The care and feeding of a123-based packs...

Batteries, Chargers, and Battery Management Systems.

The care and feeding of a123-based packs...

Postby GGoodrum » Fri Oct 19, 2007 5:19 pm

I decided to start this thread to share my experiences with a123-based packs, in terms of what is really required to get the most out of the cells, how to make packs, how to charge/balance them and how to protect them in use.

The a123-M1 cells were developed by an MIT-spinoff called a123Systems. They use the same LiFePO4 chemistry as many of the emerging Chinese variants, which is inherently much safer than any of the Cobalt and/or Maganese-based Lithium cells. What sets the a123-M1 cells apart is their proprietary use of nanotechnology in the electrode design. This allows these cells to be discharged at very high rates, and also allows them to be charged at very high rates as well, mainly due to a very high "C" rating, and low internal resistance. They are also extremely safe. Unlike othre chemistries, these will not explode if you over-charge or over discharge the cells.

Each cell has a totally usable capacity of 2.3Ah, a nominal voltage of 3.3V per cell and a fully-charged voltage of 3.65V per cell. They are rated for 30C continuous discharge (70A) and over 50C (120A) for up to 10 second "bursts". Most of the Chinese LiFe cells have max discharge rates around 2-3C. What that means is that in order to pull 40-50A out of them, you need packs that are 20Ah, or better. With a123 cells, you could easily pull 40-50A out of a "single-p" 2.3Ah pack. In practical terms, though, you won't get much range on 2.3Ah, but the point is that it doesn't have to be 20Ah. I think 90% of the users out there don't need more than 10Ah, 90% of the time.

When using a123 cells paralleled together, in order to get higher capacities, this also raises the "C" rating even further. In a 4p/9.2Ah configuration, the continuous discharge rating goes up to 280A and the burst current rating to a whopping 480A. :shock: This flattens out the discharge "curve" even more. The power delivered is very consistant, all the way up to the end of the capacity, and then it drops off quickly. With every other type of battery I've tried, you can "feel" the performance under load drop as you get closer to the end of the capacity. With these, the power pretty much feels the same 10 seconds from the end as it does in the beginning. If the cells in the pack are well balanced, power at the end will drop so fast, you will think you blew a fuse. Therein lies the only real "weakness" with using a123 cells in an EV application. The only way I'v found to kill these things is by over-discharging. This, of course, is applicable to all Lithium-based batteries, but it is actually easier to do with a123-based packs, due to the fact you don't get any notice that the end of capacity is near.


How to get a123 cells...

You can get bare a123-M1 cells directly from the manufacturer in development kits, but these are quite expensive at about $18 a cell. Many in the RC world, including myself, have tried to get a123Systems to sell us bare cells at some sort of reasonable price, but to no avail. The best I was able to do was to get them down to about $12 a cell, but that was for a quantity of 50,000. These cells were originally developed for Black and Decker's DeWalt brand pro line of 36V power tools. DeWalt bought 10 million a123-M1 cells last year, so even if you added up the total potential number of cells used in the RC and ebike/scooter applications, it would be a statistically miniscule portion of DeWalt's production run. :)

Image

DeWalt uses 10 M1 cells in each 36V power pack. I'm not sure why they call them 36V because the nominal voltage per cell is 3.3V. The only time this pack is anywhere close to 36V is fresh off the charger. Anyway, DeWalt sells these packs retail at about $160 each. They also include them, however, as part of combo packages where they will have a couple of power tools and maybe a light, along with two of these packs and a charger, at quite a cost savings over the individually priced items. There are a number of ebay power tool dealers who will break these combo packs apart and sell the pieces separately, including the packs. What this means is that you can buy these 10-cell DeWalt packs off eBay for about $90-$100, shipped. That makes the per cell cost of about $10, much more reasonable. To find these , just do a search on eBay for DeWalt 36V DC9360 and you should get a list of what's currently available.


Building, charging and balancing packs...

Inside the DeWalt packs the 10 cells are connected together in series via welded tabs. The 10-cell bundle is held in place with a pair of molded plastic end caps that also contain some springloaded wires that are used for individual cell moitoring and balancing. There is an embedded battery management system (BMS) that is part of the top half of the case. In order to get at the cells you need to use a Torx T-10 security tool to remove 7 screws that hold the top half of the case to the bottom portion.

Once the screws are removed, you need to unplug the two 5-pin plugs that connect the BMS to the plastic end cases.

Image

There is a temp sensor taped to one of the cells and there are three wires soldered to the output tabs of the 10-cell pack.

Image

You can simply cut the wires off, being careful not to short out the tabs.

Image

You can remove the 10 cells, with the end caps by using a flat-head screwdriver to gently lift under the cells on either end, a bit at a time.

Image

Once out of the main case, the end caps will actually simply pop off. Be careful handling these, however, as the cells are loose at this point, simply held together by the welded tabs. I would be careful handling these, as they tend to want to flop around a lot, and you should remove all rings, etc. Once I caught a loose pack of these, as it was falling off the bench, and the main out tabs arc-welded themselves to my ring and finger. :oops:

Image

As you can see below, one of the tabs is quite a bit narrower than the others, which makes this a last resort fuse, if all other means of protection fail. I have no idea high much current it would take to break this connection. I've dead shorted 10-cell packs for up to several seconds and although the results were quite spectacular, this connection never broke. I'm guessing it would take something north of 100A for 10 seconds, or so.

Image

In any case, to keep the cells from flopping around, I like to use a hot glue bead between the cells.

Image

In the RC world, there are several popular methods for building packs. The easiest of these is to take a 10-cell pack, like the one shown above, add balancer plugs, main power leads and the simply shrink wrap the whole package together

Image

Image

This form factor doesn't always fit the models very well, so othre methods are used as well. Here's one that uses clusters of four cells, inline:

Image

I've used this method to make some 16s2p eBike packs, but it is a very time-consuming process.

Image

Image

Initially, these didn't have balancing plugs on them, but I have since gon back and added them. I also think that for EV applications, it is much better to connect cells in parallel first and then connect the parallel blocks in series. The reason for doing this is that you can eliminate the need for any sort of current limiting, beyond what the controller already has, and simply add a big fuse. As I have said in a couple other threads, I recently had a short in a harness that cased a dead short on an 18s5p pack. It melted 12 guage wires and caused the plastic plugs and insulation to catch fire. It took me about 10 seconds to blow out the fire and get the connections apart. To my amazement, not only were the cells fine, they weren't even knocked out of balance. :)

One of the advantages of using a123-based packs is that you have a lot of flexibilty in how you organize packs, both from a total voltage point-of-view, and in the capacity you need. What really drives configuration, more than anything is how the packs will be charged and balanced. If packs are made in multiples of 4 cells, you can use standard SLA-type chargers. What a123, and other LiFe cells need is to be charged at whatever max rate the cell will handle, until the voltage reaches 3.65V. At that point, the voltage needs to be held there while the current is gradually reduced until it gets down below about 1A. or about 10% of the max rate. You can actually charge an a123 cell/pack in as little as 5 minutes, if you have a big enough charger. It is not usually practical to have a charger that can do more than 10-20A, but still, this high charge rate ability is unique to a123 cells. The max charge rate for Chinese LiFe cells is rarely above about 0.5-1C. Anyway, it turns out most SLA chargers use the same sort of constant current/constant voltage (CC/CV) charging profile that is best for a123 cells. For a 12V SLA, a typical SLA charger will peak out at around 14.5.V. The cutoff voltages for 24V, 36V and 48V SLA chargers are about 29V, 44V and 58V, respectively. The optimum 3.65V/cell CC/CV cutoff voltage for a a 4-cell a123 pack is 14.6V, and for 8-cell, 12-cell and 16-cell packs, the equivalent voltages would be 29.2V, 43.8V and 58.4V, respectively.

What ia standard SLA charger will not do, however, is balance the cells in the pack. SLA-type cells will self balance, at the end of the charge cycle. a123 cells, and most other Lithium-based batteries, need to be "manually" balanced. In the RC world, we have a number of automatic balancers that will typically work with packs from 5-10 cells. Most of these have small microcontrollers that look at the voltage of each cell. If they are all above a certain level (usually around 3.0V...), it will start drawing off 150mA from each of the cells that are above the cell with the lowest voltage. This is exactly how the popular Thunder Power 10-cell TP-210V and AstroFlight 6-cell "Blinky" work. Either of these can work while the pack is being charged, but most of the balancing occurs in the CV mode, as the current is being reduced. Thunder power also has the matching 10-cell TP-210V charger that will charge a 10-cell pack of pretty much anything, up to a 5A rate. It can be linked to the TP-210V balancer which allows the charger to monitor and display the voltage of each individual cell. If it detects that the cells are out of balance by more than .3V per cell, it will automatically swith to a "Balance Charge" mode, where it cuts the charge current to .3A, until the cells are balanced again.

The trend these days with RC chargers is to either integrate the balancing and charging functions, or to charge the cells independently. The FMADirect BalancePro HD uses an integrated cable that will independently charge up to 6 a123 cells at a max 10A rate. It has a USB port that can be connected to a PC for monitoring the charge process, but no display on the unit itself. This would be a great solution if it were good for more cells, and had an integrated display. The new Schulze 7-36-8 will actually charge up to 14 a123 cells in series at a max rate of 7A, but oddly, it will only auto-balance packs up to 8 cells.

In any case, thought needs to be put into how a123 packs will be balanced and charged before a final pack and/or subpack size is selected. In my case, for my own use, and for the packs I'm doing for my local conversion business, I decided that doing a 10-cell subpack configuration makes the most sense. This is for a couple of reasons. The main one is because I've found it is far easier, and a lot quicker, to build packs based on how they come out of the DeWalt packs. I find I can save even more time by making use of the plastic end caps with all the existing balancing connections. With pre-made harnesses for standard RC balancer plugs, I can completely teardown 4 DeWalt packs, hot-glue them together, plug in the balancer harness, solder main power leads and shrinkwrap the whole pack with heavy-duty rubberized shrink tubing in well under an hour. That ends up being a 10s4p, 33V/9.2Ah, subpack that weighs 7 pounds 9 ounces.

Image

Image

With the test pack shown above, I actually configured it as two 10s2p packs, stuck side-by-side. The balancer wires for each 10s2p pair are brought out separately. One side has a match set of connectors, which allows both sets to be paralleled together by connecting the 2nd set of balancer plugs into this matching set. An external main power harness is used to connect the main leads in parallel. I did it this way because it allows me to disconnect the 2nd set of balancer plugs from the first, and wire these two 10s2p subpacks into a 20s2p configuration, and use it as a "booster pack" to add capacity to my normal 20s4p configuration I'm trying to standardize on for most of my bikes.

For now, my charging and balancing solution is to use a pair of TP-1010C/TP-210V charger/balancer combos, one with each 10s4p subpack. It takes about two hours to completely charge and balance the complete 20s4p setup. Since the TP-1010C, like all RC-based chargers, are designed to run off 12-15VDC, I have them connected to a 25A power supply I got at Radio Shack.


Protection...

As I said earlier, the only current limit protection I think that a123-based pack need is a big fuse. Most, if not all, ebike controllers already have current limiting functions built-in. That simplifies a123 BMS requirements significantly. All that is really required is protection against over-discharging the cells, which is the only way to really kill these things. I know most controllers also have LVC functions that can be "tweaked", but LV protection really needs to be done at the cell/cell block level. Otherwise, you might have a case with a grossly out-of-balance pack where most cells are well above the limit and one is below and thus damaging the cell.

Some have been using the BMS in the DeWalt packs, mainly by cpying the resistor network in one of the tools in order to enable it. The problem is these seem to be overly tempermental, and, more importantly, I haven't seen anyone be able to pull any more than about 20A through them. As I said, I don't hink we need any additional current limiting, just LVC protection at the cell/block level.

Another way to simplify this even further is to not include a positive cutoff at the pack level, which would require multiple high-power FETS, heat sinks, driver logic, etc. What Bob Mcree came up with in the Technical Forum's schematic thread, to satisfy my desire for as simple and as low-cost a circuit as possible, is one that uses two chips and a resistor for each channel that simply drives an opto-coupled output low when the cell voltage drops below 2.7V. The outputs from multiple circuits can be ganged together and then wired into the "Brake Inhibit" line on any Clyte controller. That way if any block of cells falls below the cutoff, the controller will cut power.

Image

The reason for 2.7V is because the chip comes pre-calibrated to this value. The next one down is 2.1V. It isn't all that critical because when you get to that point, the voltage is goig to drop very quick. I have noticed under load, my a123-based setups will stay rock solid at around 3.1V per cell with a 40A load, all the way up to a few seconds before you are at the end of the capacity (usually 2250-2300 mAh per cell...). At that point, if you keep the load on, the voltage will drop a very quick rate. It would literally only take about 15-20 seconds to completely kill a cell if the load is not removed. What an efective LVC needs to do is to detect the start of that rapid fall. I think anything in the 2.1-2.9V will work fine. I believe using 2.7V is perfect, as it should give you a bit more "get home" capability if the load is backed off, which will cause the voltage drop to be reduced for the lesser load, which in turn raises the the voltage in each cell above the cutoff limit.

I have now had boards made for a 10-cell version of Bob's simple LVC design. Two are shown below:

Image

They have matching connectors and can be plugged into the 10s4p subpack's balancer plugs. Two of these boards are required for my 20s4p setups.

Image

I have now also completed another version of this LVC board that also incorporates the Dewalt balancer plug-to RC balancer plug cross-coupling, which eliminates the need for the hand-crafted balancing harness. This will significantly reduce my pack constuction time.

Image

Two of these boards will be used with each 10s4p subpack, but only one will be populated with the LVC parts.

As I said, I'm really doing this for my own use, plus my local work, but I eventaully will probably add both boards, and 10s4p pack construction kit to my RC website. In the meantime, if anybody is interested in getting any of these, for a nominal cost, PM or email me.

-- Gary
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby CGameProgrammer » Fri Oct 19, 2007 6:37 pm

Great thread. By the way, DeWalt calls them 36V because that's exactly twice 18V, which is the standard powertool-pack size. That makes it less confusing to consumers.
User avatar
CGameProgrammer
10 kW
10 kW
 
Posts: 727
Joined: Fri Mar 30, 2007 7:26 pm
Location: San Diego, CA

Postby Doctorbass » Fri Oct 19, 2007 7:10 pm

Great post GGoodrum!

You did a nice job about the LVC. Keep us informed about the other PCB and pcb file availlable! I could exchange you some pcb for some recent dewalt cells i will get 4 more pack that was returned from customer the next day after they bought those... a weak cell problem again... but on 4 pack it's 36 GOOD cells.. for few box... :wink: let me know if interested on some cells.. since i can't have acces to the equipment to make my own pcb :cry: .. you could help me with your's.

Doc
________________
-Fastest speed record from now: 113 km/h measured on GPS
-Fastest ebike 106km/h on flat and managed to enter in the 19.875 sec on the 1/4 mile drag racing !
-0-70km/h in 5sec X5 5303 on 24"
TORQUE SETUP:
-Succeded to haul a 19200 pounds schoolbus!
-Team Konion Member
113kmh Giant___http://www.evalbum.com/3406
Mongoose____http://www.evalbum.com/1947
E-trike_______ http://www.evalbum.com/3776
http://twitter.com/DocbassMelancon
I speak FRENCH and english
User avatar
Doctorbass
100 GW
100 GW
 
Posts: 7132
Joined: Sun Apr 08, 2007 2:24 am
Location: Quebec, Canada East

Postby Doctorbass » Fri Oct 19, 2007 7:29 pm

Examinating how you wired the balance connectors on the original dewalt pack, i guess that you should watch your cells very carefully manytime. the reason why i say that is that the max current on those connector/spring/rod is around only 2A... so if in your pack, one cell become weak, the same cell on the other paralleled pack will push current thru the balance connector to this weak to compensate and that may melt the balance spring...

When i built mine i wired the cell in parallel with AWG16 wire to encure they will not have problen even if one cell become week.

see the pic:
Attachments
P1020314_800x450.jpg
2 dewalt pack paralled (Doc)
P1020314_800x450.jpg (81.77 KiB) Viewed 34544 times
________________
-Fastest speed record from now: 113 km/h measured on GPS
-Fastest ebike 106km/h on flat and managed to enter in the 19.875 sec on the 1/4 mile drag racing !
-0-70km/h in 5sec X5 5303 on 24"
TORQUE SETUP:
-Succeded to haul a 19200 pounds schoolbus!
-Team Konion Member
113kmh Giant___http://www.evalbum.com/3406
Mongoose____http://www.evalbum.com/1947
E-trike_______ http://www.evalbum.com/3776
http://twitter.com/DocbassMelancon
I speak FRENCH and english
User avatar
Doctorbass
100 GW
100 GW
 
Posts: 7132
Joined: Sun Apr 08, 2007 2:24 am
Location: Quebec, Canada East

Postby Doctorbass » Fri Oct 19, 2007 7:33 pm

this pics could help too about balancing tab on the dewalt pack:
Attachments
connections et leur bons voltages_800x600.jpg
These are the reference voltage to locate the cell balance wire on the dewalt tab. 3.5V, 7V, 9.5V.... etc

3.5 is te voltage i normaly measure after a pack is carged and balanced.
connections et leur bons voltages_800x600.jpg (80.9 KiB) Viewed 34322 times
________________
-Fastest speed record from now: 113 km/h measured on GPS
-Fastest ebike 106km/h on flat and managed to enter in the 19.875 sec on the 1/4 mile drag racing !
-0-70km/h in 5sec X5 5303 on 24"
TORQUE SETUP:
-Succeded to haul a 19200 pounds schoolbus!
-Team Konion Member
113kmh Giant___http://www.evalbum.com/3406
Mongoose____http://www.evalbum.com/1947
E-trike_______ http://www.evalbum.com/3776
http://twitter.com/DocbassMelancon
I speak FRENCH and english
User avatar
Doctorbass
100 GW
100 GW
 
Posts: 7132
Joined: Sun Apr 08, 2007 2:24 am
Location: Quebec, Canada East

Postby GGoodrum » Fri Oct 19, 2007 10:33 pm

Doctorbass wrote:Examinating how you wired the balance connectors on the original dewalt pack, i guess that you should watch your cells very carefully manytime. the reason why i say that is that the max current on those connector/spring/rod is around only 2A... so if in your pack, one cell become weak, the same cell on the other paralleled pack will push current thru the balance connector to this weak to compensate and that may melt the balance spring...

When i built mine i wired the cell in parallel with AWG16 wire to encure they will not have problen even if one cell become week.

see the pic:


Actually the load is mainly carried by 12-gauge leads that are connected to each 10-cell block. The balancer lines basically just keep each block of parallel cells at the same voltage level. Most of the heavy lifting is done by the four main power leads.

Also, it has been my experience that these cells just don't randomly fail. In fact, the only I've ever seen any fail have been as a result of over-discharging. In that case, all the cells that were paralleled died at the same time.

-- Gary[/code]
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby recumbent » Sat Oct 20, 2007 1:32 am

These cells are similar to lifepo4 chemistry, & we can use "sla" chargers, okay got it.

Great post, so much to learn, thanks.
Ebike in use:
LWB Recumbent bike, Sram i-9 internal gears.
Power: 58V, 16 amp/hr, lipo batteries.
max speed: 52 km/h (30 mph),
max range: 40 km's, incl stop&go few times, minimal pedaling.
New build: Delta recumbent trike
recumbent
100 kW
100 kW
 
Posts: 1095
Joined: Fri Sep 07, 2007 5:32 pm
Location: Okanagan valley Canada

Postby Ypedal » Sat Oct 20, 2007 7:33 am

You Sir !!!! RULE !!!

Thank you ! 8)
ES site status page:
http://www.ypedal.com/ES/ES.htm
----------------
http://www.ypedal.com
User avatar
Ypedal
100 GW
100 GW
 
Posts: 13628
Joined: Wed Dec 27, 2006 12:55 pm
Location: Moncton NB, Canada

Postby knoxie » Sat Oct 20, 2007 8:13 am

Hello

Thanks very much indeed for a fantastic thread!! I love reading new threads on here, there are a lot of lurkers and reapers on the forums its so nice to have people actually contribute something good and worthwhile like this.

I have been looking in to the A123 packs for a while also, however I have so many good lipo packs laying about I cant really justify playing with them until these expire which may not be for a few more years yet.

This thread is fantastic it explains so clearly how to go about doing it, still I think a lot of people will take you up and order the packs and balancers from you direct, bobs little circuit is a neat addition also.

I don't fully understand why A123 don't want to sell to the domestic market, there must be a huge market out there for ready built RC packs and the like? what are they waiting for? it just doesn't make sense, I did read somewhere a while back that chevron had a big share in the company I don't know how true this is but it may explain the companies reluctance to arm electric vehicles with these amazing batteries.

Thanks again for a most wonderful thread, I do hope that people see the light and get some of these packs from you, I certainly would if I wasn't sat on so much already, with their amazing output it would be a neat feature to incorporate a boost button on the controller to enable 80-100A in bursts should you need it, be handy on a scooter for instance!

Thanks

Knoxie
User avatar
knoxie
100 MW
100 MW
 
Posts: 2629
Joined: Wed Jan 03, 2007 5:54 pm
Location: UK

Postby knoxie » Sat Oct 20, 2007 8:16 am

Oh

And another thing, have you done any videos or got any nice pictures of your bikes? I dont think I have seen em unless I missed them? video would be cool, what kind of performance are you getting from your A123 powered rides?

Knoxie
User avatar
knoxie
100 MW
100 MW
 
Posts: 2629
Joined: Wed Jan 03, 2007 5:54 pm
Location: UK

Postby Jozzer » Sat Oct 20, 2007 9:54 am

Main reason I imagine Knoxie are to do with the fact they can sell as many as they can make through dewalt. I wouldn't mind betting Dewalt has exclusive deal for the moment. If they don't hurry up with bigger cells then people like LifeBatt are going to beat them too it!
performance from them really couldn't be better. I get less than 5% sag at 5c or so (2.5v sag from a 66v 8AH pack under 45Aish). Nick and I tried to make a video the other day, but it didn't come out very well. Bout time you brought your camera down here for a fotoshoot and KMX race Knoxie! :lol:
Best of all, you can feasably run a very short range pack of a couple of kilo's to power even an X5 at max power!
Mazda MX-5. 300KW power. Soliton 1 controller, 11" Kostov motor, 20KW/H Turnigy Lipo for 60-100 miles range. 120mph top speed.
Hudson Kindred Spirit 3 wheeler. Twin Agni setup, 300KG 80KW. 100mph top speed (maybe more, but no-one has the guts to try!)
Aprilia RS125, Agni motor, 600A 96v Kelly controller, 6kw/h Turnigy pack.
www.Jozztek.com
User avatar
Jozzer
100 kW
100 kW
 
Posts: 1441
Joined: Thu Apr 19, 2007 8:32 am
Location: Brighton UK

Postby CGameProgrammer » Sat Oct 20, 2007 9:56 am

knoxie wrote:I don't fully understand why A123 don't want to sell to the domestic market, there must be a huge market out there for ready built RC packs and the like? what are they waiting for? it just doesn't make sense, I did read somewhere a while back that chevron had a big share in the company I don't know how true this is but it may explain the companies reluctance to arm electric vehicles with these amazing batteries.

That's not true at all. They already have a contract with GM to supply batteries for the Volt, or at least some of them (GM is working with several battery manufacturers), and of course DeWalt is the domestic market.

The truth is they won't deal with small customers. A123 is not a vendor; they're a manufacturer/designer, and it's simply good business to only work with large volume clients and let them handle the details of dividing that up among a million consumers.
User avatar
CGameProgrammer
10 kW
10 kW
 
Posts: 727
Joined: Fri Mar 30, 2007 7:26 pm
Location: San Diego, CA

Postby DeeL2003 » Sat Oct 20, 2007 10:04 am

CGameProgrammer wrote:
knoxie wrote:I don't fully understand why A123 don't want to sell to the domestic market, there must be a huge market out there for ready built RC packs and the like? what are they waiting for? it just doesn't make sense, I did read somewhere a while back that chevron had a big share in the company I don't know how true this is but it may explain the companies reluctance to arm electric vehicles with these amazing batteries.

That's not true at all. They already have a contract with GM to supply batteries for the Volt, or at least some of them (GM is working with several battery manufacturers), and of course DeWalt is the domestic market.

The truth is they won't deal with small customers. A123 is not a vendor; they're a manufacturer/designer, and it's simply good business to only work with large volume clients and let them handle the details of dividing that up among a million consumers.


True. Plus, A123 already sells RC-based packs even before the Dewalts came out. Unfortunately, the RC packs and Dewalt packs still have not driven down prices. Thank God for Ebay.
User avatar
DeeL2003
10 W
10 W
 
Posts: 69
Joined: Mon Jun 18, 2007 10:59 am
Location: Southern Cali-IE & South Bay

Postby GGoodrum » Sat Oct 20, 2007 10:19 am

Doctorbass wrote:Great post GGoodrum!

You did a nice job about the LVC. Keep us informed about the other PCB and pcb file availlable! I could exchange you some pcb for some recent dewalt cells i will get 4 more pack that was returned from customer the next day after they bought those... a weak cell problem again... but on 4 pack it's 36 GOOD cells.. for few box... :wink: let me know if interested on some cells.. since i can't have acces to the equipment to make my own pcb :cry: .. you could help me with your's.

Doc


I'm always looking for more DeWalt cells, so I'm sure we can do some horse tradin'. :)

Also, what I really need are more of 5-pin plugs that are on the DeWalt BMS units. Do you still have a big pile of failed units? If so, can you cut the plugs off?

Thanks -- Gary
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby fechter » Sat Oct 20, 2007 10:34 am

Excellent information. So good, in fact, that I stickified the thread.

I'll have to figure out a better way to collect all the 'best of' posts to make them easier to find down the road.
"One test is worth a thousand opinions"
User avatar
fechter
100 GW
100 GW
 
Posts: 10100
Joined: Sun Dec 31, 2006 3:23 pm
Location: California Bay Area, USA

Postby GGoodrum » Sat Oct 20, 2007 11:09 am

I think initially a123Systems was working through another company who was already a "Tier 1" supplier to GM. This company is doing the final packaging, I think. There was/is a second company/team that is part of the big Korean conglomerate LG, that was also being paid by GM to do Volt prototypes. As I understand it, GM was so wow'd by the performance of the a123 cells that they have since dump a bunch of money into a123Systems, to accelerate production. It's complicated, but GM and a123Systems, as a team, will supply the cells to the Tier 1 company, who then will provide completed batteries back to GM. You can read about it all here: http://www.a123systems.com/newsite/index.php#/news/news070809/.

I also heard that GM is so happy with the battery progress that they have fasttracked the Volt into the production pipeline. It is in full-scale development right now, with the first release as part of the 2010 model year, and will go on sale two years from now.

A123Systems is actually doing two new cells, both bigger than the current M1s, but not quite as big as the LiFeBatt cell. The new ones are 32mm in diameter (vs 26 fo the current ones and 40mm for the LiFeBatts...), with one being 113mm long and the other 157mm. The 113mm version is called the Ultra, and is optimized for use in hybrids that have high instantaneous power requirements. The longer 157mm version is called the HD and is optimized for high energy density. These will be optimized for plug-in hybrids and are the ones that Hymotion (now an a123Systems subsidiary...) will use in their upcoming Battery Range Extender Module (BREM) plug-in upgrade they will offer for the Prius and for the Ford Escape Hybrid. We have two Escape Hybrids in our family, so I'm really looking forward to this upgrade, which is supposed to be available by about February.

The main reason that a123Systems didn't want to deal with us in the RC world was mainly because we were not using anysort of BMS, but were just using the raw cells. Their unit that does the small packs for the RC car market, a123Racing, adds a weird voltage regulator system to their packs, which are really too small fto be of much use in the larger RC airplane and helicopter applications. Maybe with a "proper" ebike-tailored BMS, a123Systems could be convinced to sell raw cells to someone who will package them into packs, but then you still have the volume problem. In order to get a price per cell down to about the same level as what you can do by buying DeWalt packs on eBay ($10 per cell, shipped...), you will have to buy something on the order of 100,000 cells. :shock:

-- Gary
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby GGoodrum » Sat Oct 20, 2007 12:13 pm

knoxie wrote:Oh

And another thing, have you done any videos or got any nice pictures of your bikes? I dont think I have seen em unless I missed them? video would be cool, what kind of performance are you getting from your A123 powered rides?

Knoxie


Actually, I'm finally going to be doing one today. :) I figured out how to mount a small tripod to the frame of my Townie and have mounted my small HD camera to it. What I will also be using is a new beta V3 version of the EagleTreeSystems' MicroPower E-Logger unit. I've been helping them test this new version which now will be much better suited for use on an ebike. The big change they made is that the upper voltage limit has been raised from about 50V to a bit over 80V. I tried out a v2 unit on one of my bikes, but I had to wire it into only one subpack, and then use the voltage cal program in the PC application to adjust the voltage back up in what the application displays. Anyway, the V3 unit eliminates the need for this, and so I can simply put the unit in between the controller and my 20s a123 pack.

This is basically a WattsUp on steroids, but with the display remoted, and with the ability to record current, voltage, several temps, if you want it, and also RPM. The LCD panel can be setup as a live display of any number of the parameters being recorded. There are three ways to get RPM into the unit. They have a magnet/hall sensor that is similar to existing inexpensive speedometer setups, an optical sensor and a new brushless motor sensor that has a single wire that connects into one of the 3 motor wires. I'm first going to try the brushless sensor, but I also thik that it will work to simply pick off one of the motor hall sensor inputs and send that directly into the MicroLogger. The application has a way to enter an "adjustment" on the motor/wheel rpm input so that it displays MPH in the application and on the PowerPanel.

The application not only lets you do all sorts of graphing and plotting of the recorded data, but it also has a replay mode, with "steam gauges" and digital readouts. What I do with my RC helicopter fligts is to capture these replays, using a screen capture program, and then overlay it onto the the video of the flight, completely syncronized. I plan to do the same thing with the video I shoot today on the bike. :) Here's a screen shot of what this looks like:

Image

If you want to see the whole video, it is here, but the I have to warn you, the file is pretty big (85mb..). The pilot is one of the top-ten "3D" RC helicopter pilots in the world, as is his brother. This was at an all-electric RC event that is held in SD every year.

Anyway, what I want to overlay onto the bike video is voltage, amp draw, watts, Ah used and MPH. These are the same things that get displayed live on the PowerPanel, which I have mounted the handle bars.

I'll post pictures and the completed video later, probably in a new thread.

-- Gary
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby xyster » Sat Oct 20, 2007 12:51 pm

Cool video. Reminded me of when I was a kid and would build and launch model rockets with a payload of bugs as "astronauts" in the transparent nose cone section. :D
Ebike: 5304/20", 72V 35A controller, 33AH 80V 20s15p (18650 sized cells) DIY lithium-ion pack
http://endless-sphere.com/forums/viewto ... 47&start=0
Scooter: '06 Stealth s1000, 48V 30A, 4x10ah SLA
http://endless-sphere.com/forums/viewtopic.php?t=148
Ebike: '06 Currie Mongoose, 32V 35A, 32V 22AH hybrid SLA/Li-ion pack
http://endless-sphere.com/forums/viewtopic.php?t=1010
User avatar
xyster
1.21 GW
1.21 GW
 
Posts: 3580
Joined: Tue Jan 02, 2007 8:39 pm
Location: Visualize Rural Sheep

Postby Doctorbass » Sun Oct 21, 2007 1:04 am

GGoodrum wrote:
Doctorbass wrote:Great post GGoodrum!

You did a nice job about the LVC. Keep us informed about the other PCB and pcb file availlable! I could exchange you some pcb for some recent dewalt cells i will get 4 more pack that was returned from customer the next day after they bought those... a weak cell problem again... but on 4 pack it's 36 GOOD cells.. for few box... :wink: let me know if interested on some cells.. since i can't have acces to the equipment to make my own pcb :cry: .. you could help me with your's.

Doc


I'm always looking for more DeWalt cells, so I'm sure we can do some horse tradin'. :)

Also, what I really need are more of 5-pin plugs that are on the DeWalt BMS units. Do you still have a big pile of failed units? If so, can you cut the plugs off?

Thanks -- Gary


I have around 15 of those.

pm me
________________
-Fastest speed record from now: 113 km/h measured on GPS
-Fastest ebike 106km/h on flat and managed to enter in the 19.875 sec on the 1/4 mile drag racing !
-0-70km/h in 5sec X5 5303 on 24"
TORQUE SETUP:
-Succeded to haul a 19200 pounds schoolbus!
-Team Konion Member
113kmh Giant___http://www.evalbum.com/3406
Mongoose____http://www.evalbum.com/1947
E-trike_______ http://www.evalbum.com/3776
http://twitter.com/DocbassMelancon
I speak FRENCH and english
User avatar
Doctorbass
100 GW
100 GW
 
Posts: 7132
Joined: Sun Apr 08, 2007 2:24 am
Location: Quebec, Canada East

Postby GGoodrum » Sun Oct 21, 2007 2:09 am

fechter wrote:Excellent information. So good, in fact, that I stickified the thread.

I'll have to figure out a better way to collect all the 'best of' posts to make them easier to find down the road.


I'm honored, thanks. :)
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby GGoodrum » Sun Oct 21, 2007 2:19 am

Doctorbass wrote:
I have around 15 of those.

pm me


PM sent, thanks.
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby knoxie » Sun Oct 21, 2007 1:42 pm

Hi

Yes cool looking forward to the video, I did a video in February doing just the same thing, great to hear they have upped the voltage limit that was the only limitation for me, I used one of those a lot very neat little device.

Here is the video I did

http://video.google.co.uk/videoplay?docid=5147649837318141523&q=electric+bike+dash&total=1&start=0&num=10&so=0&type=search&plindex=0

It works superb.

The A123 batteries really are the future, I hope A123 do get a dealer to distribute them direct, I agree about the BMS issue as overdischarge being the problem that it is, its still quite puzzling for me why GM are back on track with electric cars though after the EV1? its odd that they should ditch it so publicly then come back to it.

I suspect the Chinese will be trying their level best to make a nano based cell as well as is always typical with the Chinese, take someone else's idea and do it cheaper, I am sure this will happen in time which is a shame in some ways(for A123) but will make the cells cheaper for everyone, the Chinese have no qualms about patent infringements generally.

I sill get a little nervous when car and oil companies get their financial teeth in to emerging and apparent battery technology like this though, the last thing the big oil companies want is the wonder battery, A123 packed cars would certainly be enough power and range for most people.

Looking forward to watching your video.

knoxie
User avatar
knoxie
100 MW
100 MW
 
Posts: 2629
Joined: Wed Jan 03, 2007 5:54 pm
Location: UK

Postby knoxie » Sun Oct 21, 2007 1:57 pm

Jozzer wrote:Main reason I imagine Knoxie are to do with the fact they can sell as many as they can make through dewalt. I wouldn't mind betting Dewalt has exclusive deal for the moment. If they don't hurry up with bigger cells then people like LifeBatt are going to beat them too it!
performance from them really couldn't be better. I get less than 5% sag at 5c or so (2.5v sag from a 66v 8AH pack under 45Aish). Nick and I tried to make a video the other day, but it didn't come out very well. Bout time you brought your camera down here for a fotoshoot and KMX race Knoxie! :lol:
Best of all, you can feasably run a very short range pack of a couple of kilo's to power even an X5 at max power!


Hello

Yes I must get and do a photo or videoshoot with you sometime, I got together with Steve Head but that was over a year ago now! quite amazing how time has flown, the KMX is still in great shape and I can arm it with 72V of lipo in no time, I have to go a little easy with it when its warm as the motor runs pretty hot! a little video race of 2 x electric KMX would make a great video if we could get some external and on board shots as well.

Dont want to hijack the thread here though so will leave it at that for now, would love to see some video of your setup and packs and performance etc etc.

Cheers

Knoxie
User avatar
knoxie
100 MW
100 MW
 
Posts: 2629
Joined: Wed Jan 03, 2007 5:54 pm
Location: UK

Postby GGoodrum » Sun Oct 21, 2007 3:02 pm

Great video, Knokie. :) I'm hoping ot get mine done today, but the winds are blowing pretty strong today in Southern California. It is causing havac up in Malibu at the moment, with a major brush fire, but even down here, where the Santa Ana winds are usually so strong, it is blowing strong enough to even move my 250 lbs around a lot more than I'd like. :D

In the meantime, while I wait for the winds to die down a bit, I'm working on gettig the magnetic RPM sensor mounted. I need to try this first, just to get a baseline that I have the application calibrated (what gets displayed on the Power Panel, and in the application will be speed, in either MPH, or km/hr...) I also want to use theis same input, once I know it works, with a direct tie-in to one of the hall sensors on the motor. Finally, I have one of ET's brushless motor sensors I want to try as well, but I need to know how many poles the Clyte 5304 motors have, does any one know? In any case, ias soon as I can safely get out there, I will give this a whirl on my Townie, which is now setup with a 20s6p a123 configuration.

-- Gary
User avatar
GGoodrum
100 MW
100 MW
 
Posts: 2980
Joined: Thu Sep 20, 2007 2:03 pm
Location: South Orange County, CA

Postby CGameProgrammer » Sun Oct 21, 2007 8:10 pm

GM developed the EV1 for one reason only: to meet California's proposed strict zero-emissions guideilnes for around 2% of the cars sold there. But the car company lobbyists succeeded in getting that dropped, so everybody stopped working on electric cars.

But the success of hybrids has proven there is genuine consumer demand for such cars, and the latest and greatest batteries from companies like A123 are significantly better than NiMH, particularly in terms of lifetime, which is what the car companies care about most because it's what the average consumer cares about most.
User avatar
CGameProgrammer
10 kW
10 kW
 
Posts: 727
Joined: Fri Mar 30, 2007 7:26 pm
Location: San Diego, CA

Next

Return to Battery Technology

Who is online

Users browsing this forum: DrInnovation, xlh1460 and 8 guests