I have a device that contains five separate power supplies. They are installed into industrial locations with a 24V battery system and an alternator to charge the batteries. The current voltage protection consisted of a ERZ-V20D390(
formatting link
). They suppress voltage at 39V. Occasionally the power supplies are getting fried.
I need to go with something faster, I think. Maybe the 5KP30CA (
formatting link
)?
It looks like it can handle more voltage and has a 1.0ps response time.
I would love to get some input on this? Any other ideas? Is this a good substitution?
You do not state if the alternator produces excess voltage, or if the frying comes from an alternate source.
If it is alternator over-voltage, by too high speed, and you have no option of field control, then you could use a zener resistor divider, and a power transistor shunt.
Other options are to redesign the supply as switched mode and with a higher input voltage range. Or to use batteries with a lower series resistance. You may find a super capacitor power resistor filter does the job.
Sounds like a wind turbine system. Use of a power shunt is next to essential in such systems, if you do not have a very large battery array, which never fully charges. Have you tried a wind shild?
You don't need anything faster. You have not properly identified the current path of a destructive anomaly. For example, if you assumed the current was incoming on one 24 volt wire and outgoing on the other, then a protector shunting that current between two wires means no destructive current through your supply. However, the transient current could also be down either or both 24 volts wires and outgoing through the electronics. A protector across those 24 volt wires sees no voltage while a massive voltage destroys the power supply.
An autopsy would help to identify the incoming and outgoing path through damaged power supplies. That 'find the problem before fixing it' solution is necessary.
Protectors were more than fast enough. Chances are you have installed a solution before learning the current path of a destructive transient.
And after detecting the current path, using some series inductance for slowing down any extremely fast transients, so that ordinary protectors can handle them, should be a viable option in this case (DC feed).
Not slow transients down because protectors are too slow (as others have assumed). MOV application notes discuss how to use less MOVs without damage. Transients are slowed so that fewer joules need be installed in a protection circuit. Does an inductor alleviate supply damage? Of course not. With or without that protector, a transient is still flowing through a power supply.
No transient should be anywhere near what the power supply is designed to handle (without damage). If the transient is a current source, then an inductor only means increased transient voltage and the same destructive current still flowing through that supply.
Effective protectors divert so that a transient does not pass through and harm a power supply. All those protector devices are more than fast enough. But again, a transient must first be identified before an effective solution can be designed.
MOVs and even traditional spark gaps are parallel impedance devices. In order to be useful, some series impedance is required at the feeding side.
With a healthy serial impedance, the parallel protection unit has even some chanches of survival.
From practical experience, the most likely cause of power supply damage is that for some reason or an other, some eccessive voltage will appear on the _output_ side of the power supply.
If a shunt (parallel) mode protector needs impedance to survive, the circuit has been so cost controlled as to barely large enough. MOVs are cheap. Simpler is to properly size the protection. Then MOV life expectancy increases exponentially and voltage shunted by those MOVs decreases. More joules also means the sum of energy absorbed by MOVs decreases (which is why life expectancy increased).
Second, surges are typically constant current events. Voltage increases as necessary to push that constant current through - with or without higher serial impedance. By simpling increase MOV joules, then that current even through a (theoretical) zero impedance path does not cause MOV failure.
Third, neither above point is relevant to the OP's post. OP implied protection which does not even see the transient. Without further OP information (ie autopsy of power supplies), then little that is useful can be posted. Meanwhile, every one of those shunt (parallel) mode protectors is more than fast enough. So fast that even the impedance of its two inch lead increases response time. Shunt mode response times without listing the lead length causes massive difference in response time - because shunt mode protectors are that fast.
More relevant: what the transient's incoming and outgoing current path? I suspect he has only assumed a transient is incoming on one 24 volt wire and outgoing on another. It would explain protectors between +24V to -24V do not even see that surge current. Worse, those protectors might even provide a surge current with more destructive paths through the supply. But again, only speculation due to information not provided by the OP.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.