This is what the breadboarded circuit looks like. It all runs from the 9V battery at left. The stuff at the bottom of the breadboard is dedicated to creating a -9V supply (the op amp needs both +9V and -9V). The stuff in the middle is the op amp, capacitor, and resistor for the differentiator.

The red and yellow alligator clips carry the millivolt output from the Fluke 80TK into the differentiator circuit.

The red and black leads from the banana jacks carry the output from the differentiator into the Fluke 189 DMM.

The DMM display shows a bean temp change rate of +16.1 F/min.

Jim

The less geeky can use two 9v batteries in series for ±9v

JimG
Brilliant!

JimG wrote:While my simple prototype has some minor issues...

There is a very long settling time upon first powering up the circuit. For instance, with the input shorted or when connected to a steady-state input voltage, a large negative voltage shows up on the output -- on the order of several volts. This is "large" in the sense that the expected output in these situations is on the order of a few tenths of a millivolt.

Gradually, the large negative voltage dissipates, and the output terminals read a suitably small (+/- 0.5mV) voltage.

Any suggestions to fix this?

Jim

[EDIT: C1 needs to be a bipolar capacitor. My prototype used a bipolar, but couldn't find the right symbol for the drawing.]

C1 should be NP.

You could add a switch across R1 & 2 to reset or use a FET with an RC turn off circuit

Thanks, Ian.

I understand what you suggested about the switch across the resistors, but will have to do a little research on the FET/RC option. If that would automatically perform the reset, then that would be nice.

BTW, I used a bipolar electrolytic for C1, but I couldn't find one to match in Eagle's library when I drew the schematic.

Jim

Hi folks.

I am working (just started) on a program to interface a Fuji PID to a laptop. One I have the interface program working I can easily do any real time graph (ie. change of rate) or modify anything. It'll be cross platform (Win, Linux, Mac) and free to use for any purpose without any license restrictions.

I am not a programmer myself and it may take a while since I have another unrelated project with more priority going on till April 2010.

regards

Here is the circuit from Jim. Just waiting for a 80TK thermocouple module to arrive then the roastometer will be ready to go . Jim should I put an on/off switch on a battery lead?

Yes, good idea. The battery will last longer.

Jim

Ed -

I'm looking forward to reading about your results, once the 80TK arrives.

In reference to the "settling time" issue I wrote of earlier, I understand it a little better now. It results from the fact that it takes the 80TK a little over a minute to fully charge the big capacitor (1000uF in the prototype) on the add-on circuit board.

At steady-state room temperature, the 80TK output will be ~70mV. Before the circuit can work correctly, the capacitor needs to be charged to match that steady state output. There is a finite time required to charge the capacitor to 70mV. As it is charging, the circuit senses the changing voltage, so your DMM is going to display a (sometimes large) non-zero output from the circuit.

Because the 80TK has either a very low output current capacity, or a high output impedance (maybe both), it takes a minute or more to fully charge the cap from the 80TK. Once the cap is charged, though, then the circuit behaves as expected.

Jim