Arduino & Varying Amperage

Forum 2005-2010 (read only) Hardware Interfacing
1

Hi folks. Question for the group.

I bought some analog thermoprobes for BBQ's. Just the probe part with 2.5mm connector:

Upon hooking them up to 2.5V, varying temperatures changed the resistance of the thermocouple inside. It's what I expected.

Solution Temperature (°C) Resistance (Ohms)
6 4700
26 12000
80 19300

As temperature of the solution rises, the resistance rises. So Amperage is what will be affected, not voltage. How can I measure/integrate this with an analog circuit on an Arduino Mega? Not too sure how to do this. I have 5 of them that I would like to get the temperature reading from about every 10 seconds.

Cheers

2

Well if you put in a potential divider circuit with 5v, gnd and the analog pin then the decrease in current will decrease the voltage. V=IR.

Mowcius

3

OK, but I wasn't sure if the amperage was indeed dropping. I don't have an ammeter here. Also, the measured voltage drop was negligible on my meter. Will that be measurable on the Arduino? I was familiar with the law, just wasn't sure how to apply it here, after measuring a tiny voltage drop. I thought it would have been the wrong variable to identify.

BTW I just did another run with many more points. It seems my previous measurements were wrong. I might have mixed them up.

If I CAN rely on voltage changes (I should plot them using the Arduino, actually), that would make things much easier.

Seeing as we're on the topic, which thermometer would you recommend? I bought the TMP102, but it's a difficult item to get into. I like the plug and play.

Thanks for the feedback.

Cheers

4

I was familiar with the law, just wasn't sure how to apply it here, after measuring a tiny voltage drop.

You only measured a tiny voltage drop because you in effect put a resistance in series with the voltage meter. This will show little or no drop as the resistance changes. It will be no difference with the arduino doing the measuring.
As pointed out before you need a potential divider. You put one resistor (use a 22K) from the +5 to the arduino's analogue input and the same input goes into your liquid. Then the other probe in the liquid goes to ground. You will measure lots of change like that.

5

Thanks for the reply Richard.

So the probe acts like a potentiometer, with the analog pin measuring reflected voltage (middle pin on potentiometer).

6

Hi Mike. Thanks for the reply.

I'm confused now.

The probe has a 2.5mm male jack on it. Essentially it's just a variable resistor, an incomplete circuit. What I'm doing is just putting all this in series and measuring any resistance change in that circuit. As stated, voltage was minimal. I'm not familiar with how much my meter would throw into this.

If I get what you are saying:

Arduino Analog Input - 22k Resistor - Thermoprobe - Ground

Yep, I'm confused. I think somehow I need to get the variable voltage from the probe somehow into the analog pin, stealing from the 5V to Ground, but not clear how to do this.

7

I think I got it:

5V---+
|
|
+-------- Thermoprobe ------- Analog Input
|
22k Resistor
|
GND--+

8

The probe has a 2.5mm male jack on it.

So there is two wires going to the probe:-

5V---+
|
22K resistor
|
+--------------- Analog Input
|
Thermoprobe
|
GND--+

9

OK, thanks for the diagram. Do you see any problem running 5 of these parallel from the same 5V and GND pins? I am assuming different resistances from the probes. I don't want one circuit to affect another.

10

With 22k series resistors you can run hundreds of these without overloading the supply!

If you need to determine the probe resistance when doing the maths in the Arduino you'll have to use the analog reading to determine both the voltage and current in the probe - voltage is 5V - input, current is input/22k. input being the input voltage:

  int input = analogRead(pin) ;
  float vin = input * 5.0 / 1024 ;
  float vprobe = 5.0 - vin ;
  float Iprobe = vin / 22000.0 ;
  float Rprobe = vprobe / Iprobe ;
11

Did you get the specs for the probes (or specifically, the thermistor inside the probes)?

The reason why I ask is because generally (from what I have seen), thermistors are typically non-linear in their response curves. That is, if you measure the resistance of the probe when put into boiling water (212F/100C), vs the resistance at freezing (32F/0C) - and plot a line (linear graph), the resistance at the mid-point (122F/50C) isn't what you would expect from the graph.

The graph ends up actually being a curve (logarithmic, I think), and guessing this curve would actually take quite a bit of tedium unless you have the specs (or an accurate automatic temperature probe system, which could possibly be built with an Arduino and a known curve thermistor).

Just something to keep in mind...

:slight_smile:

12

No, not yet. I emailed them, hopefully they'll get back to me soon. They are spare parts from a manufacturer that I had to chase up. From today's measurements, it is Logarithmic. I also expect all of them to vary a little. So I thought of looping and recording the data per probe, and run a known thermometer alongside, so that the Arduino can calibrate the items as it cools, instead of me sitting there adding cold water, as I did. Made for a nice graph though.

Thanks for the code, Mark! I was going to tackle that down the road, but you obviously saw where I was going. Heh.