I have built a simple differentiator circuit that outputs rate of temperature rise as measured by a thermocouple. That circuit is working, but I have a need to calibrate it against a reference.

Ideally, I would like to be able to generate a reference sawtooth signal that smoothly ramps 5mV over 15 seconds (yes, a low, slow ramp).

My online searching so far suggests that 555-based oscillator, using a constant current source might be the way to go:

http://www.circuitstoday.com/saw-tooth-wave-generator-using-ne555

My present thinking is to then scale that output by around 1:1000 using an op amp with voltage divider and feedback loop?

Before I get too far along, I would really value suggestions from this group.

Thanks in advance.

Jim

 

I have built a simple differentiator circuit that outputs rate of temperature rise as measured by a thermocouple. That circuit is working, but I have a need to calibrate it against a reference.

Ideally, I would like to be able to generate a reference sawtooth signal that smoothly ramps 5mV over 15 seconds (yes, a low, slow ramp).

My online searching so far suggests that 555-based oscillator, using a constant current source might be the way to go:

http://www.circuitstoday.com/saw-tooth-wave-generator-using-ne555

My present thinking is to then scale that output by around 1:1000 using an op amp with voltage divider and feedback loop?

Before I get too far along, I would really value suggestions from this group.

Thanks in advance.

Jim

Yep...that's the traditional way to go.

By the way...if you have a lot of gain, you can fake it with a VOLTAGE feedback....just amplify the daylights out of the "pseudo-linear" portion. This is how most function generators do it anyway.

eric

 

Just keep in mind that, if the ramp reset time is short, the derivative will be hellaciously large compared to that of the 15 second ramp..

 

I don't think the diode on pin 3 and pin 5 is needed. It would work, the wave form would be 1/3 to 2/3 of the power supply. You would need a tightly regulated power supply, of course, to control the amplitude.

 

How about utilizing the following circuit:



It have been able to generate sawtooth signals higher than 1MhZ with that circuit. However, the electrical componets have to be changed appropriately to generate a sawtooth signal for a given frequency..

 

Go with a TLC555, the TI version of a CMOS 555. It is rated up to 2Mhz.

 

Thanks for the helpful suggestions. Attached is the circuit I ended up using -- successfully.

Recall that I needed a reference signal that was changing linearly at a rate of approx. 20mV per minute.

I noted a small amount of non-linearity in the output (rate of capacitor voltage change varied from ~23 mV per min to ~20 mV per min as the capacitor voltage changed from 1.00V to 2.00V). But the reference signal was still usable.

Jim