General purpose power supply question - current limiting

[Red Squirrel]

When looking at power supplies, such as these: https://www.meanwell.com/webapp/product/search.aspx?prod=LRS-600

Is it safe to assume that if you overload them, they simply drop the voltage, or do they shut off?  In data sheet under specifications it does say "Constant current limiting, unit will shutdown after 3 sec. re-power on to recover" so I'm curious what exactly this means, will it shut down, and then turn back on but at a lower voltage and essentially adjust itself so it's not outputting more than max?  Or is there a limit to just how much current limiting it can do?  Ex: if it drops to certain voltage and it's still maxed out then it just shuts down?

Intended use is for a battery bank that would also be powering an inverter, so I'm hoping it can run in current limiting mode indefinitely if the battery bank is discharged very low. But if not then I will probably want to ensure I size it high enough.  Either that or replace my inverter with an inverter-charger. 

[thm_w]

Generally the power supply will shut off, then you have to re-cycle power to turn it on again, once that spec (105-150% power) is exceeded.
Some will hiccup and shut off, then attempt turn on, see the excessive load is still there, and shut off again. Which would also not be ideal for charging.

I would either get a LED type supply that is designed for current limiting, or get a proper battery charger that will limit current and manage the cells (better choice). These are lead acid battery banks?

[tom66]

For *many* single-output converters, without a standby supply, the most likely outcome under short circuit or overload is that the supply will hiccup.  The MeanWell one you linked seems to state it has a latching fault protection, so it may have some function built into the controller chip to ensure this.

Normally, hiccup fault is ensured by design because the circuitry on the primary is powered by a winding off the transformer.  If the main switching operation stops, that supply will disappear.  Supply will switch off, then the bootstrap process will begin again.  Result will be a brief output every second or so.  Each start up cycle is more or less like a short circuit to the PSU as the output caps are usually discharged, so only primary current will be limited by the short term OCP of the controller until the Vcc cap is emptied.  This is quite stressful for the main FET if the period is short, so avoid doing it.

Some supplies have a second independent power supply for the control IC (a good, familiar example is the humble ATX computer PSU - the 5VSTBY transformer almost certainly supplies primary-side Vcc to the rest of the system.)  In those cases, the 'standby' rail will behave like a hiccup function (and sometimes, shorting the standby rail will bring the whole system down), whereas the main rail may indeed have a latching fault detect.  Some others have a way to keep the controller IC powered all the time from the mains, or to latch this state as long as mains is present.

However, in all of these cases it is not really safe to assume the current limiting is anything other than a last-ditch attempt to save the power supply or its load from destruction.  It is very line-voltage dependent, so a particularly high mains voltage will mean a greater output power limit (because it's based on primary-side current: higher mains voltage, less current, for a given power), and it's also not particularly well toleranced, so paralleling supplies and relying on OCP to balance them is very risky business (not saying you're doing this, but I've seen people try.)  Running a supply at its power limit, or bouncing right off it, for some time will probably shorten its life. The power limit will probably vary significantly with temperature too.

To do what you want, you do really need a proper charger.  Even if you could do some kind of current limiting on the PSU, a proper charger will do things like compensate the charging voltage for temperature, which is mandatory for good lifespan of lead acid batteries.  It might even do desulfation cycles for batteries that benefit from that.

[Red Squirrel]

Good to know then so these may not be suitable for what I want to do.  I know the telecom rectifiers at work will current limit and not trip if they are overloaded so I was thinking maybe these would do the same.  They do make them in 24v but they are hard to find anywhere and I presume very expensive as well.   24v chargers are harder to find as most of them are 12v, and either way, they don't tend to push that much power. They are more for charging a stand alone battery, not keeping a load on at same time.  I would just be charging at what is a nominal float voltage for the battery, which would be lower than it should be for the temperature but the goal is to also run the loads at same time and keep the battery from discharging.  The solar charge controller will handle the proper charging in the day. 

I suppose I may be better off replacing the inverter with an inverter-charger instead then when I hook up a generator it would just run the loads off the generator while charging the batteries.  I would just want to be sure to get a good inverter generator as some of the loads are on UPS so they would not like a dirty wave form.


Edit:

Further digging I found this one: https://www.mouser.ca/datasheet/2/260/PSP_600_SPEC-1511885.pdf

This one states "constant current limiting" so wonder if this does what I think it does?  It will lower voltage if it gets overloaded?  I presume there is a limit to that, then it will probably trip anyway for gross overloads.  I might actually get this one, and if this fails, I can just get another since it supports parallel operation.  I don't imagine I will actually exceed 600w anyway but at least I have the option of adding more supplies if I do. 

[2N2222A]

Power supplies based around a current mode PWM controller have inherent current limiting, but not necessarily beyond the primary side switching transistor. The output current can still climb beyond what the diodes can handle as the output Voltage drops and the diodes can be destroyed. So some additional current sensing may be needed. This kind of power supply works well for battery charging since the battery Voltage at the beginning of the charge is only slightly below the nominal Voltage, so the current won't climb too high. The output is constant power regulated with a current mode PWM controller if it has a stable input Voltage.