DS1620 Higher Temp Resolution
Archiver
Posts: 46,084
The data sheet for the DS1620 Digital Thermometer states that a higher
resolution can be obtained by reading the slope (the number of counts
per degree C for that temperature) and reading the Count Remaining, the
actual temperature may be calculated by:
Temp = Temp_Read - .25 + ((CountPerC - CountRemain)/CountPerC)
I tend to think that the Count Remaining should not exceed the Count
Per degree C (Slope). But that is not my case. My slope is fairly
stable ?? at 75 @ 20C, but will vary up to about 125. The Count
Remaining, however is all over the place. Between 30 to 500. And there
seem to be no relationship of having a slope of 75 and Count Remaining
of 30 or 500. Needless to say, the higher resolution is unusable.
Has anyone tried using the DS1620 in this matter ? Can it be done ? Any
available BS2 code I can try ? Can anyone shed some light on anything I
might be missing ?
Bruce Snowden
resolution can be obtained by reading the slope (the number of counts
per degree C for that temperature) and reading the Count Remaining, the
actual temperature may be calculated by:
Temp = Temp_Read - .25 + ((CountPerC - CountRemain)/CountPerC)
I tend to think that the Count Remaining should not exceed the Count
Per degree C (Slope). But that is not my case. My slope is fairly
stable ?? at 75 @ 20C, but will vary up to about 125. The Count
Remaining, however is all over the place. Between 30 to 500. And there
seem to be no relationship of having a slope of 75 and Count Remaining
of 30 or 500. Needless to say, the higher resolution is unusable.
Has anyone tried using the DS1620 in this matter ? Can it be done ? Any
available BS2 code I can try ? Can anyone shed some light on anything I
might be missing ?
Bruce Snowden
Comments
Here is some information and code using the DS1620 in high resolution mode:
http://www.emesystems.com/OWL2face.htm#DS1620-HR
You are right, the count-remaining should always be less than
count-per-degree. There are a couple of fine points that might be
bugging you. For example, the chip has to be operated in single-shot
mode to use the high resolution mode.
-- regards,
Thomas Tracy Allen Ph.D.
electronically monitored ecosystems
mailto:tracy@e...
http://www.emesystems.com
>The data sheet for the DS1620 Digital Thermometer states that a higher
>resolution can be obtained by reading the slope (the number of counts
>per degree C for that temperature) and reading the Count Remaining, the
>actual temperature may be calculated by:
>
>Temp = Temp_Read - .25 + ((CountPerC - CountRemain)/CountPerC)
>
>I tend to think that the Count Remaining should not exceed the Count
>Per degree C (Slope). But that is not my case. My slope is fairly
>stable ?? at 75 @ 20C, but will vary up to about 125. The Count
>Remaining, however is all over the place. Between 30 to 500. And there
>seem to be no relationship of having a slope of 75 and Count Remaining
>of 30 or 500. Needless to say, the higher resolution is unusable.
>
>Has anyone tried using the DS1620 in this matter ? Can it be done ? Any
>available BS2 code I can try ? Can anyone shed some light on anything I
>might be missing ?
>
>Bruce Snowden