EEVblog Electronics Community Forum

Products => Test Equipment => Topic started by: KD0RC on December 09, 2018, 06:31:08 am

Title: Acceptable power supply power-on overshoot
Post by: KD0RC on December 09, 2018, 06:31:08 am
I just bought a cheap lab power supply (Korad KA3005P).  It works as advertised so far, but in looking at the output on an oscilloscope, I have about 100mV of overshoot when I enable the output. The voltage rises very smoothly (looks like a capacitor charging curve), but continues past the set voltage by 100mV or so, then looks like a damped sin wave that goes a few cycles until it settles on the desired voltage.  The first half cycle of the overshoot lasts somewhere around 5ms.  I get this whether I have a load or not.  It happens at any voltage, is not influenced by OCP or OVP settings, and is not influenced by the current set point.

Is 100 mV a reasonable amount or is this supply defective?  If I set my scope to 5V/Div, the curve looks perfect (like in Dave's re-review of the unit).  Once I get to 100mV/Div, the ringing overshoot is very obvious.  I am just trying to gauge whether this is more or less normal for this particular supply.

Thanks,
Len
Title: Re: Acceptable power supply power-on overshoot
Post by: tautech on December 09, 2018, 06:47:42 am
Ideally you want zero overshoot and many PSU's can achieve that however in practice a project will have additional bulk capacitors that while charging can help reduce overshoot. If you look at datasheets for IC's and max Vcc we'd not normally run supply voltages too close to max so to have some leeway to accommodate variations in power ON behavior.
Check it with a few different loads for piece of mind.
If you like the PSU and there's no other unexpected behaviors, live with it.

Datasheets and a scope are your friends.  ;)
Title: Re: Acceptable power supply power-on overshoot
Post by: Whales on December 09, 2018, 09:37:11 am
Depends on what you are trying to power.  100mV out of 5V is often nothing,  100mV out of 1.2V could mean the world.  Different chips have different voltage and noise tolerances (the freq of the overshoot damping/oscills can also be important.  Slow often is better).

I've seen PSUs with much worse overshoot, to the point you disconnect them during poweron to avoid cooking your chips.
Title: Re: Acceptable power supply power-on overshoot
Post by: nctnico on December 09, 2018, 09:58:45 am
Adding an extra capacitor of 100uf or so will probably dampen the overshoot. What is more important are overshoots when powering on and off. If the PSU doesn't have a real power switch, then plugging in the mains counts as power-on.