Having problems reading 1620 temp sensor with SX
mardor
Posts: 3
I have tried....
continuous reads with Rdtmp command (0xAA)
1-shot reads with Rdtmp command (0xAA) (to eliminate internal heat effects)
1-shot reads with Rdtmp (0xAA), Rdcnt (0xA0), and Rdslope (0xA9) (higher accuracy)
In all cases the temp (as compared to 3 other thermometers within a few inches)
is 1.5 to 2.5 degrees too high (Farenheight).
I have 4 1620s and they all read high.
I have 2 SX chips and see no difference
I have also isolated the 1620 via cable from PCB (eliminate heat effects)
I have also tried tweeking the supply voltage up/down 0.1-0.2V
I have also looked at output temp (raw 9 bits)
continuous reads with Rdtmp command (0xAA)
1-shot reads with Rdtmp command (0xAA) (to eliminate internal heat effects)
1-shot reads with Rdtmp (0xAA), Rdcnt (0xA0), and Rdslope (0xA9) (higher accuracy)
In all cases the temp (as compared to 3 other thermometers within a few inches)
is 1.5 to 2.5 degrees too high (Farenheight).
I have 4 1620s and they all read high.
I have 2 SX chips and see no difference
I have also isolated the 1620 via cable from PCB (eliminate heat effects)
I have also tried tweeking the supply voltage up/down 0.1-0.2V
I have also looked at output temp (raw 9 bits)
Comments
-- Wrap the comparison thermometers in aluminum foil together with the DS1620 on the cable.
-- Avoid sources of radiant heat. A difference of 1 to 3 degrees F can occur quite naturally due to radiate sources, different absorption, and different heat capacities. For calibration purposes, it helps to keep the DUTs in a dark box with an air flow, or in a stirred fluid.
-- Are you confident in the calibration of your reference thermometers?
re[noparse]:D[/noparse]S1620
-- In one-shot mode, read temperature infrequently (once per second or less?)
-- Be sure to use a bypass capacitor 0.1 to 1 uF across the DS1620 right at its power terminals.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com