TPL0102-100: Using a digital rheostat to adjust a negative-voltage LDO

Hi Carl,

Welcome to E2E and thank you for your query. Instead of configuring the resistance between OUT and FB pins variable, I would suggest making the resistance between FB and GND pins variable. The voltage between FB and GND remain well within the operating range of TPL0102-100 and hence, you don't need different ground potential and isolator.

Hope I answered your question. Please let me know if you see any issue with the above solution.

Regards,
Uttam Sahu
Applications Engineer, Precision DACs

Thanks for the suggestion. But making the bottom of the resistor divider the digital pot causes a highly non-linear granularity of Vout adjustment of my [-6.2, -1.8]V adjustment range. It's especially bad at the lowest point in the range, where one click jumps Vout by as much as 0.4V. I need a more linear behavior with about 0.02V change in Vout per digi-pot click across the range.

Hi Carl,

I understand your concern. If you are open to use a DAC in place of a DPOT, you can look at the following E2E post:

e2e.ti.com/.../give-your-voltage-regulator-the-margin-it-deserves

This will also give you the flexibiity to use a higher resolution DAC in the design.

Regards,
Uttam

I looked at the DAC article and I'm open to this approach. What DAC parts does TI make that would do the job for my negative-voltage TPS7A3301 LDO regulator? (Vref= -1.17V, Vnom = -4V, DeltaVDAC ?, DeltaV_Margin= +/-2.5V, ...). I need four independent channels of the same circuit so, a muli-channel DAC would be attractive. -must have I2C control capability.

Hi Carl,

I am not very clear about your requirements from the above post. Could you please put them in the following format?

1. DAC Output Range: ?
2. DAC Resolution: ?
3. No. of Channels: 4
4. Interface: I2C

Prima-facie, the DAC7574 looks like a good fit for your case. In order to have bipolar output, we can use an external opamp.

Regards,
Uttam

Uttam, thanks. I'm not sure how to best translate my circuit requirements into DAC output range and resolution requirements, but the DAC7574 looks like a good place to start.  Please see the attached diagram below, which hopefully better explains my ultimate requirements.  It looks like the DAC7574 will give me much better adjustment granularity than what I need, but that's okay.

I'm not sure I understand what you mean by using an external op amp to have a bipolar output.  Is this to make the current sourcing and sinking happen through R3 for output control around the center point of the Vout range (-4V)?

Is there an EVM kit for the DAC7574?

Thanks, Carl

DAC-ControlForNegative-VoltageLDO.pdf

Uttam, thanks.  I'm working through the details of using the OPA188 circuit to provide the set-point adjustment capability.  I'll have a new sketch for the forum to look at soon.  Meanwhile, I also need a monitor A/D converter to read back each of these four channel voltages.  I'm using an LT2487 right now, but it's not working out like I had hoped.  Does TI offer a 12-bit substitute for this 16-bit Linear Tech part (four channel, I2C-interface)?  Please let me know.

Thanks,

Carl

Following the concepts of the article "Give your voltage regulator the margin it deserves" and the TI Precision Design "Bipolar +/-10V Analog Output from a Unipolar Voltage Output DAC" (SLAU525), I've developed my solution further. I feel I'm getting close but not quite there yet. Please see the new attachment. I don't think I need the V_Ctrl signal going to the LDO to be bipolar. I do need the V_Ctrl signal to be able to both source and sink current, however.

I think I should be able to simplify the DAC - OpAmp circuit.  What other options can I use for the OpAmp?  The OPA188 is not currently in my library, it's wide voltage range is not needed and I don't need the V_Ctrl signal to be bipolar.  I does need to be able to source and sink DC current, though.  Please see the attachment...

DAC-OpAmpControlForNegative-VoltageLDO.pdf

Hi Carl,

The DAC7574 can bothe source and sink current within specified limits. Please look at Fig 13 - 15 in datasheet. In case your circuit needs higher current capability than this and you don't need bipolar output, you can use an opamp at the output in a simple buffer configuration.

Regards,
Uttam

Uttam, I've simplified my circuit to use an OPA656 Op Amp (which is already in my library) in an inverting amplifier configuration to get V_Ctrl to go from 0V to -2.2V as the V_DAC goes from 0V to +5V.  Please see my new attachment.  I'm awaiting some sample parts to try it out.

Does TI offer DACs with Non-Volatile Memory so that the last setting is stored during a power cycle?  Please let me know.

Thanks, Carl

DAC-OpAmpControlForNegative-VoltageLDO26Jan18.pdf

Hi Carl,

Most of the opamps are pin-to-pin compatible and hence, you can use the same libarary for different part numbers.

The circuit you have posted looks fine. I am only concerned about C-load. If it's more than the C-load limit of the opamp you need a compensation R-C in the feedback path of the opamp.

If you need non-volatile memory to store the configuration, the existing circuit with DAC7574 is not going to work. I can think of two options to implement that:

1. Use a DPOT in the feedback circuit of the opamp in the same cicuit you uploaded in last post and drive the opamp input with a fixed voltage instead of a DAC. You need to see if the 8-bit resolution is ok for you or not.
2. Or else, you can use LMP92066 instead of DAC7574. It has an EEPROM and also supports automatic temperature compensation. It has bipolar output - hence, no need for an opamp. The limitation is that it has only 2 channels. So, you need two of these.

Regards
Uttam

Uttam, thanks.  I am awaiting sample parts to try out this solution.  I will let you know whether or not this resolved my issue.

P.S.  I am not populating C_LOAD.

Best regards,

Carl

DAC-OpAmpControlForNegative-VoltageLDO31Jan18.pdf