Charcoal bags

Charcoal bags

Search This Blog

Sunday, 26 April 2020

Nickel-Iron Battery Review

I recently converted our house to nickel-iron batteries. I’m happy with them - but it took a while to learn how to make them work well. 
Here are our nickel-iron house batteries, on our verandah


This post is about using nickel-iron batteries in a household solar power supply. I start with some theory about batteries and why nickel-iron might be good in some situations, and then describe my practical experience of depending on them in our home.
Nickel-iron batteries (often called NiFe batteries - using the chemical symbols for nickel and iron) are rare in off-grid houses (at least in Australia). I have heaps of friends and family living off-grid, and until recently every one of them and us used lead-acid batteries.
I found it difficult to find good, practical information about nickel-iron batteries while considering getting some. Actually I’ve found it difficult to find good information about all sorts of batteries - rechargeable batteries seem to be a mystery vulnerable to all sorts of misunderstandings.
This post is my little attempt to collect some useful information together. Before I describe my practical experience with my nickel-iron battery, I outline my understanding of the usual types of off-grid solar house batteries, and how they compare.

Batteries: a major energy cost

Storing energy in batteries takes a lot of energy. Most of this energy is “embodied” in the batteries in the thousands of processes required to get them working: mining, refining and shaping metals, manufacturing plastics from oil, providing the economic needs of the many different people who design, make, sell, transport, or install them. The best way we have of measuring this energy - and its associated CO2 emissions - is the money we spend on batteries (see my blog post on how money is energy).
How much energy - or money - a solar house battery uses is very variable. A small (low cost) battery that lasts a long time will use less money and energy per year than a big battery that dies young.
Battery cost is greatly affected by how you use your solar energy. If you mostly use electricity when the sun is shining, you can get by with a small battery, with shallow discharging, that lasts a long time. If you want to do electric cooking at night after a cloudy day, you’ll need a big battery that gets deeply discharged and may not last so long.
To get a feel for the cost of battery storage, consider a traditional lead-acid solar battery just like I’ve had in some of my off-grid solar systems: 12 volts (V), 1000 Amp hours (Ah), costing about $4000 and lasting maybe 10 years. Note that this doesn’t include the cost of transporting or installing them, or building a suitable enclosure, all of which are real energy and money costs.
The daily discharge cycle of lead acid batteries should be only about 10% from full, to get a long life, so each day this battery could comfortably cycle 100Ah at 12V, which is 1.2 kilowatt-hours (kWh). Doing this every day for 10 years gives 3650 cycles, which would total 4380kWh. $4000 divided by 4380kWh gives nearly $1 per kWh of energy stored for a few hours.
Generating this energy from solar PV panels is much cheaper than storing it. If you install 1kW of panels for $2000, that generate maybe 4kWh per day and last for 20 years, their energy costs around 10c/kWh - 1/10th the cost of storing this energy in batteries (I recognise I’m ignoring various real costs such as regulators and inverters - I’m trying to get the principle here).
To summarise: storing energy in batteries costs way more than generating it. If you can reduce your battery size and increase its life - mostly by using less energy and using it when it’s being generated by your PV - you will greatly reduce the energy and money cost of your home energy system.

Why nickel iron?

Most of my life I’ve lived here at Mt Glorious, off grid, depending on photovoltaic (PV) electricity and battery storage for our electricity supply. My family and I have learnt to live with the limitations of off-grid PV power, limiting our daily consumption through frugality and efficiency, using power when the sun shines and carefully limiting electricity use when it’s cloudy.
For nearly all this time we’ve depended on lead-acid batteries for electricity storage: starting with an old car battery from the kerbside rubbish collection, then old Telecom batteries, and finally a few sets of big, brand new batteries costing 1000$ of dollars. I’ve become comfortable with the mysteries of lead-acid batteries: the need to cycle them from the top (keep them nearly full), the importance of having enough PV current to give them a good, bubbling equalisation and de-stratification every few weeks, and the lurking dangers of sulphation if they get discharged too low for too long. Lead-acid batteries are like chainsaw chains: they’re never as good as when they’re new, and time gradually fritters them away. If you get 10 years out of them, you should be grateful - 15 years is exceptional, 5 years is not unusual.
As part of our ongoing efforts to create a resilient household, I’m now trying nickel-iron batteries. This is for 2 main reasons: durability and deep-cycle ability (which are sort of the same thing). By all accounts, nickel-iron batteries have a long life expectancy - supposedly several decades. Closely related to this is a nickel-iron cell’s ability to be deeply discharged, and left at a discharged state for a long time, without causing long-term damage. This is in contrast to lead-acid batteries that are damaged by deep discharges and time spent discharged.

Amp-hour capacity

The capacity of a battery in Amp-hours (Ah) is one of its most important characteristics, but Ah means very different things in different types of battery. The ability to tolerate deep discharge without damage has a huge effect on the meaning of a battery’s amp-hour capacity.
For clarity: Ah is a measure of battery energy storage, calculated by measuring the current in amps (A) from a battery, for how long in hours (h) it can provide current, until considered to be discharged. The amount of Amps multiplied by the number of hours gives amp-hours (Ah) - e.g. if you draw 10A for 5 hours, you’ve used 50Ah.
Batteries are sold with an Ah rating: for example our last lead-acid house batteries were 1300Ah cells, meaning you could theoretically discharge them at 13A for 100h and then they’d be flat (but you’d never do this!). An important thing to know is that you get more Ah from a battery if you discharge it slowly - more energy is wasted inside the battery if you draw a higher current. Batteries are given their Ah rating based on discharging in a particular number of hours, e.g. my old batteries were rated at 1300Ah if discharged over 100h, but only around half that if you discharged them in 10h. This is shortened to saying the battery has a C100 rating of 1300Ah and a C10 rating of 650Ah. 

Lead-acid: the dangers of sulphation!

If you want them to last a long time, lead acid batteries should be kept above 90% full in daily cycling, and never discharged below 70% when you have a long cloudy period (70% full is often described as 70% SOC - state of charge). This is because lead-acid batteries gradually sulphate according to how long and deeply they are discharged - insoluble lead sulphate crystals grow on the lead plates and stop the charge-discharge reaction.
When our lead-acid house batteries got below 80% SOC due to heavy cloud, we’d start worrying. We’d plan to run a backup generator and charge the batteries up. We weren’t running the backup because the lights were about to go out - our lead-acid batteries were still 70% or 80% SOC with enough storage to run our house for days. We were running the backup because we were worried about shortening our battery life due to sulphation.
So in reality, when you buy 1000Ah of lead-acid batteries, you’re getting maybe max 250Ah of useable storage if you want to get a long life from them, less than 250Ah if you are discharging at high currents. This doesn’t mean that lead-acid batteries are bad - their popularity is because they are relatively cheap, deliver high currents easily, and last a reasonable time if looked after well. 

Deep discharging

Most other battery chemistries don’t have the problem of being damaged by deep discharge in the way lead-acid batteries do (though they all have their own problems). For example Lithium batteries can deliver their full Ah capacity repeatedly, though I understand they can be more durable if cycled through a smaller capacity - e.g. between 10% and 90% full. Nickel-iron batteries, I am assured, do not suffer harm from deep discharge, nor from being left in a discharged state for a long time. In practical terms, this means that 200Ah of nickel-iron (or lithium) batteries will provide the same daily cycling ability as 800Ah or more of lead-acid. Nickel-iron solar batteries also tend to be given an Ah rating based on a 5 hour discharge, while lead-acid solar batteries tend to be rated on a 100h discharge. This artificially inflates the Ah capacity rating of the lead-acid battery.
Overall, this means you can’t usefully compare batteries directly on cost per Ah, because some batteries can be cycled deeply while others can’t, a battery delivers different Ah depending on what current you draw from it and batteries have very different life expectancies which greatly affects cost per year of service.

When the sun goes away

Having batteries that can sit in a deeply discharged state without damage is very attractive to me. When our mountain goes into the clouds for a few weeks, it would be great if we didn’t need to worry about the batteries being damaged by getting down in charge. We can stop all the big loads and just run fridge, lights and some electronics. We can leave the battery nearly flat, do some backup charging if we are really running out, and wait for the sun to charge the batteries fully again. This is the promise of nickel-iron batteries - no need for sulphation anxiety!

Equalisation, balance charging and regulation

Any normal household solar power system uses some sort of regulator (also called a charge controller) between the solar panels and the battery, to control how the battery is charged. Mostly the regulator protects the battery from over-charging, but some regulators also give some information to the users about the battery state of charge. Depending on the type of battery, this charge regulation is more or less complex, largely regarding the problem of keeping each cell at the same state of charge as its sisters in the battery.
For clarification: a “battery” is a group of “cells” joined together, usually in series. However english language now tends to use the word battery to describe a single cell: e.g. a single AAA battery for a torch is really a single cell; a 12V car battery is made of 6 x 2V cells joined together in series.
Each cell in a battery is an individual, each having slight differences. As a battery cycles up and down over time, slight differences between cells can accumulate and become big differences in state of charge. One or two cells in a battery might lag behind the others, gradually becoming discharged while their sisters are full. In a lead-acid battery this could cause a cell or two to gradually sulphate and die young, wrecking the whole battery. To prevent this trouble, from time to time all the cells need to be brought up to being all completely full at the same time. This is called “equalisation”, and different battery chemistries require different solutions to achieve this. I’ll run through how different house battery types get equalised, so we can compare nickel-iron batteries to other types.

Equalising flooded lead-acid batteries

Flooded lead-acid batteries (traditional cells with a liquid acid electrolyte that sloshes around in the cell case) are easy to equalise. Every few weeks, a higher than usual voltage is given to the battery, for a few hours, when it’s already full. This pushes a higher than usual (for a full battery) current through all the cells in series: in effect the battery is being over-charged. When a flooded lead-acid cell is full - the reaction of lead and sulphuric acid is complete - the electrical energy being pushed through it can no longer be stored as chemical energy. The energy has to go to waste somewhere, and in a flooded cell the energy is consumed in splitting water molecules into hydrogen and oxygen gas, that bubbles up through the electrolyte and out through the cell vent. Gassing like this happens to a lesser extent even when lead-acid cells aren’t full: the fuller they are, the less charging energy is stored and the more energy is wasted as gassing. This gas has a lot of energy in it, which can be released in an explosion if you give it a spark - that’s why you keep sparks and flames away from lead-acid (or nickel-iron) batteries.
While the charge current is flowing equally through all the cells in series, the full cells are losing the energy from the electric charging current as gas, and any cells that aren’t quite full can continue to store chemical energy - charge up - using the same current. Thus all the cells in the battery can gradually equalise - get full - at the same time, even though they may have started at different states of charge.
While this equalisation charge is happening, the flow of gas bubbles rising up between the lead plates in the flooded lead-acid cell has another important purpose: stirring up the electrolyte. The sulphuric acid in these batteries is heavier than the water it is dissolved in, and with time the acid can settle down to the bottom, making the electrolyte more acidic at the bottom of the cell and more watery at the top - this is called “stratification”. This is bad for the cells: the over-strong acid at the bottom can damage the lead; and bad for energy storage: the weak acid at the top doesn’t react as much with the lead so doesn’t store so much energy. Testing the electroyte in a lead-acid cell with a hydrometer (that measures the density of the acid and thus the SOC) can sometimes show low-density electrolyte (like a discharged battery) when the batteries are fully charged, just because of stratification: the batteries are full but the acid has sunk to the bottom.
To avoid stratification, the solar power system must deliver enough current to adequately stir the electrolyte. In my experience a battery will need PV panels that can give a charge current in Amps, of about 1/20th the Ah rating of the battery. This means a 1000Ah battery needs PV charge of at least 50A, when the midday sun is shining and the regulator is equalising, to keep its electrolyte stirred. This also means that bigger isn’t always better with flooded batteries: if your battery is too big for your PV array, you may not be able to keep the electrolyte stirred, and your battery may die young.

Key points for flooded lead-acid batteries

Here are a few key things for flooded lead-acid cells:
  • Flooded lead-acid batteries need to be periodically equalised to keep them equally full, by holding them at an extra-high voltage for a few hours. This is usually done automatically by the regulator, which can often be programmed to suit the particular battery type.
  • Flooded lead-acid batteries are good at losing energy as gas when they are over-charged. Making this gas uses up water so the batteries need topping up. The regulator limits the charge voltage so that batteries don’t need topping up too frequently.
  • Because they can lose energy so easily, a flooded lead-acid battery can be equalised by simply pushing extra current through all the cells until any lagging cells catch up with their sisters.
  • Lead-acid cells need enough charging current  to give them a good bubbling and keep the acid well-mixed. For this, they need enough PV panels charging them.

Sealed lead-acid batteries

Sealed lead-acid (SLA) batteries, often titled valve regulated lead acid (VRLA) batteries, are common in houses. A lot of off-grid solar energy installers recommend sealed lead-acid batteries because sealed batteries don’t need regular topping up with demineralised water, and installers are fed up with customers who don’t maintain their batteries.
When SLA cells are working normally, the hydrogen and oxygen gas produced at their plates is re-combined to produce water and heat inside the cell. This system can only handle so much gas, so it’s very important not to over-charge sealed lead-acid batteries and overload the battery with gas. SLA cells also don’t have the problem of stratifying acid because their acid is trapped in gel or a glass mat, so they don’t need to stir their electrolyte with gas bubbles like a flooded cell. It’s very important to carefully regulate the charging of sealed lead-acid batteries, maintaining the right voltages and limiting gas production. Sealed lead-acid cells can’t handle overcharging by losing energy as vented hydrogen and oxygen gas. In a gel cell, excess gassing can cause the gel - containing the acid - to permanently lose contact with areas of the lead plates.
Sealed lead-acid batteries do need equalising by periodically holding them at a higher voltage for longer than their normal charge cycle, but they do this with lower voltages than flooded batteries, carefully controlled by a regulator.

SLA key points

The key things about sealed lead-acid batteries are:
  • SLA batteries can’t lose energy by venting gas.
  • SLA batteries (especially gel batteries) can easily be damaged if they are charged at too high a voltage, so they need strict control of charging voltage by their regulator.
  • SLA batteries can lose some energy by internal gassing and recombination. This produces heat, which needs to be limited.
  • Because SLA batteries can lose some energy internally, they can be equalised by carefully charging them all until any lagging cells catch up.
  • SLA batteries don’t need topping up.

Lithium batteries

I love lithium batteries. They’re great on our electric bikes: they’re compact and light (compared to other battery types), they deliver high currents and some types last a long time. They aren’t damaged by being kept at a low SOC - as long as they’re not too flat. They make e-bikes into the marvellous transport machines they are. 
Here's a 36V lithium iron phosphate bike battery: 12 x 3.3V cells in series, with the BMS on top, with thin wires going down to keep an eye on each cell. It also needs a special charger that gives the right current and limits to the right voltage
However lithium batteries are complex. This is because lithium cells are very fussy about how they’re charged. They are damaged by charging to too high a voltage, or discharging to too low a voltage. Lithium batteries need their cells balanced, like other battery types, but this can’t be done by over-charging all the cells in series like lead-acid batteries, because this risks taking some cells to a damaging voltage. Lithium cells can’t waste the surplus energy of being over-charged by gassing - like a flooded lead-acid cell - or heating - like a sealed lead-acid or nickel-metal-hydride cell. Over-charging damages them, so you can’t do it.
To keep lithium batteries within their safe bounds, they are managed and balanced by an electronic circuit, often called a battery management system (BMS), that is wired on to the battery. The BMS controls charging and discharging of the battery, shutting the battery’s output down if the voltage gets too low, or shutting down the charge if the voltage gets too high. The BMS also continuously measures the voltage of each cell, so it can shut down the whole battery output if one cell’s voltage gets too low. To equalise - or balance - the cells, the BMS has circuits that slowly discharge the fullest cells during charging, so the less full cells can catch up. This occurs every time the battery is charged to full, so only tiny adjustments are needed.
A lithium battery is very much dependent on its built-in BMS, as well as needing a solar charge regulator like any other battery type. It is a complex electronic device, as well as a chemical energy storage. If there is an electronic failure in the BMS, from component failure, lightning surge, etc., the parts and skills to repair it will be needed for the battery to provide power again.

Lithium key points

The key points for lithium batteries are:
  • Lithium batteries are easily harmed by being charged at too high a voltage, because they aren’t able to lose any energy as gas, and not much by heat.
  • Because of this, lithium batteries can’t be equalised by charging the whole set and waiting for lagging cells to catch up.
  • Each cell needs to be individually controlled in voltage, with a BMS that gives individual lagging cells extra charging time and current.

Charging and equalising nickel-iron batteries

Each of the batteries described above - flooded lead-acid, sealed lead-acid and lithium - are more complex in their construction and require more careful and complex regulation than the previous type. Compared to all of them, nickel-iron batteries have the simplest charging needs.
Nickel-iron cells are flooded with electrolyte (like flooded lead-acid), so they can easily lose surplus energy as hydrogen and oxygen gas. They are so tolerant of gassing that they hardly need to be regulated. The main purposes of having a regulator on a Nife battery is not to protect the battery, it’s to reduce the amount of topping up with water required, and to avoid system voltages that are too high for the loads (especially the inverter).
This tolerance to over-charging means they can be equalised by simply charging the whole battery and letting any lagging cells catch up.
Nickel-iron cells are not damaged by being left at a low SOC, so a lagging cell won’t be harmed. However cells don’t get much change to lag, because of the over-charging nickel-iron batteries tend to get on any sunny day.

Nickel-iron key points

To summarise:
  • Nickel-iron batteries easily lose energy as gas
  • Equalisation is not usually an issue when Nife batteries are charged generously
  • Nickel-iron batteries are tolerant of charging without a regulator (if currents aren’t too high) but a regulator reduces the need to add water and avoids voltages too high for the loads

Battery voltage

Different battery chemistries produce different voltages. Lead acid cells - of all types - produce about 2V. Lithium cells in solar house systems (lithium iron phosphate) produce about 3.3V, but the chemistry in lithium 18650 cells (like torches and many bike batteries use) produces about 3.7V. Nickel-iron cells produce about 1.2V each - so you need lots of them. Remember the voltage of rechargeable cells varies a lot, it rises when the cells are being charged, rises even more as they get full, and falls when being discharged.

Efficiency of off-grid batteries

Nickel-iron batteries gas a lot, even more than flooded lead-acid. This indicates a potentially low efficiency for these batteries - a lot of electric charge goes into them that is lost as gas, and won’t be coming back as electricity. This lack of efficiency could be a problem in some situations, where energy supply is very limited or expensive, but I don’t think it’s a real issue for most off-grid houses.
Battery charge efficiency is not usually a significant concern for off-grid household solar power systems. Normal off-grid solar systems are tremendously wasteful of the potential energy output of their PV panels. This is because enough panel power is installed for the house to survive normal times of less sunlight, like cloudy weather. That means that on sunny days, a typical off-grid house has filled its batteries before noon, and the PV panels are mostly switched off for the rest of the day, except for loads that directly use the power. Batteries that are more efficient, such as sealed lead-acid or lithium, tend to reduce (taper) their charging current earlier in the day, so the energy they don’t waste inside their cells through gassing, is instead wasted by the regulator turning off the power of the PV panels.
For another way of looking at off-grid efficiency: if you add an extra PV panel to help address a power shortage during your few annual cloudy weeks, that panel’s power will be wasted for the sunny rest of the year. 

I recognise that battery efficiency can be very useful sometimes: in cloudy weather when the batteries are low and there is limited sunlight for a few days, efficient batteries can help you make the most of what you can harvest.
I argue that we should be cautious about the pursuit of technical efficiency energy- producing and using devices. More efficient panels, batteries or loads need to be looked at in the big picture. The real question for me is what technology and what behaviour results in the lowest total, long-term money and energy cost? This is our best indication of damage to the environment and how much of our life’s effort will need to be spent earning money.

What if the regulator breaks down?

I seem to spend half of my life fixing things that have broken down: I know this is how things are so I plan for everything to break down and need fixing. I haven’t had a solar regulator break down in this house in decades, but I know this could change today - or more likely in the next thunderstorm. So I think it’s worth thinking ahead about how your solar battery would cope with a failed regulator.
The most vulnerable battery to breakdown would be lithium. A lithium battery uses a solar regulator to control the voltage from the PV panels, then has its internal BMS to regulate charge to each cell - there are lots of vulnerable components that could break down. To some extent the PV regulator and the BMS back eachother up: if either fails, the other will protect the lithium cells to a large extent, until the system is repaired. However, failure of either the PV regulator or any part of the BMS would probably put the battery out of action until a fairly major repair is done.
The next most vulnerable battery would be the sealed lead acid battery. Without a fully-working regulator, the charge from a PV array could easily kill an SLA - especially a gel battery - in one sunny day (I haven’t tried this - let me know if you prove me wrong).
A flooded lead-acid battery is reasonably resilient. A modestly-sized PV array could be used to charge a flooded battery without a regulator, if some attention was paid to voltage and the PV switched off at a reasonable time. Over-charging would normally simply result in extra water loss.
Nickel-iron easily wins the regulator resilience race. Unregulated on a sunny day, Nife batteries could create annoying high voltages and bubble a lot, but (unless they have a really oversized PV array cooking them) they won’t come to any harm. I believe some people run them without regulators on purpose.


I’m in no position to make my own claims about nickel-iron battery durability, because my personal experience is so short, but manufacturers’ claims and the accepted wisdom is pretty consistent: they last for decades, 30 or 40 years easily. This makes a huge difference to the energy, materials and money cost per year, or per energy delivered.

Electrolyte replacement

Nickel-iron batteries do need major maintenance every 7 to 10 years: electrolyte replacement. I’ve never done this.
Their electrolyte is potassium hydroxide, a strong alkali. However the atmosphere we live in tends to make water acidic: carbon dioxide from the air dissolves in water to make carbonic acid. That’s why rainwater is quite acidic and makes stalagtites in limestone caves, and is why increasing CO2 in the atmosphere is making seawater more acidic and damaging shelled sea creatures. The battery electrolyte absorbs CO2 from the air and is gradually neutralised - gets less alkaline. When this makes the battery lose its mojo, you need to tip out the electrolyte and put in a fresh batch. This would be a significant job, I imagine taking a day or 2. The good thing is that potassium hydroxide is not toxic, and it can safely be diluted onto earth and plants.

My experience of nickel-iron batteries

That’s enough theory about solar house batteries. Now to my actual experience with nickel-iron. 
All the PV electricity for our house comes from these panels on the shed - about 1.5kW of old 12V panels, mostly Kyocera. You can see a lot of wood energy is also stored in this shed.

Purchase and delivery

I bought my 24V, 200Ah nickel-iron battery from David Bartlett at, in August 2019. They cost $5214, delivered to a Brisbane trucking depot, arriving less than a week after I completed payment. The 20 cells were packed in a plywood crate approx 1m x .5m x .5m, weighing about 240kg, which was forklifted onto my friend’s ute.
Cells were individually packed in plastic bags, each filled with electrolyte, but we had no spills. Sometimes NiFe cells are delivered dry, and the installer must mix the potassium hydroxide electrolyte and fill the cells - this takes a day or 2. The package also included 20 cell joining straps, made of nickel-plated steel, plus plastic strap covers (to reduce the risk of accidental short-circuits, e.g. by dropping a spanner onto the top of the batteries). Each cell weighed only about 12kg, so they were easy to lift out of the crate and carry into place.


To fit into my available space, I arranged the 20 x 1.2V cells in 3 rows. I was able to use the same battery box we’ve used for nearly 20 years, this being the 3rd set of batteries to live in it (and before that I had several sets of previous batteries - see why I’m thinking about battery life?). 
20 x 200Ah 1.2V nickel-iron cells in 3 rows. This box previously held our lead-acid house batteries: 6 x 1300Ah 2V cells - that took about the same volume.
The set came with joining straps that connected adjacent cells, but the rows on different shelves needed to be joined with heavy conductors, to carry high discharge currents without a big voltage drop. I used pieces of 1/2” copper pipe, hammered flat and with a hole punched in each end, as conductors between rows. I calculated the pipe had a cross section of about 30mm2, which gives a current density of a little over 1A/mm2 when 40A (1kW) is being drawn from the batteries (10A/mm2 is considered to be a maximum current density). If I drew 80A, for a 2kW load on a sunny day, I could expect to draw around 40A from the batteries and 40A from the PV panels. 
Here's a copper strap made of flattened 1/2" copper pipe, between 2 cell links, joining 2 rows of cells
The Ironcore 200Ah batteries have 20mm diameter, nickel plated, threaded steel studs as their positive and negative poles, so you need connectors with 20mm holes in them. This is a big hole, beyond the normal capability of backyard workshop drill presses (I don’t normally drill metal bigger than 13mm, my drill press chuck goes up to 16mm, and I don’t know if my drill press would go slow enough for 20mm if I had a 20mm drill bit that fit in it). I made some extra connectors by punching holes in flattened 1” copper pipe and drifting the holes to size - a blacksmithing process.

Behaviour of the new NiFe batteries

In practice, nickel-iron batteries are quite different from the lead-acid batteries I’m used to, in two main ways: they have a wide voltage range, and they have limited ability to carry current. I suspect that both of these characteristics are because the nickel-iron chemical reaction is slower than other chemistries.

Voltage range

The first obvious difference between Nife and lead-acid batteries is the wide voltage range between full and empty, and between charging and discharging. Battery voltage in our system ranges from 19V to nearly 34V (lead-acid range would be around 24V to 30V), sometimes covering this over short periods - e.g. if I switch on a 700w electric jug on a morning when the battery is deeply discharged, the voltage might drop from 29V to 19V. A normal day might range from about 25V to 33.8V.

Early challenges

It took me a while to learn how to use our new nickel-iron battery, especially given the lack of good information about how to set a regulator for them.
For the first few months, I limited the maximum voltage to 31.5V, because this was the highest our DC fridge and freezer could tolerate (with 12-24V Danfoss compressors). While the 2019 drought was on with endless blue skies, this worked alright, but when the 2020 rainy season arrived and we went into the clouds, the battery couldn’t keep things running through the first sunless day - it dropped under 19V under a small load, and got too low to run the DC fridge. I quickly realised I must have been running the battery at too low a voltage, so what I thought was a full battery was actually nearly flat.
One of the clues I had that I’d given the battery too low a charging voltage, was that the cells were losing so little water. I hadn’t topped them up since new, nearly 6 months, because the electrolyte levels were going down so slowly.
Another clue was the battery’s intolerance of high discharge currents. In the morning, before the sun was on the panels, a 700w electric jug (that’s a really low powered jug) would drop the battery under 19V and set off the inverter’s low voltage shut down - with mood-enhancing alarm sound.
Clue 3 was the low charging currents I noticed from the PV supply. The battery could be rather flat, but might only take 10A from panels that could supply 40A (this was because I had set the charge voltage too low on the regulator).
Once I worked out how to set the regulator to suitably high charging voltages (see below), our nickel-iron battery was transformed. It’s now a bundle of energy: the voltage rarely dips below 24V at night, usually staying above 25V, even at low SOC. The lowest I’ve seen the SOC since upping the charge voltage was 17% (according to the regulator). At this SOC, the battery still keeps everything running fine (although we are very electricity frugal in cloudy weather when the SOC gets that low), and the low SOC causes no harm to the battery at all. The morning electric jug is no problem at all (but we only use it if we expect a sunny day!).
At the higher voltage settings, the cells gas more, and I need to top them up more frequently - once every month or 2. This is not a problem - if you can’t remember to top up your house battery, you probably forget to put petrol in your car and water your lettuces too.

Regulator settings

I use a Plasmatronics PL60 regulator, one that we used for 12 years on our flooded lead-acid batteries. I really like these PL regulators: good quality, made in Australia, programmable, and they count Ah and estimate SOC so you have some idea of how flat your battery is in cloudy weather - many regulators don’t. 
Here's our Plasmatronics PL60 - nearly 20 years of working 24/7. On its left are DC and AC circuit breakers.
When I realised my Nife problems were due to low charging voltages, I found I couldn’t set the PL60 to the high voltages I needed. I called Plasmatronics (Australian made, by people who know their product!) who, although they hadn’t any experience with Nife batteries, came up with a work-around. On their advice, I went into the regulator settings menu and used the temperature compensation setting to add about 2V to all the settings - tricking the regulator into thinking the batteries were very cold and needed higher charging voltages (that’s what lead acid batteries need in cold climates).
If you’re using a PL regulator, this is what I did:
SET -> REG -> TCMP -> change setting from 0 to 6. This adds about 2V to charge voltage settings
SET -> REG -> BMAX -> set max boost voltage to 31.8, giving a real life boost voltage of 33.8
SET -> REG -> EMAX -> set equalisation voltage to 31.8 - giving 33.8V
SET -> REG -> ABSV -> set absorb voltage to 31.0 - giving 33.0V.
SET -> REG -> ATIM -> set absorption time to 4 hours.
SET -> REG -> FLTV -> set float voltage to 28.0V - giving 30.0V
Some settings could be reduced if there is too much water loss by gassing - I think this is likely in sunny seasons. Also settings might need to be taken down a notch if the voltage sometimes goes over the inverter’s maximum for a moment - this has happened to me.
The maximum charging voltage is about 33.8V - just under 1.7V per cell. I’ve set this maximum to avoid the inverter having a high DC volts shutdown - maximum DC input to our inverter is 34V, and it shuts down and alarms if it gets to 34V (a Selectronics SE22 with 20 years service here). I think an inverter running on a 24V nickel-iron battery would need a high DC voltage capability of at least 34V.

The 19 cell option

I have read that some people use 19 cells, instead of the standard 20 cells, for a 24V system. This reduces the overall voltage by about 1.2V at night (when the battery is discharging) and by up to 1.7V in the day when the PV is charging.
This strategy would solve a lot of the problems I’ve had with regulators (getting a high enough voltage to charge the battery properly) and fridges (having too high a voltage for the compressor). I haven’t tried it yet because I’m reluctant to lose the 1.2V in cloudy weather when the battery voltage goes low. You could reasonably say that most of the problems I’ve had with my nickel-iron battery have been because of my sticking with 20 cells. I’ll keep the 19 cell option in mind.
If you can’t get a regulator or inverter that can cope with the high voltages of a 20 cell Nife battery, dropping a cell is a reasonable solution - that I haven't tried.

Why we should make hay when the sun shines

Using power when it's sunny, and not using it when it's dark or cloudy, is fundamental to using solar energy. This might seem obvious, like making banana bread when you have a ripe bunch. However cheap fossil-fuel-based energy has made us blind to the variability of most natural resources, and perhaps somewhat entitled to having as much energy as we want whenever we want.
As I described at the beginning of this post, storing solar energy costs about 10 times as much as producing it. You can imagine one kWh of solar electricity, stored in a battery and used at night, is worth about 11 times as much (the cost of generating plus the cost of storing) as a kWh used in the day. Even if some revolutionary battery storage cost 1/10 as much as it currently does - costing the same as PV generation - night power would still have twice the cost of day power.
Of course, off-grid power is much more complex than this. If you’re having a sunny week, each afternoon your regulator might be wasting nearly all your panels’ production because your battery is full and you’ve done all your big electric jobs for the day. That wasted solar energy has zero value (but it did cost money). At the other extreme, if you’re in the middle of a rainy month, your battery is nearly flat, you have a freezer full of beef, your generator has broken down and you insist on vacuuming the carpet, that little bit of power might cost you a fortune.
These sorts of limits affect energy systems at all sorts of scales. Here’s a fascinating story about what happened to Tasmania in 2016, when they had a hydro power drought and lost their backup power. I note that in the face of this crisis, the Tasmanian government didn’t raise the price of electricity to reduce demand, and have kept the cost of backup diesel generators a secret - perhaps they saw a political imperative not to challenge people’s sense of entitlement to cheap energy.
If our goal is to reduce the cost of our off-grid solar power, or if it’s to reduce our carbon emissions, we should maximise our direct use of PV power while it’s being produced, and minimise our dependence on battery storage. It’s the same: by reducing our long-term cost, we reduce our long-term carbon emissions, and we usually get a more resilient system.

How to manage demand

Successful living with solar power is like growing a vegetable garden. You need to work within a philosophy, build systems, and develop your skills. On a sunny day the gardener might water her lettuces and shade her seedlings, while the solar power user might run the washing machine, pump water and make tea with the electric jug. I think there’s no way around learning to become conscious of the energy supply and matching tasks to it.
As an example, let’s look at how we cook in our house. If there is sun on our PV panels, and our batteries are reasonably full, we cook on a portable benchtop induction cooker, and boil our 700w Birko electric jug. If it’s cloudy and we have a medium amount of cooking to do, we’ll light a charcoal fire. If we want to make a little expresso coffee or re-heat a cup of tea, we can use our LPG gas ring (our 9kg LPG refill is now over 5 years old). For major cooking tasks in the daytime, or dinner every evening, we light our big Rayburn slow combustion cooker - burning wood. We didn’t design our PV system for electric cooking, but we can do most of our daytime cooking using solar electricity that otherwise would be wasted. This electric cooking saves us time and saves charcoal or wood fuel. The LPG is just a very cheap luxury, because we don’t use it for serious cooking.
When it’s sunny, we pump water, cut firewood with the electric chainsaw, do vacuuming, run the washing machine, use electric saws and planes in the workshop, etc.. Our loads need to stay within the power capacity of the inverter: there is a limit to our maximum power (in watts) as well as our supply of energy (in kilowatt hours). We can use up to 4kWh these days.
In cloudy weather we reduce loads. No electric cooking, essential pumping only, minimal workshop machines. We can stay under 1kWh on these days.
A few times a year we have extended heavy cloud with nearly no solar input. If the fridge and freezer are doing low voltage shutdowns, we'll turn them off overnight - even a cloudy day generally provides enough solar input to get them to run. We'll leave the inverter turned off most of the day, turning it on for limited periods to charge computers, grind some coffee or briefly run a bench grinder.

Daylight drive

While I was searching around the web for information about nickel-iron batteries, I found a mob called Living Energy Farms (LEF). LEF take the principle of minimising dependence on batteries to its logical conclusion: many of their home and farm machines are run by a micro-grid that uses DC panel power, direct to loads, without battery storage. They call this “daylight drive” because work can only be done when there is sunlight on the PV panels. They have about 1400w of panels, (maybe 6 x 230w) connected in series, that drive 180V (nominal) DC motors they have installed on a range of machines. Their system is exceptionally resilient and low cost, but would take a fair bit of technical work to set up.
Here’s a brief description, and a video.

Backup charging

In the 6 months since I raised the regulator voltages, I've only done backup charging once - but we've had a short rainy season this year.
I don't currently have a good backup charging system in place, partly because we manage pretty well without one. So far we’ve used our old Christie charger, a direct charging unit with a small Honda motor and a large car alternator, made to produce DC power straight into a 12V battery. With the nickel-iron battery, we use the Christie charger to charge 7 or 8 NiFe cells in series, then move the clamps along to charge another 7 or 8 cells. This doesn’t add up to 20 cells, so some cells get much more charging than others. This is a feature of the resilience of Nife cells: they will easily balance out their states of charge in sunny weather.

DC loads

I like to run our lights, fridge, freezer and a few other small things on DC circuits, direct from the battery. I know the modern way is to run everything at 230V through the inverter, but I think it’s more efficient, safe and resilient to have DC circuits as well.
Having DC circuits is efficient. When we go to bed at night, we turn off our inverter (its in our bedroom, so that’s easy). This saves about 25Ah of energy overnight (over 0.5kWh), that would otherwise go into running the inverter and powering various little things that are left plugged in, like cordless phones etc.. The fridge and freezer can still run on their DC circuit, and the DC lights can still be turned on. 25Ah may be trivial in sunny weather, but when we’re in the clouds for a month, it makes a huge difference to our energy balance.
Using DC circuits is also more resilient. Like I keep saying: everything breaks down; and this includes inverters. A good inverter costs a few thousand dollar$, most of us don’t keep a spare in the shed, and it could take days or months to get a new one (perhaps more if a virus has shut down China) and get an installer to visit. With DC lights and fridge, a broken-down inverter won’t leave us in the dark with rotting food - although of course other failures could put us in that situation.

DC lights

25 years ago, it was normal for off-grid houses to have DC lights, running straight off the house battery. Now it’s unusual, because PV panels are much cheaper and people are much richer - lights usually run on AC from the inverter. So it’s become harder to find good quality DC lights for a house.
Lately we’ve been buying LED bulbs with a wide voltage range: the seller states 24-36V, but we’ve had no trouble when house battery voltage has dropped to 19V. We’ve been buying from The bulbs are quite expensive (over AU$20 each), but we’ve had no failures. We mostly use 15w bulbs in our living area. 

Here's the 15w LED light that lights our dining table - from 12vmonster
There are lots of “LED Corn Bulbs” available on ebay at low prices, but these have repeatedly proven to be very short lived. 
These "corn bulbs" from ebay don't last long - sometimes only minutes
Another thing we have done for lights is to make up lights using cheap little constant-current constant-voltage (CC CV) power supplies from ebay. These cost only $2 - 3 each, take an input voltage of up to 35V, and can be adjusted to give an output from 1V to 30V. I use them to run some of our old 12V lights, and they are also very good for running LEDs.

DC fridge and freezer

We use a DC fridge and a DC freezer. These are very efficient. They also allow us to leave our inverter turned off at night, or when we go away from home for a few days, and if we had an inverter breakdown they could keep running. I’ll write a separate post about the tricks of running them on the nickel-iron battery.


If you're planning an off-grid power system:
  • learn to be aware of your solar supply and how full your batteries are
  • sometimes it's sunny, sometimes it's not: match your loads to your energy supply
  • installing and replacing batteries is a major, ongoing cost
  • expect everything to break down some time
  • DC lights and fridges can make your system more resilient to breakdowns and cloudy weather
  • your expectations are your greatest challenge to living off-grid!
Here are some key points if you’re considering a Nife battery:
  • you’ll need to set unusually high regulator settings - check that your regulator can go high enough
  • your inverter will need to work at higher than usual voltages
  • or you’ll need to drop down to 19 cells (in a 24V system)
  • you’ll need to top up with distilled/demineralised water
  • you’ll need to plan on replacing the electrolyte every 7 - 10 years
  • the best thing is you don't have to worry about damaging them with deep discharges
NIckel-iron batteries are very different from usual off-grid solar batteries. They won’t suit someone who isn’t engaged in their household systems - but I reckon learning how to understand and look after the systems you (and your neighbours) depend on is a key strategy for getting on in this world.
I’m very happy with our new nickel-iron batteries. I’m expecting them to be reliable and long-lasting, as long as I keep up the distilled water to them and can replace the electrolyte when needed. I think they’ll be an asset to a low-cost, resilient household. 

Extra info and links

Edison battery

This website: has some information about nickel-iron batteries, including a historic brochure from the Edison Storage Battery Company, which can be downloaded directly from this link:

Damn the matrix

My friend Mike Stasse was the first (and I think only) person I personally know to get a nickel-iron battery for his house. His blog posts about his experience are worth a look:

Living Energy Farm

These people use radical but well tried and tested systems for managing energy:

Tuesday, 25 February 2020

Baumr SX45 (Chinese 4500 45cc chainsaw) Review

An extremely cheap small-medium sized chainsaw manufactured with a sound design, but poor quality control. If you get a good quality bar and chain for it and have some chainsaw mechanic skills, this could be an economical, capable, resilient and repairable saw.


I’ve been testing the Baumr SX45 as a chainsaw to use for light felling and limbing, mostly while thinning forest. So far I like this saw: light enough, powerful enough, reliable, easy to repair. It’s a generic Chinese product with cheap and easy access to generic spare parts.
The main problem with this model (sold with the Baumr brand) is that the chain is really way thinner gauge than the bar so it won’t work very well for long. I also think the bar is way too long. The chain and bar problem is, however, easy and pretty cheap to repair. You can expect to have other quality control problems to fix when you buy these saws.

Bush repair

The last few months of drought and terrible bushfires should be teaching white Australians that we’ve had the wrong relationship with the Australian bush. For 2 centuries we’ve turned our back on the sophisticated Aboriginal culture of forest management with cool fires, and now most of the country is a real mess. Many forests are now so thick and flammable that they can’t be simply burnt back into good condition: they need a lot of work, cutting and dealing with the fuel, before they can be safely treated with a cool fire.
Currently, the Australian way to manage the bush depends on big machinery: pushing up piles, smashing soil and plants, burning waste in huge, hot windrow fires. I don’t think this is the way to fix the forest - but wait and see how much pressure comes, after these bushfires, to send the bulldozers around every house, town or road that has a risk of fire.
In forests around human settlement, we can step down fuel loads with selective cutting of undergrowth and thinning of trees. This produces lots of fuel that needs careful disposal, so as not to damage living trees or soil with hot fire - hopefully some fuel can be burnt productively, like heating houses, cooking food, making charcoal. Then the forest might be ready to re-start cool burns and re-establish the more open, less flammable forests of Aboriginal times. 

Here's a patch of badly overgrown native forest near Stanthorpe, Qld. The old stringybark is crowded by 60 year old black cypress trees. In a fire, the understorey would burn so hot it would incinerate the older stringybarks.
Here's a nearby forest area that has been thinned by chainsaw and then given a cool burn. A new ground cover is starting to grow.

Small saws for forest thinning

Doing this sort of thinning work over the last few years, has introduced me to small chainsaws. Previously I used bigger saws, when I was focussed on milling larger trees for timber. Now, spending many hours thinning, I have a new interest in small saws, light for lifting up and down all day, just powerful enough for the small wood I’m cutting.

The economic thinning saw

My favourite thinning saws have been the Stihl 024: 42cc engine, running .325” chain. These are light, reliable and repairable, with easy access to non-genuine parts (here's my blog post about them). However, small professional saws are very expensive new (the Stihl MS261, the smallest professional Stihl sold in Australia, costs $1500), and not available as a non-genuine copy or kit (some bigger Stihls are - see my post about making a kit chainsaw from Even a suitably sized Stihl consumer saw like the MS231 costs $750 - a good quality unit that will work well straight out of the shop, but not cheap or easy to repair (few non-genuine parts). For an explanation of the difference between professional and consumer chainsaws see my post about the Huztl MS360 kit.
Minimising what we spend to do what we need to do is important: minimum spend is minimum carbon emissions, minimum need to earn money and maximum time to do what needs to be done.
So I’ve been looking at what are available as cheap but good enough thinning saws - good enough means reasonable to use, reliable and cheaply and easily repairable. The Baumr SX45, from ebay, cost AU$91 delivered. I wanted to find out what they’re like, and if they could make a good thinning saw, so I bought one (and then another!).

The generic Chinese chainsaw

Here's the SX45 with a 13" GB bar. Note the metal bar oil tank below the muffler, with a slightly corrugated surface - this is one of the characteristic marks of the Chinese 45cc chainsaw

The Baumr SX45 is a 45cc petrol chainsaw, based on a standard Chinese model that is marketed under many brands. In Australia brands include Baumr, Giantz, Rockwell, etc., and there are a range of other brand names used overseas for the same basic saw. The well-known Chinese chainsaw parts online shop sells this saw as a Joncutter G4500. There are various minor variations in the cladding of the saw under different brands, but the essential parts appear to remain the same.

Is the SX45 a copy?

I expected, when I bought the SX45, that it would be a copy of a big brand saw - probably a Husqvarna (the Baumr SX82 appears to be a close copy of the Husqvarna 372). It does have some Husqvarna design characteristics (like having the carby on the same chassis as the motor), but as far as I can see it isn’t a close copy and doesn’t have interchangeable parts with any particular big brand saw. For example, the piston diameter of 43mm is not shared with any major brand saw I can find, and the clutch drum doesn’t match any saw I can find either (internal drum diameter of 68mm, bore 15mm).

The professional saw

Roughly speaking, chainsaw designs are divided between professional and consumer. Professional saws have a metal (I hear it’s a magnesium alloy) crankcase coming in 2 halves with a vertical joint in the middle. The metal crankcase casting also forms a chassis for the saw and includes the bar oil tank. The chain bar studs are threaded into the metal chassis, and the cylinder is bolted down onto the crankcase. This design is strong and easy to repair and is used by many manufacturers for their professional saws. Big brand manufacturers like Stihl and Husqvarna sell professional saws for 50% to 100% higher prices than their consumer saws, which is worth it to professional users.
Consumer saws are quite different: they have a plastic frame with a metal engine inside, using only just enough cast metal to form a cylinder and crankcase. The chain bar studs are screwed into a plastic moulding. These saws work well but as they age they’re less durable and more difficult to repair.
Despite its low price, the SX45 has a professional type of design, with a metal chassis and easily unbolted cylinder. I’m not saying it’s a “professional saw”, but the design gave me some hope it’s a unit that could be maintained and repaired for a long and productive life.

Spare parts

Machines wear and break. Any tool I depend on needs to be repairable, which includes having access to parts at reasonable cost.
Parts for this saw are easily available from several sources including Aliexpress, Ebay and Huztl. They are often listed with titles including “4500 chainsaw” or “Chinese 45cc chainsaw”. All the usual parts are cheaply available, including clutch drums, clutches, ignition modules, cylinder and piston kits, carburettors, etc..
Huztl stocks some parts, including outer panels, clutch cover, etc., for their Joncutter 4500 which I expect would fit the SX45, but strangely not the clutch drum - one of the most frequently changed parts on a chainsaw.
This easy availability of spare parts indicates that there are a lot of people - somewhere - repairing these saws. That’s a good sign for its essential quality as a machine.

First impressions

I was pleased with the saw, on arrival. It seems strongly built and I like the metal chassis. The saw started easily after fuelling and oiling, and was tuned slightly rich as a new saw should be - this increases lubrication of the piston and cylinder and reduces the top speed while running the saw in. After a few tanks of fuel the carby can be leaned out slightly, to get a little more power and make a bit less smoke.
With full tanks and a 13” bar with chain, the SX45 weighs 6.5kg. This is 5% heavier than a Stihl 024 at 6.2kg, but the SX45 also has an engine with just over 5% more cc’s.
The rewind starter is a bit strange: it seems to have 2 springs. The action of pulling the starter cord does not directly turn the crankshaft: a second spring seems to take up the tension, then crank over the motor with a delayed effect. This is a bit annoying to me - you can’t really feel what the compression is like. Some new Stihl consumer saws have a similar mechanism, maybe it compensates for the ongoing decline in consumer arm strength. I’m sure I can modify it if it breaks or annoys me too much.


If you’re going to buy cheap saws like this, you can expect to have some problems to solve - a bit like if you buy a 2nd hand saw, but better because you have a warranty if the problem is bad enough. I accept that if I buy a cheap saw or a kit saw, I’ll need to give them some attention.
There are 2 types of quality problems with these saws. Firstly there are the predictable and consistent ones from bad planning or design: the main one is
the bar gauge (a problem I expected and which I expect all these saws to have). Secondly are the random problems of poor quality control, for example I had an off-centre clutch drum (a real surprise, but I expect it is rare).
All the problems I’ve had so far were easily remedied.

Chain bar gauge problem

The few Baumr chainsaws I’ve seen share a problem common with many Chinese chainsaw bars: the bar groove is way too wide for the chain. This is a design/planning problem, that I think all these saws can be expected to have, even though it could be solved without being more expensive.
The SX45 is sold with a chain that is 1.5mm (.058”) gauge (the thickness of the drive links that slide along in the bar groove). However the groove in the bar is way too wide for this, making the supplied chain very sloppy in the supplied bar. This would probably work okay for a while, but would soon enough give trouble as the chain leans to one side and the bar has trouble following in the cut.
There are two effective solutions to this bar guage problem. The best solution for most people would be to get another good quality bar with matching chain - I give more detail at the bottom of this post.
The cheap and more troublesome solution is to install a chain with a bigger gauge into the supplied bar: a 1.6mm (.063”) gauge chain fits easily. After the paint was worn out of the groove by the chain (in the first session of cutting) the 1.6mm chain was loose and the bar groove needed closing. This isn’t hard to do: have a look at my video on hammering and dressing chainsaw bars:
Periodic hammering and dressing is part of using cheap Chinese bars - get used to doing it if you don’t want to buy expensive bars (which will eventually need the same treatment if you want them to have a long life).

Bar length

Here's the SX45 with supplied bar (soon to be cut short)
The SX45 is supplied with an 18” bar, carrying .325” pitch chain. I reckon 18” is way too long: a saw this size can’t comfortably pull a chain through 18” of wood.
For my thinning saws I like a 13” bar which carries a 56 drive link (DL) chain (.325” pitch). A short bar gives the saw more power (less bar friction), makes it lighter and safer and the shorter chain gets better lubrication. 

One of my usual 13" bars above, the SX45 supplied bar below
I cut down the supplied 18” bar, and made it into a 13” bar (see my post about how to shorten bars here) which fits the 56DL, 1.6mm gauge chains I use on my 024s. The new, short bar is great.

Little quality problems

There have been a few little problems with the SX45 that I’ve fixed, but not everyone might have them. Here they are.

Off-centre clutch drum

I was pleased to see that the clutch drum supplied had a rim sprocket. This is the “professional” type of clutch drum, where the part that drives the chain - the rim sprocket - can easily be replaced without wasting the whole drum. 
See the rim sprocket in there behind the clutch?
Here is the rim sprocket on top of the matching clutch drum
However, when tensioning the chain, I found that the chain would go alternately tight and loose as the chain was pulled around. After a bit of investigation, I found that the bearing hole in the clutch drum was not centred with the splined surface, it was out by about 0.5mm. This was enough to make proper chain tensioning impossible, and promised to result in bad bar wear over time (have a look at my video of it here). Edisons (who sell the saw) was quick to replace the clutch drum (but sadly with a spur/star sprocket which doesn’t take replaceable rim sprockets). I found suitable rim sprockets on Aliexpress and ordered one in for later.
To change sprockets on the SX45, you need to remove the centrifugal clutch - this saw has an “outboard” clutch, with the sprocket behind the clutch and clutch drum (modern Stihls and many Huskies have an inboard clutch so you can change sprockets without removing the clutch). Removing the clutch is done by spinning it clockwise off the crankshaft (it is threaded on with a left hand thread). To do this you need some sort of tool to engage with the clutch and turn it. Suitable tools are easily available on Ebay, Aliexpress or Huztl (here's a link to the Huztl clutch tool). I usually drive off clutches using an impact driver. 

 Here's a video on how to remove the clutch

This is a clutch removal tool sold by Huztl for some Husqvarna models. It works fine for the SX45, driven by an impact driver and a 3/8" socket adapter that fits straight into the back of it.
Here's a tool sold on Aliexpress specially for removing Chinese chainsaw clutches. It's made for manual clutch removal where you screw the plug on the right into the sparkplug hole to lock the motor, and turn the 3-pronged tool with a spanner. I haven't tried it yet but I expect the 3-prong tool would be easily driven by an impact driver and a socket, thus avoiding the need to lock the piston with the plug.
Here's a clutch removing tool I made from an M12 connector nut with some pieces of spring steel welded on. It was easy to drive with an impact driver and 19mm socket

These are sold on ebay as "socket adapters" and are great for removing chainsaw clutches along with a suitable socket or clutch tool

Air filter joint problem

The air filter has an O ring that tries to seal it to the carby, but I found this let in a lot of sawdust, which you could find stuck to the rear inside face of the carby throat. That’s not good: in time, especially if working in dry, dusty conditions, this would risk causing wear to the engine.
I tried adding an extra O ring, but that didn’t help. What did help was to cast a gasket in the filter: I carefully cleaned the filter in the sealing area with isopropyl alcohol and degreaser, dried it, then squeezed in a layer of polyurethane sealant (I would have used silicone if I had some). I then sprayed a light layer of oil on the carby (to stop the polyurethane from bonding to the carby), pressed the filter into place and tightened down the filter cover. Since then I’ve had no more dust getting in.

Oiler vent problem

On the second SX45 I bought, I had trouble with the bar oil flow. It didn’t oil well at first - it seemed to have trouble priming the pump. I tested it the usual way by revving the motor while pointing the bar at a suitable surface and watching for oil to spray onto it. Loosening and retightening the oil tank cap seemed to provide a momentary flow of oil, that soon dried up. This suggested a blocked oil tank vent: if air can’t get into the tank, the oil has trouble getting out. Same happens with chainsaw fuel tanks. 
This is the oil pump chamber of the SX45 - note the bar studs on the right. The red arrow points at the little rubber duck bill oil tank air vent - it lets air in when the oil is pumped out. There is usually a foam plug covering the duck bill.
It took me a while to find the vent, hidden in behind the oil pump cover, behind the clutch.
The vent was behind a little foam plastic plug (and a random piston pin clip apparently stuck in the hole on the assembly line). It was a little rubber “duck bill”, that wasn’t able to quack - the end was blocked. I cut a sliver off the end of the duck bill and pushed it back into place. After that, the saw oiled fine.

Bar pattern

The chain bar on this saw uses the Husqvarna small chainsaw bar mount pattern: the slot for the bar studs, the holes for the chain tensioner, and the oil feed holes match the pattern in smaller Husqvarna saws. This is important to know if you want to buy a good quality bar for this saw (like the GB bar).
Here’s a link to a catalogue of chainsaw bar mount patterns:

Opening the engine

Here's the cylinder and carby of the SX45
After giving the saw a first run to show that it worked, I opened it up and removed the engine cylinder. I wanted to: 1. have a look, 2. measure the bore and stroke (to calculate if it really was 45cc, and know what parts are compatible) and 3. check and perhaps modify the piston pin clips.

The SX45 chainsaw engine is really very easy to repair. It took a few minutes to remove the top cowling, unscrew the carby, unbolt the cylinder, and get down to the piston. 
I measured the bore and stroke:
Cylinder dia:    43mm
Stroke:        31mm
Piston pin diameter: 11mm
This works out at 45 cubic centimetres (some of these cheap saws don’t have the cc’s they are sold as, so it was interesting to check).
The piston pin clips are a very common source of failure on cheap chainsaws. These are the spring wire clips that hold the piston pin in the middle of the piston. Sometimes they aren’t well installed, and fall out when the saw is nearly new - this happened to a neighbour’s Baumr SX82. 

This is what the piston pin clips look like in the SX45
This is what the clips look like after the centre tail is cut off
Sometimes piston pin clips give trouble after many hours of operation. Chinese chainsaw parts and saws seem to use piston pin clips with a central tail, making the clip look like an “e”, whereas genuine Stihl clips are simply shaped like a “c”, without the tail. The central tail makes installation and removal much easier than it is with the “c” shaped clips (you can grab the tail with pliers), but it seems this tail can waggle with the motion of the piston (going up and down at over 100 times per second) and eventually fatigue and break. On the chainsaw forums I’ve read a few reports of kit saws whose piston pin clips have had a piece break off and jam between the piston and cylinder after quite a few hours operation - a pretty major failure.
I now use the technique of other people building Chinese chainsaw kits, and cut the tails off my piston pin clips (I use a cutting disc in an angle grinder). Any time I’m opening up a chainsaw cylinder, I also check that the clips are tight in their grooves. The SX45 had “e” shaped clips so I took them out and cut off their tails.
Re-assembling the SX45 was quick and easy. A generous taper on the bottom of the cylinder skirt makes it easy to insert piston and rings (make sure the rings are aligned with their pins).

What to do to an SX45

These are my recommendations for upgrading an SX45 into a long-lasting, hard working chainsaw:
- Put a reasonable bar and chain on the saw - the setup supplied will be too sloppy to work for long (the bar groove is way too wide for the chain). I recommend a 13” bar. See below for more details. 
- Cast a silicone or polyurethane gasket to seal the air filter to the carby
- Check the carby tuning. The second SX45 I bought was tuned a bit lean on the high speed jet (adjusted with the H screw). Lean tuning like this risks burning the engine. See carby tuning instructions at the end of this post.

- Buy a 4mm T-handle allen key for servicing the saw - a small L-shaped allen key is supplied but is clumsy to use.
- Buy or make a clutch removal tool so you can remove the clutch to change sprockets or service the oil pump.
- If you’re game: open the engine, cut off the tails on the piston pin clips and check they’re well seated.
- be prepared to replace the fuel hoses when they perish. 

Using a 13" GB bar

I bought a good quality (Australian made!) GB bar from (13” bar to fit Husqvarna, .325” pitch, 1.5mm (.058") gauge, just over $50 with a chain). I was a bit surprised to find that this bar was a little wide gauged (the slot for the chain was a little wide) from new, and fit a 1.6mm (.063") gauge chain. I am now running 1.6mm gauge chains in it.
I don't know if all their bars are like this, but I recommend you try a 1.6mm gauge chain in the bar, perhaps after running with a 1.5mm chain for a while, and perhaps plan on using 1.6mm gauge chains in the longer term. 
FYI as a rule, Stihl chainsaws with .325", 3/8" and .404" pitch use bars and chains with a 1.6mm gauge (.063" in American language), and Husqvarna uses 1.5mm (.058"). If a 1.5mm gauge bar is worn, it can be good to use 1.6mm gauge chains, but most bars will eventually need hammering and dressing.
With all chainsaws, it's best to start with 3 chains, and rotate them as they wear out. When all the chains' teeth are sharpened away, get 3 new chains and change the drive sprocket on the chainsaw.
13" bars for this saw take a chain with 56 drive links (56 DL). If you do a bit of chainsawing, I recommend learning how to punch and rivet chains yourself, so you can shorten or lengthen chains, as well as being able to buy a roll and make up loops cheaply from new.

Tuning the carburettor

Here's what the manual has to say about tuning the carby. Pretty good instructions really:

The adjustment screws are generally located on the starter cord side of the unit. The high speed and low speed mixture screws are generall marked "L" and "H", respectively. In the adjacent image, idle speed is (A), low speed mixture is (B) and high speed mixture is (C).

Factory Settings

Each adjustment screw has a general "factory setting", which is measured from the adjustment screw being rotated "IN" (right / clockwise) until fully seated (but not tight). From this point, the setting is made by counting the number of full rotations of the screw "OUT" (rotate left / anti-clockwise). Factory settings are:
    •    Idle Speed – Approximately 5 turns out.
    •    Low Speed Mixture – Approximately 2.5 
turn out.
    •    High Speed Mixture – Approximately 2 turns out. 
Use the factory settings as the basis for tuning. Set all adjustment screws to factory settings, then test the engine before further tuning. Use a suitable flat-blade screwdriver and ensure that the screwdriver is properly engaged with the adjustment screw before rotating.


    1.    Start and allow the engine to idle uintil it is warmed up – tuning a cold engine will result in rich running when the engine is warm. If the engine does not idle, use the throttle to keep the engine running ("blip" the throttle; do not run the engine continuously at high speed).
    2.    Adjust Idle Speed – Rotate the adjustment screw one quarter (1⁄4) turn at a time – rotate "IN" (right / clockwise) to increase idle speed; rotate "OUT" (left / anti-clockwise) to reduce idle speed. Set the speed so the engine idles as fast as possible without engaging the clutch. Never set the idle so the saw chain rotates when the throttle is released. If the engine will not idle, adjust the low speed mixture (step 3).
    3.    Adjust Low Speed Mixture – Rotate the adjustment screw one quarter (1⁄4) turn at a time – rotate "IN" (right / clockwise) to lean the low speed mixture; rotate "OUT" (left / anti-clockwise) to richen low speed mixture. Rotate the screw IN until the engine begins surging or wants to stop – this is the lean adjustment position. Make a note of the number of rotations of the screw to reach the lean adjustment position. Then, rotate the adjustment screw OUT – the engine should start running better. Keep rotating the screw OUT until the engine starts to load up – this is the rich adjustment position. Make a note of the number of rotations of the screw to reach the rich adjustment position and compare it to the lean adjustment position. Then, rotate the screw IN to a position where the engine idles best – it should be about midway between the rich and lean position settings. At this point, you may have to re-adjust idle speed (step 2).
    4.    Adjust High Speed Mixture – Rotate the adjustment screw one quarter (1⁄4) turn at a time – rotate "IN" (right / clockwise) to lean the high speed mixture; rotate "OUT" (left / anti-clockwise) to richen high speed mixture. Rotate the screw OUT until the engine begins slowing and running roughly at full throttle. Then, rotate the adjustment screw IN – the engine should start running better. Keep rotating the screw IN until the engine reaches maximum speed. Then, rotate the screw OUT one eigth (1/8) to one quarter (1⁄4) of a turn to richen the air/fuel mixture for engine cooling purposes. 


So far my only breakdown of an SX45 motor has been very simple to repair: mud wasp nests in the muffler. The 2 round pipe exits for the muffler suit mud wasps very well. 
Cooked mud wasp nests in both muffler exits - easily remedied with a screwdriver