Here's the data sheet for the new GP LifeFO4 pismatic batteries. They are actually ideally sized for the Vectrix battery box. As you can see, they are 45Ah and of these 46 would fit in two rows of 23 side by side terminal facing up with lot of room to spare. This lloks like a perfect arrangement. either 39 or 40 of these would give a 125V or 128V/45Ah pack at just 64kg!
http://liionbms.com/pdf/goldpeak/GP45EVLF.pdf
Type : Rechargeable Lithium Ferro Phosphate Model No.:GP45EVLF
Prismatic Batteries
Nominal Dimension : T = 30.0mm (max.)
L = 225.0mm (max.)
W= 95.0mm (max.)
Nominal Voltage : 3.2V
Capacity : Min. 45Ah at 22.5A (0.5C) discharge to
2.0V at 20℃
Charging Method : 22.5A (0.5C) to 3.65V for 2.5hrs at 20℃
Charging Termination
Control
: Taper Current 0.05C (2.25A) at 3.65V
Max. Discharge Current : 450A (10C)
Service Life : >1000 cycles
Weight : ~ 1.6kg
Internal Resistance : < 2mΩ
Ambient Temperature
Range
: Charge : 0 to 60℃ (conditionally)
Discharge : -30 to 70℃
They also have a 30Ah prismatic set. see here
These battery's specs are nice indeed.
Internal resistance is just 2mΩ, current NiMH have 14mΩ. That means much battery less heat and loss of capacity due to creation of heat.
With less weight (90kg(NiMH)->64kg) and higher capacity, I expect the range to double from current NiMH battery pack.
39 or 40 cells would NOT be compatible with the current Vectrix operation, as the operating voltage range seems to be 153 to 125 volts, with below normal performance at lower than 125 volts. If I were installing these new cells, I would use 42 cells, allowing a maximum charged voltage of 154.3 volts, and full performance down to 126 volts-should provide GREAT range, and allow at least 80% discharge, while maintaining realistic highway speeds! The new cells, with the 20 degree C. higher tolerance, and less resistance per volt, should greatly reduce/eliminate the HEAT problem, so my question is: How can I trade-in my old Ni-Mh pack for the new cells? I am willing/able to do the rest of the work, and/or pay a "Nominal fee",as the present performance/range is just too unpredictable/unreliable!--Bob Curry
PS:What would be my IDEAL Vectrix?--One with reliable instrumentation, turn signals with better, LOUDER warnig/self cancelling, able to cruise on slightly hilly roads, carrying 225 pounds, at 50-60 miles per hour for at least 60 miles, and having an auxillary Li-FePO battery, of 20-45 AH,(I already installed on on my bike) to power accessories/lights--this would allow use of a good stereo, GPS/telephone, Heated Gloves/Clothing, auxillary lights/hazard lights that could still operate if main motor battery system malfunctioned. This extra battery should charge from the main battery thru a "current limited" charger, and also charge at a higher rate when connected to the AC mains. AC mains charging rate should be user selectable, to allow use of a low amperage AC source, such as a 100-125 volt circuit that could only provide 8 amperes, or a faster rate requiring 15 amperes at the same voltage, and a FAST charge rate with 200-250 volts required, with a current selection of 15 or 30 amperes from the higher AC voltage supply. Useng larger wheels, say 16 inches, would also be nice, tires would probably last longer and tolerate bad pavement better.(It would be GREAT if the new manufacturer is reading this!)-Bob
Robert M. Curry
how about 300mm tall 60ah cell?
or am I just being greedy? :)
That would be nice if you can find them. I think it is interesting that the GP LifePo4s are the only ones I know o that fit exactely into the Vectrix battery box. Being only 95mm wide and 30mm deep makes a fit of two rows up to 23 each easy into the 230mm wide box. Unfortnately all of the larger capacity thunderskys and other manufacturers are either too wide or too thick. It is true that it should be possible to fit an ideally sized 60Ah pack into the space available. My take is that a vastly improved 45Ah pack at 128V with reduced weight would probably yield an extra 40% range.
Why only 40% range increase?
With 50% more capacity over the NiMH pack (30Ah for NiMH and 45Ah for LiFePo4), and 30% less weight over NiMH and only 1/7 of the internal resistance of the NiMH, at least 50% increase in range is achievable. In my opinion, it's gonna be around 75% increase or more in range over the NiMH pack.
my bad, your right. Now if you want to be adventuresome, the 30Ah packs may give a full 60Ah pack. Their size is :
T = 30.0mm (max.)
L = 175.0mm (max.)
W= 95.0mm (max.)
so if you lay them on their side, two next to each other you can fit 22 cells for each 175mm lenth hence 4 stack for a total of 88 cells. That would be two parallel sets of 30Ah 128V for a full 60Ah pack. I would have to look at the battery box to see if the interconnects allow for such tight spacing. So 45Ah is an easy fit, 60Ah possible.
A possible real world range, just a dreamland estimation:
NIMH battery: 70% effective capacity, due to hi security margins and plenty of heat lost due to internal resistance.
LIFEPo4 battery 95% effective capacity with BMS: Security margins highly reduced, low internal resistance, low energy looses due to heat.
3.75kwh*0.7= 2.62kwh effective, giving 60-70 km real world range.
(40 cells 128V)5.76kwh*0.95= 5.47kwh effective, proportionally giving 125-145 km real world range.
Effective range doubles!
Speaking about BMS I am wondering how are they going to pilot a bms with the existing electronics , even 46 cells instead of 102 to control during charge and discharge will need some hardware changes imho .
mabye are they going to fit the bms in the new pack ?
what do you think ?
regards
jean mi
vectrixhoper
As I suggested in my previous posts: balancer can be self sufficient and can be powered of the cells it's balancing and can be integrated into the pack. That was my plan if I wouldn't get a new battery, when my original failed.
38x 40Ah Kokam LiPo cells
3x Schulze balancers in series, powered by the cells it's balancing
Other than that, the new software only has to watch for charge/discharge cut-off.
That's all that's needed.
it would be great to buy a fucked vectrix and just try your system not , just for the fun of it .
the only problem is for the electronics to be able to warn the driver that it is time to stop the bike cause one of the cells has passed or is nearing it's minimum voltage while the others haven't and while the total pack voltage has not got under the cut-off minimum voltage .
regards
jean mi
vectrixhoper
Hey Handy t'as vu cela, sorry did U see this
http://datasheets.maxim-ic.com/en/ds/MAX11080.pdf
this 4 dollars chips monitors 12 cells and it is chainable to 31 others
you could play with it wothout a problem .
I am going to buit a 12 volts thundersky batt pack for boating purposes (ligth, ecosonar, electric tea pot etc etc etc)
I think we need to play with tiny projets before making something bigger .
integrated BMS seems to be the best solution .
if ever we could get a mean of changing the charging datas of the vectrix we could manage to make our own packs ... I think that in the near future many lads will do something by themselves don't you think ?
regards
jean mi
vectrixhoper
No need for that. Balancer takes care of that. Balancer keeps all cells at the same voltage all the time. It can't happen, that once cell can already be fully discharged while others are, let's say, 20% full. And since it's self sufficient and powered by the cells it is balancing, it can be completely independent of the application it is embedded in.
All, that Vectrix (as a bike), has to do, is to monitor charge/discharge cut-off and that's it. It's a win/win solution.
on charging, these cell modules http://www.evworks.com.au/index.php?product=BMS-CM060-V6 shunt current once the cell reaches 3.65v. In a bad over voltage situation, it will signal the master unit http://www.evworks.com.au/index.php?product=BMS-MCU-TS90-EV2 if the cell gets to 4.1V. which could shut off power to the charger with it's inbuilt 240V relay.
at a low cell voltage of 2.5, the module will also signal the master unit. I am thinking that the low voltage switching relay could be used to fault out the factory bms on the V to give the rider the normal battery fault indication.
If the thundersky cells would fit and deliver wat the V needs, I would have ordered this and the cells from EVPower yesterday. Still looking for the right cells :(
ha magic balancers so
vectrixhoper
ha magic balancers so
vectrixhoper
in my opinion and I hope to be wrong , the problem is that all those tiny systems won't be tuned the same .
in order to get a signal in an A to D converter or in a tension comparator one needs to devide the voltage to be measured with a resistor bridge prior to get it in the chip, all the resistors are different, so the divisions will give different results.
I hope that those tiny sexy shunts can be calibrated properly on a tension reference to be sure that each cell will be shunted at the same voltage level, if not it will be a mess .
even if we tune finely each device there will be a drift in time so the tuning will have to be done again .
the only good way of designing such shunt would be to use a very accurate voltage reference at it's nominal value and to use a comparator to triger the shunnt transistor so that there is no distortion, you take the voltage of the cell and the voltage of the reference (4.1 volt in your case) and as soon as the comparator (op amp for example) realise that both tensions are the same it would triger the transistor that would shunt the cell by resistance .mabye is it the way those systems are designed, we need details in fact
good to have to think a bit to those problems
vectrixhoper
i have been using that BMS system for ages.
thats version 6, i have used versions 2 and 5.
basic design is the same across all versions (so the bits are inter-changeable), just the newer ones are more automated in manufacturer, and alot more sturdy.
The BMS is completely analgue, so no AD converters or computers involved at all (at least for version 6, version 7 might have a digital master)
it consists of a module on each cell, and a master.
the modules do 3 things:
shunt at ~3.6v
give a "good" signal (Short circuit on signal line) unless:
below 2.5v
above 4.2v
in which case the signal line will go open (indicating bad).
version 2 master just opens a contactor or relay when the signal line goes open (disconnecting charger and or load, depending on how you have set it up)
version 5 and newer master also have a second function that limits controller output (by intercepting the throttle line) to hold the lowest cell above 2.5v on discharge.
rumour has it, version 7 will have a CAN interface, so the controller can decide (think Vectrix here).
Matt
Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km
Unfortunately that is not true of currently available BMS systems, as far as I know.
They may be able to make small adjustments to the voltage level of cells at the top or the bottom of the charge/discharge curve, but they cannot even keep up with the typical charging currents. The charger usually gets switched to a lower poweroutput mode once the first cell reaches the set upper voltage limit, so that the BMS can shunt all the current around that cell that is full first.
At the bottom end, the BMS has no chance at all to keep the cells balanced whilst they are being discharged on the road in normal use. Hundreds of amps are drawn, and the BMS can only shuffle a few amps at best.
So any significant capacity differences between cells (and these are pretty much unavoidable) make it impossible for the LiFePO4 BMS to do it's job, if it cannot communicate with the charger and the motor controller.
This information may be used entirely at your own risk.
There is always a way if there is no other way!
You're not familiar with balancers for Remote Control electric models. When we're charging our LiPo batteries, they are, from the minute, they are turned on, balancing all the cells it is connected to - from zero charge to full charge. They constantly take or give from cells. Charger doesn't even need to charge and balancer balances cells.
You can use a completely unbalanced pack, connect it to a balancer without charger being present and it will balance all cells to the same Voltage.
Schulze balancer, I was watching at, balances at max. 5A of current, for up to, I think 13 cells, but this balancer can be connected to a series with the same one or more of them so they work concurrently. It is powered by the cells it is balancing and it is balancing all the time it is turned on. And it is not working just by stopping the charge to some cell, but it is adding current to the cell that needs more capacity and taking away from the cell that has to much capacity.
5A might be enough to balance during charging, if it starts early enough, but during a fast discharge it will be hard to keep up.
But those specs are better than I expected, need to look into that. I built the M-BMS to be capable of shuffling 20A through the tabs once a suitable balancer is available! And guess what, I have 13 low capacity cell....
But probably these Schultze balancers are not suitable for NiMH cells, or are they?
This information may be used entirely at your own risk.
There is always a way if there is no other way!
I forgot it's specs.
But, bare in mind, 5A for balancing is a lot. Compared to 90A-100A continuous discharge at full throttle it might not seem enough, but remember, all it has to do is balance the imbalance between the cells - balance the difference between cells, that have different internal resistance. If you have such a balancer installed with a new pack it should be more than enough.
But it's completely useless for already damaged pack, since if you already have one cell, that only has 50% of capacity compared to other, it will never be enough.
But having the balancer from the beginning, it should balance good enough during charge and discharge, even 5A is enough; remember, the only current it has to transfer is the minor differences between cells.