AC bridge variable resistor drive

Hello g_simulator,

Try 1 fixed 75Ω resistor in parallel with another 75Ω resistor in series with a 5k digital pot. This will give you the range of 37.5Ω to 74Ω.

I know Analog Devices has digital pots @ 5k with 256 positions, both volatile and non-volatile memory, in I2C and SPI digital formats. Check here.

Mike

Mike, thanks for the help! the problem that i have is that using the 75 ohms in series and one in parallel as you have shown i only get about 30 of the 256 positions that fall into my range and they are not continuous nor very evenly distributed. the "image" below is from a spreadsheet that calculates every one of the 256 positions. It is hard to see but i can email the file to you if that would make it clearer for you. Even if i adjust the resistor values (say change r2 to 5k and r1 to 63) to better fit my resistance range, it is such a non-linear distribution that it make it unworkable.

Don't worry about the spreadsheet, I made up my own and plotted R3 values against the "output" resistance. Then I tried 250 ohms for R3 and it was way more linear. If you could find a 250 ohm dig. pot, would that resulting decrease in nonlinearity be enough?

Mike

A 250 ohm digital pot would be great but i can not find anything smaller than a 1k. Any ideas? Maybe i could put multiple 1k's in parallel to get 250 ohms...

Try a transformer to match the rheostat to indicator resistance. The driver xfmr from a 1960's push pull output transistor radio should work well.

i am not sure what you are suggesting. Attached is a simple diagram. are you saying that i can replace Rvar with a small transformer?

You probably need to account for the reactance generated by the 400 hz through Rvar

Yes, exactly. Newark's 02F450 (Paralax SP29) transformer with a 10k digital pot on its 10k winding will reflect a 500 ohm change on its 500 ohm winding. If you use the centre tap and either end of the 500 ohm winding you will reflect 125 ohms. At $20 and 2ml volume it may be acceptable, the old transistor radio trantformer may suit you better.

I would try the transformer with a manual pot, note the resistances required for zero and FSD on the instrument and buy the IC to best suit. There will be a little winding resistance so an external minimum resistance in series in the bridge will be lower. The pot should be directly across the winding for minimum errors.

Good luck!

How about a wire wound rheostat? Like proliferate junk boxes. Tony

my design needs to be something that replaces Rvar but that is controllable either through say a D/A or a direct digital interface. would a rheostat be controllable this way?

Are the results not linear because wire wound resistors are being used?

no, the proposed solution was to use a digitally controlled 5k pot. when "scaled down" (placing a fixed resistor in parallel with the pot) to the correct range of 38 to 62 ohms it becomes highly non-linear.

Try placing an op amp in place of the drive that is across the bridge.

The input to the op amp will be the previous output to the drive.

Then select the proper gain for the desired output of the op amp which will become the final output to the drive.

Very linear, very scalable, very simple.Of course, this requires a small external power supply.Will this work for you?

HTRN

i don't have access to the drive across the bridge, that is inside the instrument. all i have access to is the two wires that go to Rvar that i am trying to "create"...

Try this digital potentiometer that has 6 output pots, each individually adjustable via

Serial Peripheral Interface, with 255 steps of adjustment each.

http://www.arduino.cc/en/Tutorial/SPIDigitalPot

HTRN

From Sleepy,

How many taps do you need or is it to be continuously variable?

I asumed that the n tap version would be best in that you would have n solutions each of which had a separate physical configuration - of you could use a switch? Rotary even.

Then each tap could be a physical resistor (s) of your choice - and completly stable, calibrated as tightly as you wished and so on.

But this may not be what you are looking for,

Good Luck

Sleepy

the guage has a range of -60 to +60 deg C, so it would be nice to have at least 120 taps so that i could drive any temperature reading in 1 degree increments. If it could be an analog solution it would provide full resolution. Is there a way to use some sort of an op amp design that would look like a resistor that could be tied to a D/A for control?