Intermittent electrical power loss

Having said all the above, I feel that you have a wiring problem, rather than a regulator one. After all, the regulator shouldn't cause total loss of electrics.

Having said all the above, I feel that you have a wiring problem, rather than a regulator one. After all, the regulator shouldn't cause total loss of electrics.

The bloke at the Jab factory suggested that a bad connection on the regulator's yellow reference wire could confuse the regulator into allowing the battery to cop a 17v spike. Perhaps this causes the battery's internal BMS to disconnect until the charge voltage stabilises. I put this to the LiFePO4 battery shop and they claim the battery wouldn't disconnect.

I hope to make one more test flight to induce the failure mode. I have set up a separate voltmeter so I can monitor battery voltage to see if that's where the power cut is coming from.

Despite assurances from the seller that this battery is a drop-in replacement for lead acid batteries, it seems that continuous charging after it's topped up may cause long term damage.

I believe the Jab generator has permanent magnets. Does that mean it doesn't need current from the battery to set up a charging field?

If I disconnect the battery when it becomes fully charged, will that damage the AC generator?

-

As a adjunct to this thread...lithium batteries in aircraft. One of my friends who is a genius guru..he used to make the satellite television transmitters on the satellites that give you rural folk your TV signals. He is also a pilot and being retired and a guru he has made a new regulator that can be used on lead acid and lithium batteries designed for the crap Rotax generator system of 40v spikes. The proto so far looks excellent. We have run it on my test rig I use for the replacement CDI modules I make for the Rotax engines. My current test rig cant get up to cruise RPM as I currently only use it for start rpm firing measurement and advance and retard rpm measurement so it doesnt need to go to cruise at the moment as I have a different test rig that runs the modules at cruise rpm to test them but next week I have a new servo coming to drive the rig and it will be able to drive the rotax generator to full speed and we can load it with different batteries on his new regulator. The pulses coming in are wicked in frequency at full power rpm and we can then load the generator up to the full 18 amps and see how the new regulator performs. If all goes to plan and I am sure it will most likely we will be making them as a replacement. The regulator design is very different and will protect the battery from over voltage and overcharging

Old K - I'm not sure where to start - but I'll start on the Jab alternator.

1. Every engine-driven alternator produces an AC current, which is rectified into DC by about half a dozen diodes, and the charging current is then controlled by a voltage regulator.

Every automotive/aviation alternator ever made has a VR - it is sometimes mounted on the outside of the rear of the alternator, other times it is mounted internally.

Current-design alternators often have the brushes and VR combined into one unit. You remove the VR and the brush assembly comes out with it.

2. The Jab alternator does use permanent magnets, so it does not need the exciter wire from the battery for electromagnets, which is the normal design for nearly all other automotive alternators.

3. You must never totally disconnect an alternator whilst the engine is running. To do so will create a massive voltage spike which will fry the alternator, and possibly, many other electrical components in the system.

The Jab community website discussion (below) gives some information on the Jab alternator setup. Their alternator regulating and controlling system is different to the standard automotive arrangements.

The use of a crowbar module (usually a box of combined electrical trickery) is not common, but I can see what they're trying to achieve. The Wiki mob describe crowbar circuitry devices quite well.

https://jabiru.net.au/community/engines/voltage-regulator-functionality-and-connections/

https://en.wikipedia.org/wiki/Crowbar_(circuit)

Onto the LiFePo4 battery.

1. It's a lie that all LiFePo4 batteries can be a total drop-in replacement for L-A batteries. This has caused a lot of grief to a lot of people, particularly because of the highly specific charging requirements of LiFePo4 batteries.

2. A standard automotive/aviation alternator and regulator arrangement is inadequate for charging LiFePo4 batteries. 

3. A 12V L-A battery has 6 cells at approximately 2V per cell. In practice, the L-A cells run around 2.1 to 2.2V. They need a fast initial charge, then a float charge - and they can cope with being overcharged fairly constantly.

At the worst, an L-A battery receiving a constant overcharge will boil the acid solution in the cells. At the very worst, the cells will boil dry, and you will end up with a buggered L-A battery. It takes a lot to do that, they're a pretty durable and forgiving device.

L-A batteries like being charged at around 13.8 to 14.2V. Alternators usually max out at around 14.5V. You can charge at a higher voltage, but anything over about 14.7V creates gassing and heat in an L-A battery.

Alternator regulators are set up to charge L-A batteries - to provide an initial fast charge, then reduce the charging rate to a float charge.

The L-A battery provides the "cushion" in an electrical system - taking current from the alternator, even when charged, to ensure the alternator doesn't blow, with a complete current draw shutoff, and a severe voltage spike.

4. A 12V LiFePo4 battery has 4 cells at 3.2V each. It needs a charging rate of around 14V to 16V, and it is more tolerant of higher charging voltage than an L-A battery.

5. An LiFePo4 cell will be damaged if the voltage over the cell falls to less than 2.5V, and damaged if the voltage over the cell exceeds 4.2V.

Here's the important bit -

6. The cells of an LiFePo4 battery do not auto-balance at the end of the charge cycle. The cells in an LiFePo4 battery are not 100% identical. Therefore, when cycled, some cells will be fully charged or discharged earlier than others.

The differences in charge between cells will increase if the cells are not balanced or equalized from time to time.

In an L-A battery a small current will continue to flow even after one or more cells are fully charged (the main effect of this overcharging current is decomposition of water into hydrogen and oxygen).

This current helps to fully charge other cells that are lagging behind, thus equalizing the charge state of all L-A cells.

The current which flows through a fully-charged LiFePo4 cell however, is virtually zero, and lagging cells will therefore not be fully charged.

Over time, the differences between cells may become so extreme that, even though the overall battery voltage is within limits, some cells will be destroyed due to over-voltage, or under-voltage.

An LiFePo4 battery therefore must be protected by a BMS that actively balances the individual cells and prevents under-voltage and over-voltage.

The type of BMS fitted to LiFePo4 batteries can vary widely, according to manufacturer and according to how the manufacturer thinks the battery will be used. The BMS of these batteries can also vary widely, according to how cheap the manufacturer is.

Most of these BMS use dual MOSFET's to switch the charging current according to the cell demand/requirements.

The MOSFET's are coupled with an Integrated Chip with additional electronic control devices built into the chip to prevent voltage spikes, indulge in cell balancing, and control other electrical parameters. We're getting into high-end electronics here.

Suffice to say that the simple basic engine alternator and regulator is not up to the job of charging LiFePo4 batteries - a dedicated LiFePo4 charger really is required to match the electronics and requirements of the LiFePo4 battery.

As regards your intermittent and transient power loss - I'm struggling to envisage exactly what is causing it, but I suspect it's related to the peculiar Jab alternator charging arrangement, and the crowbar module - and the LiFePo4 battery possibly causing a sudden and major drop in amperage draw, as it rapidly reaches full charge - thus causing a voltage spike, and activating the crowbar module, and making it shut off the electrical power.

I'm not sure if the Jab crowbar module stays shut off, or reconnects the power once the voltage spike ceases. One would imagine the Jab electrical designers provided the security of the latter.

Without an inspection of the aircrafts complete wiring and charging circuitry devices, and what they comprise exactly, and how they operate - by someone highly skilled in electronics, I think you will be shooting in the dark.

The "crowbar" is simply like a switch in between the alternator supply and the battery. (There are two types of scr regulator, the most common is series type)

When the regulator senses 14.2volts at the battery, it open circuits the path.

In that way, it functions as a simple "overvoltage cutout". When this happens, your electrics are expected to run off the battery.

The big electrolytic capacitor (as recommended by Rotax, not sure if Jab have), is important to smooth out the spikes that the regulator produces as it transitions from "switch closed" (direct connection between alternator and battery), and "switch open", (fully charged). That phase is characterised by continuous short spikes of full charge voltage.

None of these circuits I have seen are capable of providing any charge unless the battery voltage falls below a threshold. Usually internally fixed at about 14.2 volts.

As a adjunct to this thread...lithium batteries in aircraft. One of my friends who is a genius guru..he used to make the satellite television transmitters on the satellites that give you rural folk your TV signals. He is also a pilot and being retired and a guru he has made a new regulator that can be used on lead acid and lithium batteries designed for the crap Rotax generator system of 40v spikes... If all goes to plan and I am sure it will most likely we will be making them as a replacement. The regulator design is very different and will protect the battery from over voltage and overcharging

Old K - I'm not sure where to start - but I'll start on the Jab alternator.

...I'm not sure if the Jab crowbar module stays shut off, or reconnects the power once the voltage spike ceases. One would imagine the Jab electrical designers provided the security of the latter.

 

Without an inspection of the aircrafts complete wiring and charging circuitry devices, and what they comprise exactly, and how they operate - by someone highly skilled in electronics, I think you will be shooting in the dark.

Home-builders like me, with limited expertise, are trying to dovetail hardware from different sources. It's a pity we can't get full and frank advice from manufacturers and suppliers - but I understand why they need to be cautious in giving out advice.

In the longer term, I hope a charge-management unit like Mark Kyle describes becomes available. With so many Li batteries being fitted to homebuilts, bikes, cars, RVs, boats, etc. there sure will be a market for it

In the shorter term perhaps I can fit more position lights to absorb some of the alternator's output that would otherwise be overcharging the Li battery. A simpler option is to put my old SLA battery back in.

Another, messier, option is to fit the lightest SLA I can find upstream of the Li battery, to absorb the charge coming from the system and act as a "buffer" to protect it. Any thoughts?

While any terminals should ideally be 'bright' and certainly free of oxidation, cutting them back with a coarse abrasive is not always the best policy:

First, if the terminal has an anti-oxidation plating, you may well cut that away.

Second, if the resulting surface is scored, you are actually reducing the surface area that is in contact.

Grease/petroleum jelly to reduce oxidation is an excellent idea.

And with a wooden aircraft, I would be attaching all my terminals to a piece of metal, then attaching that to the aircraft. That is, I would not be relying on a screw, or bolt, that passed through wood, to maintain pressure on the terminals.

Old K - Your suggestion of a lightweight SLA battery upstream of the LiFePo4 battery is definitely one way to offer some protection for the charging system and avoid voltage spikes.

However, to me, it appears the best idea is to go back to the original SLA battery, and dispense with the LiFePo4 battery, until you can figure out an onboard dedicated charging arrangement for the LiFePo4 battery.

 I don't think you can put a battery such as yours in parallel with a lead acid one. I'd stick with AGM. Safest thing in a plane and a crappy permanent magnet alternator is hardly sophisticated enough. The conventional alternator is easy to regulate with diodes and most self excite  from residual magnetism, while some don't and you can't get them on line if the battery is dead flat.. Nev

I made a suggestion, but then decided it was not going to work, now I can't delete this message (sigh...)

There's nothing wrong with "crappy" permanent magnet alternators.

Unfortunately at this stage there seems to be nobody making a good regulator for them.

Thanks everyone for the advice. After lots of phone calls I have a Power Mate regulator on order. Based on advice from other Jab flyers who have used them with Li batteries, it should be more gentle on my system.

I also have some terminal blocks to tidy up all my connections.

I plan to give the new installation a good workout next weekend, starting with Rylestone Fly in.

Let us know about how the Powermate performs Old K. I too have a LiFe setup and switch the charging off as soon as the voltage gets to 14.2 volts for the 13.2 volt hobbyking battery. The battery is then isolated electrically.

I would like to make this turnoff automatic but can't figure out how to do it without a residual drain. If I can't find a way to do this, a warning buzzer might be a good idea, because I am quite capable of forgetting to switch the charging off.

There are some big capacitors, and with the battery switched out, the alternator thinks that these are the battery. During charging, the capacitors are in parallel with the LiFe battery. This system has worked ok on the Jabiru for some years now, and the discarded odyssey battery works fine on a farm buggy. I just have the standard Jabiru regulator and it charges the capacitors to nearly 15 volts. A couple of times a year, I put the LiFe battery on a balance-charger.

Bruce I've been much influenced by your setup, but capacitors scare me. The Jab/Camit forum has been a great source of info. Several of them use Li batteries in their Jabs; one has been going well since 2013. That experience, and assurances from Lithiumax about their Battery Management System, have calmed my fears. The Power Mate regulator is apparently pretty gentle on batteries, although not intended for Lithiums.

It limits the charge to 14.2v and 8A which is well within the capabilities of the Lithiumax BMS.

I've replaced all my dodgy connections with terminal blocks and am now just waiting for the new regulator to arrive. I've cleared a spot for it around the corner from the firewall, so it gets lots of cooling air.

Update: I replaced all my dodgy connections with terminal blocks, and double-checked everything was working. The new Power Mate regulator didn't arrive until 0830 Friday. I rushed to the airport and installed it as per instructions. Packed away car, tools, loaded plane, topped up fuel, preflighted, closed hangar, strapped in, hit starter.

Silence. Much later, found what I thought was the problem and flew over the Range to Rylestone. Had a great time there and next morning a bunch of us flew west for a few days.

My voltage kept dropping until every electrical device had quit. The battery wasn't getting charged. Landing at Wings out West, we crashed Dan Crompton's L1 course. While Darren Barnfield took over, Dan put my Li battery on a charger. Meanwhile, david2ayo from this forum appeared and helped me bypass my old wiring. He knows this stuff!

The Power Mate kept the Li battery at 13.34v from then on and there hasn't been an electrical glitch since.

Thanks everyone for your interest and assistance!

Was there a particular wire that had gone open circuit?

Was there a particular wire that had gone open circuit?

At WoW, David2ayo helped me bypass a pair of wires that used to connect the previous regulator to the bus.

I will now remove them, because it all seems to be working well with connections direct to battery and earth.

Good that you seem to have a cure. It really is a good idea to have an idea of what you are doing before you start work, even if it is only as little of thinking of electrical systems in the same way as plumbing. Of course if you don't understand plumbing you are back to square zero.

Good that you seem to have a cure. It really is a good idea to have an idea of what you are doing before you start work, even if it is only as little of thinking of electrical systems in the same way as plumbing. Of course if you don't understand plumbing you are back to square zero.