7
\$\begingroup\$

I need to come up with a method to provide a microcontroller-adjustable, regulated current ranging from 0-2A at ~24V to a load in a closed-loop fashion. I have been researching this issue for awhile and have found a few ways that it could be achieved, each has its issues. I was hoping to lay out my thought process and different approaches and get feedback as to whether I may have overlooked something obvious or if I am going about this entirely the wrong way. Thank you.

Method 1 - Purchase a constant current LED driver IC

This seemed like the most obvious solution. It would have the benefit that I would not need to setup external current-sensing to implement closed-loop control as would be needed with the other two choices. However, I had trouble finding LED drivers that could handle 2A at 24V on a single channel as I browsed DigiKey. If any exist I would be very happy to have a part number.

Method 2 - Convert a voltage regulator into a current source

Unable to find a suitable LED driver I did some more research and came across this instructables which looked like it would achieve what I needed. The problem I ran into here was that this design and others like it use a small resistance (on the order of 10's og ohms) potentiometer for current adjustment. My system, in order to be closed loop, would need to use a digi-pot which do not exist at such low resistances due to limitations of their construction.

Method 3 - Construct using an Op-Amp and Transistor(s)

Finally, after the previous two failed attempts I went back to the drawing board. I came across other StackExchange questions like this one, the accepted answer for which used an op-amp and some type of transistor (the comments on the answer go through a variety of different choices from MOSFETs to Darlington pairs). This seems like it would work for my purposes but I have limited experience in this area, is there a particular transistor type that would be far superior in this application? Also, will I need to take into account the gain of the transistor or is that handled with the feedback from the sense resistor?

To conclude, my question is two fold:

  1. Is there a way to achieve my stated goal using one of first two methods I mentioned which I may have overlooked?
  2. Is there a particular transistor type that I should be using for method 3 if my goal is to be able to set the current for the load anywhere in the 0-2A range, not simply an on-off switch?

Thank you very much for taking the time to read my question and provide answers. Do not hesitate to request clarification on anything I said!

\$\endgroup\$
7
  • \$\begingroup\$ Do you need to source current as a positive voltage, or sink current to a gnd or negative voltage? This is only an issue if one of the supply rails is to be earth grounded. If it is an ISOLATED low-voltage supply then earth ground is not needed. \$\endgroup\$ – user105652 May 30 '16 at 21:53
  • \$\begingroup\$ as Sparky said above, if the current source is high-side or low-side (sink) then it's quite important to know this. \$\endgroup\$ – KyranF May 30 '16 at 21:54
  • 1
    \$\begingroup\$ What exactly is the proposed load? How much 'compliance' is necessary (what's the maximum voltage that could possibly apply across the load at 2A) and what is the minimum ? \$\endgroup\$ – Spehro Pefhany May 30 '16 at 22:17
  • \$\begingroup\$ To answer Sparky's questions, the Vcc side of the circuit is tied to the positive rail of a 24V regulated supply connected to wall power which has plenty of overhead for my application, the other side will be tied to the ground rail of that same supply. As far as sinking or sourcing the current, either strategy should be acceptable, I'm really just interested in creating the simplest circuit. Spehro, the load is a series of loops of magnet wire which are being used to induce a specific magnetic field, known as Helmholtz coils. For our purposes we are making the simplifying assumption that the \$\endgroup\$ – carpenma May 31 '16 at 5:00
  • 1
    \$\begingroup\$ Method 3, exactly as in stevenvh's answer, with a TIP110, as he say, should work. A mosfet may require some compensation in the loop, a circuit with BJT has better chances to work from the first try without oscillating. However, you have to attach a huge heatsink to TIP110. Max power consumption will be 24V*2A = 48W in the worst case (short circuit). \$\endgroup\$ – dim May 31 '16 at 8:36
1
\$\begingroup\$

http://www.digikey.com/product-detail/en/texas-instruments/DRV103H/DRV103H-ND/390444

http://www.ti.com/lit/ds/symlink/drv103.pdf

Adjusting PWM resistor you can change the output current

\$\endgroup\$
1
  • \$\begingroup\$ That part will suffer from similar issues as the Darlington pointed out in mkeith's comment above. According to the datasheet it can only output in the range of 10-90% duty cycle leaving a 10% dead zone at the top and bottom of the range. \$\endgroup\$ – carpenma Jun 1 '16 at 0:43

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.