0
\$\begingroup\$

I am trying to precisely control current limit of the LT3081 IC following this paper, page 12, schematic on page 14, but by using a microcontroller.

I'm controlling the output voltage with a 10kOhm, 8 bit, digital potentiometer, which I feed with 1.2mA constant current to make it regulate the output voltage between 0 and 12 V.

I'm trying to limit the output current from 1 mA to 1A, idealy with 1 mA step, without using more expensive 10 bit digital pots. By using a similar configuration as shown in the schematic for higher accuracy (pot + paralel res + series res), using one 10k, 8 bit, digital pot and considering the wiper resistance, I got to a point where I can control the current limit with a 7 mA step at lower values, and 2 mA steps as I get closer to 1 A limit.

Spreadsheet for 10k 8b values: link

Equation for current limit(from datasheet): RILIM = ILIMIT/360mA/kΩ + 450Ω

Simplified equation: RILIM = ((ILIM*1000)/0.36) + 450

Simplified schematic:

schematic

simulate this circuit – Schematic created using CircuitLab

My question is: Is there any other way for a more precise current limit control without using a higher step count (bit) digital pot?

Improve this question
\$\endgroup\$
4
  • \$\begingroup\$ How do you expect to get 1000 steps in the output of a device with only 256 different settings? \$\endgroup\$ – The Photon May 9 '18 at 16:25
  • \$\begingroup\$ Maybe you could get the two additional bits by switching in series resistors using MOSFETs? Just getting a 10b digital pot would be the best option, though. \$\endgroup\$ – Hearth May 9 '18 at 16:35
  • \$\begingroup\$ Yes, I agree that 10b would be the best option. I need 1mA step upto 100mA, after that 10mA step is ok. Im currently testing 1k 8b + 10k 8b in series, but i think i will just go with 10 bit 10k and close this question. \$\endgroup\$ – Lazyboy May 9 '18 at 16:38
  • \$\begingroup\$ The Imon pin is a current you have to sink, and you don't need to sink it into the output. Look at a the datasheet examples on current sharing on page 22. You could provide you own current measurement and either sink a current from Imon or sink a current from Set and perform foldback that way. \$\endgroup\$ – Jack Creasey May 9 '18 at 19:13
1
\$\begingroup\$

An 8 bit pot gives you 2^8 possible taps (256) so at best you could manage slightly under 4mV steps.

A 10 bit pot has 2^10 steps (1024) so in theory could give you your 1mA resolution in a suitable circuit.

You are going to need a digipot that can handle 12V on the element, not a given as most of them are limited to within the logic supply voltage.

Were it me, I would be thinking in terms of setting a fixed current limit with this mechanism, then doing the current limit by using the Imon pin and an opamp or such to pull the set pin down on overcurrent, this has the advantage that such things as foldback limiting become possible, and both your control signals are now ground referenced and can be trivially generated.

Improve this answer
\$\endgroup\$
2
  • \$\begingroup\$ Yes, I have to rethink a part of a design, Thank you! \$\endgroup\$ – Lazyboy May 9 '18 at 17:34
  • \$\begingroup\$ Trying to create a floating digital pot is just too hard, you can use either SET or Imon to control the current limiting but don't try to sink these currents into the output. \$\endgroup\$ – Jack Creasey May 9 '18 at 19:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.