I have a circuit using LD39015 1.2V Low Drop-out voltage regulator.

See attached image.

The value of VCC is 3.3 Volts. The goal is to supply an ADC channel with a constant stable voltage so I can use that to reverse calculate the VCC. And yes I have looked at both Vref IC and Vreg IC options. But while testing the Volt Reg option, I am seeing a 1.189 volts at the output of the voltage regulator with R2 = 10k, 1.191 V for a R2 = 1k and 1.11V for R2 = 100k Ohm. First of all why am I not getting exact 1.2V rated output from the regulator given it satisfies the min input voltage and the dropout voltage. Secondly how does changing the resistance value change the output of the regulator even when the voltage regulator gets sufficient input voltage. Thanks.

  • 1
    \$\begingroup\$ How accurate is your DMM? Is it calibrated? Have you gotten the same reading with multiple meters? \$\endgroup\$
    – Jim
    Nov 6 '18 at 15:14
  • 2
    \$\begingroup\$ If you want a constant stable voltage, it's better to use a voltage reference IC, like for example the ADR3412, which is specified to output 1.2V ± 0.1%, much better precision than your voltage regulator. \$\endgroup\$
    – Hearth
    Nov 6 '18 at 15:20
  • \$\begingroup\$ @Felthry Well that looks like a great option. But I am looking for something <$0.5. \$\endgroup\$
    – VicJay
    Nov 6 '18 at 15:59
  • \$\begingroup\$ @Jim I tested with a DMM from a different Manufacturer and it seemed to give the same value of 1.189 Volts. \$\endgroup\$
    – VicJay
    Nov 6 '18 at 16:00
  • \$\begingroup\$ @VicJay Unfortunately, you may have to shell out a dollar or two more for a reference if you need this level of precision. Engineering is all about managing tradeoffs. Perhaps you could software trim each individual board? \$\endgroup\$
    – Hearth
    Nov 6 '18 at 17:11

First of all why am I not getting exact 1.2V rated output from the regulator given it satisfies the min input voltage and the dropout voltage.

The 1.2 volt regulator (LD39015) has a +/- 2% tolerance on its output hence it could be as low as 1.176 volts. Regarding you altering the value of resistor in series with the input to the regulator you are almost certainly dropping enough voltage to take the input terminal into the regulator beyond its low-drop-out-point.

The device has a quiescent current of 18 uA and, with 100 kohm in series with the input terminal, the voltage formed across that resistor is 1.8 volts. Subtract that from 3.3 volts and the voltage at the input to the regulator is 1.5 volts.

  • \$\begingroup\$ Thanks for the reply. Yes I did have that same question. So to check that I tested with three samples of the IC. All three seemed to have the exact same output of 1.89V at R2 = 10k. So that begs the question, if it is indeed Tolerance coming into play shouldn't it vary with samples? \$\endgroup\$
    – VicJay
    Nov 6 '18 at 15:18
  • \$\begingroup\$ 1.89 volts or 1.189 volts? 3 items from the same batch will be usually very close in value but all 3 can be +/- 2% from 1.2 volts. Use a voltage reference for better initial accuracy (fractions of 1%) and better drift (sub 10 ppm/degC) if needed. \$\endgroup\$
    – Andy aka
    Nov 6 '18 at 15:21
  • \$\begingroup\$ @VicJay Not necessarily. It could be variation with load current; have you tried adding a different load? But like Andy said, three devices from the same batch will probably be near-identical; you'd have to buy them at different times and test across different batches if you wanted to see the full spread. \$\endgroup\$
    – Hearth
    Nov 6 '18 at 15:24
  • \$\begingroup\$ @Andyaka Sorry I meant 1.189 Volts. \$\endgroup\$
    – VicJay
    Nov 6 '18 at 15:57
  • \$\begingroup\$ @Felthry There is no load after the Voltage regulator. It is directly feeding in to the ADC of the Attiny44a chip. The current determining factor I am guessing here is the value of R2 and Yes i tried varying it from Zero ohms to 100 Kilo ohm, but none of them got me higher than 1.191V. \$\endgroup\$
    – VicJay
    Nov 6 '18 at 16:03

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.