Laptop powerr supply instabilities

[rr100]

I have a laptop power brick that behaves strangely, to say the least. The laptop works fine (is a really old Compaq Presario) but look in the picture what I'm getting at only about 500mA load. If I increase the load the swings are getting even much larger!

At first I suspected the dummy load, I'm using this: https://www.eevblog.com/forum/projects/another-dummy-load-%29/

But I realized that I have a diode (checked, still there) to protect the dummy load in case of messing up the connection. So whatever it does it can't feed back anything, the spikes (and again they can get even much larger). Any idea what I'm looking at? Bad caps in the power supply (it is like 12 years old)?

[10101]

Just guessing, but maybe bad output eletrolytic capacitor ?

[dom0]

Yep

[amyk]

It's probably not a problem. They don't need tight regulation or low ripple since the laptop itself has a set of buck regulators to step down the voltage even further.

[rr100]

It might be, also might be that they have some bigger caps in the laptop itself. Still, that's pretty extreme and I'm pretty sure I've seen even 38V peaks at some higher current (still around 1A I think). I stopped because I need to think how much the dummy load can take.

Also it is some interaction with the dummy load (see link above for "which one") which is just a simple analog setup:

shunt-> opamp amplified voltage ->mosfets

With a resistive load (even well above 1A) it is perfectly clean.

[Halvmand]

That thing is oscillating. I had a similar experience with a home build load. Bodged a (I think) 1 or .1 uf ceramic between drain and ground and it was stable.
Is your load the exact same as in the schematic?

[codeboy2k]

It's probably not a problem. They don't need tight regulation or low ripple since the laptop itself has a set of buck regulators to step down the voltage even further.

I sort of agree with what you're saying, but I sure hope their laptop supply is better than that.  It seems more likely that the output cap is bad and should be replaced.

For a comparison, my Dell laptop supply, rated 18.5V @ 3.8A, is actually a very good supply.  It's 18.7V unloaded, 52mV ripple, and 18.3V @ 2A load, and 56mV ripple.  I can see the 60Khz switching spikes.

You can see waveform captures for it in in this message from another thread:
https://www.eevblog.com/forum/beginners/power-adapter-tester-for-male-barrels/msg458790/#msg458790

[Halvmand]

I'm pretty new to electronics, but to be clear about something....

The PSU is good when resistively loaded
The PSU is bad when the electronic load is used. 8 Vpp

Would the output cap of the PSU really have this impact on stability?

[DanielS]

The last time I had gross output stability issue in a switching supply, it turned out to be dead power filtering caps on the PSU's auxiliary output that powers primary-side circuitry. The regulator worked fine under heavy resistive loading but was going completely nuts with capacitive loading. After replacing the primary-side auxiliary caps, the supply was perfectly stable under any load I threw at it.

[rr100]

Yes, I'm using the load from the first post. Plus a diode so as I said there's no excuse for the spikes. The load itself might be oscillating  as well but still there's no excuse.

It is probably the caps - as I said it is rather old and it's been really, REALLY used. That P3 went strong from something like 2001 to 2011 and it's been used all day long, sometimes it was left in bed on while watching a movie (lucky it didn't start a fire). I had to replace the panel as the hinges and back-light were gone, also the whole palm rest/touch-pad as they were really, really worn out but still survived. It's been retired as defective but in fact only the power connector from the motherboard is flaky, otherwise it still works.

Too bad for the power supply, I really liked it how small it is. I'm pretty sure it's only the caps but I don't think I'll be able to put it back if I crack it open.

[rr100]

For what is worth I've got now the "Dell mother connector" I've been waiting for a while and I've been able to testt some of the Dell PSes I have. Incluiding a "new" one (rectangular as opposed to "squeezed bath sponge shape they've had for 10+ years or so) plus a "very new one" for the fancy ultrabook-like notebooks they have.

Guess what, they're doing ALL THE SAME! Sure, the load is oscilating, nothing you can't fix with enough capacitors in the right place, RIGHT.

But still the power supply shouldn't shoot 27+V into the load no matter what happens! It seems this is some common issue for switched power supplies, not a particular fault in my 15+ years old Compaq power supply.

[amyk]

My guess is that the laptop PSUs are not designed for direct use with such loads. If you look at the power input circuitry in a typical laptop, the filter choke and capacitors are there.