• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Analog Thermometor

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

pinkles

Member
Joined
Jun 11, 2009
Does anyone know a worklog or tutorial or article about building/using an analog thermometer in a computer build? Think like steampunk style!

If you have any ideas of how to convert an analog thermometer from kitchen or car(?) use for use in a computer please let me know!
 
I like the idea. Thinking from the top of my head, you would need to use an electronic analog meter; something along the lines of a VU meter from an old stereo head unit. To actually measure the temperature you would need to use some sort of thermocouple which will produce a voltage which varies with the temperature it is exposed to. Then, you could use a variable resistor (potentiometer) to "calibrate" your temperature meter.

I know it's a pretty vauge suggestion, but it should get you started. I remember a tutorial a while back on making an analog HDD meter in very much the same way. The same basic concepts should apply, just with some minor tweaking to measure temp instead of HDD usage. I'll see if I can dig up that old article.
 
Dude, that would be awesome if you could find that. I have been looking for an excuse to venture into some small electrical project.
 
http://www.scribd.com/doc/3216005/Make-an-analogue-hard-drive-meter

There's a link, although it's not the article I was originally thinking of. The one thing I don't like about this article is that there's no wiring diagram. He goes through the steps but never gives an electrical schematic. I should be able to work something up though. How electrically inclined are you?

Also, if you do a search on ebay for "VU meter" there's some good round bezel, very vintage looking, analog meters that come up.
 
So; I've been doing some more digging on this and I'm quickly realizing that my inital idea of using a thermocouple was not quite correct. What we actually want to be using is a thermistor; a thermistor based sensor will be cheaper and likely more accurate for the small temperature tange we'll be measuring in.

http://www.facstaff.bucknell.edu/mastascu/elessonsHTML/Sensors/TempR.html

The above link is the reference info that I'm pulling from. What we essentially need to do is build a bridge circuit. Using the nomenclature from the above site, our thermistor will be Rs, and our analog VU meter will be driven Vout (+/-). Vc (+/-) can be pulled from the 12VDC line of the PSU. What we need to do now is "tune the bridge circuit" for the thermistor and VU meter we'll be using. Work is pretty slow this week so I might be able to work on this in my spare time and do a DIY tutorial.
 
So if I am understanding this right...

We feed power through this thermistor, through a potentiometer to fine tune the reading on the VU meter, and into the VU meter. The thermistor changes the power flowing through it according to the temp it is experiencing and boom, an analog thermometer!
 
So if I am understanding this right...

We feed power through this thermistor, through a potentiometer to fine tune the reading on the VU meter, and into the VU meter. The thermistor changes the power flowing through it according to the temp it is experiencing and boom, an analog thermometer!

That's more or less exactly the plan. It's a little more complicated since you've got the bridge circuit which has the two parallel branches with the VU meter across them. Essentially though, you've got the gist of it.

I've been bowsing VU meters on Ebay and I think I've found a pair that look pretty suitable for the job. I'm going to go ahead and buy them but they're coming from Taiwan, so who knows how long they'll actually take to get here. Meanwhile I'll start going though the bridge circuit calculations and work up a list of the resistor values needed to (hopefully) make this work.
 
Ah, after briefly reading about bridge circuits I am more than a little confused where they come in to play in this project! The null thing in the center of the bridge would change depending on the resistors around it. And, ohhh, that makes sense! Ahh okay, I see!
 
Triple post >.<

Ok, here is what I am imagining! The green ovals are themistors/resistors. I based this off a wheatstone bridge design.

The more I look at this the more I think something might explode it I actually wired this up...
 

Attachments

  • messywiring.jpg
    messywiring.jpg
    46 KB · Views: 351
Well, I went ahead and bought the following on Ebay:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=140430889104&ssPageName=ADME:L:COSI:US:1123

Two meters for ~$21 after shipping isn't bad; if I can get them to work... I'm thinking a steampunk case theme with a lot of round gauges on the front panel. If I can figure everything out, I plan to use one of these for CPU temp, as we're discussing here, and the other for a HDD activity meter. I'm also working out ideas for a gauge for the static pressure in my water cooling loop, and some sort of sight glass type device to indicate reservoir water level. But, I disgress.
 
That sounds like a sweet case :D I was reading more about bridge circuits and I came across this...

Most electric interference and noise problems can be solved by shielding and guarding. A shield around the measurement lead wires will intercept interferences and may also reduce any errors caused by insulation degradation. Shielding also will guard the measurement from capacitive coupling. If the measurement leads are routed near electromagnetic interference sources such as transformers, twisting the leads will minimize signal degradation due to magnetic induction. By twisting the wire, the flux-induced current is inverted and the areas that the flux crosses cancel out. For industrial process applications, twisted and shielded lead wires are used almost without exception.

Because this will be in a computer will we have any problems with interference from the components?
 
Because this will be in a computer will we have any problems with interference from the components?

It's a possibility since all of the components inside a case do tend to throw off some EMI. This is especially true for something like a water pump if you've got a water cooling system. I would say that twisting the sensor wires certainly won't hurt. Installing a grounded cable shield is probably overkill.
 
Well, at this point the project is at a standstill until I actually receive the gauges that I purchased. Once I receive them I will be able to hook up power supply and a variable resistor to determine the voltage required to drive the meter to certain positions. Once I know the voltage range that the meter functions across, then I'll be able to select an appropriate thermistor and size the other resistors in the bridge circuit for the normal temperature range that the meter will be sensing.
 
Here is a simple schematic I threw together, tell me if I am going in the right direction!

Do you want to work through a hypothetical situation so I can see the formulas we will be using? Like, what do you think will push the VU meter all the way through its range? My guess (total shot in the dark here) would be like 10V. So if R2/Thermistor=Potentiometer/R1 what would our next step be? I think that is the right formula for a wheatstone bridge....


I used a program called EAGLE for making the schematic. I don't understand how to run simulations with it though :/
 

Attachments

  • thermostat.jpg
    thermostat.jpg
    34.9 KB · Views: 311
Pinkles; The schematic looks good, you've got the idea down. I had been planning on making and posting a schematic, but I've not gotten around to it in the past couple of days.

I've got some hand calculations mostly worked out; I can probably finish them up in my slow time today if work isn't too busy. Once I get them finished I'll scan and .PDF them and post them up for everyone to look at.
 
Looking at your schematic, what you are really doing is creating two voltage dividers and measuring the difference. I haven't looked into your project completely, except for the last schematic. The equation you should also consider is:

Vout = Vin*R2/(R1+R2)

Vout = the point between Resistor 2 and Thermistor or Potentiometer-Resistor 1.

for example if you used a 5Kohm for your Resistor 2, and 10Kohm for your Thermistor, then

Vout = 12V*10K/(10K + 5K) = 8V

Likewise, you would want to set up a similar divider on the othe side of your VU meter. When both of the 'voltage divider' values are equal, then you will have 0v across the meter. From the specs listed on the eBay auction, the meter you bought would register 0 decibels on the gauge with 1.22v across the terminals. One other thing to consider is how much current the meter actually consumes when its working, and to choose values for the resistors so that the meter gets enough power.
 
Keep in mind Thermistors act inversly to temperature, higher temp = less resistance, for needle deflection purposes.

They do make analogue dial gauges with remote sensors, but the lead length might be a bit long. Also depending on where you want the sensor might be tough to locate with the industrial ends on it (big and bulky, not going to fit under a heatsink).

Example: http://www.ttnet.net/show_html.jsp/...033378/item_no/61/itno/TL545/dtno/010/type1/A (first image I found to demonstrate what I mean, I was lazy)

Edit: more examples http://www.wika.ca/products_TM_en_ca.WIKA?ActiveID=23804

You could also use a brass gauge and tarnish it for a rustic look?
 
Last edited:
Back