Nut upsuser - Nov 2008 - compaq r3000 triggering the fuse when power is reconnected

All,

I have a fairly old Compaq R3000h which is overdue for new batteries. 
Before I go and splash out a few hundred francs, I'd just need a quick
sanity check - 
when I disconnect the power (UPS under load), it cuts out pretty much
right away, so the batteries are flat (even though the charge indicator
says full).  When I reconnect the power, the surge triggers a 16A mains
fuse.  This sounds to me to be due to the batteries being completely
flat and thereby causing a major surge on startup?  Sounds reasonable?


/Per Jessen, Z?rich

Citeren Per Jessen <per at computer.org>:
> I have a fairly old Compaq R3000h which is overdue for new batteries.
> Before I go and splash out a few hundred francs, I'd just need a quick
> sanity check -
> when I disconnect the power (UPS under load), it cuts out pretty much
> right away, so the batteries are flat (even though the charge indicator
> says full).
That's usually the tell tale sign that the batteries are well past  
their 'due to replace by' date. When was the last time you ran a deep  
discharge test on them?
> When I reconnect the power, the surge triggers a 16A mains
> fuse.  This sounds to me to be due to the batteries being completely
> flat and thereby causing a major surge on startup?  Sounds reasonable?
Depends on whether or not anything is attached at that time. If a  
number of computers is attached to the outlets, the combined inrush  
current may already be enough to trigger a circuit breaker. The same  
goes if it is not the only device connected on that group (maximum  
load of a 16A fuse is about 3.7 kVA).

So while I agree that it looks like the batteries need replacement,  
this may not fix the above mentioned problem with the circuit breaker  
popping. Note that some UPS'es (and I believe the Compaq R3000h is one  
of them) allow for staged outlet switching. In that case it might make  
sense to not start all systems at the same time (if you're not already  
doing that).

Best regards, Arjen
-- 
Please keep list traffic on the list

Citeren Per Jessen <per at computer.org>:
>> That's usually the tell tale sign that the batteries are well past
>> their 'due to replace by' date. When was the last time you ran
a deep
>> discharge test on them?
>
> Not sure - I'll have to admit my ignorance and ask what is that?
Never mind, I noticed that the upscode2 driver doesn't seem to support  
this command (yet), so the chances that you actually did a runtime  
calibration are fairly slim (unless you used something else than NUT  
to control it).

[...]
> I have two of these r3000, both with almost exactly the same load
> attached - each is about 2.2kVA.  The other box does not trigger the
> fuse when power is reconnected.
Nominal load depends on the amount of VA (or W) that a UPS sees. It  
makes no difference if this is one load drawing 2.2 kVA or ten each  
drawing 220 VA, the total load will be the same.

Surge (inrush) current is something different. Usually, PC/server  
power supplies will peak around 20..50 A (each), with essentially no  
dependency on the actual load. So there is a huge difference if you  
have one or ten systems connected when it comes to inrush current. In  
many cases, five PC/server power supplies all connected to the mains  
at the same time will be enough to trip the circuit breaker. Adding  
just a single additional one may tip the scale. When it comes to  
tripping fuses, you may be fine if you stay just below a certain limit  
and will only notice a problem if you (just) exceed it.

So basically the only way to check for differences, is to disconnect  
all loads. Don't leave the mains connections attached, the power  
button on a PC/server is just a soft switch, it doesn't actually  
disconnect the mains.

Best regards, Arjen
-- 
Please keep list traffic on the list

Arjen de Korte wrote:
>> I have two of these r3000, both with almost exactly the same load
>> attached - each is about 2.2kVA.  The other box does not trigger the
>> fuse when power is reconnected.
> 
> Nominal load depends on the amount of VA (or W) that a UPS sees. It
> makes no difference if this is one load drawing 2.2 kVA or ten each
> drawing 220 VA, the total load will be the same.
Sure, I understand that - these two actually have virtually identical
2.2kW loads, except UPS#2 has one extra PC hooked up too. 
> Surge (inrush) current is something different. Usually, PC/server
> power supplies will peak around 20..50 A (each), with essentially no
> dependency on the actual load. 
20-50A startup surges??
> Adding just a single additional one may tip the scale. When it comes
> to tripping fuses, you may be fine if you stay just below a certain
> limit and will only notice a problem if you (just) exceed it.
In my case, UPS#1 has the lowest load and is the one that trips the fuse
(16A electronic) - UPS#2 has the extra PC (std desktop) and has no
problem with the fuse.  


/Per Jessen, Z?rich