accurate tuning of RC time constant

[opamp741]

Hi all,
I want to acheive a 2% accurate RC time constant with only one external components..can someone plz help me with the tuning techniques available to acheive this kind of accuracy.
thanx

 

[Btrend]

I got a idea,
u need a bandgap (Vref=1.25V) and a external resistor, to generate the reference current Iref, then use this Iref to generate a
Ich=Iref*(b0*2^-8+b1*2^-67+...+b7*2^-1) ........ which can be tuned through b0~b7 (8 bit resolution).
then implement a Cap (Cx) either poly cap or MIM cap.
==> charge Cx through Ich
==> whenever Vref=Ich*T/Cx reach, reset the charge to dischage very quickly.
==> u got what u want T=Vref*Cx/Ich.

if Cx has variation upto 10%, u can tune ur Ich to cancel this variation

hope this will help u

 

[opamp741]

i just got somewht similiar concept from US Patent 5638031. I also come accross the data sheet of LTC1799.. they r also using similar idea..
But the big problem is tunning. How to tune the current with respect to capacitor variation.

plz help..

 

[Kral]

Use an accurate voltage source.

Fully discharge the capacitor by shorting it to the reference common. Keep the short applied long enough for the effects of dielectric absorption to dissipate.

Apply the reference voltage, and measure the time required to reach (1-1/e) times the reference voltage (approximately 0.63212 Vref). An accurately calibrated comparitor is useful for this.

Scale the resistor value to achieve the desired Time constsnt.

Your main problems will be:
Capacitor variation with temperature.
Capacitor insulation resistance decrease with temperature.
This can be mitigated by choosing R << Rinsulation.
Capacitors with the lowest temperature coefficient are:
Polystyrene (Only good for temperature <85 deg C)
Teflon (Expensive and physically large)
Polyporopylene
Polycarbonate (Smallest physical size, but as good as above)
NPO or CGO Ceramic (Low temp coeffecient, but high dielectric absorption)

Avoid Tantalum, Aluminum. They age, and have high dielctric absorption.

Added after 2 hours 26 minutes:

Correction to my previous reply:
The statement Polycarbonate (Smallest physical size, but as good as above) should read Polycarbonate (Smallest physical size, but NOT as good as above)