Posted by Jim Wilkins on June 9, 2018, 12:57 pm
I ran through the same exercise with my Summit ~45 liter fridge and
APC1400 UPS. An HP digital storage scope and Fluke current probe
showed the starting surge to be 12A for 300mS. The Peak function on a
UT61E meter gave 11.29A. The resistance at the plug is 10.8 Ohms,
which gives about the same result with 120V AC.
The UT61E is a somewhat sketchy implementation of a very good intent.
The backlight is reportedly present and can be enabled by hacking.
The UT210E is also worth a look.
The Bayite PZEM-061 approximates a KAW for 80-260VAC and up to 100A.
I think a suitable load sensing circuit could be built by connecting
the fridge to the COM contacts of a DPDT relay and sensing the
thermostat closure on the NC side, then turning on the inverter and
energizing the relay with battery DC to power the fridge from the NO
contacts. The relay could stay closed for a fixed delay that's longer
than the fridge run time to avoid the need to sense the shutoff, or
you could use a current transformer and more intelligent (less
My APC1400 UPS doesn't respond to remote serial port turn-on commands
when on battery power, unless I tap into the front panel pushbutton
and beeper and build a controller, so I bought a DC powered Alpicool
fridge/freezer. which gives me considerably more frozen food (ice
cream!!) storage than the Summit, and can travel in the car.
Apparently later APC models actively discourage being (mis)used in
solar battery systems.
Posted by Johnny B Good on June 10, 2018, 2:28 am
On Sat, 09 Jun 2018 08:57:25 -0400, Jim Wilkins wrote:
That strongly suggests that the initial surge is being limited by the
motor's coil resistance which suggests an even higher power factor in
excess of 90%, perhaps even 95% or higher (the windings inductance will
plummet once the iron has been pushed into saturation by such extreme
This is an important factor with inverter gensets which, AFAICS, all
quote wattage ratings at unity power factor with never any mention of a VA
rating, unlike most UPSes. Basically, the limit is effectively the VA
limit and you'll only see the wattage match that VA rating with unity
power factor loads. It's rather handy that my fridge startup load just
about mimics a 1KW resistive load (and your Summit -45 litre fridge, a
1.44KW load for 300ms). :-)
I did follow the link and read through the review. Interesting though it
was, I've already got enough DMMs for my immediate needs for the time
being, so that was enough of a side trip thank you very much. :-)
That looks to be cheap and handy way to monitor whole house electricity
consumption. The only problems I see is that it effectively duplicates
the electricity meter and, unless there is a clamp on CT option, you'll
have to interrupt the live or neutral wire to fit the CT and tap onto the
live and neutral connections to provide both power and monitoring of the
This isn't quite as bad as it immediately looks since it can be fitted
after the CU isolator switch with the power cut off. Also, it could be
used to monitor a single lighting circuit or ring main or even several
(there seems to be plenty of space in the CT to thread all the relevant
wires in a typical domestic installation through it) which would save it
from merely duplicating the electricity meter's function. You could use
it to monitor the kitchen ring main and/or the electric cooker and/or
oven energy use for instance.
The only thing I'd suggest is that the live feed be taken off a lighting
circuit fuse for safety. I don't fancy taking the risk of a bus bar live
to neutral short via the line voltage connection wires.
There are plenty of more expensive alternatives which can transmit
their readings to a remote data logging unit which data can then be
perused with a PC which, for those interested in such metering projects,
would be a more useful and convenient feature than having to approach the
CU just to check the readings each time, especially if it's located in an
awkward to access spot (under the stairs or at the bottom of a steep
flight of irregular stone steps into the basement).
Still, I suppose the live and neutral feed can be extended easily enough
and ditto for the CT connections so, with a little more effort, it could
be placed somewhere more convenient to keep tabs on its readings. Unless
there's an auto phase reversal correction feature built into it, there's
a 50:50 chance the CT wires will need to be swapped round, an action best
done with the power shut off to avoid the risk of excess voltage in the CT
I'm assuming you're now addressing the issue of handling those once
every 4 or 5 hours, 2 or 3 seconds duration overloading events when the
fridge stat kicks in. That seems a more complicated solution to getting
the UPS to temporarily take over the load from the genset than my own.
I'm guessing the DC powered Alpicool fridge/freezer is designed to run
off a 12v car or leisure battery supply. Since it can freeze stuff, I'm
guessing it must be using a VFD compressor motor with proportional
control of the compressor speed to maintain the required temps, neatly
eliminating the issue of compressor startup surge. You can clarify the
details if I haven't got them spot on.
The statement, "My APC1400 UPS doesn't respond to remote serial port
turn-on commands when on battery power" doesn't on the face of it make
sense since 'being on battery power' is normally shorthand for the UPS
supplying mains voltage using 'battery power' to drive its mains voltage
inverter to serve the protected load.
I can only guess you're referring to an automated shutdown instigated as
a result of a signal from the UPS monitoring S/W in whichever PC was
chosen to issue similar shutdown requests to any other PCs that are also
running off the same UPS protected supply which will complete the shutdown
sequence by issuing a final delayed shutdown request to the UPS so as to
preserve its battery pack from any further unnecessary depletion.
Even if this leaves the UPS in some sort of idle state, waiting for
return of supply so it can recharge its battery pack and ultimately
restore mains power through to the load after short refractory period, it
wouldn't make a lot of sense to allow a restart until at least after
resumption of supply so I can understand the logic in not responding
'when on battery alone with the inverter shut down' to such a startup
request via its remote interface.
Otoh, you might be able to pre-empt the refractory period on return of
supply via the remote interface but only if the battery state is above a
minimum state of charge (typically, from memory, 25% is the minimum
requirement for a UPS restart).
I'm guessing the issue of later APC models actively discouraging being
misused in solar battery systems is to do with using the solar charged
battery as a stand in for its regular battery pack or as a supplementary
battery in parallel to the UPS's own battery pack.
If you're talking about sub 2KVA UPSes such as the SmartUPS700, these
rely upon the very limited run time of their pair of 7AH SLAs to save
their under specced transformers from burning out. The earlier models
just relied on battery voltage monitoring alone to decide when to give up
and shut down since the installed batteries were incapable of burning out
the transformer due to their deliberately limited capacity.
The later models might now be including transformer temperature
monitoring, a feature formerly deemed unnecessary, to force a shutdown
regardless of the battery state; a feature that would spoil the plans of
any solar panel power enthusiast trying to do things on the cheap.
The larger 2KVA and up UPSes with seperate battery box will usually
allow a doubling or tripling up of battery capacity just fine since their
transformers are rated for indefinite run times, even if this does rely
upon forced cooling to ensure this. I should imagine it's these large
UPSes that will be the most amenable to such solar powered setups but if
it were me, I'd be inclined to improve the UPS's ventilation just to be
sure to reduce the thermal stress of extended inverter run times.
 There may even be whole house energy monitors with several CT
channels to allow individual circuits to be monitored independently but
I'm only guessing since I've never had enough interest to check out the
availability of such kit. I'm only assuming the existence of such since
that seems such a blindingly obvious feature to have on a more expensive
and comprehensive energy monitoring system.
Johnny B Good
Posted by Jim Wilkins on June 10, 2018, 5:47 pm
It may have a PTC current limiter.
The scope showed constant current for the 300mS.
I bought it for datalogging, along with several less expensive
Except for starting surges DVMs are fast enough to capture most normal
battery and AC power events and the optical isolation suits them to
placement anywhere in the systen without short circuit or common mode
concerns, a problem I had to solve when instrumenting a prototype
electric vehicle. Optically isolated DVMs can accurately measure the
voltages at the battery and at the load and across a shunt in the
negative or positive lead between them with no interference issues.
The TP4000's accuracy with Type K thermocouples is unimpressive but if
you import the data file into a spreadsheet it can apply slush and
teapot calibration corrections. The interface program applies
linearizing corrections to the raw readings that appear on the DVM.
My older, thicker laptops have CardBus and Expresscard slots for
serial or USB port expanders that allow up to four meters per
computer. The separate datalog files can be synchronized by aligning
the time stamps in a spreadsheet. Integrating voltage and current to
total wattage is a simple summation.
I wired it to the outlet strip that distributes emergency power from
my UPS. Normally the strip is plugged into the stove 120V outlet to
serve various cooking appliances and the meter's large backlit display
reminds me to shut them off.
When a night ice storm threatens I plug the APC in to charge through a
KAWez, and the fridge into it through the PZEM-061. Then it becomes
obvious that the APC draws at least twice as much wattage as it
supplies to the fridge, even after the battery float current drops
low. During fridge-only run time tests the VAC-1100A that monitors
battery charge and discharge shows a similar result.
I can shut down the fridge once all its contents fit the Alpi (or a
cooler in the woodshed or the car, weather permitting) and then have
time and battery life to shovel through the snow and ice and fallen
branches to the back of the house and clear enough space for a
generator, again weather permitting. Sometimes we are directly
downwind of Hudson's Bay.
US wiring doesn't use ring mains.
The breakers for individual circuits plug directly into and cover the
two hot busbars. The busbar contacts are interleaved such that each
connects to every second breaker in the column, so a double breaker
supplies 240V and a single one supplies 120V from whichever side of
the center-tapped pole transformer happens to be under it. Typically
each 120V circuit is a series string and each 240V circuit drives a
single load, though the rules and practices and level of inspection
have changed over time.
Since there's relatively little I can do to fine tune appliances
beyond using them less I just plug them into a P4460 KAWez and record
the monthly cost of operation.
That might address the issue of the UPS's wasteful idle current by
allowing the controller to turn it off when not needed, at night when
the fridge is the only load. The APC1400 only complains about the
starting current when the batteries are low.
You hit it. It uses a Chinese copy of the Danfoss/Secop compressor,
this one I think:
I used the APC command set revealed in apcupsd documentation to write
a status and control program that runs on this laptop.
The serial cable is nonstandard and battery power appears on one of
the pins. I suspect the port was meant for production line testing.
APC is unhappy that their proprietary protocol was leaked and won't
Starting the APC1400 on battery power alone requires pushing the Start
button until the beeper sounds, then quickly releasing it, a
technician rather than an operator mode. Unlike other front panel
functions there isn't a serial port command that duplicates it. I can
run it only as needed on batteries during the day but not of course
when I'm asleep.
The APC1400 900W, 1400VA pure sine UPS has a controlled fan and an
Anderson Powerpole connector on the rear panel that serves as a
battery series disconnect, but can easily be reconfigured as an
external input. I leave the method to those able to figure it out for
themselves. The older model I have can be programmed for the number of
VRLA external battery packs, or their wet deep cycle equivalent. I
left the float voltage at the 13.6V AGM level to avoid hydrogen and
equalize the batteries outdoors periodically.
I worked for a former APC engineer for a while and learned a lot about
the problems of hacking a UPS for solar power. He said what you did
about their cost-cutting design.
I've tested it to full load but the fridge draws only around 100W.
However it pulls nearly the full 1400VA to start, and blips the Low
Battery alarm as they get low. .
Posted by danny burstein on June 3, 2018, 3:02 am
Maybe, maybe not. There are all sorts of "invisible", for want of
a better term, electrical loads that are big enough to be a
a: in the US, the common standard way natural gas stove/oven
combinations operate these days is that the stove top is lit
by an electrcial spark ignitor. Noisy, but low current and
it shuts off after a couple of seconds.
the OVEN is different. It's very common for that gas to get
ignited by an electrical glow plate, which
1: pulls 500 watts
2: keeps drawing all that power the entire time the oven's
gas flame is lit.
b: refrigerator freezers may only pull 100 to 200 watts these
days (often even less),
when they go into auto-defrost mode, the heater can easily
pull 500 watts.
So you really need to check for these issues.
What would also be nice is for other manufacturers to pick up
a design Yamaha used to offer. They had a line of electric (battery)
start generators which... kept the battery in the circuit.
This allowed (numbers for illustration) their 2 kw generator
to, for a short period (as in when your well pump started)
provide 3 kw.
Hence you could (again, numbers for illustration) measure
out 1.5 kw in peak demand and get a 2 kw unit, and not worry
about the extra power requirments when motors kicked in.
Otherwise, you might have needed the extra capacity of
a larger, more expensive, and less fuel efficien 5 kw unit.
Knowledge may be power, but communications is the key
[to foil spammers, my address has been double rot-13 encoded]
Posted by Johnny B Good on June 3, 2018, 5:02 pm
On Sun, 03 Jun 2018 03:02:34 +0000, danny burstein wrote:
"hidden" I think is the word you were looking for. :-)
We have a gas hob which uses a mains powered spark igniter just like
that. That's been a common feature of gas hobs in the UK for the past 20
years or more. An alternative to that is a battery powered spark igniter
which our previous gas hob, purchased some 25 years ago, used (for the
first couple of years anyway until it failed). A new battery didn't fix
it so we resorted to the old fashioned way of lighting the burners until
we replaced it some two or three years ago.
That's an interesting use of electricity to supplement the heat
contribution to a gas oven. I suppose it has the merit of eliminating the
risk of a flame blow-out and the potential for a gas explosion. Since we
moved house some 35 years ago, we've only had all electric (built in)
ovens. I'm not sure what the practice is with gas ovens in the UK since
they're not as common as they used to be back in the days of free
standing gas cookers.
Very good points raised, particularly in regard of auto-defrosting
heater elements in fridges and freezers. It's rather like overlooking the
reason for not powering your 25 watt (average power consumption) laser
printer via the protected outlets on a UPS due to the 500W fast warm up
drum heater that only draws power for 5% of the time to maintain drum
temperature after the initial half minute run up. It may only average 25
watts but it's those 2 or 3 second bursts of 500W that'll trip your 450W
700VA UPS out every time.
We've got a seperate under the counter fridge and chest freezer, neither
of which is blessed (cursed) with such defrost heaters. Modern fridges, I
understand, usually work smarter rather than harder with their automatic
Since the ideal temperature in a fridge is between 5 and 8 deg C (41 to
46.4 deg F), the (or part of the) cold plate can be placed low down at
the back just above moulded gutters that lead to a drain hole which drips
the condensate onto the can of the compressor, the heat of which
evaporates it back to the atmosphere from whence it came.
Eventually, that cold spot humidity collector will frost up enough to
warrant a proper defrost by hand operation but this is a small price to
pay to eliminate a counter-productive heating element inside something
that's meant to be kept cool at *all* times.
Freezers otoh, are a different kettle of fish in regard of keeping frost
build up at bay, unlike a fridge which isn't meant to even drop as low as
0 deg C (32 deg F), they have to contend with cabinet temperatures
typically around the -18 deg C mark (0 deg F) so frost build up is
inevitable, requiring either manual intervention or else a heating based
automatic defrost cycle.
If you're "cheap" like me, you won't tolerate the expense (both capital
and running costs) of a combined fridge-freezer with its automated
defrost cycles, opting instead for seperate fridge and the classic chest
freezer guaranteed to not have any such encumbrances.
I don't believe it's practical to incorporate automatic defrost in a
chest freezer anyway (BICBW so please feel free to correct me if needs
be), besides which, the chest freezer format remains a popular option
because of its improved efficiency at keeping its contents frozen for
long periods of time even when the lid is opened up two or three times a
day to add to or retrieve from its contents.
A top opening lid offers minimum disturbance to the freezing cold air
within, especially if opened and closed slowly, unlike the vertically
hung door of a fridge-freezer where it's a race to add to or retrieve
from its contents before all the freezing cold air has dropped onto the
kitchen floor to be replaced with warm air. The conventional domestic
fridge suffers this problem too but we're only dealing with chilled air
rather than the more expensively produced freezing cold air of a freezer
so the price of convenience in this case is far better value.
I guess, if you're using an older fridge freezer instead of a seperate
chest freezer and fridge, you're not going to be dicking around with a
cheap 1KW/1.2KW peak @ unity power factor inverter genset like the
Parkside model I've been discussing here.
Otoh, if you've recently pushed the boat out on an even higher
efficiency "Analogue Compressor" fridge-freezer (in truth an inverter VFD
compressor type) which doesn't suffer from the compressor startup surge
loading effect, a cheap inverter genset might do the trick, assuming a
smarter auto-defrosting algorithm that doesn't involve half kilowatt
heaters to brute force the defrost cycle.
I'm not sure about this but you'd think if the manufacturers were going
to the expense of a digital VFD setup, they'd have put more thought into
designing a less brute force approach to the question of auto-defrost so
as not to compromise their A+++ ratings.
Anyway, as you pointed out, it's the hidden vampire loads you need to be
aware of when trying to size an emergency power generator for home use,
an issue I'd overlooked simply because our fridge and chest freezer are
only cursed by the compressor startup surge issue and the oven is all
electric so totally out of the running. The gas hob will likewise not be
part of the emergency power load since the burners can still be lit the
old fashioned way just like any such gas hob (the flame out gas cut-off
valves are self powered from the thermo-couple sensor current making them
independent of a mains voltage supply anyway).
Good for Yamaha for "thinking (just a tiny bit) out of the box".
Although none of that aught to be "Patentable" (it's just an obvious use
of existing technology after all), I wouldn't mind betting they got their
patent lawyers to give it the good old "College Try" to, if not stop the
competition dead in its tracks, then at least slow it down a little. The
competition might be contesting or working around Yamaha's patents as we
discuss the matter.
BTW, well pumps are the very rare exception in this ever so green and
pleasant land of ours, here in Blighty. That's yet another difficult
startup load most UK citizens rarely have to include in their emergency
back up power plans for the relatively small proportion of the population
that have actually given it any thought at all. :-)
Britain is such a small and self contained kingdom compared to the US,
that it would take a peculiarly extreme disaster to cut off water
supplies on any significant scale. In the event of nuclear Armageddon
most everyone's problems will likely be over in a flash unlike in the
remoter parts of the US where Survivalism does at least make some sense.
That's not to say the far northern parts of Scotland, in particular the
northern isles, would lend some credibility to the Survivalist philosophy
but that's an even tinier fraction of the UK population compared to the
US case where Survivalism might pay its promised dividend.
 If I can work out how to apply the existing switching based voltage
converter and VFD technologies to get the most utility from a basic ICE
with integrated PM three phase multi-pole flywheel alternator machine,
then it can hardly be "patentable"... surely?
 Just imagine it! A world swept clean of a 7 billion surplus of the
"Chattering Masses" onto which so much of the planet's resources have
been squandered. Now *that* would be something worth surviving for, a
world where meeting a fellow human being could be considered a rare and
welcome treat rather than the all too commonplace irritation it can often
Don't get me wrong. I'm not advocating the mass culling of 99.9% of
Humanity, who like 99.999% of Humanity, myself included, have sleepwalked
into this predicament of overpopulation that now threatens our very
existence through making unsustainable demands on the planet's resources,
since I'm likely to be in that 99.9% group destined to make way for the
hopefully grateful 0.01% of Humanity that does survive the seemingly
inevitable nuclear Armageddon.
From what I see in the news of world events, particularly political
events where the politicians (and electorate) seem to getting stupider by
the hour, I can well imagine a real "Doomsday Clock" in various offices
of "The Golden Rulers" dispersed around the several underground nuclear
bunker complexes, showing not the the last few seconds representation of
all human history to date but actual years and months left to go (perhaps
just months if recent events are any guide) before the puppet strings are
to be pulled to initiate the final nuclear Armageddon to make way for a
new hope for Humanity to have another go at getting it right this time.
Like the published doomsday clock, I can imagine it being 'tweaked'
every now and again to take account of delays in the many and varied
preparations towards the actual doomsday event.
I know this seems like yet another conspiracy theory. It can certainly
be viewed that way but, in all honesty, this is my way of hanging onto
the hope that there is purpose behind the seemingly inevitable
destruction of human civilisation rather than it just being the result of
pure dumb stupidity alone with no other purpose than MAD.
Quite frankly, I'd rather my assessment was right rather than merely a
comforting delusion. Regardless of events, death awaits us all anyway so
I might as well entertain myself with my best guess at what's really
going on as we seemingly edge closer and closer to our ultimate demise.
It's a viewpoint not unlike that of a famous philosopher/scientist who,
when asked if he believed in the existence of God, explained that his
belief in God was in the nature of a bet on his existence being true in
which case he'd reap the reward of everlasting life in heaven rather than
merely have his soul extinguished at death if he'd chosen to deny God's
existence and it turned out to be true - if his belief in the existence
in God proved a false one, he'd be no worse off anyway. Unfortunately, I
can't recall which famous philosopher/scientist it was who said this
(also, I've had to paraphrase the actual quote).
Johnny B Good