More cheap temperature sensors

[Nofossil]

I finally got around to playing with the National Semiconductor LM35 temperature sensor, These guys are really inexpensive and quite accurate: Less than $2, and typical peak error of less than 0.8 degrees C.

They have three wires - power, output, and ground. Power can be anything between 3v and 35v. Output is 10mv/degree C, so any meter becomes a temperature gauge just by shifting the decimal point. With no external components, measures 2 to 150 degrees C. With a couple of resistors, can be converted to Fahrenheit output and/or extend the range to -55 degrees C.

This one is simple enough for a non-electronics type like me to get working, though it looks like a bit more fussing may be necessary if using it with long cables. I'll do experiments and update.

In my application, effective resolution is 0.1 degrees Fahrenheit and usable range is about 35 to 300 degrees Fahrenheit.

For any fellow computer geeks, I've got one set up on channel 0 of my development system. If I have it running, you can see it at http://www.cedarlakedesign.com:7261/cgi-bin/7260client. Right now it will show you the temperature of my lab workbench.

Source code for the applications that read the temperature and create the web page are at http://www.nofossil.org/armsw/7260devel/

 

[Nofossil]

Here's a photo of the almost-world-famous Nofossil electronics workbench and a closeup of the development system.

The TS7260 development system that I mention above is in the extreme bottom right corner of the photo. The computer consists of a TS7260 CPU board, a TS9700 8 channel analog input board, and a DIO-64 64 channel digital I/O board. It's running in this picture. I've got it bolted to an aluminum plate that's held to the workbench frame with magnets for easy tinkering. The LM35 temperature sensor is hidden behind the soldering iron cable, but I also include a close-up of the computer with a yellow circle around the LM35. The orange wire is 16vdc from the same wall-wart supply that powers the computer.

This is the same board set that I use for my heating system controller.

 

[Hansson]

Thats a nice setup.I am going to build my own system when i got the time.

Then I want to have a graph of my electricity meter too.

 

[Medman]

Nice setup! I have something similar at home. I am surprised to see an oscilloscope in your shop - that's quite a tool for a "non-electronics" guy to have!
As a technologist responsible for over 2000 electronic medical devices in a hospital, I can tell you that an o-scope is quickly becoming a tool for dinosaurs. Of my entire fleet, only two devices now require an o-scope to perform quality assurance or repair. It has come to the point that each time I need it, I have to get the manual out to learn how to use it again.

Thanks for all the great info!

 

[mtfallsmikey]

Is that a Tektronix scope?

 

[ManiacPD]

At the risk of going slightly off topic, I am going to try a Maverick ET-73 grill/smoker wireless thermometer. I've added the link where I bought it for review. For my application I thought this would be a good start for monitoring temperature on the send and return loops. I'll be firing up in the next few weeks so we'll see if it does what it's supposed to do. It was under $50 with shipping and I can use it in the summer to make sure I don't overcook the T-bones!

http://www.partshelf.com/maet73.html

Maniac

 

[SteveJ]

nofossil,

I am so jealous of your setup...

Some of us have to have a virtual lab using online simulations, for example for K thermocouples for the high temsp stuff:

http://designtools.analog.com/Thermocouple/Main.aspx

 

[Nofossil]

I can tell you that an o-scope is quickly becoming a tool for dinosaurs

That would be me ;-)

It's a Tek 2246. The DOD dumped a ton of them on the used market a few years back. I picked it up off of eBay for a song. I use it mostly for audio circuits. I used to manage a team of engineers (if anyone can truly be said to manage an engineer) and a lot of the work we did was at the intersection of hardware and software. I learned to use a scope in self defense.

I am going to try a Maverick ET-73 grill/smoker wireless thermometer.

Those wireless thermometers are a great way to give yourself a simple and effective monitoring system. My brother has a fleet of them. Only comment - don't put the sensor in water for prolonged periods.

Some of us have to have a virtual lab using online simulations

A chunk of my life has been spent doing simulation software. It's good stuff, but sooner or later you have to find out what happens in the real world. When that time comes, us paleolithic types want to have our trusty oscilloscopes close at hand.

Thanks for the kind comments. I felt I should share a bit of the behind-the-scenes work that I'm doing in case it might help others working the same issues. I have a mountain of work in front of me, but getting my lambda sensor installed is on the list - more stuff to share later.

It's interesting that as far as I can tell from my logs, no one has clicked through on the link to my development system :-(

 

[Willman]

It’s interesting that as far as I can tell from my logs, no one has clicked through on the link to my development system :-(

In my case, I have enough to unerstand on your regular pages, let alone a development page. Is that where you have pictures of all the houses you have built ?
Will

 

[SteveJ]

nofossil,

I already stole (oops borrowed, looked at) all you dev code and have started modifying the server code

I am still working out the new temperature sensors and the thermocouple circuitry.

Oh yeah, after I get all the controls done, I will start working on the plumbing...

Thanks for yet another great post,
Steve

 

[cguida]

Hello Nofossil,

Very pleased to hear about the LM35. Last spring I sent off for several of them, but they are still in the package waiting for me to get around to trying to figure out how to use them. Three lines, a battery, and a meter.... even I might be able to do that! Any thoughts on how to get water temperature from the outside of a copper pipe? Could you just tuck the IC iinside some pipe insulation? Or is there some thermal paste you could use to get good, consistent contact with the copper? Also, is there a practical way to seal the whole IC up and submerge it into a hot water tank?

On a slightly different subject...
My $35 Arduino MPC/LM35 control system exists so far only in my dreams. I've been busy finishing the PBVermont Memorial Hot Water Storage Tank. Pleased to report that the tank is complete and 'on line'. For control, we are just turning the circulators on and off by hand. One thing we notice right off -- to actually get heat into the tank (via 40 plate Flat Plate Heat Exchanger), there has to be a significant temerature difference across the heat exchanger -- 10 to 20 degrees -- the more the better. Apparently you can't just leave the circulators running and expect the tank to charge up with only a few degrees temperature difference. This apparent property of flat plate heat exchangers is causing us to make a zillion trips up and down the basement stairs. New miracle weight loss work-out!

Somewhere on this forum I thought I read that you were using a flat plate heat exchanger, and had experienced something similar -- i.e. no temperature gradient; no heat transfer. Have I got this right? What do you think is the right (or minimal) temperature difference to shoot for in order to actually get the tank temperature to rise? Can you recommend threads that deal with this?

Also, is there a thread that deals with flat plate heat exchangers, tank turbulance, and stratification? I thought I read somewhere that if you are pumping water in and out of a stoage tank to transfer heat as in a flat plate situation, stratification is not an option, and that the temperature in the tank will be kind of homogonized. Have I got this right?

Looking forward to hearing more about using the LM35!
Regards,
Christopher Guida

 

[free75degrees]

Also, is there a practical way to seal the whole IC up and submerge it into a hot water tank?

I spent some time, trial, and error getting this to work. I used a differnet device, but i imagine that the package isn't much different. Here is what I tried:
- 1/4" copper tube with one end sqeezed shut then sealed with solder. The device slid inside with a wire protruding. The wire then connected to rubber underground cable and the connection selaed with shrink wrap then liquid electrical tape. The liquid electrical tape was rated to 200*. It worked great at room temp water but then failed miserably at 170*.
- Same 1/4" copper tube with sensor inside but this time with the open end in a brass compression fitting that mates to brass reducing fittings to a 1/2" NPT to pex adapter. The same rubber cable as above connects to the sensor and feeds through a 1/2" pex line. The pex extends aboce the surface of the water so that no wire is exposed under water and the pex is shrink wrapped to the wire above the water level. I still have one of these in service as the bottom of tank sensor, but the other ones failed.
- Same as above but with the compression fitting and copper tube replaced with 1/4" brass nipple into which the sensor is inserted. The end of the nipple is capped. These work great, and i doubt they will ever fail.

 

[BrownianHeatingTech]

You know, I think I'm going to give these contraptions a try.

I have three installations going right now using pressurized tanks. Two tanks on one of them, and three tanks on each of the other two. So, I have eight tanks, and two sensors per tank (upper and lower), for balancing purposes - there are one or three operating sensors per tank assembly, but I like having extra information about what is going on, when balancing flow.

I can buy a whole bunch of LM35's and build a two-temp display interface with a selector switch, for much less money than 16 pre-built sensor/display units. Just include an connector for the interface box, and when I show up to check the system, I can connect onto it and cycle through the tank temperatures with a three-position switch.

The displays on the operating sensors (Ranco ETC and internal boiler-control sensors) are more than enough information for the end-user...

Joe

 

[webbie]

I'd bet there is a market for some multi-channel wireless systems with the higher temps of wood, pellet and boiler systems taken into account.

Wireless is really the key.

I noticed that some Euro pellet stoves allow control by cell phone......that is cool.

For those of you who want to watch or develop, it seems that this new standard alliance:
http://www.zigbee.org/en/index.asp

is coming up with the idea of wireless systems which are very user friendly.....no config, etc.
The first products are just hitting the market now.
Homeheartbeat
Ingrid systems

 

[Nofossil]

Replies to several posts:

Flat plate heat exchangers, like all heat exchangers, need a temperature difference to work, and the more the better. Using them with storage creates another tradeoff - higher flows give you more heat transfer and less stratification, lower flows give less heat transfer and better stratification.

My development system is still running - the link in the first post will give the temp on the floor of my development lab via an LM35 on a 25' cable.

My solution for submerged temp sensors is 3/8" soft copper flattened and soldered on one end. Welded a 3/8" nipple through the tank wall, and connected the soft copper with an NPT to compression fitting. Works so far - only two years, though.

I've seen the Zigbee hardware. I think there's an interface for it that works with my controller. High cost per channel, though. Would be good in applications where physical cables are difficult.

My system is also controllable by cell phone, if you have a cell phone that can display web pages. Have to be careful to restrict access to control functions, though. There's a real risk there. It's scary enough that people out on the net can tell when someone is taking a shower here.....

 

[BrownianHeatingTech]

Any good sources for the LM35's?

Having trouble finding them, other than by ordering them straight from National, at a too-high price...

Joe

 

[Nofossil]

Any good sources for the LM35's?

Having trouble finding them, other than by ordering them straight from National, at a too-high price...

Joe

As with all things electronic, Digi-key is a good source. Need a $25 order to avoid a handling charge, but unfortunately that's never been a problem....

 

[cguida]

Re: where to get LM35 (34)

I got mine from www.jameco.com

Actually, I got LM34, which is the Fahrenheit version.

Enthusiasts for this sort of thing should take a look at Arduino PLC - a capable, open-source $36 STAMP-like controller that you program via PC USB cable.

http://www.arduino.cc/

Let me know what you think. I hope to have something to report later this winter after I get caught up on other pressing stuff. Meanwhile, I am pleased to report that my (nominal) 900 gallon tank is north of 60 degrees. (We shifted to a higher class of junk wood -- almost dry!). Still -- alot of trips to the basement with out a control system...

 

[BrownianHeatingTech]

As with all things electronic, Digi-key is a good source. Need a $25 order to avoid a handling charge, but unfortunately that's never been a problem....

Heh. I remember why I stopped getting their catalog, now. Thousands of pages...

Re: where to get LM35 (34)

I got mine from www.jameco.com

Actually, I got LM34, which is the Fahrenheit version.

Enthusiasts for this sort of thing should take a look at Arduino PLC - a capable, open-source $36 STAMP-like controller that you program via PC USB cable.

http://www.arduino.cc/

Let me know what you think. I hope to have something to report later this winter after I get caught up on other pressing stuff. Meanwhile, I am pleased to report that my (nominal) 900 gallon tank is north of 60 degrees. (We shifted to a higher class of junk wood -- almost dry!). Still -- alot of trips to the basement with out a control system...

Thanks. More good info

Joe