In a regulated power supply set to CV mode, what is the purpose of OVP?

[SharpEars]

Most quality regulated power supplies offer Overvoltage Protection. This, however, seems redundant in Constant Voltage mode, since the point of that mode is to regulate the voltage to a fixed (constant) setting. My question thus is why is the Overvoltage Protection necessary? Shouldn't the Constant Voltage mode already prevent the voltage from exceeding the user supplied voltage settings sans momentary bursts due to transient responses to changes in current load?

Thanks for any clarification on this topic.

[free_electron]

this is to trigger the internal crowbar circuitry.

if an external force were to try to lift the power rail the internal crowbar will kick in.

let's say you have a 5 volt supply and a relay turns off. this relay is missing the back-emf diode : this could induce a large voltage spike on the power rail frying the electronics.  the OVP will kick in clamping the rail.

[electr_peter]

OVP may seem redundant at first, but it is very useful.
OVP gives two key advantages:

• if something terribly wrong happens inside power supply, it will kill the output rather than the load. In other words, it is very good insurance policy for those rare cases.
• If somebody (most likely you) accidentally enters 30.00 instead of 3.00 or rotates a knob too far and presses Enter from a reflex, OVP will save the day.

Former case is very rare (if power supply has OVP, then it is pretty good quality from the start).
Latter case is more important - anybody can make a mistake, it is just a question of time before something happens. Remember that you are just few button presses from an over-voltage disaster. Nobody wants to break plates and glasses, but they break all the time anyway...

[mikeselectricstuff]

It's for when you're adjusting the voltage when looking at something other than the PSU's meter.

[AF6LJ]

In short; OVP is there to protect your load, because everything fails sooner or later.

[EEVblog]

• If somebody (most likely you) accidentally enters 30.00 instead of 3.00 or rotates a knob too far and presses Enter from a reflex, OVP will save the day.

This is its most useful purpose.
When you are powering an expensive prototype board, the last thing you want is to accidentally blow the board.
As an example, I worked on a ridiculously expensive demonstrator military board, it was a hand made one-off that took 9 months of work to get to that point.
If we fried that board it literally would have been the loss of a $20M contract because wouldn't have had time or parts to make another one.
We searched the whole company for supplies that had OVP before we even thought of powering it up.

[EEVblog]

this is to trigger the internal crowbar circuitry.

Some PSU's with OVP don't have crowbar circuitry, it's just a software interlock function to prevent a PEBKAC

[VK5RC]

Don't some linear psu fail with the power transistor as a short exposing the connected device to the full rail voltage,  with  the transistor buggered no regulation occurs?

[BravoV]

According to my Kenwood lab psu manual, the OVP level (adjustable) serves two purposes :

1. As mentioned above, to protect if the voltage adjustment pot accidentally turned above the safe voltage, as this PSU is adjustable mechanically using pot.
2. To shutdown the whole psu once voltage rail is reaching the set OVP voltage externally, eg. the output is set to 3.3 Volt, and by accident the external circuit that the psu is powering suddenly raised the power rail say to 5 Volt, this also will trigger the OVP mechanism, if its set say at > 3.3 volt.

Mine has the solenoid driven main power switch, once the OVP triggered, the main power switch will turn off mechanically, making a loud clicking sound as the internal selenoid is pulling the switch lever down.

Picture of the main switch attached below.

[retrolefty]

Don't some linear psu fail with the power transistor as a short exposing the connected device to the full rail voltage,  with  the transistor buggered no regulation occurs?

 Absolutely. That''s why the better ones always included an internal crow-bar as standard or at least an option. I worked and repaired many large Lambda 5 volt PSs made in the 70/80s.

[ConKbot]

If you have a nonlinear or otherwise difficult load, it can kick in while running that.  I was setting up an improvised shaker table with a seat shaker transducer (bolts to a home theater chair so you can feel the bass more) a class D amplifier reference eval kit from international rectifier we had laying around, and 2 power supplies, one for +25 V, one for -25 V.

When we hit resonance, the transducer became a very reactive load, and the output mosfets sent the current back into the supply rails, voltage shot up, and tripped the OVP at 33v, and it shut down before things started to fry. There wasn't much in the way of big bulk capacitance, just a few thousand uf on the eval kit.

A few power resistors on the rails as a static load, and the quick and dirty shaker was working again.

[SeanB]

Only issue is that OVP actually needs to be tested to ensure it will function as designed. OVP can fail silently as open circuit, and then when the PSU does it's own failure then the whole lot goes bang.


Had to fix a computer that had had that happen. I still have the one board in a frame on the wall. Replaced every component in the 5V rail and the crowbar circuitry as well, as a preventive measure fixing that PSU, including the resistors.

[VK5RC]

I learnt a  lot about OVP "repairing" an old HP 6632 psu which ultimately was missing an output capacitor (hard to pick on 66xx psu as half the board isn't populated and the circuit diagram covers 4 models, WTF!!!!!) , it had an 8uS pulse to the rail V+ upon switching the output on, which then triggered the OVP. This is where a good DSO with lots of memory was very helpful.
After, I have a lot of respect for those units, and they can be bought for not crazy money, but the fan is a bit noisy.