Hybrid Car – More Fun with Less Gas

refrigerator on inverter - Page 2

register ::  Login Password  :: Lost Password?
Posted by Jim Wilkins on September 13, 2018, 3:06 pm
 
<ads> wrote in message  

I've put together a roughly similar backup system, two ageing and two  
new 12V 105Ah batteries and 500W of solar, plus a quiet 2500/2200W  
inverter generator for meals and recharging and a loud 3750W one (also  
bought cheap and barely running) for larger hard-starting motor loads  
like the a/c and washing machine. I heat and cook with a wood stove  
that doesn't need electricity and dry laundry outdoors.

I made a 24V 25A fast charger from a buzz box arc welder transformer  
and a Variac to rapidly recharge the batteries to just below the  
gassing level during breakfast, then solar can take over for the rest  
of the day. I can fast charge them again if needed during lunch and  
supper.

This regulates the charging voltage and current when I'm away and  
can't watch the battery monitor sitting atop a computer speaker. I  
bought it before the 20A version came out.
https://www.amazon.com/uniquegoods-Step-down-Programmable-Adjustable-Regulator/dp/B01N3YSE6S  

Once the batteries reach the 13.6V - 13.8V float level the current  
they accept falls off to where a $0, 20A PWM solar controller is  
adequate, since the DPS5015 doesn't play well with the varying output  
of solar panels. By my measurements and calculations I haven't reached  
the array size where MPPT returns more than the same $ spent on adding  
panels. My fast charger covers the condition of low batteries on a  
sunny cold day where MPPT is most productive.

The higher voltage absorption and equalization stages are missing  
because I do them outdoors with a homebrew linear adjustable  
regulator. AFAICT an occasional week or two below full charge but in  
active use shouldn't hurt the batteries.

I bought a $0 HF steel service cart to run the inverter generator on  
in bad weather, protected by corrugated roofing on the upper tray. The  
rest of the time it's an outdoor mobile workbench and tool tray.

This closeout 4.3 cubic foot Magic Chef refrigerator drew at the rate  
of 260 KWH / year at room temperatures in the upper 70's to low 80's,  
and is now at 190 KWH per year at 65-70F.

The 12/24VDC portable Alpicool C20 is somewhat less efficient but  
makes up for that by not needing a true sine inverter running  
constantly and consuming more battery amp-hours than its load. It can  
store almost a second week's worth of frozen food when a major storm  
threatens, then freshly cooked food by setting it to refrigerate. Here  
in New England we risk both tropical hurricanes and ice storms.

Without added insulation the Alpi ran constantly and couldn't reach  
its -18C freezer setpoint in the car on a low 80's day. Covering the  
cold compartment with a winter parka let it cycle.

I'm finishing up an insulated box for it. The first attempt used a  
moving blanket inside 1" styrofoam but condensation soaked the blanket  
and it's unmanageable inside plastic bags. Rev 2 has a waterproof yoga  
mat liner protecting the rigid foam. The flexible lid cover is still  
the blanket.

The last time I had to repair roof damage during a long ice storm  
outage I had only a 500W generator. It proved adequate to power a  
saber saw to fit a new rafter and recharge a cordless panel saw to cut  
the plywood sheathing along rafter centerlines and make patch panels.

-jsw  



Posted by ads on September 14, 2018, 1:37 am
 
On Thu, 13 Sep 2018 11:06:47 -0400, "Jim Wilkins"


I'm using EPSolar/EPEver Tracer 3210A and 4210A MPPT controllers. With
the bigger solar panels (250W, 60 cell, 30V Vmp or higher), the Tracer
controllers go into MPPT mode at about 5 watts of solar input. Even at
low panel power (5W from a 250W panel early/late in the day) the
voltage difference becomes a current multiplier: 30V panel / 13V
battery = 2.3 so 2.3 * 0.1A = 0.23A into the batteries.

The Tracer controllers don't track sun changes as fast as the MidNite
controllers, but there's a 2x or more difference in price (30Amp
Tracer is around $00-$20 -and they probably went up with the China
tariffs).

I did the math on a solar tracker vs another panel vs an MPPT
controller and the MPPT won out after the fourth 100 watt panel.  I've
since found a couple of semi-local (hour's drive or so) sellers who
have good deals on either used or "new, never installed" panels.  The
used panels are typically a group removed from service so you can get
a number of identical panels.  The new, never installed panels are
basically "leftovers".  The most recent "300 watt" panels I bought
were a 300 watt ReneSola and a 310 watt Jinko ($25 each,
craigslist.org - go to For Sale and search for solar panel ) but the
specs are close enough for them to work fine in parallel or series,
depending on need.

The inverters, controllers, a few 100 watt solar panels and the wire
were puchased new, but some of this project was on a "How cheaply can
it be done?" basis.  The six 90AH AGM batteries are from a big UPS in
a corporate data center - they were changed out by time, not wear, so
tested as new when I got them for $5 each over a year ago (I should
have bought 3 times as many as I did but loose $$ were hard to come
by).  Five of the six batteries still read 12.72v or better after 24
hours at rest (factory spec for fully charged) and number six misses
that by 0.02 volts at 12.70v.  

I have a 0-16V, 0-20A power supply that I can use for charging when a
gas generator is running.  The supply can be set for current-limited
constant voltage output and bring the batteries to whatever charge
level they need, although this page:
https://marinehowto.com/how-fast-can-an-agm-battery-be-charged/  
tells me I probably need a much bigger charger to get the batteries
fully charged quickly - seems that those last few per cent of charge
can make a big difference in battery life and the batteries should be
charged as long as needed to ensure the proper levels are reached very
soon after discharge.

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus  


Posted by Jim Wilkins on September 14, 2018, 1:29 pm
 <ads> wrote in message  

Thanks, that's a good informative article that falls in the range of  
what I've read from battery manufacturers and my own long-term  
experiments and experience as an industrial battery tech. I use 1% of  
Ah capacity as the charging cutoff at the high end of the float  
voltage range simply because it's easy to figure.

I chose flooded batteries as most cost-effective (and tolerant of  
experiments) for my emergency backup system although I would have  
bought AGMs or Lithiums for daily cycling. The few recommendations  
I've found on how long they can be left at partial charge suggest that  
a month may be too long, two weeks perhaps OK as long as they are  
equalized afterwards. It helps that room and battery temperature here  
rarely exceeds 60F in winter.

I've found that I can get typically 10+ years of useful life from a  
non-maintenance-proof flooded battery that's topped off every month or  
two, while similarly cared for AGMs fail sooner and are much less  
receptive to desulfation attempts.
http://www.thelongestlecture.com/agm-battery-life-expectancy/  
https://www.thehulltruth.com/boating-forum/159661-life-expectancy-absorbed-glass-mat-batteries.html  

I used to collect the changed-out 12V 18Ah AGM batteries from the  
company's exit lights and test them for lifespan by measuring Ah  
capacity and discharge current at 10.5V. The HF carbon pile tester is  
great for measuring maximum current but doesn't have the power  
dissipation capacity or stability for longer, lower current discharge  
for which I use large rotary rheostats, or an inverter and 120VAC load  
for constant power discharges with automatic low voltage cutoff.

The battery in my truck lasted 13 years before its discharge current  
fell uncomfortably close to the starter's draw in cold weather. In my  
home solar system the SLI31MDC battery from 2008 is down to ~50%  
capacity, the 2011 battery considerably better. I bought two new ones  
for the 24V inverter, trading in the dead AGMs, and use the older ones  
for 12V loads and experiments.

DVMs with optically coupled data outputs are excellent for assembling  
a multichannel battery testing datalogger that doesn't short-circuit  
through the data lines. My pesonal setup is an older laptop that  
accepts PC Card and ExpressCard COM port expanders and these meters
https://www.amazon.com/Tekpower-TP4000ZC-Interfaced-Multimeter-Computer/dp/B000OPDFLM  

I've been using it with thermocouples to watch the Magic Chef cycle  
and set the turn-on temperature below 40F. The suggested setting lets  
it rise above 45F. I'd rather use a little more electricity than risk  
spoilage. The temperature graph shows very clearly how much opening  
the door for more than a second or two affects internal temperature.  
Recording turn-on and turn-off with a clamp-on amp probe on the power  
cord wasn't really necessary, the freezer temp reacts in less than my  
60 second sampling interval.

The laptop is easier than a desktop to set up in odd places like atop  
the refrigerator or beside the water heater. The separate data log  
files can be time-stamp aligned in a spreadsheet.

-jsw



Posted by ads on September 15, 2018, 8:01 pm
 On Fri, 14 Sep 2018 09:29:16 -0400, "Jim Wilkins"



Nice meter for the price and I like that it uses AA cells - need to
add the meter to my wishlist.  There's another meter on amazon that
has "optical RS232" so that might also be a possibility - but it uses
a 9 volt battery and the AAAA cells in those don't last long.


---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus  


Posted by Jim Wilkins on September 16, 2018, 1:06 am
 <ads> wrote in message  

If I started over without my old Radio Shack PC-Interface meters and  
with newer computers I'd buy USB meters. I stayed with RS-232 because  
I can access and decode my COM port meters with QBasic though I never  
wrote a program that took action based on the readings after buying a  
low voltage disconnect module for battery discharge testing.

The raw type K thermocouple temperature reading on the TP4000 may be  
off a little. The datalogging program applies a linearization  
correction and can convert to Fahrenheit.

There are at least two versions of the datalogging program. The last  
meter I bought came with Hong Kong Digitek software which allows  
installing multiple instances of the program in separate folders. Each  
instance controls one meter.  I cut and paste their outputs together  
in a spreadsheet.
There's also this, which I haven't done much with yet:
http://www.ultradmm.com/  

-jw  



This Thread
Bookmark this thread:
 
 
 
 
 
 
  •  
  • Subject
  • Author
  • Date
please rate this thread