0
\$\begingroup\$

I'm looking at calculating the DC Accuracy of a couple LDOs and came across some good resources - but I'm having trouble linking their example calculations to the "real world."

[TI LDO] https://www.ti.com/lit/an/slva072/slva072.pdf
[Analog LDO] https://www.analog.com/en/analog-dialogue/articles/understand-ldo-concepts.html

The example (interestingly, the exact same example in both articles) uses the parameters:
3.3V LDO over 0 to 125C temperature span
±100 ppm/°C resistor temperature coefficient
±0.25% sampling resistor tolerance
±10 mV output voltage change due to load regulation
±5 mV output voltage change due to line regulation
1% reference accuracy

  1. I get that the sampling resistor parameters come right from the resistor datasheet.

  2. The Line/Load regulation errors (I imagine) come from my system? If I'm supplying voltage with a USB port it would be the variation in the USB port voltage supplied, and if my load is a couple LEDs the amount of voltage variation over those LEDs?

  3. The reference accuracy - I figured that comes from the LDO datasheet, but I'm not sure exactly which parameter it refers to. Is it just the "Fixed Output Voltage Accuracy" or is there another parameter I haven't connected?
    [ADM7172] https://www.analog.com/media/en/technical-documentation/data-sheets/ADM7172.pdf

Thanks so much for any guidance, and/or additional resources.

\$\endgroup\$

2 Answers 2

Reset to default
0
\$\begingroup\$

The Line/Load regulation errors (I imagine) come from my system?

The line and load variations come from your system. The errors come from how the power supply responds to them.

The reference accuracy - I figured that comes from the LDO datasheet, but I'm not sure exactly which parameter it refers to. Is it just the "Fixed Output Voltage Accuracy"

Yes. At least that's the best stand-in you have in this case. The app note was probably originally written for an adjustable regulator, and someone didn't go back and edit it, or they just figured you're just supposed to know.

There's going to be details different in every chip, but you can take the action of a fixed-output regulator as having an internal voltage divider down to whatever the chip's reference voltage is -- pretty much exactly like you'd do with an adjustable regulator outside the chip, only pulled inside.

At any rate, "Fixed Output Voltage Accuracy" rolls up the reference inaccuracy and any inaccuracies in the internal divider (or other mechanism for changing the feedback/reference relationship) into that one number.

\$\endgroup\$
2
  • \$\begingroup\$ Ah, maybe the regulation definitions is where I'm getting confused. I started reviewing this [Load Regulation]<techweb.rohm.com/knowledge/dcdc/dcdc_sr/dcdc_sr01/1551>. If I understand you correctly, in my particular case, it's the response of the LDO to these variations. So in the example datasheet under: REGULATION (simplified) Line ∆VOUT/∆VIN : +0.1 %/V Load ∆VOUT/∆ILOAD : 0.3 %/A I would use the 0.1% and 0.3% in the calculations, as that's the response of the regulator ? \$\endgroup\$
    – vkaiser
    Commented Mar 5, 2021 at 18:08
  • \$\begingroup\$ Pretty much, yes. \$\endgroup\$
    – TimWescott
    Commented Mar 5, 2021 at 18:46
1
\$\begingroup\$

Load regulation is defined for some rated step load with the tolerance error reflecting the impedance ratio of output to load resistance.

line regulation is just dVout/dVin, and in both cases reflected in % or absolute values. Step dynamic response depends on the quality of your cap ESR and layout inductance.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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