• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 654 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS7A24: Using adjustable regular with a floating ground to make a constant current source

Mark Meyerhofer
Prodigy 40 points

Part Number: TPS7A24

I've been using a similar circuit with other manufacturers' regulators to create a constant current source for use as a measurement stimulus current.  When the circuit is activated, the fixed voltage output and a precision resistor set up a constant current that can then be used to stimulate another part of the circuit and measure an implied resistance from a voltage that is measured at the stimulus point (simple I x R).  The accuracy is better than 2% using a 1% - 1.25% regulator in this way.  Can the FAEs comment on whether the TPS7A24 will function in this way?  I was intending on using the TPS7A24 in another part of my design and being able to use it in this area would be helpful rather than to use some other manufacturer's regulator, if the part will work.  As I've said, I've used this approach successfully for many years, so I do know that some regulators can function when configured as constant current sources.  The circuit is shown in the attached schematic, as it would be used:

  • 0
    Prodigy 40 points

    One correction to note.  Under the time t = t4, it should read Vccs = 1.25V instead of Vout.  Vout is the output voltage of the regulator and floats above Vccs by 1.25V (it would therefore measure as 2.5V).

  • 0 in reply to Mark Meyerhofer
    TI__Mastermind 37885 points

    Hi Mark,

    This looks to be a pretty clever circuit.  You are correctly floating the GND of the LDO, which is what you have to do in order to operate an LDO with a GND based internal reference voltage as a current source. While we haven't built this and tested it (so I can't be 100% sure it will work), on first glance the circuit looks like it will work as you describe.

    I would double check the connection from DGND to AGND.  If it is galvanically isolated then most of the voltage from +5VA will theoretically drop across the galvanic isolation and not the 10k resistor when Q16 is fully ON.  I assume you have connected DGND and AGND someplace, like a net tie used in a star ground, and these aren't really floating with respect to each other.

    If you have any further doubts, I would grab an EVM and modify it to test this to confirm all operating modes work to your satisfaction.

    Thanks,

    Stephen

  • 0 in reply to Stephen Ziel
    Prodigy 40 points

    Steven,

    Thank you for reviewing the circuit.  I can't take all the credit for the circuit, since a basic version of it appeared in the old National Semiconductor Linear Applications Handbook, published through the 1980's and 90's.  I've made some adaptations from that simple circuit to allow it to be controlled by a CPU/MPU and other things specific to my own application.

    Yes, you are correct about the AGND to DGND connection at an un-shown off-page connection; I was doing some quick cutting and pasting and I didn't bother with that important piece.  Since I have this working with other regulators, I just wanted to confirm that the floating ground wouldn't necessarily be problematic for this device.  The rest of the connections to the part are tied as in a normal application, so I figured that those connections would "convince" the part into believing it was properly biased and function as a regulator (regulating between its "ground" and Vout) .  In reality, it floats until a solid current path is established through R224 (when it's connected) and this is when meaningful voltages are established.  My fall back position is that the SOT23-5 footprint (and other similar package footprints) will allow me to substitute other regulators if I find that something prevents it from running as expected; this will avoid a hardware board re-spin in that case.

    If the forum support ticket allows it, I'll add a confirmation note once I've got the new hardware in place and had a chance to verify the correct operation.

    For now, the ticket can be considered complete and can be closed.

    Thank you.

    Best regards,

    Mark Meyerhofer