Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Non-linear loads on inverters
14 views
Skip to first unread message
bob prohaska
May 13, 2022, 11:24:12 AM5/13/22
to
I'm thinking to set up an inverter/charger and battery as an
improvised UPS for a small cluster of electronics that provide
the Internet service for my home. The inverter I'm looking at is:
https://www.amazon.com/Ampinvt-Inverter-Charger-Frequency-Batteries/dp/B098QL2VBZ/ref=cm_cr_arp_d_pl_foot_top?ie=UTF8&th=1
It's a pure sine pulse width modulating inverter with a 60Hz transformer.
On the face of it the problem looks fairly easy. Total draw is
less than fifty watts, so an inverter/charger of a few hundred
watts seems to be ample. But, there's a catch, maybe:
Each device is powered by a wallwart, which uses a capacitor-input
DC-DC converter. That means that of those fifty watts, all the current
is delivered in a short pulse at the beginning of each half-cycle. The
power factor, as reported by a kill-a-watt, is very close to 1, but
that's not the whole story.
If anybody has done something similar I'd be curious to know how it
worked out. It seems clear I need to de-rate the inverter for this
sort of load, but by how much? I think 800 watts capacity will be
more than enough for a 50 watt average load but am not certain.
Would a high-frequency-type inverter, with explicit output filtration,
be a better choice? Upsizing the inverter carries penalties in both
cost and efficiency that I'd like to minimize.
Thanks for reading,
bob prohaska
improvised UPS for a small cluster of electronics that provide
the Internet service for my home. The inverter I'm looking at is:
https://www.amazon.com/Ampinvt-Inverter-Charger-Frequency-Batteries/dp/B098QL2VBZ/ref=cm_cr_arp_d_pl_foot_top?ie=UTF8&th=1
It's a pure sine pulse width modulating inverter with a 60Hz transformer.
On the face of it the problem looks fairly easy. Total draw is
less than fifty watts, so an inverter/charger of a few hundred
watts seems to be ample. But, there's a catch, maybe:
Each device is powered by a wallwart, which uses a capacitor-input
DC-DC converter. That means that of those fifty watts, all the current
is delivered in a short pulse at the beginning of each half-cycle. The
power factor, as reported by a kill-a-watt, is very close to 1, but
that's not the whole story.
If anybody has done something similar I'd be curious to know how it
worked out. It seems clear I need to de-rate the inverter for this
sort of load, but by how much? I think 800 watts capacity will be
more than enough for a 50 watt average load but am not certain.
Would a high-frequency-type inverter, with explicit output filtration,
be a better choice? Upsizing the inverter carries penalties in both
cost and efficiency that I'd like to minimize.
Thanks for reading,
bob prohaska
Jim Wilkins
May 13, 2022, 2:20:41 PM5/13/22
to
"bob prohaska" wrote in message news:t5lt6q$uja$1...@dont-email.me...
I'm thinking to set up an inverter/charger and battery as an
improvised UPS for a small cluster of electronics that provide
the Internet service for my home. The inverter I'm looking at is:
https://www.amazon.com/Ampinvt-Inverter-Charger-Frequency-Batteries/dp/B098QL2VBZ/ref=cm_cr_arp_d_pl_foot_top?ie=UTF8&th=1
It's a pure sine pulse width modulating inverter with a 60Hz transformer.
On the face of it the problem looks fairly easy. Total draw is
less than fifty watts, so an inverter/charger of a few hundred
watts seems to be ample. But, there's a catch, maybe:
Each device is powered by a wallwart, which uses a capacitor-input
DC-DC converter. That means that of those fifty watts, all the current
is delivered in a short pulse at the beginning of each half-cycle. The
power factor, as reported by a kill-a-watt, is very close to 1, but
that's not the whole story.
------------------------
Is that worse than a rectifier that draws all its current at the peak? I
haven't had any sine inverter issues with older laptop power bricks.
It looks like an interesting first attempt. If designing (hacking) one I
would change the always-on fan to on/off thermostat or variable speed
thermistor control, like a laptop, and vary the fixed 10% (80W) ECO mode
sensing which might not detect my freezer's <40W DC inverter compressor. You
have to consider what will be cycling or running continuously on inverter
power overnight. My APC1400 pure sine UPS uses as much power internally as a
compact refrigerator, halving battery run time. 50W doesn't seem like much
until you get it from expensive batteries.
One answer mentioned a new version coming soon, that accommodates solar
input, though you can connect a solar controller directly to the battery as
I do. The APC's taper charge and additional current from a solar controller
do not conflict.
I run two solar controllers into one battery without any issues, in fact the
MPPT's manual suggests it to increase capacity. I also run two controllers
from one solar array to separate batteries. The current goes to the battery
with lower voltage, then divides as both rise. My homebrew linear regulator
charges a backup battery while the solar powered fridge is off and draws no
current when the fridge is on, due to the LM350's internal voltage drop
being higher than the PWM controller's.
These are an inexpensive way to monitor your solar array voltage and current
instead of guessing and hoping:
https://www.amazon.com/dp/B08HQM1RMF/ref=twister_B08HQLQ2BX?_encoding=UTF8&th=1
The panel voltage has diagnostic value, such as if the controller is letting
battery voltage leak back to the panels at night, or there's an open or
short in the outdoor wiring. It gives some idea of the PWM duty cycle and
how close you are to the array's capacity.
I prefer more voltage precision on the battery side, to better estimate
State of Charge, and ideally a ammeter that shows and sums charge and
discharge current like Amazon's SIN9020s / Drok 200320 (long URL). Yesterday
I finally figured out the obscure calibration instructions for it. Both of
mine were inaccurate as received.
The generator mode looks useful. A 4-stroke engine slows down between power
pulses, varying the frequency enough that my APC rejects or barely tolerates
generator AC on its least sensitive setting, so I bought an inverter
generator.
I'm thinking to set up an inverter/charger and battery as an
improvised UPS for a small cluster of electronics that provide
the Internet service for my home. The inverter I'm looking at is:
https://www.amazon.com/Ampinvt-Inverter-Charger-Frequency-Batteries/dp/B098QL2VBZ/ref=cm_cr_arp_d_pl_foot_top?ie=UTF8&th=1
It's a pure sine pulse width modulating inverter with a 60Hz transformer.
On the face of it the problem looks fairly easy. Total draw is
less than fifty watts, so an inverter/charger of a few hundred
watts seems to be ample. But, there's a catch, maybe:
Each device is powered by a wallwart, which uses a capacitor-input
DC-DC converter. That means that of those fifty watts, all the current
is delivered in a short pulse at the beginning of each half-cycle. The
power factor, as reported by a kill-a-watt, is very close to 1, but
that's not the whole story.
Is that worse than a rectifier that draws all its current at the peak? I
haven't had any sine inverter issues with older laptop power bricks.
It looks like an interesting first attempt. If designing (hacking) one I
would change the always-on fan to on/off thermostat or variable speed
thermistor control, like a laptop, and vary the fixed 10% (80W) ECO mode
sensing which might not detect my freezer's <40W DC inverter compressor. You
have to consider what will be cycling or running continuously on inverter
power overnight. My APC1400 pure sine UPS uses as much power internally as a
compact refrigerator, halving battery run time. 50W doesn't seem like much
until you get it from expensive batteries.
One answer mentioned a new version coming soon, that accommodates solar
input, though you can connect a solar controller directly to the battery as
I do. The APC's taper charge and additional current from a solar controller
do not conflict.
I run two solar controllers into one battery without any issues, in fact the
MPPT's manual suggests it to increase capacity. I also run two controllers
from one solar array to separate batteries. The current goes to the battery
with lower voltage, then divides as both rise. My homebrew linear regulator
charges a backup battery while the solar powered fridge is off and draws no
current when the fridge is on, due to the LM350's internal voltage drop
being higher than the PWM controller's.
These are an inexpensive way to monitor your solar array voltage and current
instead of guessing and hoping:
https://www.amazon.com/dp/B08HQM1RMF/ref=twister_B08HQLQ2BX?_encoding=UTF8&th=1
The panel voltage has diagnostic value, such as if the controller is letting
battery voltage leak back to the panels at night, or there's an open or
short in the outdoor wiring. It gives some idea of the PWM duty cycle and
how close you are to the array's capacity.
I prefer more voltage precision on the battery side, to better estimate
State of Charge, and ideally a ammeter that shows and sums charge and
discharge current like Amazon's SIN9020s / Drok 200320 (long URL). Yesterday
I finally figured out the obscure calibration instructions for it. Both of
mine were inaccurate as received.
The generator mode looks useful. A 4-stroke engine slows down between power
pulses, varying the frequency enough that my APC rejects or barely tolerates
generator AC on its least sensitive setting, so I bought an inverter
generator.
Bob F
May 13, 2022, 6:43:51 PM5/13/22
to
thinking about powering the low voltage devices directly from the
batteries rather than sticking an inverter and wall wart into the
circuit. It would probably be a lot more efficient.
bob prohaska
May 13, 2022, 8:29:07 PM5/13/22
to
Jim Wilkins <murat...@gmail.com> wrote:
> ------------------------
> Is that worse than a rectifier that draws all its current at the peak? I
> haven't had any sine inverter issues with older laptop power bricks.
>
I believe it's exactly the same, so if a pure sine inverter works for you
> ------------------------
> Is that worse than a rectifier that draws all its current at the peak? I
> haven't had any sine inverter issues with older laptop power bricks.
>
it has a decent chance of working in my application.
> It looks like an interesting first attempt. If designing (hacking) one I
> would change the always-on fan to on/off thermostat or variable speed
> thermistor control, like a laptop, and vary the fixed 10% (80W) ECO mode
> sensing which might not detect my freezer's <40W DC inverter compressor. You
> have to consider what will be cycling or running continuously on inverter
> power overnight. My APC1400 pure sine UPS uses as much power internally as a
> compact refrigerator, halving battery run time. 50W doesn't seem like much
> until you get it from expensive batteries.
>
I could find. I won't be able to use "intelligent" load sensing, since the
network machinery is always on. That motivates the smallest inverter possible
to minimize parasitic losses. $250 for an experiment seems justifiable.
> One answer mentioned a new version coming soon, that accommodates solar
> input, though you can connect a solar controller directly to the battery as
> I do. The APC's taper charge and additional current from a solar controller
> do not conflict.
>
space.
> I run two solar controllers into one battery without any issues, in fact the
> MPPT's manual suggests it to increase capacity. I also run two controllers
> from one solar array to separate batteries. The current goes to the battery
> with lower voltage, then divides as both rise. My homebrew linear regulator
> charges a backup battery while the solar powered fridge is off and draws no
> current when the fridge is on, due to the LM350's internal voltage drop
> being higher than the PWM controller's.
>
> These are an inexpensive way to monitor your solar array voltage and current
> instead of guessing and hoping:
> https://www.amazon.com/dp/B08HQM1RMF/ref=twister_B08HQLQ2BX?_encoding=UTF8&th=1
> The panel voltage has diagnostic value, such as if the controller is letting
> battery voltage leak back to the panels at night, or there's an open or
> short in the outdoor wiring. It gives some idea of the PWM duty cycle and
> how close you are to the array's capacity.
>
> I prefer more voltage precision on the battery side, to better estimate
> State of Charge, and ideally a ammeter that shows and sums charge and
> discharge current like Amazon's SIN9020s / Drok 200320 (long URL). Yesterday
> I finally figured out the obscure calibration instructions for it. Both of
> mine were inaccurate as received.
>
> The generator mode looks useful. A 4-stroke engine slows down between power
> pulses, varying the frequency enough that my APC rejects or barely tolerates
> generator AC on its least sensitive setting, so I bought an inverter
> generator.
an educational setup. I'm on-grid in a late 50's suburban area near Sacramento.
Lots of big trees among overhead wires. The system isn't expected to get a lot of
exercise, but it will get some. The goal is just to learn something about what works
and what doesn't, against the chance I'll need to get serious in the future.
Thanks for writing!
bob prohaska
bob prohaska
May 13, 2022, 8:44:07 PM5/13/22
to
Bob F <bobn...@gmail.com> wrote:
>
> It seems like most such devices I see all want about 12V. I would be
> thinking about powering the low voltage devices directly from the
> batteries rather than sticking an inverter and wall wart into the
> circuit. It would probably be a lot more efficient.
Most devices want 5V, that's basically the same. Clearly, a DC
>
> It seems like most such devices I see all want about 12V. I would be
> thinking about powering the low voltage devices directly from the
> batteries rather than sticking an inverter and wall wart into the
> circuit. It would probably be a lot more efficient.
system will be more efficient. The sticking point is grounding. If I
use a single power supply everything ends up sharing a common ground
and all the wiring turns into an antenna. With about a dozen gadgets
in the "data center" it's a fairly nightmarish collection of cables.
The wallwarts provide a degree of isolation that's hard to duplicate.
It might not be required, it does offer a degree of robustness.
There's a very old photo here:
http://www.zefox.net/~fbsd/pi_cluster
The collection has grown considerably since....
Thanks for writing,
bob prohaska
David Lesher
May 13, 2022, 10:08:09 PM5/13/22
to
bob prohaska <b...@www.zefox.net> writes:
>Each device is powered by a wallwart, which uses a capacitor-input
>DC-DC converter. That means that of those fifty watts, all the current
>is delivered in a short pulse at the beginning of each half-cycle. The
>power factor, as reported by a kill-a-watt, is very close to 1, but
>that's not the whole story.
There are CHEAP, low power, DC->DC converters available.
>Each device is powered by a wallwart, which uses a capacitor-input
>DC-DC converter. That means that of those fifty watts, all the current
>is delivered in a short pulse at the beginning of each half-cycle. The
>power factor, as reported by a kill-a-watt, is very close to 1, but
>that's not the whole story.
Run them right off the batteries, each set to the needed
load voltage.
Ground loops may be an issue. I don't know which DC->DC
converters are transformer-coupled. Some will be, I'd think. I
--
A host is a host from coast to coast...............wb8foz@panix.com
& no one will talk to a host that's close..........................
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433
bob prohaska
May 14, 2022, 12:30:41 AM5/14/22
to
David Lesher <wb8...@panix.com> wrote:
> bob prohaska <b...@www.zefox.net> writes:
>
>
>>Each device is powered by a wallwart, which uses a capacitor-input
>>DC-DC converter. That means that of those fifty watts, all the current
>>is delivered in a short pulse at the beginning of each half-cycle. The
>>power factor, as reported by a kill-a-watt, is very close to 1, but
>>that's not the whole story.
>
> There are CHEAP, low power, DC->DC converters available.
> Run them right off the batteries, each set to the needed
> load voltage.
>
I did look at cheap DC-DC converters. IIRC ground isolation
> bob prohaska <b...@www.zefox.net> writes:
>
>
>>Each device is powered by a wallwart, which uses a capacitor-input
>>DC-DC converter. That means that of those fifty watts, all the current
>>is delivered in a short pulse at the beginning of each half-cycle. The
>>power factor, as reported by a kill-a-watt, is very close to 1, but
>>that's not the whole story.
>
> There are CHEAP, low power, DC->DC converters available.
> Run them right off the batteries, each set to the needed
> load voltage.
>
wasn't a common feature. But, that's not really the point. I'm
trying to stick with standard, off-the-shelf plug in commodity
parts.
> Ground loops may be an issue.
Jim Wilkins
May 14, 2022, 9:14:05 AM5/14/22
to
"David Lesher" wrote in message news:t5n2u7$q3v$1...@reader1.panix.com...
bob prohaska <b...@www.zefox.net> writes:
>Each device is powered by a wallwart, which uses a capacitor-input
>DC-DC converter. That means that of those fifty watts, all the current
>is delivered in a short pulse at the beginning of each half-cycle. The
>power factor, as reported by a kill-a-watt, is very close to 1, but
>that's not the whole story.
There are CHEAP, low power, DC->DC converters available.
Run them right off the batteries, each set to the needed
load voltage.
Ground loops may be an issue. I don't know which DC->DC
converters are transformer-coupled. Some will be, I'd think. I
----------------------
This might help, or if not you haven't lost much:
https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
The one I have displayed a clean sine wave up to its 350W shut off. It's no
load power is ~10W and it has an on-demand fan.
I've tried to convert to as much 12V DC power as economically practical. My
computers are all laptops that can use 12V auto/air adapters and much of my
external storage is USB-powered portable drives (>30TB). The Bestek can
power the USB hub and AC-only monitors but not the printer. Since I don't
have cable TV my Internet is cellular -> iPhone -> USB tether or WiFi, so
there's no network wiring. That leaves refrigeration which also qualifies
for a UPS if your grid isn't reliable, and the printer which doesn't.
Most DC-powered refrigeration is wildly expensive, except for Alpicool and
clones. I have an older camping-sized C20 which is occasionally flaky, but
still usable, and a larger T60 that hasn't had a problem (yet). 100W of
solar is enough to operate them all day, with grid or battery power at
night. The automatic switching between AC and DC required some electronic
knowledge and testing although it's very simple. My main fridge is an
AC-only compact Magic Chef that the APC1400 UPS can start.
If you can find a beige APC1400 or similar free/cheap (mine were $0 and $20)
they make decent stand-alone UPS power sources if you rewire the rear
connector for two (24V) external batteries. It's an Anderson SB50 and you
have to match the color, they are keyed. The APC may need reprogramming and
the serial connector pinout isn't standard. APC told me to derate to 750W
for continuous use. Open them to remove the dead batteries by pulling the
front cover off by the finger notches.
I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
measure and understand the starting surges etc to match the power sources to
the equipment, without overbuying an expensive inverter and large battery
bank.
bob prohaska <b...@www.zefox.net> writes:
>Each device is powered by a wallwart, which uses a capacitor-input
>DC-DC converter. That means that of those fifty watts, all the current
>is delivered in a short pulse at the beginning of each half-cycle. The
>power factor, as reported by a kill-a-watt, is very close to 1, but
>that's not the whole story.
There are CHEAP, low power, DC->DC converters available.
Run them right off the batteries, each set to the needed
load voltage.
Ground loops may be an issue. I don't know which DC->DC
converters are transformer-coupled. Some will be, I'd think. I
This might help, or if not you haven't lost much:
https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
The one I have displayed a clean sine wave up to its 350W shut off. It's no
load power is ~10W and it has an on-demand fan.
I've tried to convert to as much 12V DC power as economically practical. My
computers are all laptops that can use 12V auto/air adapters and much of my
external storage is USB-powered portable drives (>30TB). The Bestek can
power the USB hub and AC-only monitors but not the printer. Since I don't
have cable TV my Internet is cellular -> iPhone -> USB tether or WiFi, so
there's no network wiring. That leaves refrigeration which also qualifies
for a UPS if your grid isn't reliable, and the printer which doesn't.
Most DC-powered refrigeration is wildly expensive, except for Alpicool and
clones. I have an older camping-sized C20 which is occasionally flaky, but
still usable, and a larger T60 that hasn't had a problem (yet). 100W of
solar is enough to operate them all day, with grid or battery power at
night. The automatic switching between AC and DC required some electronic
knowledge and testing although it's very simple. My main fridge is an
AC-only compact Magic Chef that the APC1400 UPS can start.
If you can find a beige APC1400 or similar free/cheap (mine were $0 and $20)
they make decent stand-alone UPS power sources if you rewire the rear
connector for two (24V) external batteries. It's an Anderson SB50 and you
have to match the color, they are keyed. The APC may need reprogramming and
the serial connector pinout isn't standard. APC told me to derate to 750W
for continuous use. Open them to remove the dead batteries by pulling the
front cover off by the finger notches.
I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
measure and understand the starting surges etc to match the power sources to
the equipment, without overbuying an expensive inverter and large battery
bank.
Jim Wilkins
May 14, 2022, 9:41:54 AM5/14/22
to
"bob prohaska" wrote in message news:t5nb9f$rtv$1...@dont-email.me...
I'm
trying to stick with standard, off-the-shelf plug in commodity
parts.
----------------------
My system is off-the-shelf except for solar battery charging which is only
slightly custom, mainly adding breakers, fuses and meters in boxes and
interconnecting them. I couldn't find DC substitutes for everything so the
computers all normally run on grid power without a UPS, as I estimate the
depreciation cost of cycled batteries to be higher than the grid per KWH
consumed. The laptop's batteries act as a UPS for themselves and USB-powered
external storage. 10 minutes of remaining battery life is enough to switch
to the UPS when the grid fails. They are the thicker and heavier
desktop-replacement business models that accept second internal hard drives,
so they have an SSD to boot from and a Terabyte HDD for storage.
I'm
trying to stick with standard, off-the-shelf plug in commodity
parts.
My system is off-the-shelf except for solar battery charging which is only
slightly custom, mainly adding breakers, fuses and meters in boxes and
interconnecting them. I couldn't find DC substitutes for everything so the
computers all normally run on grid power without a UPS, as I estimate the
depreciation cost of cycled batteries to be higher than the grid per KWH
consumed. The laptop's batteries act as a UPS for themselves and USB-powered
external storage. 10 minutes of remaining battery life is enough to switch
to the UPS when the grid fails. They are the thicker and heavier
desktop-replacement business models that accept second internal hard drives,
so they have an SSD to boot from and a Terabyte HDD for storage.
bob prohaska
May 14, 2022, 12:30:57 PM5/14/22
to
Jim Wilkins <murat...@gmail.com> wrote:
>
> This might help, or if not you haven't lost much:
> https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
>
> The one I have displayed a clean sine wave up to its 350W shut off. It's no
> load power is ~10W and it has an on-demand fan.
>
Now you're making me think 8-) That often complicates matters.......
>
> This might help, or if not you haven't lost much:
> https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
>
> The one I have displayed a clean sine wave up to its 350W shut off. It's no
> load power is ~10W and it has an on-demand fan.
>
My "data center" draws about 50 watts steady-state. If we guess that all
the power is delivered in the leading quarter cycle of the waveform that's
about 200 watts recurrent peak power. In fact, the power has to come from
the second half of the quarter cycle, since the waveform voltage is too low
in the first. That puts the recurrent peak near 400 watts. I can't find
any overcurrent-versus-shutdown-time statement among the answered questions,
for some reason there's a problem posting a new question at the moment.
Still, it's an interesting idea. I'd have to buy a separate battery charger
that can deliver both the inverter draw and battery recharge current, but
that isn't out of the question. It clearly is too small to run my fridge.
> I've tried to convert to as much 12V DC power as economically practical. My
> computers are all laptops that can use 12V auto/air adapters and much of my
> external storage is USB-powered portable drives (>30TB). The Bestek can
> power the USB hub and AC-only monitors but not the printer. Since I don't
> have cable TV my Internet is cellular -> iPhone -> USB tether or WiFi, so
> there's no network wiring. That leaves refrigeration which also qualifies
> for a UPS if your grid isn't reliable, and the printer which doesn't.
>
grid is usually up.
> Most DC-powered refrigeration is wildly expensive, except for Alpicool and
> clones. I have an older camping-sized C20 which is occasionally flaky, but
> still usable, and a larger T60 that hasn't had a problem (yet). 100W of
> solar is enough to operate them all day, with grid or battery power at
> night. The automatic switching between AC and DC required some electronic
> knowledge and testing although it's very simple. My main fridge is an
> AC-only compact Magic Chef that the APC1400 UPS can start.
>
> If you can find a beige APC1400 or similar free/cheap (mine were $0 and $20)
> they make decent stand-alone UPS power sources if you rewire the rear
> connector for two (24V) external batteries. It's an Anderson SB50 and you
> have to match the color, they are keyed. The APC may need reprogramming and
> the serial connector pinout isn't standard. APC told me to derate to 750W
> for continuous use. Open them to remove the dead batteries by pulling the
> front cover off by the finger notches.
>
have become relatively scarce in my area. There's some selection in the SF Bay
area, but the cost of driving and the uncertainty of finding something useful
makes the search rather unattractive.
> I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
> measure and understand the starting surges etc to match the power sources to
> the equipment, without overbuying an expensive inverter and large battery
> bank.
danny burstein
May 14, 2022, 12:55:59 PM5/14/22
to
In <t5olfv$esp$1...@dont-email.me> bob prohaska <b...@www.zefox.net> writes:
[snip]
>> I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
>> measure and understand the starting surges etc to match the power sources to
>> the equipment, without overbuying an expensive inverter and large battery
>> bank.
>One is blind without a DSO. I should probably do the same if I get serious.
Just hop over to the local Heathkit outlet and...
ummm...
never mind
--
_____________________________________________________
Knowledge may be power, but communications is the key
dan...@panix.com
[to foil spammers, my address has been double rot-13 encoded]
[snip]
>> I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
>> measure and understand the starting surges etc to match the power sources to
>> the equipment, without overbuying an expensive inverter and large battery
>> bank.
>One is blind without a DSO. I should probably do the same if I get serious.
ummm...
never mind
--
_____________________________________________________
Knowledge may be power, but communications is the key
dan...@panix.com
[to foil spammers, my address has been double rot-13 encoded]
bob prohaska
May 14, 2022, 1:51:02 PM5/14/22
to
danny burstein <dan...@panix.com> wrote:
> In <t5olfv$esp$1...@dont-email.me> bob prohaska <b...@www.zefox.net> writes:
>
> [snip]
>
>>> I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
>>> measure and understand the starting surges etc to match the power sources to
>>> the equipment, without overbuying an expensive inverter and large battery
>>> bank.
>
>>One is blind without a DSO. I should probably do the same if I get serious.
>
> Just hop over to the local Heathkit outlet and...
>
> ummm...
>
> never mind
>
8-)
> In <t5olfv$esp$1...@dont-email.me> bob prohaska <b...@www.zefox.net> writes:
>
> [snip]
>
>>> I had to buy a Tekna digital storage scope and Hantek AC+DC current probe to
>>> measure and understand the starting surges etc to match the power sources to
>>> the equipment, without overbuying an expensive inverter and large battery
>>> bank.
>
>>One is blind without a DSO. I should probably do the same if I get serious.
>
> Just hop over to the local Heathkit outlet and...
>
> ummm...
>
> never mind
>
Jim Wilkins
May 14, 2022, 2:49:19 PM5/14/22
to
"bob prohaska" wrote in message news:t5olfv$esp$1...@dont-email.me...
Jim Wilkins <murat...@gmail.com> wrote:
>
> This might help, or if not you haven't lost much:
> https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
>
> The one I have displayed a clean sine wave up to its 350W shut off. It's
> no
> load power is ~10W and it has an on-demand fan.
>
Now you're making me think 8-) That often complicates matters.......
My "data center" draws about 50 watts steady-state. If we guess that all
the power is delivered in the leading quarter cycle of the waveform that's
about 200 watts recurrent peak power. In fact, the power has to come from
the second half of the quarter cycle, since the waveform voltage is too low
in the first. That puts the recurrent peak near 400 watts. I can't find
any overcurrent-versus-shutdown-time statement among the answered
questions,
for some reason there's a problem posting a new question at the moment.
Still, it's an interesting idea. I'd have to buy a separate battery charger
that can deliver both the inverter draw and battery recharge current, but
that isn't out of the question. It clearly is too small to run my fridge.
----------------------
You could buy one Bestek to test, load it up with your data center and if it
stays on then add more loads to determine the margin. I can find an overload
trip point with a resistive load, a hotplate and a Variac, but I don't know
exactly how AC switching sources sense overload, especially with poor power
factors. It may be continuous power, temperature rise, instantaneous
amperage, a combination, or something else I haven't heard of. I was a tech
who tested complex circuits but designed only simple ones.
The combination I'm working on uses AC plus solar to charge a battery. The
worst cases to plan for are when the grid returns late at night after the
battery has discharged and draws high charging current, and when the input
AC fails with 12V on the charger output. A simple rectifier charger can
handle that but one with a regulator may not. I haven't found an inexpensive
12V switching supply with voltage adjustment and a current limit so I've
been adding a buck or buck-boost converter after a cheap or the OEM supply
to limit the current and adjust the voltage to the battery's float level,
which is as high as it can be left permanently. A series diode cures the
backfeed problem, unless the converter is rated to charge a battery.
This is the converter I use with the Alpicool T60's power supply. I like
meters to see what's going on when I experiment.
https://www.amazon.com/DROK-Converter-5-3V-32V-Regulator-Transformer/dp/B078Q1624B?ref_=ast_sto_dp&th=1
The Alpicool power supply gets hot if left at 5A so the converter's current
limit is set to 4.5A. In normal operation the T60 draws about 2.5~3A but its
starting surge exceeds 4.5A and comes partly from the battery. The
converter's output voltage is 13.8V which puts 13.4V on the AGM battery
after a series Schottky diode that allows trying other supplies that lack
backfeed protection. The float voltage could be higher if there was no
additional solar input, or can be set for other types of battery such as
Lithium.
Jim Wilkins <murat...@gmail.com> wrote:
>
> This might help, or if not you haven't lost much:
> https://www.amazon.com/BESTEK-300Watt-Power-Inverter-Adapter/dp/B07KQ4Q2L5
>
> The one I have displayed a clean sine wave up to its 350W shut off. It's
> no
> load power is ~10W and it has an on-demand fan.
>
Now you're making me think 8-) That often complicates matters.......
My "data center" draws about 50 watts steady-state. If we guess that all
the power is delivered in the leading quarter cycle of the waveform that's
about 200 watts recurrent peak power. In fact, the power has to come from
the second half of the quarter cycle, since the waveform voltage is too low
in the first. That puts the recurrent peak near 400 watts. I can't find
any overcurrent-versus-shutdown-time statement among the answered
questions,
for some reason there's a problem posting a new question at the moment.
Still, it's an interesting idea. I'd have to buy a separate battery charger
that can deliver both the inverter draw and battery recharge current, but
that isn't out of the question. It clearly is too small to run my fridge.
You could buy one Bestek to test, load it up with your data center and if it
stays on then add more loads to determine the margin. I can find an overload
trip point with a resistive load, a hotplate and a Variac, but I don't know
exactly how AC switching sources sense overload, especially with poor power
factors. It may be continuous power, temperature rise, instantaneous
amperage, a combination, or something else I haven't heard of. I was a tech
who tested complex circuits but designed only simple ones.
The combination I'm working on uses AC plus solar to charge a battery. The
worst cases to plan for are when the grid returns late at night after the
battery has discharged and draws high charging current, and when the input
AC fails with 12V on the charger output. A simple rectifier charger can
handle that but one with a regulator may not. I haven't found an inexpensive
12V switching supply with voltage adjustment and a current limit so I've
been adding a buck or buck-boost converter after a cheap or the OEM supply
to limit the current and adjust the voltage to the battery's float level,
which is as high as it can be left permanently. A series diode cures the
backfeed problem, unless the converter is rated to charge a battery.
This is the converter I use with the Alpicool T60's power supply. I like
meters to see what's going on when I experiment.
https://www.amazon.com/DROK-Converter-5-3V-32V-Regulator-Transformer/dp/B078Q1624B?ref_=ast_sto_dp&th=1
The Alpicool power supply gets hot if left at 5A so the converter's current
limit is set to 4.5A. In normal operation the T60 draws about 2.5~3A but its
starting surge exceeds 4.5A and comes partly from the battery. The
converter's output voltage is 13.8V which puts 13.4V on the AGM battery
after a series Schottky diode that allows trying other supplies that lack
backfeed protection. The float voltage could be higher if there was no
additional solar input, or can be set for other types of battery such as
Lithium.
Jim Wilkins
May 15, 2022, 1:23:32 PM5/15/22
to
"bob prohaska" wrote in message news:t5olfv$esp$1...@dont-email.me...
Still, it's an interesting idea. I'd have to buy a separate battery charger
that can deliver both the inverter draw and battery recharge current, but
that isn't out of the question. It clearly is too small to run my fridge.
--------------------
that can deliver both the inverter draw and battery recharge current, but
that isn't out of the question. It clearly is too small to run my fridge.
Let us know what you find for a battery charger. Your use for one is
different from normal, more what a solar controller does, which is to float
charge the battery until a load draws it down, then temporarily raise the
voltage to recharge it faster.
My APC1400 charges a depleted battery at 6A until the voltage reaches float,
which I set to 13.6V to avoid gassing, then taper charges with slowly
decreasing current. The battery is considered to be fully charged when the
charging current drops to 1% of the Ah capacity, 180mA for the 12V 18Ah AGMs
they come with. I've read that a lead-acid battery reaches the float voltage
at around 70~80% of full charge.
Lithiums are different, read the manufacturer's recommendations. I can't
justify spending four or more times as much for similar performance, my lead
batteries last over 10 years. I've repaired industrial Lithium battery
controllers but don't trust my ability to design one. They aren't simple.
https://www.ti.com/lit/ds/symlink/bq25798.pdf?ts=1652633595625&ref_url=https%253A%252F%252Fwww.ti.com%252Fproduct%252FBQ25798
https://www.oreillyauto.com/how-to-hub/battery-charger-maintainer-differences
I design and build my own chargers that can be manually set to either charge
or maintain the common types of battery, including power tool packs, so I
can't recommend a good one to buy. The ones I installed on the Alpicools use
grid power to float the battery and solar power at higher voltage to
recharge it. With adequate solar power the grid consumption is less than 1
Watt. Today is overcast so they take all available solar power and make up
the rest from the grid. The batteries may supply some power when the fridge
is on and recharge when it's off.
bob prohaska
May 15, 2022, 10:19:25 PM5/15/22
to
Jim Wilkins <murat...@gmail.com> wrote:
> "bob prohaska" wrote in message news:t5olfv$esp$1...@dont-email.me...
>
> Still, it's an interesting idea. I'd have to buy a separate battery charger
> that can deliver both the inverter draw and battery recharge current, but
> that isn't out of the question. It clearly is too small to run my fridge.
>
> --------------------
>
> Let us know what you find for a battery charger. Your use for one is
> different from normal, more what a solar controller does, which is to float
> charge the battery until a load draws it down, then temporarily raise the
> voltage to recharge it faster.
>
As I think it over, it does not seem possible for a battery charger to
> "bob prohaska" wrote in message news:t5olfv$esp$1...@dont-email.me...
>
> Still, it's an interesting idea. I'd have to buy a separate battery charger
> that can deliver both the inverter draw and battery recharge current, but
> that isn't out of the question. It clearly is too small to run my fridge.
>
> --------------------
>
> Let us know what you find for a battery charger. Your use for one is
> different from normal, more what a solar controller does, which is to float
> charge the battery until a load draws it down, then temporarily raise the
> voltage to recharge it faster.
>
accurately charge a battery in parallel with a load. The charger needs
an independent measurement of battery current, along with voltage, to
get an estimate of SOC. Three-terminals-and-shunt appear to be required.
Certainly possible, but not with off-the-shelf units I've found.
I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
That device unloads the battery when AC is available, so it at least has
a chance of getting the SOC right.
I'll start by experimenting with an old Delco Voyager deep cycle lead
acid battery. If the inverter behaves well I'll look into LiPO4, though
the prices do make me wince. I've had a LiPO4 starting battery in one
of my motorcycles since 2011 and the damn thing still starts. Lead acids
didn't last more than a couple of years with reasonable if not meticulous
care.
Thanks very much for your insights!
bob prohaska
Jim Wilkins
May 16, 2022, 9:49:37 AM5/16/22
to
"bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
As I think it over, it does not seem possible for a battery charger to
accurately charge a battery in parallel with a load. The charger needs
an independent measurement of battery current, along with voltage, to
get an estimate of SOC. Three-terminals-and-shunt appear to be required.
Certainly possible, but not with off-the-shelf units I've found.
[[[[[[[[[[
You are right, the charger can't determine SOC if some of its output goes
into the load, that takes a separate meter monitoring +/- battery current.
This does the job for my two DC refrigerator AC/solar/battery power
supplies.
https://robojax.com/sites/default/files/docs/robojax_SIN9020S-manual.pdf
One is an identical Drok 200320. It looks like I bought the last of them
from Amazon. Here it is with the optional protection relay on AliExpress,
which I've never bought from.
https://www.aliexpress.com/item/32784107301.html?spm=a2g0o.seodetail.topbuy.1.6b475219KrfbDR
The current sense is in series with the battery and estimates SOC from the
difference between charge and discharge currents. It displays Ah and
percentage of discharge, 0.00AH and 0% is fully charged. You program OAH
with your estimation or measurement of battery Ah capacity and then it
displays State Of Discharge as a percentage of it. Although remaining
capacity would be more useful, discharge is easier and more certain to
measure after you zero the meter at known full charge. By stopping at zero
it ignores extra charge current that goes to gassing and replacing
self-discharge.
Mine needs external Vext from a small DC-DC converter to remain accurate as
the battery voltage approaches full discharge. Calibration requires loads in
the range of 1A to 4A at 12V.
The charger and load are in parallel and you'd have to measure their
currents separately if you wanted to know it for setup and normal operation.
I have a KAW on the AC supply and this on the solar input from the panels.
https://www.amazon.com/dp/B08HQM1RMF/ref=twister_B08HQLQ2BX?_encoding=UTF8&th=1
The SIN9020s shows load current from the battery when the AC supply is off,
and charging current when the load is disconnected.
With experience you can judge SOC well enough from only the battery voltage
when idle or under load. The meters I mentioned provide that experience and
detect problems. You know actual Ah capacity only if you've recently run a
full discharge test and even then it's not consistent between runs, and
varies with the load.
]]]]]]]]]]
I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
That device unloads the battery when AC is available, so it at least has
a chance of getting the SOC right.
I'll start by experimenting with an old Delco Voyager deep cycle lead
acid battery. If the inverter behaves well I'll look into LiPO4, though
the prices do make me wince. I've had a LiPO4 starting battery in one
of my motorcycles since 2011 and the damn thing still starts. Lead acids
didn't last more than a couple of years with reasonable if not meticulous
care.
Thanks very much for your insights!
bob prohaska
----------------------
Good luck. I buy and integrate separate components and modules to learn,
plus I have the equipment to fabricate sheet metal enclosures for them, so I
don't know what is available as fully packaged plug-and-play. Generally it
won't do what I want anyway, such as the charger/inverters that can't
measure SOC or detect a 40W load. I've been building custom electronics for
industry and the government since the 70's.
Adequate battery care appears to be topping them off every month or two, and
equalizing occasionally if you have the equipment for it. Modern vehicles
may need more frequent attention due to higher key-off battery drain. An
Interstate battery lasted 15 years in my truck, until its measured starting
current became marginal for a cold start. I traded a pair of marine
batteries after ~12 years when their remaining capacity was barely enough to
run a refrigerator overnight.
Yesterday I got around to completing the battery backup circuit for one of
my DC refrigerators, and set up a recently bought and restored APC1400 UPS
to power the error check of a new 5TB external hard drive. A few hours later
they were on battery, disconnected from the grid, for thunderstorms. That's
lucky timing!
As I think it over, it does not seem possible for a battery charger to
accurately charge a battery in parallel with a load. The charger needs
an independent measurement of battery current, along with voltage, to
get an estimate of SOC. Three-terminals-and-shunt appear to be required.
Certainly possible, but not with off-the-shelf units I've found.
You are right, the charger can't determine SOC if some of its output goes
into the load, that takes a separate meter monitoring +/- battery current.
This does the job for my two DC refrigerator AC/solar/battery power
supplies.
https://robojax.com/sites/default/files/docs/robojax_SIN9020S-manual.pdf
One is an identical Drok 200320. It looks like I bought the last of them
from Amazon. Here it is with the optional protection relay on AliExpress,
which I've never bought from.
https://www.aliexpress.com/item/32784107301.html?spm=a2g0o.seodetail.topbuy.1.6b475219KrfbDR
The current sense is in series with the battery and estimates SOC from the
difference between charge and discharge currents. It displays Ah and
percentage of discharge, 0.00AH and 0% is fully charged. You program OAH
with your estimation or measurement of battery Ah capacity and then it
displays State Of Discharge as a percentage of it. Although remaining
capacity would be more useful, discharge is easier and more certain to
measure after you zero the meter at known full charge. By stopping at zero
it ignores extra charge current that goes to gassing and replacing
self-discharge.
Mine needs external Vext from a small DC-DC converter to remain accurate as
the battery voltage approaches full discharge. Calibration requires loads in
the range of 1A to 4A at 12V.
The charger and load are in parallel and you'd have to measure their
currents separately if you wanted to know it for setup and normal operation.
I have a KAW on the AC supply and this on the solar input from the panels.
https://www.amazon.com/dp/B08HQM1RMF/ref=twister_B08HQLQ2BX?_encoding=UTF8&th=1
The SIN9020s shows load current from the battery when the AC supply is off,
and charging current when the load is disconnected.
With experience you can judge SOC well enough from only the battery voltage
when idle or under load. The meters I mentioned provide that experience and
detect problems. You know actual Ah capacity only if you've recently run a
full discharge test and even then it's not consistent between runs, and
varies with the load.
]]]]]]]]]]
I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
That device unloads the battery when AC is available, so it at least has
a chance of getting the SOC right.
I'll start by experimenting with an old Delco Voyager deep cycle lead
acid battery. If the inverter behaves well I'll look into LiPO4, though
the prices do make me wince. I've had a LiPO4 starting battery in one
of my motorcycles since 2011 and the damn thing still starts. Lead acids
didn't last more than a couple of years with reasonable if not meticulous
care.
Thanks very much for your insights!
bob prohaska
Good luck. I buy and integrate separate components and modules to learn,
plus I have the equipment to fabricate sheet metal enclosures for them, so I
don't know what is available as fully packaged plug-and-play. Generally it
won't do what I want anyway, such as the charger/inverters that can't
measure SOC or detect a 40W load. I've been building custom electronics for
industry and the government since the 70's.
Adequate battery care appears to be topping them off every month or two, and
equalizing occasionally if you have the equipment for it. Modern vehicles
may need more frequent attention due to higher key-off battery drain. An
Interstate battery lasted 15 years in my truck, until its measured starting
current became marginal for a cold start. I traded a pair of marine
batteries after ~12 years when their remaining capacity was barely enough to
run a refrigerator overnight.
Yesterday I got around to completing the battery backup circuit for one of
my DC refrigerators, and set up a recently bought and restored APC1400 UPS
to power the error check of a new 5TB external hard drive. A few hours later
they were on battery, disconnected from the grid, for thunderstorms. That's
lucky timing!
Jim Wilkins
May 16, 2022, 1:42:12 PM5/16/22
to
"Jim Wilkins" wrote in message news:t5tkpe$qeq$1...@dont-email.me...
Briefly. to measure all currents there's a voltmeter and bidirectional
ammeter in the battery cable and optional unidirectional ammeters in the
charging and load leads. The battery meter also displays the integration of
discharging minus charging amps, so ideally at full charge it reads 0.00 AH
and at full discharge it shows the battery's Amp-Hour capacity, which you
can enter as OAH to make the percentage correct.
Briefly. to measure all currents there's a voltmeter and bidirectional
ammeter in the battery cable and optional unidirectional ammeters in the
charging and load leads. The battery meter also displays the integration of
discharging minus charging amps, so ideally at full charge it reads 0.00 AH
and at full discharge it shows the battery's Amp-Hour capacity, which you
can enter as OAH to make the percentage correct.
Jim Wilkins
May 17, 2022, 9:18:44 AM5/17/22
to
"Jim Wilkins" wrote in message news:t5u2di$a08$1...@dont-email.me...
I mentioned recharging a discharged battery as a potential problem, and this
morning I discovered that failure. The fridge is only keeping bread cool so
it wasn't a big problem. The power supply, a small one meant for 12V
cameras, had tested OK at 5A output into a voltage and current regulator in
the cool basement but at 74F in the kitchen its PTC thermistor overload
tripped some time overnight, while trying to keep up with a very weak
battery.
I use inexpensive buck or buck-boost converters to add precise voltage and
current control to inexpensive/surplus power supplies, to create battery
charger/maintainers that do what I want. When I was the battery tech at
Segway I learned how to charge almost any type of battery with a metered
adjustable laboratory power supply like this:
https://www.amazon.com/gp/product/B0852JZQZR/ref=ask_ql_qh_dp_hza
That one is just a random hit with a short URL, not a recommendation.
Usually the battery maker's data sheet provides a voltage and current limit
setting that will properly charge the battery. It may not be a fast charge.
The charging voltage range and maximum current are often printed on the side
of AGMs.
For this use a power supply needs adjustable output voltage and current
limiting, backfeed protection from battery voltage if the grid goes down,
and an output that turns on automatically when the grid returns. It doesn't
really need built-in meters or convenient adjusting knobs, I can use
separate meters and a screwdriver adjustment pot. Lab supplies are rather
expensive to dedicate as a battery charger so I imitate their essentials
with cheap power bricks and separate regulators.
Here's an example of matching the supply to a custom battery pack:
http://www.mastechpowersupplies.com/fast-charging-a123-mastech-hy5020e.html
"These power supplies are designed with over-voltage and reverse-voltage
protection , so are protected from occasional user errors, and are ideally
suitable for fast charging Lithium batteries like A123 Cells, or any other
batteries."
A lab supply is a versatile piece of test equipment, not just a power
source. They can curve-trace a diode, measure the resistance of transformer
and motor windings, find short circuits, reform old electrolytic capacitors,
and many other things. Their function is an important part of industrial
semiconductor testers, used to measure an IC's DC input and output
parameters such as Vol under load and a MOSFET's Rds(on). For that the
voltage and current can be all four combinations of positive or negative.
They can measure the resistance of one inch of 10 AWG wire, it drops 5mV at
5A. When I bought 200 feet of 10AWG solar PV wire from Amazon I checked its
resistance with a lab supply to be sure I hadn't been cheated with
undersized wire. Like Radio Shack some things there don't quite meet spec.
I mentioned recharging a discharged battery as a potential problem, and this
morning I discovered that failure. The fridge is only keeping bread cool so
it wasn't a big problem. The power supply, a small one meant for 12V
cameras, had tested OK at 5A output into a voltage and current regulator in
the cool basement but at 74F in the kitchen its PTC thermistor overload
tripped some time overnight, while trying to keep up with a very weak
battery.
I use inexpensive buck or buck-boost converters to add precise voltage and
current control to inexpensive/surplus power supplies, to create battery
charger/maintainers that do what I want. When I was the battery tech at
Segway I learned how to charge almost any type of battery with a metered
adjustable laboratory power supply like this:
https://www.amazon.com/gp/product/B0852JZQZR/ref=ask_ql_qh_dp_hza
That one is just a random hit with a short URL, not a recommendation.
Usually the battery maker's data sheet provides a voltage and current limit
setting that will properly charge the battery. It may not be a fast charge.
The charging voltage range and maximum current are often printed on the side
of AGMs.
For this use a power supply needs adjustable output voltage and current
limiting, backfeed protection from battery voltage if the grid goes down,
and an output that turns on automatically when the grid returns. It doesn't
really need built-in meters or convenient adjusting knobs, I can use
separate meters and a screwdriver adjustment pot. Lab supplies are rather
expensive to dedicate as a battery charger so I imitate their essentials
with cheap power bricks and separate regulators.
Here's an example of matching the supply to a custom battery pack:
http://www.mastechpowersupplies.com/fast-charging-a123-mastech-hy5020e.html
"These power supplies are designed with over-voltage and reverse-voltage
protection , so are protected from occasional user errors, and are ideally
suitable for fast charging Lithium batteries like A123 Cells, or any other
batteries."
A lab supply is a versatile piece of test equipment, not just a power
source. They can curve-trace a diode, measure the resistance of transformer
and motor windings, find short circuits, reform old electrolytic capacitors,
and many other things. Their function is an important part of industrial
semiconductor testers, used to measure an IC's DC input and output
parameters such as Vol under load and a MOSFET's Rds(on). For that the
voltage and current can be all four combinations of positive or negative.
They can measure the resistance of one inch of 10 AWG wire, it drops 5mV at
5A. When I bought 200 feet of 10AWG solar PV wire from Amazon I checked its
resistance with a lab supply to be sure I hadn't been cheated with
undersized wire. Like Radio Shack some things there don't quite meet spec.
Jim Wilkins
May 20, 2022, 1:27:24 PM5/20/22
to
"bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
--------------------
I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
which is quite good for a true-sine inverter. What do you see?
The front panel image on Amazon shows an MPPT solar charging function,
apparently a feature of the new model, though the text doesn't mention it.
Does yours have it?
In general it appears to be an integrated package of the features I've
assembled separately, with the voltage and current meters I like to monitor
performance and diagnose problems, but apparently without battery SOC. I
assume the generator mode permits greater voltage and frequency deviations
before declaring an AC line fault and switching to battery. My APC1400 has
an adjustment for that but it doesn't extend to accepting a poorly regulated
generator. It's fine with an inverter generator. Electric starting is
typically on generators considerably larger than 800W.
Good luck with it. I hope the warnings of nationwide rolling blackouts don't
come true, but if they do your food won't spoil.
https://www.city-journal.org/energy-policy-failures-may-lead-to-summer-blackouts
I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
which is quite good for a true-sine inverter. What do you see?
The front panel image on Amazon shows an MPPT solar charging function,
apparently a feature of the new model, though the text doesn't mention it.
Does yours have it?
In general it appears to be an integrated package of the features I've
assembled separately, with the voltage and current meters I like to monitor
performance and diagnose problems, but apparently without battery SOC. I
assume the generator mode permits greater voltage and frequency deviations
before declaring an AC line fault and switching to battery. My APC1400 has
an adjustment for that but it doesn't extend to accepting a poorly regulated
generator. It's fine with an inverter generator. Electric starting is
typically on generators considerably larger than 800W.
Good luck with it. I hope the warnings of nationwide rolling blackouts don't
come true, but if they do your food won't spoil.
https://www.city-journal.org/energy-policy-failures-may-lead-to-summer-blackouts
bob prohaska
May 23, 2022, 11:01:00 PM5/23/22
to
Jim Wilkins <murat...@gmail.com> wrote:
> "bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
>
> I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
>
> --------------------
>
> I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
> which is quite good for a true-sine inverter. What do you see?
>
I think that's in "eco" mode, which switches too slow for my UPS-style
> "bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
>
> I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
>
> --------------------
>
> I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
> which is quite good for a true-sine inverter. What do you see?
>
application. In "ac priority" mode, which does switch fast enought, ~15W.
Some of that might be battery charge, but likely not much.
> The front panel image on Amazon shows an MPPT solar charging function,
> apparently a feature of the new model, though the text doesn't mention it.
> Does yours have it?
>
the unit I got.
> In general it appears to be an integrated package of the features I've
> assembled separately, with the voltage and current meters I like to monitor
> performance and diagnose problems, but apparently without battery SOC. I
> assume the generator mode permits greater voltage and frequency deviations
> before declaring an AC line fault and switching to battery. My APC1400 has
> an adjustment for that but it doesn't extend to accepting a poorly regulated
> generator. It's fine with an inverter generator. Electric starting is
> typically on generators considerably larger than 800W.
>
can't be remote started and I didn't think about power quality issues.
It's worth further investigation. Mine's a flat twin synchronous machine,
so I'd expect the power quality to be decent, but I don't know for sure.
> Good luck with it. I hope the warnings of nationwide rolling blackouts don't
> come true, but if they do your food won't spoil.
> https://www.city-journal.org/energy-policy-failures-may-lead-to-summer-blackouts
>
bob prohaska
Jim Wilkins
May 24, 2022, 8:45:34 AM5/24/22
to
"bob prohaska" wrote in message news:t6hhpa$beb$1...@dont-email.me...
Jim Wilkins <murat...@gmail.com> wrote:
> I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
> which is quite good for a true-sine inverter. What do you see?
>
I think that's in "eco" mode, which switches too slow for my UPS-style
application. In "ac priority" mode, which does switch fast enought, ~15W.
Some of that might be battery charge, but likely not much.
[[ I've seen newish lead batteries draw less than 1% of C at 13.6V float
voltage when fully charged, perhaps .2~.3A for a 100Ah Marine and .1A for a
12V 18A AGM, but each is different and it increases as they age. Lithiums
should draw much less, though they are still too expensive for me to
experiment with.
My DC-powered Alpicool freezers are good battery test stations to measure Ah
capacity with the SIN9020s by shutting off the grid and solar inputs. It's a
convenient way to discharge test older and second-hand batteries without
wasting their energy, and the solar input recharges them for free. ]]
Didn't pay much attention to the ramification of generator mode. My genny
can't be remote started and I didn't think about power quality issues.
It's worth further investigation. Mine's a flat twin synchronous machine,
so I'd expect the power quality to be decent, but I don't know for sure.
-----------
Since you don't expect to need a generator often or for long at a time you
could use it with a separate charger to keep the battery up. I haven't had
problems charging a battery from multiple sources simultaneously, as long as
the sources can withstand a higher voltage on their outputs, but anything
connected to a battery should be able to. The battery should give you enough
time to manually switch over.
I'd add at least a voltmeter to show if the battery is getting low or the
charging voltage has risen into the gassing range. These give you a +/- 10mV
battery voltage readout without having to install a panel meter. A meter
that reads to 0.1V and could be off 1 or 2 counts isn't really good enough
for float voltage.
https://www.amazon.com/INNOVA-3721-Battery-Charging-Monitor/dp/B000EVWDU0
I like these 19.99V LCD panel meters to show battery voltage on fused leads
connected directly to the battery terminals.
https://www.aliexpress.com/item/1005002084389556.html
My system is sized to run the refrigerators overnight on battery in case of
night ice or lightning storms. Hopefully I'll be able to go out and set up
the genny in the morning.
I made a priming and tank emptying attachment for my Honda EU1000i from a
rubber stopper that fits the fuel filler opening and a rubber siphon bulb to
lightly pressurize the tank. After pumping the tank as low as a kerosine
drum pump will go the bulb and stopper force the remaining gas out the carb
drain tube, so I can safely store the Honda indoors instead of risking a
fall on ice to fetch it from the shed. The bulb and stopper prime the carb
after refilling the tank. The stopper pops out to limit pressure.
https://www.manualslib.com/manual/719234/Honda-Eu1000i.html?page=60
Jim Wilkins <murat...@gmail.com> wrote:
> I read your Amazon review of it. Luckysolar claims its idle power is 6-8W
> which is quite good for a true-sine inverter. What do you see?
>
I think that's in "eco" mode, which switches too slow for my UPS-style
application. In "ac priority" mode, which does switch fast enought, ~15W.
Some of that might be battery charge, but likely not much.
voltage when fully charged, perhaps .2~.3A for a 100Ah Marine and .1A for a
12V 18A AGM, but each is different and it increases as they age. Lithiums
should draw much less, though they are still too expensive for me to
experiment with.
My DC-powered Alpicool freezers are good battery test stations to measure Ah
capacity with the SIN9020s by shutting off the grid and solar inputs. It's a
convenient way to discharge test older and second-hand batteries without
wasting their energy, and the solar input recharges them for free. ]]
Didn't pay much attention to the ramification of generator mode. My genny
can't be remote started and I didn't think about power quality issues.
It's worth further investigation. Mine's a flat twin synchronous machine,
so I'd expect the power quality to be decent, but I don't know for sure.
Since you don't expect to need a generator often or for long at a time you
could use it with a separate charger to keep the battery up. I haven't had
problems charging a battery from multiple sources simultaneously, as long as
the sources can withstand a higher voltage on their outputs, but anything
connected to a battery should be able to. The battery should give you enough
time to manually switch over.
I'd add at least a voltmeter to show if the battery is getting low or the
charging voltage has risen into the gassing range. These give you a +/- 10mV
battery voltage readout without having to install a panel meter. A meter
that reads to 0.1V and could be off 1 or 2 counts isn't really good enough
for float voltage.
https://www.amazon.com/INNOVA-3721-Battery-Charging-Monitor/dp/B000EVWDU0
I like these 19.99V LCD panel meters to show battery voltage on fused leads
connected directly to the battery terminals.
https://www.aliexpress.com/item/1005002084389556.html
My system is sized to run the refrigerators overnight on battery in case of
night ice or lightning storms. Hopefully I'll be able to go out and set up
the genny in the morning.
I made a priming and tank emptying attachment for my Honda EU1000i from a
rubber stopper that fits the fuel filler opening and a rubber siphon bulb to
lightly pressurize the tank. After pumping the tank as low as a kerosine
drum pump will go the bulb and stopper force the remaining gas out the carb
drain tube, so I can safely store the Honda indoors instead of risking a
fall on ice to fetch it from the shed. The bulb and stopper prime the carb
after refilling the tank. The stopper pops out to limit pressure.
https://www.manualslib.com/manual/719234/Honda-Eu1000i.html?page=60
Jim Wilkins
May 24, 2022, 9:43:36 PM5/24/22
to
"Jim Wilkins" wrote in message news:t6ik1c$9q3$1...@dont-email.me...
>....Lithiums should draw much less, though they are still too expensive for
>me to experiment with.
Maybe not any more. Amazon has 12V 18A LiFePO4s with many good reviews for
around $75. I just ordered a Weize for $69.99 with coupon, not that much
more than the Power Sonic AGM I had been considering.
https://www.power-sonic.com/blog/how-to-charge-lithium-iron-phosphate-lifepo4-batteries/
>....Lithiums should draw much less, though they are still too expensive for
>me to experiment with.
Maybe not any more. Amazon has 12V 18A LiFePO4s with many good reviews for
around $75. I just ordered a Weize for $69.99 with coupon, not that much
more than the Power Sonic AGM I had been considering.
https://www.power-sonic.com/blog/how-to-charge-lithium-iron-phosphate-lifepo4-batteries/
bob prohaska
May 27, 2022, 7:06:49 PM5/27/22
to
bob prohaska <b...@www.zefox.net> wrote:
> Jim Wilkins <murat...@gmail.com> wrote:
>> "bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
>>
>> I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
>>
>> --------------------
>>
>> I read your Amazon review of it.
Unfortunately it'll be necessary to amend the review at least slightly.
> Jim Wilkins <murat...@gmail.com> wrote:
>> "bob prohaska" wrote in message news:t5scbb$n1d$1...@dont-email.me...
>>
>> I've decided to purchase the packaged Ampinvt 800 watt inverter-charger.
>>
>> --------------------
>>
>> I read your Amazon review of it.
The inverter and transfer switch seem to work admirably, but the charger
is more crude than I expected. Far as can be told it's only a single
stage power supply, with only current and voltage limits. The "custom"
battery type just lets me set a single, fixed output voltage. I've
queried the folks at ampinvert.com and have received prompt but
uncomprehending replies.
The Delco Voyager battery I'm using wants a 10-15 amp taper charge
ending at 15 volts followed by "no" float charge, presumably a resting
voltage of 12.8 volts. Delco's instructions explicitly warn against
leaving the battery on trickle charge.
The obvious workaround is to manually set a 15 volt charge, wait for
the charge to complete and then lower the voltage to a safe float value.
In a situation where I'm using a generator, then I'll put the charge
voltage back to 15 volts, knowing it'll never be on the genny long
enough to reach saturation.
As an instructional device I think the inverter was a reasonable buy.
It incorporates a satisfactory inverter, transfer switch and crude
charger in one package with the essential control logic. I'm fairly
sure it'd cost more than I paid to assemble those functions from parts.
It wasn't a wonderful bargain. But it promises to be reasonably useful.
Jim Wilkins
May 27, 2022, 11:25:03 PM5/27/22
to
"bob prohaska" wrote in message news:t6rli7$oah$1...@dont-email.me...
----------------------
A constant voltage and current limit setting is fine as long as you keep an
eye on the voltage and time. The higher absorption voltage just hastens the
charge, it isn't essential unless you are cycling the battery daily. Trickle
charging isn't inherently bad, the problem is cheap units that slowly
electrolyze away all the water. The answer is measuring and intelligently
controlling the voltage and current, but I've never seen adequate meters on
a commercial battery charger and most people wouldn't know what the numbers
mean anyway.
15.0V seems high for a flooded battery, up in the equalizing and desulfating
range, but it's what they recommend.
https://images.wrenchead.com/digi/DCL/MarineRVDeepCycle_ChargingInstructions.pdf
This is the charging sheet for my East Penn batteries:
https://www.eastpennmanufacturing.com/wp-content/uploads/Renewable-Energy-Charging-Parameters-1913.pdf
"Float Charge – The optimum voltage level at which a battery string gives
maximum life and full capacity."
I fully charged one, then removed the fillers and stretched clear tape over
the openings to contain acid spatter, and observed the gassing rate while
gradually stepping up the voltage.
At 13.6V the liquid surface was calm. At 13.75V it was mostly calm, with a
few bubbles. At 13.78V all cells were noticeably but very lightly bubbling.
At 13.8V all cells were bubbling continuously. YMMV. FWIW, I've read that
the gassing rate isn't a concern below 14.2V. At that voltage the surface
was continuously disturbed. Acid spatter on the tape is another sigh of
gassing that's easier to watch.
Based on that experiment I set my baseline charging voltages to 13.6V and
occasionally raise it to around 13.8~14.2V with solar or other temporary
source. 13.6V is enough to quickly bring the battery up to ~70-80% full,
they say, and then a 100Ah marine battery takes up to a day to continue to
full charge where the current falls below 1A, 1% of C. East Penn defines the
endpoint as a steady current, decreasing less than 0.1A per hour.
My APC UPS charges its internal battery at 6A until it reaches the 13.6V
float, then the current slowly tapers down. As received they were set
somewhat higher, which Internet advice says was a known problem that
shortens battery life, although the recharge is quicker.
http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage
"Using a "test set" of batteries, the float voltage was measured at 27.94 V.
Reference data from the Power-Sonic web site calls for a 2.25 – 2.30 V float
voltage per cell. This translates to 27.00 – 27.60 for a 24 volt pack."
I made the custom serial cable and set the APC to my preferences, such as no
alarm.
A constant voltage and current limit setting is fine as long as you keep an
eye on the voltage and time. The higher absorption voltage just hastens the
charge, it isn't essential unless you are cycling the battery daily. Trickle
charging isn't inherently bad, the problem is cheap units that slowly
electrolyze away all the water. The answer is measuring and intelligently
controlling the voltage and current, but I've never seen adequate meters on
a commercial battery charger and most people wouldn't know what the numbers
mean anyway.
15.0V seems high for a flooded battery, up in the equalizing and desulfating
range, but it's what they recommend.
https://images.wrenchead.com/digi/DCL/MarineRVDeepCycle_ChargingInstructions.pdf
This is the charging sheet for my East Penn batteries:
https://www.eastpennmanufacturing.com/wp-content/uploads/Renewable-Energy-Charging-Parameters-1913.pdf
"Float Charge – The optimum voltage level at which a battery string gives
maximum life and full capacity."
I fully charged one, then removed the fillers and stretched clear tape over
the openings to contain acid spatter, and observed the gassing rate while
gradually stepping up the voltage.
At 13.6V the liquid surface was calm. At 13.75V it was mostly calm, with a
few bubbles. At 13.78V all cells were noticeably but very lightly bubbling.
At 13.8V all cells were bubbling continuously. YMMV. FWIW, I've read that
the gassing rate isn't a concern below 14.2V. At that voltage the surface
was continuously disturbed. Acid spatter on the tape is another sigh of
gassing that's easier to watch.
Based on that experiment I set my baseline charging voltages to 13.6V and
occasionally raise it to around 13.8~14.2V with solar or other temporary
source. 13.6V is enough to quickly bring the battery up to ~70-80% full,
they say, and then a 100Ah marine battery takes up to a day to continue to
full charge where the current falls below 1A, 1% of C. East Penn defines the
endpoint as a steady current, decreasing less than 0.1A per hour.
My APC UPS charges its internal battery at 6A until it reaches the 13.6V
float, then the current slowly tapers down. As received they were set
somewhat higher, which Internet advice says was a known problem that
shortens battery life, although the recharge is quicker.
http://www.jjoseph.org/notes/apc_smartups_battery_float_voltage
"Using a "test set" of batteries, the float voltage was measured at 27.94 V.
Reference data from the Power-Sonic web site calls for a 2.25 – 2.30 V float
voltage per cell. This translates to 27.00 – 27.60 for a 24 volt pack."
I made the custom serial cable and set the APC to my preferences, such as no
alarm.
Jim Wilkins
May 30, 2022, 2:05:17 PM5/30/22
to
"Jim Wilkins" wrote in message news:t6k1k5$ht2$1...@dont-email.me...
I just ordered a Weize for $69.99 ...
-------------
It's 12V, 18Ah, a drop-in replacement for an AGM. Their 100Ah battery costs
$380 with coupon. I paid $105 a few weeks ago for a 100Ah flooded marine
lead-acid.
So far it seems promising and is easier to manage than flooded or AGMs since
it doesn't need to be recharged after use. I put it in an old Harbor Freight
5-in-1 power pack, changed the analog voltmeter to a digital one and
substituted thick leads with Anderson PP45s for one of the lighter outlets.
The first discharge cycle running my freezer gave 15.9Ah but it may not have
been fully charged to start, as it arrived late in the afternoon and solar
power was fading away.
This morning my recharge setting was 14.0V at 5A. The current dropped to 0
when it reached full charge. They allow 14.4V but warn against leaving it
fully charged for a long time. I need to further research the voltage
setting for leaving the battery continuously connected to power as a DIY
UPS. I think it's between 13.4 and 13.65V. Lower may extend the lifespan,
higher will recharge faster.
https://battlebornbatteries.com/charging-battleborn-lifepo4-batteries/
"Our batteries do not require a float charge. But they can be in float mode
as long as the float is around 13.6, it is OK for the batteries to float
around that voltage."
I just ordered a Weize for $69.99 ...
-------------
It's 12V, 18Ah, a drop-in replacement for an AGM. Their 100Ah battery costs
$380 with coupon. I paid $105 a few weeks ago for a 100Ah flooded marine
lead-acid.
So far it seems promising and is easier to manage than flooded or AGMs since
it doesn't need to be recharged after use. I put it in an old Harbor Freight
5-in-1 power pack, changed the analog voltmeter to a digital one and
substituted thick leads with Anderson PP45s for one of the lighter outlets.
The first discharge cycle running my freezer gave 15.9Ah but it may not have
been fully charged to start, as it arrived late in the afternoon and solar
power was fading away.
This morning my recharge setting was 14.0V at 5A. The current dropped to 0
when it reached full charge. They allow 14.4V but warn against leaving it
fully charged for a long time. I need to further research the voltage
setting for leaving the battery continuously connected to power as a DIY
UPS. I think it's between 13.4 and 13.65V. Lower may extend the lifespan,
higher will recharge faster.
https://battlebornbatteries.com/charging-battleborn-lifepo4-batteries/
"Our batteries do not require a float charge. But they can be in float mode
as long as the float is around 13.6, it is OK for the batteries to float
around that voltage."
Jim Wilkins
Jun 1, 2022, 7:17:05 AM6/1/22
to
Weize sent me a technical data sheet that says cell balancing requires at
least 14.2V, and suggests storing at 100% SOC. When charging the voltage
settles to a good indication of SOC after a 15 minute rest.
least 14.2V, and suggests storing at 100% SOC. When charging the voltage
settles to a good indication of SOC after a 15 minute rest.
0 new messages