What is output overshoot voltage of SMPS?

[cupoftea]

Hi,
If the following offline PSU is on full load, then it is suddenly no-loaded, then to how high will the vout overshoot?
Why is this crucial measure not in the datasheet?

Meanwell GS220 Offline SMPS

https://www.meanwelldirect.co.uk/wp-content/uploads/sites/55/2018/04/GS220-Series-Specification_MWD.pdf

Why does no manufacturer ever specify this for any of their offline PSUs?

 

[KlausST]

Hi,

I agree, I good datasheet will show a scope view of load regulation.

But there is an overvoltage protection (105% to 135%). If it activates, it must be cycled to allow power to restart. I guess customers would complain if they had to do a power cycle on a sudden load disconnect (power off).

Klaus

 

[cupoftea]

Thanks, yes, and i wonder if that quoted Output OV protection is actually for an externally applied overvoltage , which has to be present for a second or so....in order to trip the PSU off....i suspect the transient from a sudden no loading, will not actually trip the output overvoltage shutdown....105% is too close to the rated voltage......i am certain this power supply will go above 105% of rated voltage when suddenly no loaded....but i have never tested it

 

[andre_luis]

then it is suddenly no-loaded, then to how high will the vout overshoot?

I guess that for a SMPS which was properly designed on its feedback control characteristics it should not matter regardless of the full to no load transient ( not the opposite due to softstart ), their inner output capacittors should assure at some extent the specified output voltage load regulation stated on datasheet. However, whether this power supply actually meets the stated specification or not, it is another question.

 

[cupoftea]

Thanks, ive often seen OffTheShelf 48Vout Offline PSUs transiently going up to 60V when suddenly no_loaded.

 

[Calm down]

Hello buddy, if you are using a power supply without a DC cable, it will definitely indicate forward and reverse overshoot, usually ± 3% -5%. But this is an adapter, and the extended DC line is equivalent to a line resistor and inductance, which will amplify the overshoot. The longer and thinner the cable, the more severe it becomes, so adapters with DC cables do not label this data. That's what I do.

 

[KlausST]

The longer and thinner the cable, the more severe it becomes,

thinner cable --> more overshot: I can´t agree.

And cable (wire) thickness has very low influence on inductance. The overshot cause by the stray inductance of a cable is very fast (needs fast switch OFF of the load) but still conains low energy. (compared to the overshot caused by slow SMPS regulation)

Klaus

 

[Calm down]

thinner cable --> more overshot: I can´t agree.

And cable (wire) thickness has very low influence on inductance. The overshot cause by the stray inductance of a cable is very fast (needs fast switch OFF of the load) but still conains low energy. (compared to the overshot caused by slow SMPS regulation)

Klaus

The problem caused by cable resistance and inductance is manifested in the speed of voltage adjustment. When the load changes rapidly, it will cause the voltage waveform measured by the DC port to become a square wave that decreases quickly and rises slowly, so it is mostly a negative overshoot. I think the forward overshoot (the part that exceeds the rated voltage) is more a problem with the load side than the power side.

 

[KlausST]

Hi,

The problem caused by cable resistance and inductance is manifested in the speed of voltage adjustment.

The mentioned power supply has no "sense" feature.
Thus neither cable resistance nor cable inductance is invovled in voltage adjustment (=regulation) at all.

What you call "negative overshoot" do I call "undershoot". Indeed here we need the OP´s definition.

When the load changes rapidly, it will cause the voltage waveform measured by the DC port to become a square wave that decreases quickly and rises slowly, so it is mostly a negative overshoot.

In your post you refer to the cable. But what you describe is the behaviour of the SMPS control loop.

The OP´s wrote: "is suddenly no-loaded, then to how high will the vout overshoot?" So in this case I can´t agree with your statement "it decreases quickly and rises slowly".

Overshot on sudden no-load causes:
* fast peak low energy by series inductance
* overshoot in output capacitor due to slow regulation loop. May take long time to be reduced ... because there is no load current to discharge the capacitor. Here the voltage rises rather quickly (within microseconds) and decays rather slow (maybe many seconds)

I guess there is a misunderstanding on my side of what you meant. To avoid useless discussion I will end it here, unless you show a simulation that depicts what you wrote.

Klaus

 

[cupoftea]

Thanks,
And the problem for these offtheshelf PSU's is that even if they put an overvoltage comparator in there, and set it say 5% above the converter's rated output voltage, then when the converter suddenly goes "full load to no load", the converter's vout will just go into oscillation, ..repeatedly bouncing off the overvoltage comparator reference voltage, so to speak.

 

[KlausST]

then when the converter suddenly goes "full load to no load", the converter's vout will just go into oscillation

Why? I dont understand your idea.
Did you read the datasheet or post#2?

The power supply goes OFF and stays OFF until a power cycle to restart. .. how can it oscillate?

Klaus

 

[cupoftea]

Thanks, yes, in that case, , it does, indeed...but i was speaking of another , more general case....the general case of someone trying to stop significant power supply vout overshoot following a transient....and them doing it using a comparator, and getting into trouble as described...

 

[KlausST]

Makes no sense.

You say "general"...
What supplies do it this way? (I can only imagine mature supplies with slow regulation loop)
And does it relate to the topic at all?

Klaus