The issues surrounding lithium batteries.

[Downunder]

As I was trying to imply, pilots and aircraft owners are very culturally risk adverse.

You are making statements AGAINST what aircraft manufacturers and aircraft engine manufacturers approve.

When I came into aviation, I was like you. I thought I would be getting into cutting edge technology when the exact opposite is true......carby fed engines in 2020? Who would have thought......

 

Lead acid and their derivatives have been working perfectly adequately for more than a few decades in aircraft.

Alot of aviation technology is based on a history of reliability over efficiency.......you will learn this.

Things change very very slowly....in aviation.

 

As was stated previously a fire at 8500 feet is potentially fatal.

You can't just walk away from it.....

 

Experimentation is great, but it needs to be acknowledged as such......

[spacesailor]

Question !.

What,s the difference between,

Iron ( fe ) and ion in batteries.

LiFePo 4 (lithium iron phosphate ).

spacesailor

[Bruce Tuncks]

Lead is heavy stuff. A heavy weight in a plane is dangerous. Once a lead-acid battery departed from a glider in extreme turbulence. It made a hole in the canopy just behind the pilot's head.

Yep, I am risk-averse and part of this is being weight-averse.

[Bruce Tuncks]

my understanding of the difference goes like this:

A big potential customer said that li- po's are fine but for the fire risk, so could they please make some similar batteries only with all non-flammable stuff.

LiFe batteries are the result.

[RFguy]

They're all "lithium Ion" but vary in their construction, and slight chemistry differences.

The cathodes in LiFePO4 are phosphate instead of Cobalt (LiCoO2), and the phosphate-oxygen chemical bond are tighter than the cobalt-oxygen bonds, which when the battery is damaged or stressed in some way, take your pick, means the oxygen is not liberated as easily. Also the Lithium migrates in the the LiCoO2 battery during charge and discharge which changes its stability. There are a few good writeups on this online.

 

In practical sense, the LiFePO4 battery of four cells or eight cells, closely approximates a 6 cell or 12 cell (12 or 24V ) lead acid battery. (14.4V top of charge at 25 deg C)

 

Charging is approx compatible with Lead acid batteries with the exception of :

 

1) "Boost or equalise" for lead acid must not be used. Lead acid batteries are typically deliberately overcharged to equalise the cells ! Lead -acid "smart chargers" might try and do this...

2) Temperature compensation if present (only in the good stuff with temperature sensors for the battery) should be set to zero

3) Ideally , the initial charge when battery has been over discharged (<2.7V per cell) should be limited to 0.05C until cell voltage reaches 2.7V per cell (10.8V for "12V" pack)

4) LiFEPO4 batteries can charge usually as fast as your can supply current. Charging a well discharged LiFEPO4 battery may stress the charging system, since it will likely to be have been design assumed for a lead acid battery with appreciable internal resistance. As for the aircraft alternator charging question, that requires some care and caution in order to determine if the alternator diode pack could over dissipate.

[Bruce Tuncks]

I didn't know about point 4. Should I arrange for a heavy-duty resistor in the charging circuit on my Jabiru? By this I mean something like a high-wattage 2 ohm resistor between the voltage regulator and the battery. I thought the voltage regulator would protect the alternator.

[Kyle Communications]

 

4) LiFEPO4 batteries can charge usually as fast as your can supply current. Charging a well discharged LiFEPO4 battery may stress the charging system, since it will likely to be have been design assumed for a lead acid battery with appreciable internal resistance. As for the aircraft alternator charging question, that requires some care and caution in order to determine if the alternator diode pack could over dissipate.

 

This is why our new design regulator is current and voltage limited, also the rectifier diodes are high speed rectifier diodes. As I said earlier the current Rotax regulator is only really for the older style batteries

[RFguy]

Yeah voltage regulators are just that. "voltage " regulators.

 

The cable between the Alternator diode pack output and the battery would limit the current, but not in a useful manner.

 

Kyle, be a little careful with high speed diodes, too fast and you will get undesirable ringing, possible VHF oscillation and potentially radio interference. Also, some of the high speed diodes will have undesirable (high) reverse leakage currents when really hot. It depends on the barrier construction. (yes I do this for a living...) . on the other hand, if I am telling you how to suck eggs, I apologise ! Happy to advise, comment or keep my mouth shut, which ever is most helpful.....

[Kyle Communications]

Comments and advice are always appreciated but not necessarily heeded or considered The current design is working extremely well and has been exhaustively load tested on a actual Rotax generator spinning from idle to 5000rpm. Everything is extremely stable so far.

[RFguy]

got it.

Load steps with poor condition batteries (read: high internal resistance) are the biggest trouble I have found. As long as the regulator does not overshoot too much in the control department should be OK. take a scope probe to the battery terminals, Kelvin connected. set it up to trigger on the load step you will apply (usually a transistor and load bank) , and observe what the voltage does during the load step. Oscillatory behaviour is problematic. Excessive overshoot also. Vary the series resistance (very small) with the battery, essentially place a resistance in series to simulate various battery conditions that result in higher internal resistance. At some unrealistic point, it will oscillate... The practical real world result is that because you cannot control the battery condition, the control loop needs to be slow enough not to oscillate into the worst case battery. Where the voltage is picked off for the regulator feedback matters, also. Fortunately, by the time the series resistance is high enough to be a problem for most designs, the battery wont crank the engine ! but batteries do fail, get boiled etc.

And that is probably a perfect example of old-tech works ! The regulator feedback loop is slow in old designs.

[Roscoe]

I didn't know about point 4. Should I arrange for a heavy-duty resistor in the charging circuit on my Jabiru? By this I mean something like a high-wattage 2 ohm resistor between the voltage regulator and the battery. I thought the voltage regulator would protect the alternator.

Hey Bruce, I am confused by all the technical talk going on in this thread.

My Jabiru J170 has the standard Odyssey PC625 Battery and Jab factory Voltage Regulator.

Any cause for concern?.....never had any issues!

[RFguy]

In answering that question, the answer is the extra resistance is not useful. This is because the amount of charge , or support the alternator could provide is proportional to the voltage difference between the alternator and the battery, divided by the resistance of the cable. Too much resistance and you cant charge or support the battery.

 

Kyle's idea of the voltage AND CURRENT limiting regulator is a good one.

 

I do like to see fuses in series with any battery, usually a separate charging fuse and load fuse,s (including a separate starter fuse) . But I do not know how conventional that is in aviation, I have not taken notice of this in enough aircraft.

 

In the case of the alternator/charger, ideally the voltage sense wires for the regulator should be on the battery itself, this is known as kelvin connected, so that the connectors and cable that are carrying high current do not affect the battery voltage reading. (since they would cause a voltage error equal to to the current multiplied by the resistance of the joint or cable) . The said voltage sense wires ALSO need to be fused, since they probably run off somewhere in the aircraft, and are liable to damage.

Glen.

 

I didn't know about point 4. Should I arrange for a heavy-duty resistor in the charging circuit on my Jabiru? By this I mean something like a high-wattage 2 ohm resistor between the voltage regulator and the battery. I thought the voltage regulator would protect the alternator.

[RFguy]

Hey Bruce, I am confused by all the technical talk going on in this thread.

My Jabiru J170 has the standard Odyssey PC625 Battery and Jab factory Voltage Regulator.

Any cause for concern?.....never had any issues!

Bruce ?

 

If that is what the manufacturer suggests.... anything else is experimental.

 

*** I see two immediate important problems changing to a LiFePO4 battery : ***

 

1) Changing to a lighter weight battery will of course change the weight distribution...

 

2) Starting cranking current. Airplane batteries are primarily starting batteries, which lead acid batteries do well at (unless they are cold !) .

 

I see that little Odyssey battery is rated at only 18 AH capacity (20h rate) and 540 crank amps for 5 seconds, 220A CCA . The battery is rated for more than 12x the ampere hour rating for CCA.

 

The LiFePO4 battery for the same physical weight will be rated at ~ 4x capacity of the lead acid. Although at these small capacities, the difference is more like 3x as case and hardware is a high percentage of construction.

 

Therefore for the same weight the LiFePO4 battery will have 18 x 3 = 54 AH capacity approx

 

Now, the starting current for a 1.5kW starter (according to Jab2200 info I found), will be approx 150 amps, based on 10V at the starter itself)

The LiFePO4 battery MUST be rated for that peak current.

Now, the CALB 60AH LiFePO4 batteries I use are rated for max current of 180A.

 

That sounds inside spec . However the stall current for the starter motor could be much much higher. Kyle ?

 

So, consideration must be with care ! . The standard lead acid battery for $200 seems a good idea.

----

On the upside, The LiFePO4 battery will have a capacity advantage for the same weight, this may be an advantage if the alternator stops making charge during flight , providing extended use of avionics . for the above example, most likely 4 to 5 x the non charge duration . That could be useful . The other useful thing is, that the LiFePO4 will be happy to provide that starter crank for long periods. Although the starter may overheat. Watch the spec on that. (Kyle ?) .

 

glen.

[Kyle Communications]

Output of the regulator at 11.5 amps load. The first picture is the AC output of the generator you can see how the power is used on each cycle

 

The second picture shows the current limiting working. The violet line is the current limiting clamping.

 

Current is monitored by a hall effect and used as the input for the foldback circuit

 

 

 

 

 

 

 

 

[Kyle Communications]

Most of the Lithium batteries seem to cope fine with the amount of start current due to their larger capacity the CCA spec of the batteries is also pretty good. The Lithiums batteries are a good fit for size and weight and also how much power they can deliver. The big issue is the charging from older style generator/regulator systems not taking into account the special conditions required for the new generation of batteries...its like the difference between lead acid and calcium batteries everyone thinks you can just put a calcium battery into a system that lead acids were in but in reality the charge conditions for the calcium are quite different

[RFguy]

Output of the regulator at 11.5 amps load. The first picture is the AC output of the generator you can see how the power is used on each cycle

 

Hi Kyle. Looking good. I gather the yellow trace is current sensor out and that's a step load change going into I-limit on the 2nd pic ? I notice the harmonic structure (waveshape) of the current sensor changes on the high load- is that a slew rate limit in the current sensor, or the behaviour of the alternator ?

 

I've use generally Allegro hall effect sensors in this sort of service, mainly to stop semiconductors vaporizing in fault condix where there is a bit of capacitance upstream.. IE keep the semiconductor under its SOAR curve. Hall sensor rise times of around 4uS mean it can work pretty fast...The Allegro sensors were pretty good on temperature stability.

 

cheers

[Old Koreelah]

*** I see two immediate important problems changing to a LiFePO4 battery : ***

1) Changing to a lighter weight battery will of course change the weight distribution...

Thanks for your excellent explanations, Glen, Mark and others.

The weight saving of a Li battery is significant but can lead to CoG problems. I had to move my wing back 18mm; when combined with the $400 cost of the battery, this made changing to LiFePO4 an expensive exercise.

 

Another issue is the lack of charge from the Jab alternator during warmup (no charge till aboout 1800); my voltage takes a hammering from repeated starts while doing ground testing. A couple of times I've had to recharge with a small 240v unit. Luckily it doesn't take long to top up the Li battery.

 

As previously stated, mine has a Powermate regulator and I've only done a few longer trips, so will watch this space...

Edited August 27, 2020 by Old Koreelah

[RFguy]

Mark,

why is this Alternator so lazy at low RPM ? , is this a general character of alternators designed for aviation ? I would have thought pulley sizes could have been made 'ideal' since the motors don't run over such a wide RPM compared to a petrol motor car engine. - glen

[Old Koreelah]

Mark,

why is this Alternator so lazy at low RPM ? , is this a general character of alternators designed for aviation ? I would have thought pulley sizes could have been made 'ideal' since the motors don't run over such a wide RPM compared to a petrol motor car engine. - glen

Glen if you're asking why the Jab alternator doesn't charge below about 1,800rpm, it's bolted to the flywheel, so no pulleys are involved and you can't "gear it up".

Ian Bent of CAMit sold lots of engines with a belt-driven car-type alternator.

[RFguy]

Hi Mark,

Thanks for the info. OK.... I gather because that RPM is 50% of the running RPM, that it is more of a lack of field current at low RPM, and that some sort of low RPM field boost might improve that ? IE simply to restore the field current provision, it wouldn't be increasing the max field current IE so no ill effects on the brushes, rotor windings etc.

 

glen

[Old Koreelah]

Hi Mark,

Thanks for the info. OK.... I gather because that RPM is 50% of the running RPM, that it is more of a lack of field current at low RPM, and that some sort of low RPM field boost might improve that ? IE simply to restore the field current provision, it wouldn't be increasing the max field current IE so no ill effects on the brushes, rotor windings etc.

 

glen

Glen this fairly simple mod was discussed not long ago.

https://www.recreationalflying.com/threads/jabiru-12-pole-alternator-modification-upgrade.19458/

Since my Li battery has solved most of my starting problems I haven't bothered to improve the alternator.

Edited August 27, 2020 by Old Koreelah

[Kyle Communications]

Glen

 

The waveform is from the Rotax generator. it is direct drive off the rear of the engine so whatever the engine rpm is thats is the generator rpm..there are no drive pulleys

 

The current sensor we are using the ACS770..I use them at work and they are bullet proof as you know. That waveform is direct from the generator. Thats how much of the waverform is switched at a load of 11.5 amps. We stabilize the voltage to try to maintain 14.2V no matter what speed or load the generator has. The current limit waveform you see there does have some hystersis in it of course but so far in testing under load with very low charged batteries and also full batteries the circuit is performing very well.

 

The generator and also the original regulator are basically motorbike style ones and we know that they are always crappy. This is why lead acid batteries make a better capacitor..these new batteries do not like what they are being served with from the originals. This is why i dont recommend any lithium batteries on a Rotax. the charging system is just not up to scratch for them

[RFguy]

Glen

....

The generator and also the original regulator are basically motorbike style ones and we know that they are always crappy. This is why lead acid batteries make a better capacitor..these new batteries do not like what they are being served with from the originals. This is why i dont recommend any lithium batteries on a Rotax. the charging system is just not up to scratch for them

Hi Mark. thanks for the time taken to explain. That is good advice and voice of experience for everyone. Yes ACS770 is many a favourite. I rarely deal with more than 40 amps on things I make at DC , IE no ventures into 2 and 4 ounce copper.... ! but plenty of high voltages high RF current.

 

Bruce, I have gone back and read a few of your posts, you've done a fair bit of work on this. Yeah those RC LiFePO4 (LFP) packs really pack a punch with all those parallel cells. If you charge them nicely (say at 1/5C or 1/10C the cells stay in pretty good balance.

 

Thanks for the link Old K on the 12 pole mod. Actually yesterday I did a bit of reading on what mfrs are doing - 3 phase IGBT synchronous rectifiers, dual windings in series/parallel configuration. Frankly I think a 3 phase sync rectifier with computer control is a bit complex compared to 6 or 8 diodes (8 for star connected and harmonic current sink) . But the dual winding alternator was interesting. rather like traction motors in trains, switching of the windings for speed or torque.

 

I think Mark's idea with the simple voltage/current limited regulator is a good one. And overcoming to low RPM issues would seem a better target than getting dissipation down in the diode pack.

One of the issues with low speed is the stator core saturation due to the low frequency. The output is very NON linear down the low end. It really needs a rotor designed specifically for low speed operation.

 

Tell me- in these direct drive alternator designs, what happens if a chunk of the rotor breaks and stops the rotor rotating ? I have NEVER seen it happen though in autpmotive alternators, even at crazy RPM. . Rotors are usually well balanced in ones I have seen.

 

 

cheers- glen

[Kyle Communications]

I havent heard of any coils departing and sticking the engine..the metal core wouldnt do that not looking at the design but there are heaps of the rotax generators that get burnt out. The rotax is supposed to supply 18 amps....yeah NO..it gets really hot and these new regs that some are promoting actually short the stator as the battery becomes fully charged which of course puts heat back into the coils of the generator. Apparently its common to do this on motorbike systems

 

The more glass you have in the plane now seems to be why they are toasting the generators..I was told of one CTLS in WA that has has 2 at least generators due to the load of avionics and probably autopilot and a airmaster or similar and nav lighting..just too much load. These engines were designed with analogue avionics fit out which is far less load.

 

The regulator we are making will start current limiting at around 12 amps as this is mainly for the battery charge rate we are concerned about. It maybe a issue if you want to load it up with 2 of Dynon 10 inch screens and constant speed electric driven prop and autopilot so will have to see.

 

My Rans S-21 will have a MGL Discovery 8.5 inch display and Ibox with autopilot and a few other things like USB chargers for Ipad and phone etc then most likely 2 radios and transponder and 2 analogue instruments. Hopefully soon I will have all that gear here and I will hook it all up and just see how much load that actually is so I can be sure of what current the whole system will draw

 

Mark

[RFguy]

Hi Mark. An alternative might be a 2nd-order current control- IE duration x load- IE heat load on the alternator knowing that there is some thermal inertia available. Complicates the design, I agree.

 

I was reading somewhere the Rotax alt was perm mag ? No brushes so that's cool, but I think I only replaced brushes in my car alternator every 250,000 km ! But..... no field control ? how does that work for control ? (apart from as you said, motos short the alt) .

..I guess a shunt regulator would work, but that's quite alot of extra heat, and engine kW to blow off. ...... and I am sure with a 70kW aircraft motor every 500W is precious.

 

So then the only (efficient) control you have is phase fire rectifier control....needs to be foolproof. As you know, you can't UNLOAD the alternator, that is bad ....the EMF has to go somewhere. the thing has to stay somewhat loaded as not to send the voltages appearing between the diode stacks in reverse bias beyond the limits. seems that conventional field control alternator is better suited to not wasting power. Although I guess there will always be a decent load on the alternator, so the only time it would be no load would be if all avionics was off and battery was full.... seems reasonable to dump then. ---glen.