PDA

View Full Version : Temperature Increase in DS1620



DosEdge
01-27-2007, 06:59 AM
Does the temperature reading of the DS1620 suppose to constantly rise if you allow it to continous retrieve the measurements?· The readings begins perfect, but begin to read wrong.· Temp starts at 72 degrees F, then in 20 mins the temp reads 80-82 degrees.· Is this due to power of calculating within the DS1620.· Is this heat added to the readings.·

Chris Savage
01-27-2007, 08:45 AM
As I recall the chip experiences very little self-heating. Are you powering it with a regulated 5V? Also, is it inside a case with the PINK Module? Bear in mind with the amount of current the PINK module consumes you can expect you voltage regulator to get a little hot. Add to that the heat generated by the pink and you could experience temperature rise in a case, or to some degree on the board.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support

DosEdge
01-28-2007, 01:27 AM
Basically, I have nothing enclose in a case.· The DS1620 is located on the BOE with the BS2.· The PINK, BS2, and DS1620 are all· powered through the BOE as well.· I will probably set the DS1620 mode to one shot.· I will have the BS2 to nap until I request it to read a temperature from the DS1620.· Maybe that will help lower the power between all of the devices.

TransistorToaster
05-30-2007, 07:10 AM
Hello,
I am getting a temperature somewhat offset with my DS1620. It gives me 25C, however other mercury and digital thermometers around say 22C. Do any of you have similar findings? Please send me a private message and carbon copy it on the board.
Thanks

Tracy Allen
05-30-2007, 12:43 PM
It is quite possible to see that kind of temperature difference when the other reference temperature sensors are hanging in air while the DS1620 is plugged into a breadboard. Have you looked into that possibility? Fasten the sensors together wrap them together in foil or paper/plastic. You may have to put the DS1620 on its own little carrier. Be sure there is a bypass capacitor right across the DS1620 power supply.

The internal power dissipation of the DS1620 when it is operating in "continuous" mode is around 5 milliwatts, and assuming Tja of 100 degC per watt for DIP package, that could raise the die temperature by about 0.5 degrees above ambient. To eliminate that possibility, put the chip in single shot mode and query it only once every few seconds. It goes into a microwatt sleep mode between readings.

Other things: No loads on the thermostat outputs of the DS1620 that would make it dissipate power, and similarly no pullup or pulldown resistor on the data line.

Finally, are you sure of the calibration of the other sensors? Are they ASTM certified?

It is usually not good form to cross post solicit responses on three fora at once. A processor-neutral question like this is usually best posed in the Sandbox forum.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com (http://www.emesystems.com)

Post Edited (Tracy Allen) : 5/30/2007 4:48:39 AM GMT

Chris Savage
05-30-2007, 10:42 PM
Tracy just reminded me of my first Digital Thermostat project…While prototyping and programming the unit it sat on my desk flat. Later when I mounted it to the wall it was vertical and the temperature readings were a lot different. But the biggest contributor to this was not placement of the unit itself, but rather the fact that I had the voltage regulator mounted below the DS1620. With the backlit LCD, Relay, controller, etc. it cause the regulator to get warm enough to raise the temperature by almost 10 degrees. This is called, “poor planning” and I don’t make that mistake anymore. =)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support

TransistorToaster
05-31-2007, 11:51 AM
>It is quite possible to see that kind of temperature difference when the other reference temperature sensors are hanging in air while the DS1620 is plugged into a breadboard.

I am reading room temperature, the other thermometers are not at all connected to my breadboard. Also, another sensor is me, and what I feel that the temperature is.


> The internal power dissipation of the DS1620 when it is operating in "continuous" mode is around 5 milliwatts, and assuming Tja of 100 degC per watt for DIP package, that could raise the die temperature by about 0.5 degrees above ambient. To eliminate that possibility, put the chip in single shot mode and query it only once every few seconds. It goes into a microwatt sleep mode between readings.

Ok, I'll experiment with different polling rates. I already have it in one shot mode.


>Other things: No loads on the thermostat outputs of the DS1620 that would make it dissipate power, and similarly no pullup or pulldown resistor on the data line.
I let the thermostat output lines unused. I do have a resistor on the data line, but it is in series, so no heat generation.


Finally, are you sure of the calibration of the other sensors? Are they ASTM certified?


>It is usually not good form to cross post solicit responses on three fora at once. A processor-neutral question like this is usually best posed in the Sandbox forum.
I thought about that one. You see, last time nobody at all responded. Or the ones that continued on the threads ignored my question. Probably not the best excuse. Not all people read all the sections of the board.