New LiFePO4 Cells For Use With Solar Power. Test And Charging.

I picked up 28 unused, but older LiFePO4 cells, and started setting up a test system for them.

I also picked up some BMSs (Battery Management System), 2 for 12V 50/100A, and one for 24V 50/100A systems. I may need to get one or two more.

Solar Power Charging: 

The charging system for those batteries are configured like this:

I am using my solar system in my lab, described in the previous post, i.e. 2x 100W solar panels, charge controller and 3x 30Ah / 12V batteries. This may be expanded a bit with more batteries in parallel. providing a nominal capacity of about 180Ah. Because the batteries are GEL type, I count on a total usable capacity of about 90-100Ah.

The single cell charger consists of a DC/DC buck (step down) converter capable of delivering a maximum peak current of 20A, with CV (Constant Voltage) and CC (Constant Current) settings available on the PCB. To be sure of avoiding overheating I have set those like this:

CV: 3.60V (the upper limit of charge voltage for LiFePO4 cells)

CC: 10A (Just to keep it cool)

The circuit does have a fuse in the battery lead, a 15A one. A bit of redundancy is good. The system should not make a show of spontaneous combustion ;) .

This system is now running, and the first cell is being charged. Things seem to run smoothly right now - no overheating.

More Efficient Energy Use:

The intention is to get all 28 cells tested for capacity, then match them for the best possible batteries.

For this I could just use a resistive load and waste all the energy. Here is the idea to avoid some of this waste:

When I have 4 cells fully charged I will get them balanced for voltage. This happens with 4 low-value resistors, one in series with each battery (plus connection). When connected all in parallel the 4 battery-resistor sets will slowly balance the voltage of the cells. This is the simplest way to do this, and may not keep the batteries in balance long term, but that can come later, when all cells have been fully tested.

The first 4 cells can then be connected with a BMS to form a second 12V power supply, provisionally. With this and a second charging circuit, and considering the losses in the system, this system should be capable of charging 3 cells from (near) 0 to full charge. The 4th cell can be charged from the solar system, providing a new 4-cell battery for test charging the next cells. I could also be connected as a provisional (second) solar power system.

This is a long process, but it will test the cells for reaching the full charge voltage.

Checking Capacity Of Single Cells:

The intention is expanding the 12V system for the radio station, and also create an expanded battery capacity in the lab upstairs.

I expect the 12V battery at the station to be about 300Ah (12 cells), and the lab system to be 200Ah (8 cells).

The first system should likely be the one for the lab. So a full test will have to be made of enough matched cells (8 pcs) to make the 12V - 200Ah battery with all cells the closest possible with respect to capacity.

For this a discharge system has to be made. This could be a simple resistive load with a digital meter. The meters I have seen, however, need 6.5V, so they cannot be used for testing a single cell. Hmmm!

Can I program an Arduino (or the like) and make a simple circuit for testing capacity? Of course...

The simplest solution will be making a simple constant current load (or a simple resistive load) and measure the time before the output voltage reaches the lower limit of 3V, then read out the time and switch off the load. Not too difficult with an Arduino, even with my limited programming skills.

I do not need to measure the absolute capacity, just to match the cells, so the resistive load should be sufficient. The absolute capacity can then be measured when the cells are connected as a 12V battery.

It will take a while, as I am not constantly home to monitor the process.

Also, making this with simple means, things do take time, although they battery/cell testing is not too time consuming, as other things can be done while the tests are running.

I suspect some building of supports for solar panels, and some simple receiver circuits and/or microwave stuff, is in my future, along with the battery testing - and doing some QSOs with the radios.

No time to be bored...

Update: After some hours of charging the first cell still charges with about 8A, and the voltage rises very slowly.

Solar Power Expansion.

I moved the older GEL battery system batteries upstairs into the lab:

The 4x 30Ah battery is now connected in parallel (again) and placed on the shelf next to the lab desk. 

I set up 2x 100W solar panels in series with a (max) 6A cable to the indoors. (maybe more). This is a preliminary setup, so the system can be tested (well, it has worked before, so why not again?) . A certain voltage loss is expected, but less so as the panels are in series, and the current in the cable is therefore reduced. Preliminary cable laid, and a charge regulator added, the lab 12V 60Ah system is now running and charging.

Further needed:

- improved (and shorter) cable from panels to the indoor charger/battery system

When I get more batteries I will probably move the 2x 12V/100Ah LiFePO4 batteries upstairs to use as the base 12V system. The existing lead-acid batteries can then be used together providing 24V/about 100Ah along with providing 200Ah -> about 100Ah at 12V. More solar panels will be needed for this extension, of course, as I want the charge to be the best possible.

The 12V - when finished - should be powering things like the GPS/10MHz reference generator for the spectrum analyzer, frequency counter and the transceiver test set.

The 24V part should be powering the small 24V temperature controlled soldering iron, and also a small laptop. 

When the bigger battery gets installed I can try using a (pure sine wave) inverter to power some of the 220V instruments, like the transceiver test set, spectrum analyzer and a few more.

While this system should work well in the summer time It may need some additional assistance from the mains grid in winter time. I should build some relay switching for this before the winter comes, because, of course - Winter Is Coming  (but not for several months) ;) 

For now I do have enough panels for improving the existing system with the batteries at hand, but I need to build much better supports for the panels. I also expect to start getting some larger panels, a few at a time.

Right now it is about extending the solar power system, and getting the antenna system maintained, and some indoor activity with getting some microwave (and other) stuff up and running.

There are interesting times ahead at OZ9QV. I hope it is not in the Chinese sense ;)

Microwave Day.

 In the past week end I went to the Microwave Day in the town of Horsens, Denmark.

Although I have not really built any microwave equipment yet, everyone interested is welcome, and I did also go to the previous one 3 years ago - the ones in the last two years were canceled due to the COVID situation.

I wanted to take things easy, so I did the drive there on Friday, and home again on Sunday.

There were two presentations, one about the Norwegian beacon LB2SHF and the other about the 122GHz experiments. 

After the presentations there was social gathering and microwave talk, as well as Ole, OZ2OE demonstrating his 122GHz equipment. 

There was a small flea market, and I got just a few things. Two N-connector to WR90 waveguide transitions and some ancient 12GHz LNBs with WR75 waveguide inputs.

Because I have some horn antennas for both sizes of waveguide I wanted to have this. The WR90 transitions can be used as they are, and the LNBs can be used in different ways:

1) As a simple down converter, used as-is.

2) As an "active antenna". This requires a modification, taking the output of the 10-12GHz (pre)amplifier to an SMA connector.

3) Removing all the electronics and grinding a bit of material away, the LNB can also be used as a waveguide to SMA transition.

At the flea market I also saw transverters. There was an old version of the DB6NT transverter, complete with a horn antenna for a reasonable price, even if the output power was limited to 30mW. I did not bring it home, but I should probably have bought it, as I do have another 10GHz amplifier module that could have increased the output to about 200mW. A bit of a regret, but there may be other opportunities.

A 24GHz transverter, also at a reasonable price, was seen, but again, I did not bring it home. I know the seller, however, so I could probably still find it for sale.

Then there is test equipment. I had an appointment with another participant, that he would bring a 26GHz spectrum analyzer to check and possibly buy, and I did bring it home. I can now see what the things I build and buy are doing, signal-wise.

In the afternoon there was a demonstration of the 122GHz equipment at a longer distance (5km). Not that far, but not so easy. Signals were heard both ways, though.

Some time ago I purchased a CBNL 10GHz link transceiver module, but there is not much information, so I asked if any of the others had more information.

I have found a bit, but I am worried about the simple stuff: How critical is the order in which I connect the power supply. On the on6ll.be website I found the following:

Since it is used in a remote link the "base supply voltage is 48V (see the two connections at the right side of the connector). Then I see that there is a +8V connection The question is now - do I connect the 48V first, or the 8V? What I cannot easily see is where the negative bias for the GaAsFETs is generated. From the 48V converted to 7V, or from the 8V? It is not even clear if the 8V is generated on board when the 48V is connected. 

The other parts I have found sufficient information about to proceed. The LO should be somewhere around 7.5-8GHz, and the IF for the "RF/microwave" section then uses an IF of about 2.4GHz.

The T/R switching is clear enough from the description on the website, even the illustration above.

If any reader knows I would appreciate the information. I simply do not want to make a potentially destructive test, as the GaAsFETs simply will self destruct if the drain voltage is applied before the negative gate bias, as those familiar with GaAsFETs will know. If no reader knows I will have to extend the search.

Good to be back doing social activities again, there will be a few more coming up in the next months. Some of them will be outdoors, and others indoors. The next one is only a few weeks away.