DS18S20 & DS18B20 Codes

[ishmael]

Hello. I have a problem that drives me insane. I am sure it is in simulation or fuses. Both temperature nibbles are always 255, so the result returned from read() is always -0.5. Has anyone else met this problem ?

I attached the pull-up, I use a pic16f648A, and the pin is B5 (or B4, I tried both, no result).

Please help.

[Shamber]

Hello.
Does anybody help me with alarm search procedure?

[jruibarroso]

When temperatures are above zero (>0º) display shows correct values but when is below 0º (<0º) strange values appears (~65535). Can anyone tell me how can I correct this ?

[PCM programmer]

What code did you use ? There are several versions of the driver in
this thread:
1. Han Solo
2. scanan
3. xokolatecake (2nd post)

What sensor did you use ?
A. ds18B20
B. ds18B20-PAR
C. ds18S20

I used the first one, ds18B20, and the original Han Solo code.
I configured the code for the ds18B20, as shown in his post:

[xuxa]

I am having the same problem of ''ishmael'' and I use CCS in Mplab, I've tried the 16f628a and the 16f877a, both only shows -0.5 C on display...

I'm from Brazil, my english is not perfect, thanks for any help

[ishmael]

xuxa, I solved the problem. I'm sorry I haven't posted a follow-up, it's just that the project is my bachelor's degree, and the exam is tomorrow, so I am a bit stressed. As far as I can remember, the -0.5 problem in the proteus simulation got solved when I loaded the proper configuration word onto the pic (from the end of the ccs exported lst file). If necessary, delete and add the pic and/or ds18s20, or even start a new project (maybe there are some bugs regarding applying the changes to those settings). Also pay attention to the oscillator setting, so that the pic speed you used in the C code, will match the speed set up in proteus.


And if your problem is not in simulation, remember to attach the pull-up resistor to the 1-wire bus. Upon checking my original post, I discovered I changed the port to, it's RA6 used as I/O , - pic uses internal oscillator. try using my fuse list : NOWDT, INTRC_IO, MCLR, PUT, NOPROTECT, BROWNOUT, NOLVP, NOCPD
config word : 0x3F70


Hope it will work for you too.

[xuxa]

In the project with 16f877A, I solved the problem using the same oscillator setting as I 've been using in my CCS code (20Mhz). Than the LCD only displayed multiples of 8, like: -16 C, -8 C, 0 C, 8 C, 16 C .....So I went to the ds1820.c code and changed the configuration to ds18B20, in a escale of 0.1 , (just by changing the comment of two lines):

[milkman41]

Hi all,

I'm using the original code posted by Han Solo on the first page, and I can't get a temperature other than 85 C, which I think I read was the startup temperature. I'm using a PIC16F876A with a DS18B20. The code I'm using is pretty much exactly the one Han Solo posted; all I've changed is the one wire pin, and then made some minor changes to the main code, but nothing to any of the actual one wire functions. Any help would be greatly, greatly appreciated.

thanks.

[deperkin]

I am trying to connect a ds18s20. (I am using a 4.7k pullup).

I am using the original hans solo code, with the only difference that the clock is set to 500 kHz. And I am trying to use pin D2 on an 18F4550.

I also am constantly reading a -0.5 C temperature.

Any ideas? Due to other things on the circuit I need to stick with this clock speed.

I am using compiler 4.093 as well.

I am NOT using Proteus.

If I use the updated code from http://www.ccsinfo.com/forum/viewtopic.php?t=42648&highlight=ds1820
then I constantly get 4095.0 C

I have also tried various input pins.


Thanks all!

[atai21]

I'm sorry but I don't understand what this line means:

[tejca]

[adrix]

How can I make that the output would be 0,1 deg accuracy? I'm using DS18S20, so the division by 16 doesn't work.

[Mike Walne]

Of course it doesn't work.

I don't even use these parts, but I can read the data sheet.

Best ACCURACY is 0.5C.

DS18S20 is 9 bit RESOLUTION part.

LSB represents 0.5C, so that's the best it can do.

If you want better RESOLUTION, maybe use the DS18B20 which goes to 12 bits.

Each LSB is then 0.0625C, you get nearest 0.1C by rounding off.

But ACCURACY is still only 0.5C at best.

I'm assuming you understand the difference between ACCURACY and RESOLUTION.

Mike

[stoyanoff]

I have a problem. I tried your codes but the 18B20 returned very high temperatures - 85 degrees for example. The temperature in the room is not more the 25 degrees.
Where is the problem???

[shaharyar]

Thanks jds-pic and scanan
the code worked like a charm with a single error
cheers