1
\$\begingroup\$

I am designing a circuit that requires a regulated 1.75 V source. After searching around I found the LM317-N voltage regulator and got it's SPICE model from TI to simulate it within my circuit. The issue is my simulation has the output voltage at ~1.53 V despite my calculations showing that it should be ~1.75 V with the resistor values I've chosen. I use a 12 V supply so am not violating the min differential voltage, and draw enough current at the load (50 Ohm resistor). Not sure about what I'm doing wrong here. Any help is appreciated!

schematic diagram

simulation result

raster image of some calculations for R2

\$\endgroup\$

2 Answers 2

2
\$\begingroup\$

Your Iadj part of the calculation is wrong.

Iadj is a fixed value of about 50uA (from datasheet) - it can be ignored with low value resistors such as what you are using.

The equation then reduces to Vout = 1.25 (1 + 47/220).

This results in a value of 1.517V. Close to that of the simulation.

To get 1.75V R2 needs to be about 88 Ohms.

LM317 Datasheet

\$\endgroup\$
5
  • \$\begingroup\$ Thanks for pointing out the issue with Iadj, wouldn't have guessed that by myself. After adjusting to 88, my result gets a lot closer but still slightly off. For ex. 1.76V as opposed to 1.75V. Any ideas as to where this variation might be coming from? \$\endgroup\$ Commented Apr 9 at 2:09
  • \$\begingroup\$ @JaysonOkhman You are asking for too much out of simulation and out of parts you will get and can use in an actual design. if you expect to see an accurate value of 1.750 V then you will be needing a voltmeter with a nist-traceable calibration certificate such as this one and the knowledge of how to use that equipment properly (temperature, humidity, and so on) so as to stay in operation specs while using it. This is expensive and rare. (Precision is cheaper.) \$\endgroup\$ Commented Apr 9 at 5:49
  • \$\begingroup\$ Also, the circuit that requires 1.75V will have a range of voltages over which it will work correctly possibly +/- 5% or 10%. You also have component tolerances - for example the LM317 has a +/- 1.5% tolerance. What circuit are you powering? \$\endgroup\$ Commented Apr 9 at 16:26
  • \$\begingroup\$ @JaysonOkhman If you’re only needing to supply a few milliamps at a very high accuracy, it might be better to look into using a “voltage reference” instead. \$\endgroup\$
    – Ste Kulov
    Commented Apr 9 at 20:59
  • \$\begingroup\$ @KevinWhite Thanks for explaining. I had a hunch that maybe this is as close as the simulation can get, just wanted to make sure that it wasn't some other error I made. As for the circuit, the 1.75V will be acting as a bias voltage to an opamp which will then be fed into a car's ECU. I am not sure what the tolerance of the ECU is (and don't have a way to find out) so wanted to get as close as possible. \$\endgroup\$ Commented Apr 9 at 22:00
1
\$\begingroup\$

Kevin has already addressed your main question. I just want to point out a couple of things:

To your follow-up question:

"After adjusting to 88, my result gets a lot closer but still slightly off. For ex. 1.76V as opposed to 1.75V. Any ideas as to where this variation might be coming from?"

Wow, you got much closer than I would have guessed. Probably a little too close. Typically the models take min/max values into account. Depending on which datasheet you look at, the LM317 usually has a min Vref of 1.2 V and a max of 1.3 V. The Ltspice models I've seen are usually closer to 1.3 V. That range would give you an output voltage between 1.68 V and 1.82 V. Something like that.

I also wanted to point out that you are losing a lot of power. For a linear regulator, to drop your input voltage to your output voltage you are basically burning it off as heat.
The general forumula is: Preg = (VIn-VOut)*IOut

So, to drive a 50 Ohm load, your total power is:
Ptot ~= (12-1.75)*(1.75/50) + 1.752/50 ~= 420 mW

That's a lot of power, for 1.75 V across a 50 Ohm load.

If that's important to you, you might look at something more efficient, like a SMPS.

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for bringing up the power consideration. as I mentioned, the 1.75V will be acting as a bias voltage to an opamp which will then be fed into a car's ECU. Because of this I am limited to a 12V supply. I needed something cheap, lightweight, and stable under load changes which is why I settled on a linear regulator. Does my selection make sense for the application or should I look more closely at using a SMPS? \$\endgroup\$ Commented Apr 9 at 22:09

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.