EL302P output regulation speed test

[IanB]

I was curious about how fast my EL302P power supply can react to sudden load changes and how stable the regulation would be. This was prompted by my observation that when I connect a bulb to the output the current limiting cuts in almost instantly to limit the cold filament surge current. I decided to test this with my oscilloscope to examine more closely what happens.

I did the test below with a 12 V 12 W automotive bulb. I set the power supply to regulate at 8 V and 1.5 A and measured the current with a 1 ohm series resistor in circuit with the bulb.

For test #1 I switched the power supply on first to regulate at 8 V and then connected the bulb. For test #2 I used the same circuit but I used the output on/off switch on the power supply instead of an external switch.

Test #1: Connecting load to power supply with output already on

We can see here a momentary overshoot of the current to about 3 A before recovering to the regulated 1.5 A. The overshoot spike has a duration of less than 500 µs and is caused by the discharge of the output capacitor on the power supply. Once the capacitor has discharged the current regulation is immediate and stable at the set value.

The voltage climbs as the bulb filament warms up until the power supply switches over to voltage regulation once the filament is hot enough to limit the current.



Test #2: Using the output switch on the power supply with the load already connected

This time we see that the current is well controlled with no overshoot. When using the power supply's own output switch it electronically ramps up the output voltage at a measured rate and introduces voltage or current regulation as soon as either set point is reached (in this case the current).



This shows the benefit of having a purpose designed output on/off switch on a laboratory power supply. Since all power supplies have some capacitance on the output for improved voltage regulation you are always at risk of a current spike if you suddenly short the output of the power supply while it is switched on.

[icon]

You have shown your beautiful curves, and they so pleasantly please the eye,  but how does all this even pretend to predict the price of perfume in Persia?

Does not a power supply pretense to potentially progress a proper project?

Should not a toad be trampled, a horse rode?

You've given me a lot to think about.

John

[IanB]

You have shown your beautiful curves, and they so pleasantly please the eye,  but how does all this even pretend to predict the price of perfume in Persia?

It's an illustration of a regulatory control problem. In this other thread, you can see how the makers of a different power supply completely failed to design an effective solution to the same problem:

https://www.eevblog.com/forum/testgear/comparing-two-power-supplies/

[Paul Price]

Ianb,
A Beautiful Presentation,

Thank you for explaining the purpose of your presentation.

 Information like this can really help beginners and even remind experienced engineers how power supplies should and do work.