Batteries and New Battery Technologies

That was fun watching. Eventually they went to Li-ion batteries. Here is an article I just received on that subject:

Use Cell Balancing To Enable Large-Scale Li-ion Batteries ​

Many applications need lithium-ion (Li-ion) battery packs with high cell counts. Large-scale arrays based on Li-ion batteries can provide the high voltage, current, and capacity required by many emerging markets. Yet designers of larger battery packs face many problems, including the issues around cell imbalance.
..article HERE

 

Brian
I now have a 5Ah 6S Li-Po pack. Cost USD76 and the charger cost USD38. The charger has the charging leads and a separate 6-pin plug to monitor the individual cell voltage during charge. The cells go to 4.2V each. So a 6S actually starts out at 25.2V.

The charger is amazing value. You can look at each cell throughout the charge. You can set the charge rate to whatever value within its limits. The charger will also discharge to cycle the battery. It also detects the number of cells. Very clever for the price.

The power available from this little battery is amazing. It weighs less 900g but can deliver something like 2kW. It is something to handle with care. I have already frazzled a set of leads that I accidently shorted.

I have also cooked my little Turnigy motor doing a full load stall test. The tiny 50g 80A controller I have does not have an externally set current limit (it was real cheap) and I did not take care to monitor the current when I was stalling it. It still works but one of the coils has shorted turns. It got very hot very quickly.

Rick W

 

Having used battery packs for the past 27 years in solar energy installations, I can tell you that until we have supercapacitors with the appropriate specific energy density, we are better off using low voltage battery packs (fewer cells in series) and then using efficient DC to DC converters (as found in maximum power point tracking charge controllers) to raise the voltage. Although it is true that lower voltage packs require much heavier gauge interconnects, because the cells are adjacent, this is a minimal expense. It is much more important to extend the life of the (expensive) battery pack by reducing the possibility of cell imbalance and thermal runaway. Energy wasting schemes to transfer energy from strong cells to weak cells is also a waste of money.

 

As far as I've been able to tell, very few, perhaps none, of the RC model sensorless, brushless, controllers have proper current sensing. This is one of the reasons that I chose to use a controller intended for an electric bike, as at least this has input current sensing and a programmable current limit.

These small motors will draw extremely high phase currents if presented with a high torque demand at low rpm. The phase current can exceed the battery current by a very large amount unless the controller has some form of compensation. In the extreme case of a motor held near-stall, the phase current will only be limited by the combined resistance of the motor winding, controller FET Rdson and the resistance of the cables and battery. It's fairly easy to get phase currents of greater than 100 amps on a tiny motor, which will soon cook the windings.

The cheap ($22) controller that I'm using is currently set for 15 amps battery current and 70 amps phase current, at input voltages up to about 60 volts. It controls phase current by limiting the PWM duty cycle at low rpm when the battery current limit is reached, effectively setting a maximum multiple of the battery current for the phase current.

As for batteries, then I'm presuming that the particular flavour of Li Ion batteries that you're using, Rick, are LiCo2 RC model packs. These are scarily dangerous if abused, as nicely demonstrated by the vivid collection of lithium battery fire/explosion videos littering the web. They are fine if carefully looked after, charged carefully in a safe location (fire resistant charging bags are available) and treated with the caution they deserve. The other main snag with them is their short cycle life, a few hundred cycles at most, plus their short calendar life (they lose 10 to 20% of their capacity per year at room temperature).

A far safer alternative, albeit with a slightly lower energy density, would be LiFePO4 cells. These are extremely safe by comparison, you can even drive nails through the pouches without causing a fire or explosion. Like the LiCo2 RC model battery packs, they need some form of battery management, but as they can be "built in" to a boat quite safely, it's best to fit one of the readily available full management PCBs, one that controls charge and discharge. LiFePO4 has a long calendar life (probably around 5 to 10 years) and a cycle life of well over 1000 cycles.

Jeremy

 

Jeremy
The more expensive controllers are programmable. The ones that cost around twice the USD30 I paid for the 85A controller I have. My Mars motor controllers are programmable but they cost USD200 each.

Normally stall condition is a remote risk with a prop application but I wanted to see if I could get the little toothed belt to slip a tooth - it didn't. Motor made smoke first. Ahh well a USD24 lesson.

I am blown away by the price of this stuff. My first PWM controller on an electric bike in 1975 used 20 by 2N3055 transistors that I could drive to around 5A each before the losses were too much to handle. You can imagine how big it was with all the transistors on an aluminium heat sink. It cost me about AUD30 then getting transistors in volume buys. I later learnt how to commutate SCRs. MOSFETs were only used for low power then and I do not think the IGBTs were even thought about.

The battery is so I can test things at higher power level than I can get from my two little lead acid batteries. The can produce about 200W and weigh 7kg. About 100 times lower power density than the Li-Po - hard to believe really. Of course I only want to get about 600W from it. It does that easily for short period.

The little battery is really scary. I would hate to short the actual batter leads. The large form cells with much lower power density are a safer option I believe. I am not rushing into batteries. Their development is so fast right now that it will be the last thing I get. I have almost decided I will use lithium of some form because there is a weight saving.

Rick W

 

Good info Jeremy. Here is link to the nanotech A123 (LiFePO4)with prices from an RC site which gives an overview of what is available. It may not be the most recent, since I was looking into these over a year ago. I believe up to 4 packs can be used in parallel. I haven't seen the small black and decker around for some time, so they may be discontinued. Individual Li poly cells were thin and flat rather than cylindrical, and seems like some used a soft outer case, if memory serves.

I was a member of the power assist group in HPV many years ago, and some members built bikes powered by the small RC airplane motors. There was even a keen commercial product on the market back then that used one of these motors-the chronos "Hammer". http://wantitnow.ebay.com.au/chronos-hammer-bicycle-electric-assist_W0QQadidZ330329927164 Intermittent use only in those days, but there may be something new that stands up to full duty cycle for all I know.

Porta

 

Oops: http://progressiverc.com/LiFePO4.html

 

Even the programmable RC controllers don't usually have any current limiting or control capability, Rick. Programming is limited to things like phase timing, autobraking, etc. The reason is just as you've noted, because a prop has an approximate cube law power vs rpm characteristic there's no real need to provide protection for the low rpm, high torque, condition.

Given the low price, good programming capability, built in current limiting and easy external world interface (simple pot for throttle, reverse capability, built-in "cruise" control etc) there's every reason to use one of these simple, $22, electric bike controllers. This is the one I'm using: http://cgi.ebay.com/48V-350W-brushless-controller-for-E-bike-scooter_W0QQitemZ260323431675QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item3c9c7c54fb&_trksid=p4634.c0.m14.l1262&_trkparms=|293%3A1|294%3A30

As well as the very expensive (but very good) A123 LiFePO4 cells, there are now many other Far Eastern suppliers. I have a 36 volt, 10Ah, pouch cell LiFePO4 pack that weighs just 3.5kg, including the battery management system. It comes from here: http://cgi.ebay.com/36V-10AH-LiFePO4-Electric-Scooter-E-Bike-Ping-Battery_W0QQitemZ220443955294QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item33537bd05e&_trksid=p4634.c0.m14.l1262&_trkparms=|293%3A1|294%3A30. An eBay search for "LiFePO4" will give a couple of hundred battery packs for sale, although some of the sellers might not be too good. Li Ping, the seller of the above linked battery, is extremely good, with excellent customer support.

A good source of battery and electric motor/controller information, at least for relatively low power electric vehicle requirements, is the Endless Sphere forum. Although it's focussed on electric bikes, several members are running at power levels of a kW or two, enough power for a modest boat (and far more power than I'm going to be using on my own project).

Jeremy

 

The boat application is really no different to the aeroplane unless you try to make belts skip. I want a controller that can comfortably get to 600W. I have one of the bike controllers that is rated at 250W on 24V. It has current limiting.

After looking at these little controllers I consider the field feedback speed control on the Kelly Controller and Mars motor combination to be overkill. I don't think it will be too long and we will see LV induction motors running in boats using the model plane type controllers. Boats do not need precise speed control.

The stuff I have bought is really for a RC model of the solar-wind boat. I am just playing with it on the bigger boat to get some data.

I also have a power meter now so I can get better performance measurements with the little electric motor mounted on V11J. My new motor is slightly higher power and I believe the most efficient in the Turnigy SK range.


With the battery, I was after the cheapest and smallest that would easily give me 600W for 10 minutes or so. The USD414 one is 5 times what I paid.

What I AM interested to find is "safer" lithium batteries in large form that are reliable. I want 12 to 14 cells (battery around 48V) with 100Ah at the 5 hour rate. These need to have good shelf life and capable of deep cycle.

Rick W

 

Isn't this electrode material technology really usable in these Li-ion batteries?

From posting #3

 

Rick, That little $22 bike controller will happily run a motor up to a supply current of 15 amps, with a 48 volt battery, so can deliver 720 watts with no mods. It's been given that power label to make it seem to comply with ebike power restrictions I think(!). Change the output FETs for something with a lower Rdson (which is what I've done) and it works OK up to about 50 amps, or around 2.4kW. I know of one chap who has been running one at this power level on a bike for some time, with no problems.

My experience with RC controllers on heavier loads (400 to 600 watts) has not been that good. I would guess that I've now managed to blow around 6 or 7 over the past couple of years, almost always because of the lack of current limiting capability. It only needs something to load the motor up at partial throttle for the phase currents to go sky high; something not likely in a model aircraft, but quite probable on a boat (weed, prop hitting something, accidental grounding etc). I've thoroughly abused several of these cheap bike controllers and have not even managed to get one warm, as they are nicely protected from user-abuse! The worst case condition for the controller (if it hasn't got any current limiting) is the low throttle, low speed, high torque demand condition. This is when the phase current can easily be five or more times the battery current, and it's this phase current in the output FETs that is the main problem.

I agree that LiFePO4 batteries are marginally more expensive than equivalent quality and capacity LiCo2 RC packs, but they do have at least three times the cycle and calendar life, plus they are a lot safer. Also, that price includes a charger and a full battery management system on the pack, with discharge protection as well as charge protection, so you don't need a separate LVC unit to protect the batteries whilst discharging.

It's hard to beat the LiFePO4 pouch cells for low discharge rate, long life, high cycle life applications at the moment. Given the safety issues with other Li ion chemistries, and the ensuing legal hassle that most countries will make you go though for a high capacity Li Ion pack, there isn't really much choice, it's LiFePO4 or another cell chemistry altogether if you're after 100 Ah. Cylindrical LiFePO4 (A123, Headway, BMI etc) have higher maximum discharge rates, but aren't so well "field proven" yet. 0.2C discharge rate is well within the optimum operating area for pouch-type LiFePO4 cells, and are probably the best deal going for a safe, high capacity, 100 Ah pack that only needs to discharge at 20 amps.

For higher discharge current, then they aren't so well suited. On my electric motorcycle, I'm using first generation Headway LiFePO4 cells, a 48V nominal, 40Ah pack (16S, 4P), made up from 64 10Ah cylindrical cells, welded together in parallel groups of four. These are OK for up to about 200 amps max, although they dip a bit at that level. The newer Headway cells are a bit better (lower internal R), plus they have screw terminations on the end, like the more expensive BMI cells of the same size (BMI cells are also sold under the US brand name LifeBatt, with a hefty price markup).

There are some Chinese/Taiwanese companies making rectangular LiFePO4 high capacity packs, Thundersky and HiPower spring to mind. I know of people using cells from both of these companies, but quality has been a bit more variable than for the pouch-type packs. Dead cells, swelling and early failure have been fairly common, even amongst the small hobbyist user base. Interestingly, I don't know of a single cell failure (other than caused by gross abuse) of a Li Ping LiFePO4 pouch cell pack, which at least gives some indication that the cells are pretty good. Quite a few of the more extreme electric bike enthusiasts have been abusing these packs for around 18 months or so now, frequently running them up top their current limit of around 40 to 50 amps, so I'm sure any problems would have come to light (as they have with some of the other vendors products).

If money isn't too much of an issue, then I think I would perhaps look at BMI cells, as they may prove to be a very robust solution. You have a vendor there in Australia who retails them and provides a good backup and support service from what I've heard. I believe that BMI are able to supply a BMS and cell mounting hardware, too, which would make building a pack more straightforward. You'll need 160 cells, arranged as a 16S, 10P, pack. Balance connections will have to come from each of the 16 sub-packs to the BMS, to ensure that the pack stays within limits, just the same as the RC LiCo2 packs. The BMA should handle charge control as well, I think, but a chat with Armind (sorry, can't remember his surname) over there should tell you all you need to know. Off the top of my head I think that BMI cells in that sort of quantity should be something like $20 to $25 each, perhaps a bit less. I paid around $16 each for my Headway cells, that are the same 38120 size and 10Ah capacity. Shipping costs are the crippling factor though, particularly if buying in smaller quantities.

Jeremy

 

Jeremy
That is helpful.

Regarding what I have purchased so far, it is simply to learn for the short term. I started with a couple of small outrunner motors and controllers that cost USD16 a set. I have only run one but I have the just in case I find a use.

I would be interested to know any recent history with Thundersky batteries. Their 90Ah cell is currently attractive to me but I do not want a dog.

I monitor the EVDL list but not with great interest. They sometimes discuss Thundersky but like all forums there is a lot of noise and it does not have the organisation of this site where you can avoid a lot of the noise.

I will have a look at the others you suggest. I would prefer large format cells either 50Ah or 100Ah rather than having heaps of smaller cells unless there are real safety advantages with the smaller ones. I have seen the LEV50 shown on this link in the flesh:
http://lithiumenergy.jp/en/products/index.html
They look good. I would be interested if they get into the consumer market. If I had 28 of these I would get about 10 hours operation between 6 to 7 knots. Better than 60nm without any energy collection with 50kg of batteries.

I am still doing the predictions on my boat design and piecing together the options for the energy collection, storage and propulsion system. Still early days but trying to have some low cost fun along the way.

Rick W

 

One thing I've observed for the A123 nanos is that they don't appear to need a low voltage cutoff. At least the cordless tool paks using A123 don't appear to have one, or it is set much lower than normally required. For example my measurements show that a 7.2v pak containing 2 cells is allowed to go until the tool stops turning at maybe 2v or so. This is different than in other Li systems I have toyed with. Normally running a Li system below 2.9v per cell cut off will mean death to the pack.

One reason I have been using Li cordless power tool batteries in my experiments is that they have a generous warranty of 2 years up to lifetime in some promotional cases. So I thought even though Li cells or paks can be bought for less through the internet, most have shorter warranties and don't have the extras. Extras like built in fuel guage, low voltage cut off, shock package, other safety cut offs, etc. I guess that some controllers can be set to do some of those things, though. Anyway, my Milwaukee 28v cordless pak has declined from 3AH capacity to 2AH in about 1.5 years and 100 cycles. When I tried to get replacement under warranty at the local service center they would not replace under warranty. Their definition for replacement was total failure or 80% loss of capacity. Just a heads up to anyone else considering this approach to extend the limited shelf life of Li. I have gone to the Bosch brand for my second pack- hopefully they will be better with their warranty. I am looking to get a pack that holds up well with a generous warranty. Then several packs can be snapped together for higher power applications like my e bikes or larger e boats.

Porta

 

Jeremy
Have you got links to any sites having prices for the BMI and Headway packs.

It is clear that China is the place to be for this technology. They will be the world supplier of motor cars and bikes in 10 years if the others do not get into electric. It seems they already lead on value, if not on outright performance.

I see casual car travel being reduced. Even in Australia many more people are using public transport. My three sons have never shown any interest in cars. The eldest lives in the city and has no car. He rides a bike sometimes but mostly uses public transport or walks. Other two just used spare cars at home. If they go to the city they usually use public transport.

My old car gets used to transport a pedal boat to the water maybe once a week. I get a limo to the airport because it is cheaper than fuel and paying for parking.

As a matter of interest I did my undergraduate thesis on electric cars. I realised then that they would not be a reality without much improved battery technology. It is here now.

I do not even have a photo of the battery bike made. At least don't think I have. Might drag out the old photos. I worked in the lead industry initially so managed to get the batteries donated.

Rick W

 

The A123 LiFePO4 cells are more robust than some, but do still need a BMS. I have a feeling that there is one built-in to the power tool packs, so when you see the voltage drop, you may well just be seeing the BMS cut-off. My guess is that they may have made it gradual, to emulate a "normal" battery going flat.

BMI were reputed to be using the same phosphate powder as A123, the Phostec supplied, so-called "nano" powder. I believe there is an ongoing patent dispute over the use of this, although it doesn't seem to be restricting sales.

The Australian BMI dealers name is Armin, (I misspelled it earlier) and his web site is here: http://www.lithbattoz.com.au/. He's in Sydney, I believe.

You have to deal directly with Victoria at Headway, who is generally very helpful and who speaks/writes good English. Buying from them is a bit awkward, as it involves international money transfer, with all the expense (and potential risk) that entails. I don't know if they have set up an online purchase system yet, they hadn't when I bought my cells from them. Their website is: http://headway-cn.en.alibaba.com//index.html

Thundersky had a dubious reputation when they first started selling cells, but that may have improved recently. Until I hear positive news from users that they are now OK I'd personally be a bit wary. BTW, a chap here in the UK cut open a dead Thundersky high capacity "cell". Inside the plastic case were a group of paralleled foil pouch cells, with the termination tabs bolted to the inside of the case terminal bolts. The plastic box was partially filled with what looked to be light oil, presumably to equalise pressure between the outside of the pouches or perhaps assist in heat transfer.

HiPower have a website here: http://hipower.en.ecplaza.net/ and are reputed to be good people to deal with. The cells look very similar to Thundersky ones, which makes me wonder if they share the same internal construction.

I'm afraid I know nothing about the LEV-50 cells, but they seem to be either lithium cobalt or possibly lithium manganese from the terminal voltage, neither of which are safe chemistries for use in something like a boat, in my personal view. I don't know the law in Australia, but many countries are now introducing restrictions on some types of high capacity lithium ion cells because of safety concerns.

Hope this lot helps, rather than hinders!

Jeremy