Control Feedback Question 1

[BrownPanda]

Hello, I'm new to this forum and have been reading through some of the other questions people have posted. Great community here.

I had a question about a circuit I am trying to build. Basically I am applying a sine wave to three "loads" (we can think of them as basic resistors). Now, we found out that these loads will change with temperature and other effects. What I want to be able to do is to maintain the same current across all loads by varying the voltage. I started with sensing the current across the loads by using a series resistor (.05Ohms) and amplifying it once then feeding that into a differential amplifier. I am using load one as a reference point to which all other loads will be compared to. When I go and breadboard this circuit, the current across all the loads isn't exactly the same. If i place a 30 ohm load as a reference load and a 35 ohm load as load 2, the difference is 30-40mA. What can I do to improve this circuit to achieve accuracy of less than 5%? Do I need to add a PID controller or something?

I attached an image of the circuit I have made so far.

Thank You

 

[MacGyverS2000]

Look up current mirrors... lots of useful info, and significantly more simple than introducing a controller of any sort into the equation.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[ScottyUK]

Your loop gain is quite low. If you want to drive down errors using a proportional-only loop then you'll need to increase the gain.

You may need to consider component tolerances too, depending on what you are already using, and/or incorporate trimmers to allow manual calibration.

 

[benta]

Why are you gain-limiting the difference amp (the one to the right)?

Benta.

 

[BrownPanda]

Hello,

Thank you all for the replies.

MacGyverS200: I will look into current mirros.

ScottyUK: Should I increase the gain of the inverting amplifiers, U11 & U2, or the differential amplifier, U6? Or does it not matter which amps i increase the gain on? I am using +/-0.5% resistors.

Benta: I didnt really think about increasing the gain of the differential amp. I can if i have to, this wouldnt be an issue.


If i increase the gain too much, will that have some kind of overshoot issue?

Again, thank you all for the feedback.

 

[BrianE22]

You want to control the current for 3 different loads? I'd think you'd need 3 feedbacks and 3 separate control loops (or some other way to divvy up the command to the 3 resistors).

 

[benta]

I've read your original post again, and from what I understand is that you want 3 different ohmic loads to carry the same current (sinusoidal).
I think you need a different approach here:
Three voltage-to-current amplifiers.
Take the circuit of U5, but instead of using feedback from the complementary pair, use feedback from the current sensing resistor.
Remove R1.
Connect your input through a voltage divider to the non-inverting input.

Repeat twice, then you have three channels.

Much simpler...

Benta.

 

[daveross100]

The following may help you

https://web.stanford.edu/~boyd/ee102/fdbk-static.pdf

Note: the Error you allow on your loads shall be depended on the loop gain,
that how you calculate the gain of each one of the amplifiers in the loop,
you may want to take into account the bandwidth = how much fast you want the loop to correct the error
-What is the application?
-Any question forward to:
daveross100atgmaildatcom GOOD LUCK