Month: September 2018

Ever since I built my first “Pi-Hole” with a Raspberry Pi I have had problems with finding good USB cables that will provide enough power. I have also increased my inventory to almost a hundred devices that are powered from a USB connector. The quality of the cable and the length of the cable are huge factors in having reliable device. None of the devices I am using are USB-C Power Delivery compatible which means they are 5V and some of them require 3 amps of current. After trying a number of things my solution is to provide a power supply at the device that can convert 12-24V into 5V with a USB connection. (POE is an option, but at present most of my devices are NOT POE. They are wireless only and just need power source.

There are several pieces to this puzzle that took a lot of time to resolve, like years!

By far the first item I had to find was a low voltage connector.

One day last year I saw this LED connector show up on Amazon. (I’ve bought several hundred of them since.) I have NOT found a competing product! (Yet)

They are NOT perfect!

1. I had to figure out how to load the wire into them without damaging the connector.
2. It’s actually tricky to get the wire onto all four of the IDC terminals inside it.
3. They have a very limited wire size, 22-20 works best.
4. They do hold together and I’ve not had an issue with the connection between them.
5. I LOVE how small they are and how they have a good “Snap” connection.

The next item was twin lead wire that the connectors could use.

This picture is severely enlarged.

I happened on wire that was used for garage door openers. The wire is normally run to the two sensors at the bottom of the garage door opening.

It’s available in over one-hundred-foot lengths, 20AWG and inexpensive!

Most power supplies also use very similar wire that is clearly marked. (Usually black wire) They are NOT consistent on which wire is positive/negative. For my purposes I chose the dark bar wire as positive since 80% of the power supplies are similarly marked.

The third component was the USB power supply with connector.

I’ve seen these inexpensive China power supplies for years, but getting the DC to them was the more difficult problem. Now that I have a connector and wire, they are looking much more inviting.

They have varying voltage requirements and most of them provide 3 amps with a very efficient switching power supply which is sufficient for all of the devices I currently have. (They typically do not get warm.) This particular one can take 55Volts which is insane for a 15W load. (A modern golf cart is 56 volts.) The voltage drop at 2A on 100 foot of 20AWG wire is roughly 4 Volts which would still provide 16W (3A@5V) of power. Raising the voltage to 14V drops the current to 1.5A and 18V drops it even further to a little over 1A. Increasing the wire AWG proved to be a worthless venture. (Power supplies are significantly cheaper to change out.)

This wireless router can be mounted just about anywhere, and the power requirements are so low that it could be run on one of the 24AWG pairs.

I have toyed with the idea of having a genuine $1,300 Dancer brand solar powered refrigerator to pile the essentials into during an extended power outage.  We currently don’t have any medicine or anything else out of the ordinary in our fridge that would warrant such an expensive purchase.  The most important decision making task was to measure how much power our current fridge draws.  In my case it was much smaller than I expected at 175 Watts.  (It was also strange that the fridge had three 40W incandescent lights which drew almost as much power as the compressor.) 

The compressor startup current (AKA Stalled Rotor Current to size the UPS), the defrost cycle heater and the normal duty cycle of the compressor to size the battery bank and define solar requirements.  The single most expensive purchase was the Triplite SU1000XLA UPS at about $600 New.  It had sufficient surge capacity to start the compressor and an external battery connection to keep it running all day.  

Next on the list was a storage battery capable to run the fridge for 24 hrs.  I used 175W x 24 (Hours) x .5 (Duty Cycle) =2100WH. 

Actual usage is 2370WH / 24V * 1.2 (Efficiency) * 2 (50% discharge) = 237AH of battery.  (12V 100AH AGM batteries are roughly $250 each)  To keep it running 24Hours/Day we need to recharge the batteries AND provide running power during the day.

Each 100W panel produces 350WH/Day on average so I need at least Five 100W panels.  Three of them will be charging the battery and two to run the refrigerator during the day (Need 3x panels during the winter.).  If we loose power because of a storm, I’ll have to run a tiny 600VA 20lb generator twice a day for 5 hours to recharge the batteries.  (~1Gl gas/Day) until the sun comes back out.

Two panels will likely cover most of the long term outages we have around here with 1500WH of Solar power for two days and more than 5880WH of stored power.  The last item on the list is the Solar Charge Controller which provides accurate battery voltages to protect them.

Because of the current situation with the COVID-19 we bought a DC powered freezer which can also be used as a fridge.

The good thing is that it is dual powered (AC/Battery) and draws about 85W half what our fridge in the kitchen draws. 85W x 6.9 Hours x 30 Days is $2.11/Month in Fridge mode and 85W x 10.8 Hours x 30 Days = $3.31 in Freezer mode. (Calculated) On battery it will consume 76AH/Day in Freezer mode. My solar panels generate about 1.5KW/Day and the Fridge consumes 918WH/Day in freezer mode. The solar panels generate power for 5 hours/Day so 19 Hours/Day will be on battery. 80% efficiency for round trip to battery and back for 19 hours will use 8.5H x 85W x 120% = 872W / Day on Battery during a sunny day and 1100WH on a cloudy day. Charge Amps for 872WH / 5 hours = 15A minimum into the battery during daylight. Storage needs to be 2 cloudy days x 50% battery discharge 4400 WH = 366AH (10 x 35AH battery) Required storage for AGM charge at 15 Amps is 150AH minimum.

After having the refrigerator for two years I finally put a temp sensor inside it. I was shocked how a warm bottle of water affected the inside temperatures. Now I am wondering if it matters where inside the fridge the warm water is placed.

I recently needed a UPS that could support about 30W for several hours and after looking at the specifications of consumer grade UPS units for a couple days I became very discouraged.

The scope of my research was mostly UPS units that were under 450VA.

Conversion Efficiency: I found out that most units (Even at a reasonable 60W load.) were less than 50% efficient.  (Most were about 35%)  The most logical component for the poor efficiency was the transformer and I was shocked how inefficient they were.  (Way too small for the VA rating of most of them.)

I found a video of someone attaching a large battery to one of these small units to extend the run time and the transformer melted through the case into the carpet floor and almost caught fire.   I started recording the battery size and the run time for a 60W load from the runtime charts that were offered.  The data was horrible, I just couldn’t believe that it passed UL.  The good thing is that the larger VA rating of the UPS, the more efficient the transformer became.  It did seem unreasonable to purchase a 700VA UPS to drive a 40W load!

Battery: The battery in these consumer grade units were unbelievably small, like 4AH.  I get it, batteries are heavy and the price point is low for home units.  A 350VA UPS would need roughly 25-35 Amps from this tiny 4AH battery contained inside.  Looking at a typical battery specification, the efficiency of the battery is terrible at this high current draw.

My rules of thumb after working with UPS units for 10 years is that any UPS that supports more than 100W should have a 24V battery. (It’s actually more efficient to light a 60W light bulb with a 12V battery for some reason.)  I wrote all this up including charts and posted negative review on Amazon.  Sure enough, APC came out with a much more efficient unit (~70%) about a year later.  (No 60HZ Transformer)  The complaints posted on Amazon suggest that the high frequency design is not as reliable.  If you are wondering I have five of them and they are still in service a year later with no failures. 

The UPS I ended up using was a $40 BN450M that has no transformer and generates very little heat. (So it can run a long time, but it doesn’t have any active cooling so I had to be careful.) 

The first designs removed the board and put them in a much larger case. 

The final design removed the battery and set the UPS onto a plastic tub with several much larger batteries underneath. I did end up putting a heat sink on the battery charger component since it takes several days to recharge the battery.

They keep the cameras up for about 2 hours with much larger external batteries mostly because APC included a timing “Feature” that shuts down the UPS at 160 minutes.  I also found out that the batteries are floating at approximately 55VAC. (They need to be protected from accidental contact while the UPS is connected to commercial power.)  

Boat Anchors: UPS units in the 1000-1500VA range are reasonably efficient at 100W, but they can weigh 60-80Lbs with the batteries.  The 18AH batteries in most of them should be limited to 36Amps which yields ~800W with two in series (24V) and they produce less than half their rated capacity at that current.  I wanted roughly 100 Watts for 2-1/2 hours. (250WH delivered to the load.)  With a conversion efficiency of 70% and counting the conversion losses, I need roughly 370WH of battery power.  Two 18AH batteries have roughly 432WH of energy and 370WH will leave some charge in the battery.  Keep in mind though that two 18AH batteries exceed 30Lbs and the UPS is probably 20Lbs without the batteries.

My first adventure with flying with batteries was a trip to Germany.  I constructed a home made battery for my Netbook computer to extend the run time to many hours.  It consisted of 10 “D” size Nickel Cadmium batteries that were soldered in series with small wire.  (That way the wire would keep the current low and dissipate the energy slowly.)  The batteries were in a common Radio Shack plastic box with zip cord powering the computer.  I was able to use the computer for most of the flight to Germany.  What I hadn’t prepared for was the German security folks demanding I disassemble the battery before they would let me board my flight.  From then on I didn’t try to carry anything homemade on a commercial flight.  It it necessary that I carry 200WH of battery when traveling to third world countries where the power was unreliable.  The obvious choice was a couple  8AH 12V Sealed Lead Acid batteries.  (They were the safest battery I could carry that had sufficient power for my needs.)  According to TSA all I had to do was keep the terminals from shorting out and I could fly with them.  The safest place on a commercial aircraft is in the cabin, packing batteries in checked luggage is not preferred by the airline.  As time went by it became increasingly troublesome to have these two batteries in my bags and I would get delayed trying to explain them to TSA check point.  My final flight with them was when a female TSA supervisor insisted that I check the batteries and not carry them aboard the aircraft.  (I was unable to sway her decision so I checked them on that trip.)  I wasn’t comfortable carrying two  100WH lithium battery packs in my luggage so I bought 24 individual lithium cells that were about 10WH each.  I bought six USB battery boxes that could safely carry four individual cells each.  Believe it or not, I never had anyone question me about the USB battery packs or the 24 lithium batteries I was carrying, even when returning from a foreign country.  I was confident that if a single cell went awry that it wouldn’t heat the adjacent cell sufficiently to cause it to fail.  I was pretty sure I (Or flight attendant) could safely pull an overheating cell from the USB case in flight.  I found bicycle battery packs that I put the cells into once I got to my hotel room.  I still travel with them to this day.

When I was about 12 I bought a kit and constructed a “Carbon-Zinc” Dry Cell battery.  It was crude, but it worked for a short while.  The zinc casing eventually leaked and made a huge mess.  My next battery was a 12V 4AH rechargeable wet cell Nickel Cadmium battery that came from Allied Radio.  It was the center of a lot of projects for maybe the next 5 years.  (I can’t imagine how a 12Yr old managed to bring a very hazardous battery full of acid across the country from Illinois to Nevada in the late 60’s.)  My exploration of batteries took a break until 1984 when I bought my house. I knew from my job that leaving a battery on a charger 24/7 (Think Generator starter battery.)  appeared to shorten the life of these batteries.   I also knew that running them completely empty could ruin them as well.   My next purchase was four 100AH AGM batteries (5KWH) for a 48V 5KW UPS that I bought used.  I checked the UPS batteries after a couple of years and found that two of them blew the sides off of couple cells.  (You could see the plates right down to the bottom of the battery.)  These batteries were held off the floor in close proximity to each other such that they could have caused a cascading failure of 5KWH of heat. (Fire?)  I took that UPS out of service and purchased a smaller 24V 2.5KW Trip Lite Powerverter to replace it.  I paralleled, fused and physically separated the two 100AH battery banks and put them inside a 200LB cement block cocoon to prevent a cascading failure.  They are still in service today.

The Home Depot near us advertised a small 600W (860VA) generator for $165 and some change.

Now I won’t go into a lot of useless details except to say it’s got a very small 4 cycle 40cc engine and just as small .4 gallon gas tank. (ESI860i) It is advertised that it will provide 50% load (350W) for “Up To” 3 hours. (I did not want to mix oil with the fuel so 2cycle was not an option.) Still this generator comes in at a measly 25Lbs!

My intent for this generator was to power a full size refrigerator that draws about 175W and maybe a little more during a defrost cycle. (But certainly much less than the 600W rating of the generator.)

I ordered one from THD and picked it up a couple weeks later. To my surprise it looked like they threw it off the truck. The box was heavily damaged when I received it.

Still, the generator appeared to suffer minor damage and did run for the first couple days. (I found out later that the oil level sensor was permanently damaged by the rough handling and had to be bypassed.)

I tried to break in the generator by running it at idle for an hour or so then at high idle for another hour. I put a light 60W load on it and ran it for two more hours. I filled the gas tank and hooked it up to 300W load and ran it for maybe 3 tanks about 10 hours before changing the oil.

I was happy enough with the performance to purchase two more generators while they were still on sale @ 30% off. I ran the generator as much as I could during the next few days with the 300 W load and it seemed to provide it with moderate difficulty. The generator eventually refused to run and the low oil light could not be extinguished.

The two additional generators arrived and I performed the same break-in routine I did with the first one. One of the generators was having great difficulty providing the 350W load and would trip offline at about 450 watts. (Much less than the rated 600W) So with three generators at my feet, one refused to run, one couldn’t provide much more than half its rated power and the third seemed to be running ok. The reliability was not as good as I had hoped. (My coworker had warned me about buying these cheep generators.)

The generator that performed the best still could not support a 600W load, it would trip offline about 450W. (I think the advertised capability was stretched a little with the “Up To” caveat.) After hours of research the efficiency was not as good as I had hoped, a 2200W generator could run about as long on a gallon of gasoline as this 600W one. (Which would be MUCH more desirable to have the additional surge capability.) Further, the Yamaha had a 1.1Gallon tank so it would run 10 hours without needing to be fueled or needing oil! (The surge capability would eliminate the need for a UPS to supply the additional current needed to start the fridge.)

I am still on the fence about getting a Propane powered generator, they burn a LOT of propane!

I disassembled the generator that refused to run and the oil sensor was indeed stuck, I pushed the pin out of the connector to regain use of the first generator that produced much more power then generator #3. The only major problem with #1 is that there’s something wrong with the fuel pump. It stops running at about half a tank. LIFAN sent me two fuel pumps, neither of them fit my generator. What I did learn when I attempted to replace the fuel pump was that the vacuum side was full of fuel. I dumped the fuel out and reassembled the generator with a repaired (Frankenstein) inverter module and it ran fine for several tanks of fuel.

Generator #1 and Generator #2 I am almost happy with their performance. They consume enough oil to need replenishment after 6 hours at 300W. A full tank might last three hours before running out of gasoline. (There is no sight glass, the only way to see the fuel is to take off the gas cap.) I am rather surprised at the poor performance of generator #3, I changed out the inverter module and that didn’t help anything. (I assume at this point that the alternator itself is not as efficient and cannot supply the power as the other two that seem to be able to without difficulty.)

I bought a forth generator when they went on sale again. I ran it though the same break-in process that I did with the first three. When I finally put a load on the generator it had no guts at all and tripped offline. Further research over the next couple days revealed that the exhaust had a constriction of some sort and the engine backpressure caused the engine to loose most of it’s torque.

LIFAN didn’t want me to return it to THD even though I’ve only had it for 5 days. They are sending a UPS pickup to repair the generator under warranty. So if you’ve managed to follow this, I have two running and one limping generator in the garage with a fourth being repaired under warranty. After two weeks they said there was nothing wrong with the generator and wanted a credit card to ship the unit back. $30 and three days later I found out that the generator still would cave in at 350W. (Not sure how they managed to get it to support 550W.)

LIFAN shipping came through with a new generator to replace #4 with the bad exhaust after I told them it still wasn’t fixed. (Keep in mind that I could have returned it to THD since it was only 4 days old at that point.) The new generator arrived and I didn’t like the choke lever sticking out of the box.

I pulled the maintenance cover and the air filter cover to take a look at the choke vane in the carburetor. The vane was not inserted properly and prevented the lever from rotating all the way.

I pushed the vane as far as I could and it allowed the lever to operate almost normally. The load tests on the generator passed with flying colors. The one remaining problem child was generator #4, I got out a drill and punched a very small hole into the exhaust manifold. Cranked up the generator and that hole along with the exhaust port was sufficient for the engine to run properly under heavy load.

If you’ve been keeping up, I now have FIVE running generators!

For comparison I looked online for the “Best” generator for fuel efficiency. The Yamaha 2200VA generator makes about 4KWH out of a gallon of fuel. The LIFAN ESI 860i makes only 2KWH from a gallon of gas and has severe limitations because of the small engine size. The assumption at this point is that the sales department stretched the specifications to the limits of the engineering to make it look better. Then when we (The customers) bought the generator and realized the number of shortfalls we would bring it back and buy a more expensive generator at the same store. (Using the credit from the ESI 860i)

I’ve learned a lot over the last several weeks. The last small engine I owned was a gasoline lawn mower. I don’t remember changing that oil more than once per year. This generator’s oil gets fairly dirty after 4 tanks of gasoline, say 10-12 hours. Not to mention oil had to be added at 6 hours or it would not start at 9 hours. I ran 5W20 oil in Generator #1 for about three hours with 60W load in an attempt to clean the engine since being my first of the four, it was run so long without changing the oil. The 5W20 oil wasn’t real dirty, but it was too risky to keep running the generator at such a lightweight oil.

Over the last month I’ve enjoyed playing with these generators. I had long forgotten the pleasure of working on mechanical things. In many ways they are simpler than the computers and electronics that I spend most of my working hours trying to coax them into doing what we need them to do.

Generators have always fascinated me, they can run for insanely amounts of time without needing replacement parts or adjustments. Feed them their fluids and they run most of the time. These LIFAN generators caught me off guard, I hadn’t planned on disassembling them to do simple maintenance. Normally, easy access is granted for things like the air filter. The air filter on these tiny generators requires the generator case be split in two and the motor removed from the plastic enclosure to clean the air filter! When I was talking to LIFAN, they talked as though breaking down the plastic enclosure was something they planned on the average consumer to do. I cannot imagine my daughter (who is quite talented), taking these generators apart in her apartment. (Neither of us have a useable garage for this sort of maintenance.) Being in the military I am used to having well written service manuals, LIFAN has a TERRIBLE manual. There’s NO way that these generators should run 90 hours without an oil change. I also had not counted on how finicky the output power was. They are “Rated” at 600W, but they often cave in at 540W and drop the output until the generator is shut down and restarted. All I really needed was 175W, they were rated at 600W. Where is the math on that one.

When I was a teenager I bought solar cells from a catalog and mounted them on top of my portable radio.  It worked pretty well and was popular on the beach.  I ended up in the middle of the dessert with a dead cell phone battery and had to use a 60KW generator to charge my Cell phone.  I bought and carried a USB solar panel with me after that.  Over the years those USB panels got a little bit bigger and a lot more efficient to where they’d push a couple amps so I gave them to my siblings for emergencies.  The bigger panels 80-100W were getting reasonably priced so a couple years ago (2012 ish) I started buying them a couple at a time until I had twelve of them in storage.  When I quit my full time job in 2017 I laid one of them on the roof in the winter for a couple days to gather some data.  One of my requirements is to provide 200WH of power for my CPAP.   I have enough battery for a couple days, but going without it is not an option.  After a couple days I found out that one panel in the winter sun was not going to provide enough power.  I grabbed a second panel and bolted them together making a 200W 36V array that should provide approximately 360WH of power per day in December. (The worst month.)  Since I sleep at night the power has to make a round trip into and out of a battery which has a significant cost.  But I can also live on six hours of sleep if I have to.  During the summer months those two panels produce almost 900WH/Day which can keep our refrigerator powered and have enough left over for Cell Phones and charging lanterns for the evening.  It is my opinion that all homes aught to have these two panels to provide USB power for the occupants during an emergency.  If your a bit of a Macguyver, 200WH is enough power to make a pot of coffee, toast a couple bagels or thaw waffles with a microwave.

One benefit of living in the same house for 30 years is having the knowledge gained from being there.  Our little home looses power momentarily several times a month.  Longer outages (like an hour) happen very infrequently and are usually related to a violent storm moving through our area.  Maybe twice since we’ve lived here did the power fail for more than a day.  So for me, a generator would probably die in storage before it was needed. (I had a 7KVA generator for a while and really loud!  It was also a significant amount of work exercising the engine and maintaining a supply of gasoline, so I gave it away.) 

We have three large banks of batteries in the garage  and they power three UPS units that are designed to be connected to them.   Our highest priority load is the refrigerator. The 1KVA UPS with external battery can maintain operation for about 24 hours and I’ve got two solar panels to extend that into two days with bright sunlight.  A second UPS and battery bank supports all the computers in the house and the emergency lights.  A rather large third unit can make a pot of coffee in the morning or toast a bagel on solar power.  (It also can run the microwave with some adjustments.)

In 2019 I bought a very small generator to extend the runtime of these UPS units, it took months to figure out a reliable way of connecting this tiny generator.  In the end the UPS units could not be directly powered by the generator using their power cords.  What does work well is to use power supplies and charge the batteries with the generator.

One surprising revelation was that these small generators are most efficient at 100% load.  An 860VA generator producing 500W generates about 4.4KW/Gallon of gas.  At 130W that same generator only produces about 1.8KW/Gallon of gas.  With that in mind it made sense to me to intermittently run the generator and use the UPS batteries to support the load.  The charge/discharge battery losses were easily made up with the increased efficiency of the generator at it’s maximum rating.

Using a adjustable current & voltage power supply, the charging currents and full charge voltages can be set for the generator power rating, storage battery size and type.  In my case I have enough battery to consume more charging current than the ESI860 generator can produce.   The ESI860 can run for about a little over 3.5 hours on a tank (.4 Gal) of gasoline at full load of 500W.  12 Amps (About the maximum the ESI860 would support) for 3.5 hours is 54AH at 28Volts (3780WH/Gal) into 200AH of battery.

Looking at the load side of the equation I need to be under 300W average load.  The Fridge is roughly 100W average so everything else should be less than 200W.  I can also run the generator more often or supplement the charging with solar.