How to use a 2 terminal rheostat as voltage divider?

[pavjayt]

This might be a very basic question. I am wondering if there is any way to use a 2 terminal rheostat as a voltage divider. I am trying to use AD5174 as a 5V DC divider. AD has other chips that are digital potentiometers, but most of them are of very high tolerance except AD5292, but its not designed to work at its max efficienfy for voltages below 15V. AD5174 is very good in it resistance repeatability and tolerance, but it was only a rheostat.

Any suggestions?

thanks

 

[chemelec]

Put another resistor in Series with it, Than take off your output between the Two.

 

[MikeMl]

Put another resistor in Series with it, Than take off your output between the Two.

That will affect the linearity, however. Look at the attached sim, where a 10K variable resistor is put in series with 100, 1K and 10K. The voltage at X, Y and Z is plotted vs the resistance R (the variable resistor). Note that the only way to get a linear V vs R (see V(W), dark blue trace) is to drive the variable resistance with a constant current source.

 

[pavjayt]

thanks all for your advise, I ended up using AD5292. Even though it needs calibration due to its tolerance values, I guess its minor for us.

BTW, I have one simpel question. What does they mean when they say connect to 0V? On AD5292 if you are using single supply, one should connect Vss to 0V, does that basically mean to connect to GND? I would like to use it for a 15Vsingle supply.

 

[ccurtis]

On AD5292 if you are using single supply, one should connect Vss to 0V, does that basically mean to connect to GND? I would like to use it for a 15Vsingle supply.

Yes.

If don't need a precise absolute resistance across the ends of the potentiometer (as perhaps with a voltage divider application such as yours) then you might like this one. Linearity is ±2%. Pretty good. https://www.electro-tech-online.com/custompdfs/2012/09/fn8159.pdf

 

[pavjayt]

thanks for the suggestion. It looks really good compared to AD or maxim pots. But the part that I am using it for states as follows

"Voltage control input: 0V to 5Vdc input into 10K ohm "

So, I am not sure if I can use this 100K one.

 

[ccurtis]

If the voltage divider is looking into 10K ohms, then, yes, a 100K voltage divider would have a loading problem without a buffer. But, so would a 10K voltage divider for that matter.

Maybe your resistor tolerance issue is more a problem that the part you are using is loading the divider's voltage down because your part is draining some of the current from it.

 

[pavjayt]

I am trying to use AD5292 as a replacement for an analog panel pot that we have been using in our old designs. The analog pot is of 1K and acts as a voltage divider for 10V. Now I switched the 3 pins of analog pot to digital pot (20K) pins and as soon as I connect it to my system, the voltage accross W and B pins drop down to less than 50mV even at max pot value. This looks to me that the 20K is the reason, any suggestions or ideas to get over this problem but still use AD5292 to have a digital control of 10bit resolution?

If it helps, the W pin is connected to VRef pin on AD5733.

thanks in advance

 

[ccurtis]

I can't find what a AD5733 is. A schematic would be helpful also. The spec sheet for the AD5292 shows an absolute maximum rating of ±3 mA on any pot pin. Are you within that rating? Many old designs with mechanical pots have currents far above 3 mA through them.

If the pot (voltage divider) is looking into a 10K load, as you hinted at earlier, you will be within the ±3 mA rating and you should be able to get almost 10 volts to the load with the wiper set all the way toward the +10V end, not to mention that you hinted that the input should be 0-5 volts, not 10 volts. Maybe by exceeding 5V at the input to the load, the load clamped the excess voltage and the max current on the pot pins was exceeded, possibly damaging the AD5733?

 

[pavjayt]

Sorry, its AD7533. It looks like I got it correct and its working fine now except that even if I had 15V between A&B terminals, the max voltage I can get at W is around 12V which I also fine in my application.

 

[ccurtis]

I'm glad you got it working. Of course, you realize that with the 10K ohm load that the AD7533 places on the wiper of the 20K ohm dig pot, the Ref voltage will not nearly be linear for each step of the dig pot. The original 1K pot saw relatively much less loading effect and so is much more linear in operation.