post by **Digit** » Tue Nov 21, 2006 1:14 am

Hi Peter,

My gas meter looks like this:

In the "6" of the last digit there's a reflecting area.
Now, everytime the digit "6" comes along, you can detect this with a CNY70. Pieter has a nice schematic how this can be done: see this topic:http://www.domoticaforum.eu/topic.asp?T ... erms=cny70

The output goes right to a Counter from Hobby Boards (http://www.hobby-boards.com/catalog/pro ... ucts_id=42)

Does this answer your question?

Regards,
Digit.