Lliam20789
New Member
does anyone know of a cheep and small temperature sensor that is easily adaptable to PICAXE. By temperature sensor I mean one of the ones that can sense from a distance. I'm looking and a 3-cm to 4m range.
Temperature Sensor
- Thread starter Lliam20789
- Start date
Michael 2727
Senior Member
Temp range ?
Moving / fixed object ?
Accuracy ?
Etc ?
Etc ?
Moving / fixed object ?
Accuracy ?
Etc ?
Etc ?
Lliam20789
New Member
The sensor will be on a robot so slow moving. Any range that covers average temp readings. eg., -10 to 50 degrees C.
The accuracy doesn't have to be that great. within a few degrees would be nice.
The accuracy doesn't have to be that great. within a few degrees would be nice.
The DS18B20 is such a neat 08M "mate" that naturally this should be considered, although cost may daunt? It's normally used for contact measurements - can you perhaps focus the heat source onto it with a simple lens?
I also echo the need for MUCH more info however, as numerous alternatives exist. Exactly WHAT are you trying to do ? I've seen small "fire fighting" robots that will locate & try to put out a candle- is this your need too ? Stan
I also echo the need for MUCH more info however, as numerous alternatives exist. Exactly WHAT are you trying to do ? I've seen small "fire fighting" robots that will locate & try to put out a candle- is this your need too ? Stan
Michael 2727
Senior Member
I have a US or UK Electronics magazine
somewhere here with a project that uses
an LM35 or something similar and as Stan said
it uses a lens/reflector to focus the heat
onto the transistor/sensor.
It may work with a DS18B20, and the object
would probably need to be quite a few Deg above ambient.
I have been thinking of experimenting with
this, when I'm bored one rainy day.
Try Googling Thermopile sensors, they are
in the PIR thermo sensor family.
somewhere here with a project that uses
an LM35 or something similar and as Stan said
it uses a lens/reflector to focus the heat
onto the transistor/sensor.
It may work with a DS18B20, and the object
would probably need to be quite a few Deg above ambient.
I have been thinking of experimenting with
this, when I'm bored one rainy day.
Try Googling Thermopile sensors, they are
in the PIR thermo sensor family.
Lliam20789
New Member
Hi guys.
I googled those sensors, but I'm not exactly sure what I'm looking for. I't doesn't really seem like many people have needed this concept but the info is good.
What I am trying to do is have the sensor mounted with a SRF05 on a robot that sends it's data to a computer, which then processes the data and draws a crude map. Instead of just having black lines as walls I thought it would be nice to have a colour coded arrangement. This would mean heat sources would produce, say a red line and cooler object would create a blue line.
I googled those sensors, but I'm not exactly sure what I'm looking for. I't doesn't really seem like many people have needed this concept but the info is good.
What I am trying to do is have the sensor mounted with a SRF05 on a robot that sends it's data to a computer, which then processes the data and draws a crude map. Instead of just having black lines as walls I thought it would be nice to have a colour coded arrangement. This would mean heat sources would produce, say a red line and cooler object would create a blue line.
Aha- well now you're getting into more exotic & costly "thermal imaging". Far infra red ( ~10,000 nm wavelength)- as seen in miliary & & security applications etc- is NOT easy to detect with off the shelf items, as most only respond to the near IR (~1000 nm.) The ambient temperature differences are usually too small to be measured by simple DIY gear. Allow $$$$ budgets I'm afraid.
However simple insights may be possible with a hacked passive IR (PIR)sensor- check Wiki =><A href='http://en.wikipedia.org/wiki/Passive_infrared_sensors' Target=_Blank>External Web Link</a>
However simple insights may be possible with a hacked passive IR (PIR)sensor- check Wiki =><A href='http://en.wikipedia.org/wiki/Passive_infrared_sensors' Target=_Blank>External Web Link</a>
stjosephsopunake
New Member
A 100k thermistor to +ve and a 10k resistor just happens to give degree C reading direct into a bn variable register when you do a readadc on the input pin. Serendipitous is the word… This will not even be affected by battery voltage etc.
Very nice and simple. At the most it just needs calibrating a few degree's by adding or subtracting a few counts.
For greater accuracy you could readadc10 and divide the result by 4 to get whole numbers and // (mod) then x 100 and / 25 by 4 to give a few decimal places.
Very nice and simple. At the most it just needs calibrating a few degree's by adding or subtracting a few counts.
For greater accuracy you could readadc10 and divide the result by 4 to get whole numbers and // (mod) then x 100 and / 25 by 4 to give a few decimal places.
BeanieBots
Moderator
LM35, DS18B20 and thermisters are all contact devices. As stated by manuka, you need an IR device for non-contact. Traditionally, these cost $$$ but are becomming more realistic. eg <A href='http://www.maplin.co.uk/module.aspx?ModuleNo=46033&doy=28m4' Target=_Blank>External Web Link</a>
As also stated by manuka, hacking a PIR sensor might the best option but you will need to do a lot of background reading about the active element and build your own lens which could be done using the front of the PIR sensor.
Not a simple project by any standard.
As also stated by manuka, hacking a PIR sensor might the best option but you will need to do a lot of background reading about the active element and build your own lens which could be done using the front of the PIR sensor.
Not a simple project by any standard.
For high-ish res thermal imaging like Stan says Big Bucks.
But for low res stuff and a little input from yourself you can keep the budget down e.g.:
<A href='http://www.robot-electronics.co.uk/shop/Thermal_Array_Sensors2086.htm' Target=_Blank>External Web Link</a>
(which has a servo output too).
I think this sort of stuff (thermopile) has been mentioned before on this forum. Maybe search on thermopile.
Sadly, you'll have a hard time focussing medium/far IR with your average lens (glass/polystyrene/acrylic) as it will be almost opaque. Security PIRs tend to use a polyethylene fresnel style lens to keep losses to a minimum (and prices too I daresay).
Aside: The acrylic lenses from CTP-Coil are brilliant for near IR (LED/Laser).
But for low res stuff and a little input from yourself you can keep the budget down e.g.:
<A href='http://www.robot-electronics.co.uk/shop/Thermal_Array_Sensors2086.htm' Target=_Blank>External Web Link</a>
(which has a servo output too).
I think this sort of stuff (thermopile) has been mentioned before on this forum. Maybe search on thermopile.
Sadly, you'll have a hard time focussing medium/far IR with your average lens (glass/polystyrene/acrylic) as it will be almost opaque. Security PIRs tend to use a polyethylene fresnel style lens to keep losses to a minimum (and prices too I daresay).
Aside: The acrylic lenses from CTP-Coil are brilliant for near IR (LED/Laser).
Michael 2727
Senior Member
I found the circuit described above,
it was published in Electronics Now, Aug 1998, P: 35.
( magazine long dead and burried unfortunately )
It used a BC846ATL1 Surface Mount NPN transistor, Op AMP etc.
The SMD transistor was mounted at the focal
point of a Strobe Lght type reflector to catch and focus the IR source/heat.
A little too large for the front of any Robot I think.
PIR sensors are fiddly to play with and
usually need
huge amounts of stable amplification to get
good results.
But unless you want to spend a lot of money
they may be the cheapest way to go.
it was published in Electronics Now, Aug 1998, P: 35.
( magazine long dead and burried unfortunately )
It used a BC846ATL1 Surface Mount NPN transistor, Op AMP etc.
The SMD transistor was mounted at the focal
point of a Strobe Lght type reflector to catch and focus the IR source/heat.
A little too large for the front of any Robot I think.
PIR sensors are fiddly to play with and
usually need
huge amounts of stable amplification to get
good results.
But unless you want to spend a lot of money
they may be the cheapest way to go.
How willing are you to hack an existing product?
Harbor Freight (sells cheap tools and stuff) has a non-contact digital thermometer with LCD readout for as little as $7US on sale. <A href='http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=93983' Target=_Blank>External Web Link</a> The beamwidth is proportional to distance (the aperture is about 1cm). It runs 40 hours (measurement time) on one CR2032 battery (you may be able to tap power from your circuit).
I haven't tried hacking into it (yet) but, if the reading can be accessed, it's a cheap way to get 1 degree C accuracy.
John
Harbor Freight (sells cheap tools and stuff) has a non-contact digital thermometer with LCD readout for as little as $7US on sale. <A href='http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=93983' Target=_Blank>External Web Link</a> The beamwidth is proportional to distance (the aperture is about 1cm). It runs 40 hours (measurement time) on one CR2032 battery (you may be able to tap power from your circuit).
I haven't tried hacking into it (yet) but, if the reading can be accessed, it's a cheap way to get 1 degree C accuracy.
John
It's not that expensive...you can get a pocket sized digitizing IR camera for about $40k US <img src="wink.gif" width=15 height=15 align=middle>
OR you can find the thermopile array (I think by dontronics maybe?) and it's designed for interfacing to things like picaxes and whatnot. sorta. It'll take work but it will work. It's a line of 8 independant sensors on a single chip, can detect a candle flame at like 5 meters or soemthin like that. I guess you could even do basic very low res thermal imaging by mounting it sideways on a servo and scanning back and forth.
It's out there. I jsut dont remember what it's called or who makes it. sorry!
--Andy P
OR you can find the thermopile array (I think by dontronics maybe?) and it's designed for interfacing to things like picaxes and whatnot. sorta. It'll take work but it will work. It's a line of 8 independant sensors on a single chip, can detect a candle flame at like 5 meters or soemthin like that. I guess you could even do basic very low res thermal imaging by mounting it sideways on a servo and scanning back and forth.
It's out there. I jsut dont remember what it's called or who makes it. sorry!
--Andy P
Lliam20789
New Member
Due to time I am not willing to do major hacking.
All i need is a simple sensor that can tell me the temperature of whatever is in front of it with a range of about 3m.
All i need is a simple sensor that can tell me the temperature of whatever is in front of it with a range of about 3m.
Beyond the ideas (& limitations) we have outlined above, I can only suggest you Goggle. Our collective wisdom says that you are asking the impossible given low budget, tight time scale & reluctance to "hack".
Have you tried any of the initial approaches we outlined? The spirit of this forum is to help those who help themselves...
Have you tried any of the initial approaches we outlined? The spirit of this forum is to help those who help themselves...
What your looking for is a thermopile. You can get a lot more exotic than that, but it appears your not willing to do so. I'm with Stan along the lines of "what you ask aint gonna happen".
Also, I was wrong in my last post. The TP array isn't by dontronics, it's by Devantech. Heres a link to the Jameco entry on it:
<A href='http://www.robotstore.com/store/product.asp?pid=874&catid=1600' Target=_Blank>External Web Link</a>
That will do exactly what you want, and then some. It's slightly expensive, but relatively dirt cheap. to get more, you need to spend mucho mucho bucks. To get less, well...you get a LOT less.
--Andy P
Also, I was wrong in my last post. The TP array isn't by dontronics, it's by Devantech. Heres a link to the Jameco entry on it:
<A href='http://www.robotstore.com/store/product.asp?pid=874&catid=1600' Target=_Blank>External Web Link</a>
That will do exactly what you want, and then some. It's slightly expensive, but relatively dirt cheap. to get more, you need to spend mucho mucho bucks. To get less, well...you get a LOT less.
--Andy P
Great find! This site even comes with (COPYRIGHTED) PIXACE-18X "thermal radar" software too <A href='http://www.robot-electronics.co.uk/files/TPA81%20-%20Picaxe18X.bas' Target=_Blank>External Web Link</a>
Allow US$101 for the TPS81, ~$10 18X, $10 PCB or breadboard etc etc - make it US$150 =~£60-70. There's a good breadboarded BS2 application <A href='http://www.robot-electronics.co.uk/htm/tpa81bs2p.htm' Target=_Blank>External Web Link</a>
I ran across these little darlings in 2004-5 while researching a "poor mans thermal viewer" for insulation leakage work, but never took it further. They're well thought of for robotic work, but the all up price daunts the fainthearted... Stan
Allow US$101 for the TPS81, ~$10 18X, $10 PCB or breadboard etc etc - make it US$150 =~£60-70. There's a good breadboarded BS2 application <A href='http://www.robot-electronics.co.uk/htm/tpa81bs2p.htm' Target=_Blank>External Web Link</a>
I ran across these little darlings in 2004-5 while researching a "poor mans thermal viewer" for insulation leakage work, but never took it further. They're well thought of for robotic work, but the all up price daunts the fainthearted... Stan
Lliam20789
New Member
Thankyou all for your help and I understand it must be difficult dealing with someone with so many constraints...
sorry guys.
I think this idea is out of my possibilities for now but thanks again!
sorry guys.
I think this idea is out of my possibilities for now but thanks again!