you could add a 2nd opamp to buffer the input or scale the RC for quicker response or both

I've already breadboarded a version that dispenses with the 80TK. Instead, direct thermocouple input is run through an op amp. This 2nd op amp provides both amplification and buffering, and does indeed eliminate the capacitor charging problem. (It also has introduced an unacceptable amount of noise, but I expect this to be solvable with a simple RC filter).

I hadn't thought about buffering the input from the 80TK Should work like a charm. Thanks.

Jim

Did my first test runs today. Ran the thermocouple bead through the same sleeve and just slightly below the BT bead.Turned it on while I was pre-heating. It settled plenty quick enough. It shows a -rate after dropping the beans then +rate after the turnaround. It's slightly jumpy at times, a degree or 2, but is easy enough to tell the average. I will try separating the TCs a bit to see if that quiets it some. It really gave me a different perspective of the roast profile. I like it !!!

The reading on the DMM is during ramp showing 19.6 degrees/min. rise rate
next I'll make a control panel box to tidy it up.

farmroast wrote:It's slightly jumpy at times, a degree or 2, but is easy enough to tell the average.

I'm investigating adding a simple filter. Will let you know if I have any success.

Thanks for the update. Glad it's working for you.

Jim

JimG wrote:I'm investigating adding a simple filter. Will let you know if I have any success.

Thanks for the update. Glad it's working for you.

Jim

If you note the 80TK uses several stages of op-amps at much lower gain and with some active filtering in the amp stages themselves to keep the noise threshold down. You could be seeing any number of noise sources if you try to use a single high gain amp stage.

Hope that helps you tame it.
-Cecil

I've now done 15-20 roasts using JimG's circuit. Today it was 8 back to back roasts only 2 being the same bean and they were at different levels. Was weighing, bagging and labeling at the same time to get them sent out to a friend. The rate meter was nice to be able to glance at to check the speed. My TC reacts pretty fast so I'm starting to see some of the reactions right as they're happening. I can tell that with time I'll start thinking roasts as speed profiles along with the ET/times/voltage profiles. I can on-board data log if I want to save the ET/BT data. No on site laptop needed. For my way of roasting and roaster I'm really liking it. Thanks again Jim.
Ed

If it works, well then don't fool with it. This sounds like a good data point for not fooling with it.

Thanks Ed.

-C

JimG wrote:I've already breadboarded a version that dispenses with the 80TK. Instead, direct thermocouple input is run through an op amp. This 2nd op amp provides both amplification and buffering, and does indeed eliminate the capacitor charging problem. (It also has introduced an unacceptable amount of noise, but I expect this to be solvable with a simple RC filter).
Jim

So do the rest of us lurkers get to see the circuit that eliminated the 80TK?
-Ed

farmroast wrote:For my way of roasting and roaster I'm really liking it. Thanks again Jim.

Ed, that's very nice to hear. I still have plans to tweak the circuit, but am having trouble getting a roundtuit. Knowing that it's working as-is takes a little pressure off

Jim

decaf_Ed wrote:So do the rest of us lurkers get to see the circuit that eliminated the 80TK?

Can't get it to work right. The noise from the t/c amp output (I am trying to use a LTK001) is so great that it renders the differentiation circuit useless. No luck so far with RC filtering.

Sorry.

Jim

PS - here's a link to the Linear Technology datasheet for the LTK001. I have tried it both with, and without the cold junction compensator (CJC is largely unnecessary for the differentiator). I am getting "correct" output from the LTK001, but it is very noisy.