Digital potentiometer problem and questions

The diode stops other parts of the circuit from discharging the capacitor too quickly. The capacitor is large enough that with the load of the digital potentiometer only, it will discharge slowly enough that the digital potentiometer has time to store the wiper position, starting when the supply voltage drops, and finishing before the supply voltage is so low that it can't store the wiper position.

If there are other devices supplied from the same power supply, and no diode were fitted, then the capacitor would discharge faster and the digital potentiometer might not have enough time to store its wiper position.

The C value with a diode shutoff to prevent C decay time then depends on the circuit current.
The IC specified current as ;


and autostore parameters as ;

Thus Cmin = Ic dt /dV = 2 mA * 2ms / 500 mV = 8 uF yet figure 2 in spec. showed only 3.3 uF was required.

Thanks Driver300 for the reasoning behind having the s. diode.

Tony, I'm not sure that I follow your math. How did you get from mA, ms, and mV to Farads? Anyway, it appears that you are suggesting the failure to autostore may be due to too small a capacitor. Is that what you are saying.

Thanks to both of you for your rapid responses.

GH

The data sheet isn't clear about the worst-case conditions for Autostore, so the smallest require capacitor can't be calculated. I would use a larger capacitor to give the IC more time to store.

Tony has assumed 2 mA. The typical is 1 mA and the max is 3 mA.

When the power is turned off, the voltage goes down as the capacitor discharges. When the capacitor gets to 4 V, the Autostore starts, and there has to be 2 ms before the voltage gets to 3.5 V.

The Voltage on a 1 Farad capacitor will go down at 1 V / second if a current of 1 A is drawn from it. That's the definition of a Farad.

In this case, I would take the current as being 3 mA, and the voltage has to go down by 0.5 V (or less) in 2 ms, so the rate of change of voltage can be up to 250 Volts / second. As the current can be up to 3 mA, the capacitor needed is 0.003 / 250 = 12 μF. Tony got a result of 8 μF by using 2 mA as the estimate of current.

3.3 μF may well work, because the typical current is 1 mA and the Autostore could take less than 2 ms, and it could work OK down to a lower voltage than 3.5 V.

I would use something like 22 μF to be sure. It's still a tiny capacitor and large won't hurt.

To complement the excellent replies already provided by Diver and Tony, just make sure that you use a Schottky diode. The 1N581x if you prefer THT devices, the BAT54x if you prefer SMT devices.

The present diode is a BAT54w.

I will try a larger capacitor. I have to dig around and see what I have. I know that I have some 10uf and some 100uF but don't know about in between.


Anybody know anything about programming a ATTINY85? After I get this autostore issue figured out, I need to get an ATTINY85-20 to read an external value, the x9511 wiper, and non-volatile store that value.

Sorry for not getting back with everybody sooner. It took me some time to try out the suggestion that were made.

I'm happy to report that everything is working as advertised, as it should. I used a 10uf capacitor and the X9511's autostore worked most of the time. I did a little more study of the setup I was using and found that my power source was short on voltage supplied. It should have been 5.0v but most of the time it was short of that by a tenth of a volt or so. I changed to a more accurate power supply, set it at 5.1 volts and after that it autostores every time.

Thanks to everybody for your help.