2
\$\begingroup\$

I'm trying to make a ~4V 2A regulator with the LM317. It works with minimal load but when I increase the load the voltage drops instantly to 3.4V. I'm not sure what's happening.

I'm trying to build something like this circuit:

circuit I'm trying to implement

I don't have the exact same components so my circuit actually looks like this:

schematic

simulate this circuit – Schematic created using CircuitLab

Another strange thing I've noticed is that when I add the load resistor, the sum of VCE + Vout is less than 24V (around 22V,) but the 24V power supply has it's own voltmeter and it marks 24V all the time, which shouldn't be possible as there's nowhere else between Vin and GND for those 2V to be. I thought it might be my cheap multimeter so I grabbed a better one but that one reads the same voltages.

Here are some photos of my setup:

With no load With load

I'm aware the heatsink for the transistor is smaller than it should be, but this is just testing the circuit and it takes about 30s to heat up to where you can't touch it, the LM317 is always cold so it can't be thermal shutdown right?

I'm also aware that making a 2A voltage regulator on a breadboard is a bad idea, but again I am just trying to see if the circuit works.

From what I've seen of other questions it seems this can happen when you invert the pins, this is the pinout I'm following:

Used pinout

Edit: When i made the circuit on an universal board the regulator worked like it should.

\$\endgroup\$
18
  • 4
    \$\begingroup\$ "I'm also aware that making a 2A voltage regulator on a breadboard is a bad idea"- Yes, it is a bad idea. Breadboard contacts often have high resistance that causes trouble even at less than 1A. Connect your load and measure voltages across all components and the wires between them. What do you get? \$\endgroup\$
    – Bruce Abbott
    Commented Jul 16, 2021 at 21:44
  • 1
    \$\begingroup\$ Do you intend to operate this supply with no load? Your design violates LM317 minimum load current of 3.5 mA. Your design requires about 400 ohms load or less. \$\endgroup\$
    – glen_geek
    Commented Jul 16, 2021 at 21:53
  • \$\begingroup\$ I intend to use it to supply a GSM module so minimal load is guaranteed also there's a fixed 1.2K resistor which @3.98V comes at 3.31mA so yeah, not minimal, I'll change it for a 1K so that's within requirements but I don't think the issue is there. When the load is essentially 10Ohm there's still a problem, so surely the issue is somewhere else? \$\endgroup\$
    – Santiago Perotti
    Commented Jul 16, 2021 at 21:58
  • \$\begingroup\$ Measure and see if there is a voltage difference between the different ground points on the circuit when the load is added. Maybe a high resistance somewhere. \$\endgroup\$
    – user173271
    Commented Jul 16, 2021 at 22:05
  • 1
    \$\begingroup\$ @BruceAbbott The issue was indeed the breadboard, cables had drops of ~140mV when the circuit was loaded. I made the circuit on an universal board and it works now. Since you were the first (but by no means the last) to bring this up to my attention maybe you'd like to post it as an answer so I can accept to wrap this up bow and all \$\endgroup\$
    – Santiago Perotti
    Commented Jul 17, 2021 at 1:34

1 Answer 1

Reset to default
2
\$\begingroup\$

Another strange thing I've noticed is that when I add the load resistor, the sum of VCE + Vout is less than 24V (around 22V,) but the 24V power supply has it's own voltmeter and it marks 24V all the time... When i made the circuit on an universal board the regulator worked like it should.

This suggests the problem was... your solderless breadboard! Cheap breadboards often have high contact resistance and a low maximum current rating. I have seen significant voltage drops even at currents well below 1 A on breadboards I bought from eBay, whereas a board I have had for 30 years is much better (even after I partially melted it passing too much current through the power rails one time).

For currents over 1 A it pays to solder the connections, and use thicker wire to carry the current when using stripboard too.

\$\endgroup\$
2
  • \$\begingroup\$ It would be nice if it was easier to know what breadboards were good and which ones were bad... You never see them with contact resistance ratings or anything. \$\endgroup\$
    – Hearth
    Commented Jul 17, 2021 at 15:27
  • \$\begingroup\$ @Hearth perhaps insertion force? My old breadboard is often quite hard to push thick wires into, whereas the new ones feel loose. You could measure the contact strip thickness with calipers to estimate current handling and spring tension, or of course just pass a current though it and measure the voltage drop. \$\endgroup\$
    – Bruce Abbott
    Commented Jul 17, 2021 at 15:36

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