Temperature measurement and logging

yurij

Hi - I am new to interfacing hardware to a PC, but I·need to·measure and record the temperature of 3 outdoor items every few seconds so I want to give building it a try as an introduction microcontrollers.·Which kit - or combination of kits - should i buy to build·this?·thx -yurij

metron9

60 x 60 x 24 = 86,400

86,400 / few (3?) = 28,800 times three we are back to 86,400 measurements per day.

Need to know:

How many days

How big of number for each item.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!

yurij

metron9 - thx for the interest. an observation rate of·1/minute would be fine, and i will only be looking for 12 hours. of course more would be better, but this is a minimum. thx -yurij

Mike Green

Download the "What's a Microcontroller?" tutorial from the Stamps in Class section of the website (here: www.parallax.com/html_pages/downloads/siccurriculum/documentation_sic_curriculum.asp) and the "Applied Sensors" tutorial from the same webpage. These have several examples of temperature measurement. Pretty much any of the Stamps can be used for this. Parallax sells most of the temperature sensors mentioned.

The HomeWork board would be the simplest, but it's only available in packages of 10. RadioShack used to carry the HomeWork board as a single kit and you can sometimes still find them at a discount. Other than that, the Board of Education together with any 24-pin Stamp model would be fine. I would suggest the BS2pe (or BS2p or BS2px) since they have additional capabilities that you don't need now, but may find useful in other projects.

Also, download Stamp Plot Pro from www.selmaware.com which can record data from the Stamp board and would allow you to easily plot it and analyze it.

Steph Lindsay

Thanks for pointing to those Stamps in Class books, Mike, I would have said the same thing but you beat me to it!

Just for the record, RadioShack is still selling the BASIC Stamp Activity Kit which includes a HomeWork Board and the What's A Microcontroller Parts & Text.· You can also order it directly from Parallax:

http://www.parallax.com/detail.asp?product_id=90005

Almost all of the Applied Sensors experiments will work fine on the HomeWork Board, with some care taken to the pump·power supply in the last chapters.· The temperature datalogging activities would be no problem.

-Stephanie Lindsay
Editor, Parallax Inc.

·

yurij

Can i use that kit to create something which will measure 3 different temperatures at the same time (i assume i need to buy 3 different sensors)? Is there some length limitation for the sensor, i.e. can the sensors be placed 20 feet from the rest of the circuitry (the sensors are outside and computers is inside the house) ? thx -yurij

Andy Lindsay (Parallax)

Here's a good thread about line length with the DS1620 temperature sensor:

http://forums.parallax.com/forums/default.aspx?f=5&m=177978&p=1&ord=d

Yes, you'll need to buy a couple more DS1620 sensors (assuming we are talking about measuring air temperature). The temperatures cannot be measured simultaneously, but they'll only be a few milliseconds apart, which should be fine what you're doing.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Andy Lindsay

Education Department
Parallax, Inc.

yurij

So from the replies I should buy the Education kit and as many temp sensors as i need (3)and·I can measure·the senors milliseconds apart (which is fine).·The last piece of the puzzle is that i need·better than 1/2·degree accuracy. maybe 1/10 is fine. I plan to measure the temp difference between the ambient air and a telescope mirror as the evening progresses to an accuracy of about 1/4 degree in their difference, hence the 1/10 accuracy for each sensor. i don't think thos sensors are up to it. which are? thx again

sam_sam_sam

yurij

Hi - I am new to interfacing hardware to a PC, but I·need to·measure and record the temperature of 3 outdoor items every few seconds so I want to give building it a try as an introduction microcontrollers.·Which kit - or combination of kits - should i buy to build·this?·thx -yurij

I have· temperature· humity data logger and the way that it is setup is that it only log the changes
that way you do not have so many data points you only have changes
·
My set point ..........5* F Data log change and i can change the set point any thing that i need it to be
I can also set how many seconds between data points........... every 10 seconds

This was done with Two Basic Stamps
·
Let say that you temp reading like·this
·
·10:00:30·· 75.5·················· ·start·· 75.5· The Logger Data that you would have and you could have a
·10:00:40·· 75.5··························· ·76.1·· Time Stamp for each line
·10:00:50·· 75.6················· ··End····75.4··· This way you would not use up as much memory this way
·10:01:00·· 75.7
·10:01:10·· 75.8
·10:01:20·· 75.8
·10:01:30·· 76.1
·10:01:40 · 75.7
·10:01:50· ·75.8
·10:02:00·· 75.9
·10:02:10·· 75.4
·
·I hope that you can under stand what i have put here and i hope that this help you in what you want to do

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
··Thanks for any··that you may have and all of your time finding them

·
·
·
·
Sam

Post Edited (sam_sam_sam) : 5/23/2007 3:20:57 AM GMT

yurij

I am suprised you needed to use 2 stamps. why? how do you attach it all together? seems the education kit would not be suffcient for this. and is the accuracy (1/10 increments) in the temperature measurement you posted reliable? or is it really +/- 0.5 degree and its just the illusion of accuracy. i don't really care what the absolute temperature is. i only care about the difference between 2 different sensors. and since they are independent devices, they each need to be pretty accurate. i think. thx·

Tracy Allen

Hi Yurij,

In high resolution mode the DS1620 can read out to better than 0.01 degree, and with individual calibration, it is stable enough to achieve accuracy of 0.1 degree.

It is kind of bulky though for measuring the temperature of a surface. Maybe it can be fastened to back (?) surface with a heat transfer adhesive.

Another option is a one-wire sensor like the DS18B20, which is available in a smaller TO92 package. Its performance is similar to the DS1620, but it uses a OW interface. For that you would need one of the "advanced" Stamps (BS2p, '2pe or '2px), which include the ability to read OW devices.

For measuring temperature differences, and especially for measuring temperature differences involving small objects and surfaces, thermocouples are king. They can be fabricated from very fine wire that easily equilibrates without disturbing the system under test. A pair of thermocouples can be hooked up back to back, so that their voltages cancel when they are at the same temperature. When they are at different temperatures the voltage is proportional to the temperature difference. You would use another sensor (like a DS1620) to measure the ambient temperature accurately, because thermocouples do not provide great absolute accuracy. Thermcouples produce a very low voltage, so they require a high gain amplifier and an analog to digital converter to read the temperature difference. Parallax sells an interface that would be suitable, but it uses the OW interface available only on the advanced stamps. Or you could go the route using a traditional ADC.

There are analog temperature sensors that are quite accurate or can be made so with individual calibration, for example, the AD592 or the LM34.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com