Bottom balancing?

[[we interrupt this thread to bring a big-picture view...]]

jrickard said:
And I would take exception to every point you make here. They do NOT match anything I've learned and are totally wrong.

No disrespect intended, Jack, but of course what you're hearing here does not match anything you've learned - and thus you judge it to be 'totally wrong'. That's normal!

You've bought a lot of cells. You've built a beautiful speedster, and you're working thru the Clubman build. You've somewhat recently purchased a CBA and two amps and have very recently been publishing single cell bench test data.

From a macro, Joe-public behind the wheel, auto-manufacturer point of view what you're saying about management makes sense. To borrow your 'daughter mode' comments, most folks don't care that cell 37 is a bit lower than the other 43 (or the other 99).

But - and I think this is the important part - when one builds an EV and designs a pack to power it, they're not Joe-public -- they must 'play the part' of the pack engineer, the systems engineer, the Lab Rat that - on one hand - studies the cells and defines their operating envelope, while on the other hand defines the needs of the vehicle and owner. Last step is to find a way to configure and interface the pack to the vehicle so that the pack is happy, the vehicle is happy, and the driver is happy.

To become that engineer, one has to (at least) study the chemistry, test single cells,** test multi-cell packs, instrument the pack and log road data, reduce and analyze the data, and use that to expand the understanding of the engineer and also improve the pack.

[ ** I may be wrong, but I think you're here.]

Since I'm not aware that you've logged dynamic pack data on the road, I'm not surprised to find that you don't yet understand the benefit of Ri or understand the way a pack behaves in use. I think you absolutely will get there - and I suspect you and Brain will enjoy the journey! And - most important of all - you'll raise the sorely lacking knowledge level of the folks out there in DIY land.

For comparison - the electric motorcycles and scooters imported into the 'States from China are similar to your Porsche in terms of management. A pack of TS cells, a controller with a pack level LVC cut, and a bank charger set to about 3.62V per cell. I can guarantee you that the folks with 3+ years of experience with this minimal level of pack management are flocking to ANY BMS system that promises cell-level upper and lower voltage management - because they're tired of suffering thru reduced power, reduced range, and the expense and down-time waiting for replacement cells to arrive.

Carry on!

PS: I'm still looking forward to a 'speedster travelogue' to Lamberts - just a little clip of fresh rolls flying thru the air? :D Last time I was there I flew down for lunch from Scott AFB. When one flies in, Lamberts will pick you up at the airport and will 'sneak' you to a table thru the kitchen ahead of the line that wraps around the building. Fun!
 
jrickard said:
You may live for physics, but you clearly know squat about batteries.
Unnecessary.

jrickard said:
There isn't any damaged carbon on the cathode because the carbon is on the anode.

The folks at Berkeley might disagree with you, Jack. When they build LiFePO4 cells they use, well, I'll let them tell you:

"The twelve-cm2 pouch cells contained LiFePO4 cathodes and natural graphite anodes.
The cathodes were prepared from 82% carbon-coated LiFePO4 from the University of
Montreal, 8% conducting carbon and 10% PVdF binder (Kureha). The NMP slurries of
were cast onto either bare Al current collectors or carbon-coated Al (C/Al) current
collectors. The C/Al were prepared in our lab with very thin coatings of Shawinigan
LBNL-54098 black and PVdF from the same type of slurry. These were dried extensively before
preparing the cathodes. The anodes were prepared from SL20 natural graphite (Superior
Graphite) and 10% PVdF binder (Kureha) on bare copper current collectors."

jrickard said:
There is no magic on the HVC any different from the magic on the low end, except for the fact that you're probably driving 400 amps through the cell at the low end.

Nope - no magic. Just an upper and lower limit that many find it sensible to stay within.

While pushing 400A might seem potent and manly, you're doing that with 180Ah in the Porsche (according to your entry in EV Album) for a 2.2C peak discharge rate. Folks here are running 3+C, Jack - some significantly more than that. Please understand who you're typing to before you really step on it...
 

Attachments

  • P1080816.JPG
    P1080816.JPG
    127.5 KB · Views: 864
  • P1080820.JPG
    P1080820.JPG
    109.5 KB · Views: 864
I typically run about about 10c burst discharge since my commutes are short and I don't want to carry a ton of extra weight. Low RI cells are very important, as is knowing how temperature affects RI. It makes the difference between a hot pack and a warm pack, or a sluggish (cold pack) ride and a peppy (warm pack) ride. Even if I was only discharging at 1c burst and riding long distances, I would want the lowest RI cells possible to prevent energy from going into heat instead of motion.
 
Here's what's being done commercially:

(well, four years ago, but even so...)

highlighting is mine



1 of 5 Copyright © 2007 Tesla Motors Updated: December 19, 2007
The Tesla Roadster Battery System
Tesla Motors
August 16, 2006

By Gene Berdichevsky, Kurt Kelty, JB Straubel and Erik Toomre


Summary
This paper provides details about the design of the Tesla Roadster’s lithium-ion (Li-ion) battery
pack (otherwise known as the ESS, or Energy Storage System) with a particular focus on the
multiple safety systems, both passive and active, that are incorporated into the pack. This battery
pack has been under development and refinement for over three years and is the cornerstone of
the Tesla Roadster. The high level of redundancy and multiple layers of protection in the Tesla
Roadster battery pack have culminated in the safest large Li-ion battery that we or many of the
experts in the field, with whom we’ve consulted, have seen.

Background
The battery pack of the Tesla Roadster electric vehicle is one of the largest and technically most
advanced Li-ion battery packs in the world. It is capable of delivering enough power to
accelerate the Tesla Roadster from 0 to 60 mph in about 4 seconds. Meanwhile, the battery stores
enough energy for the vehicle to travel more than 200 miles (based on EPA city/highway cycle)
without recharging, something no production electric vehicle in history can claim.

Designed to use commodity, 18650 form-factor, Li-ion cells, the Tesla Roadster battery draws on
the progress made in Li-ion batteries over the past 15 years. Under the market pull of consumer
electronics products, energy and power densities have increased while cost has dropped making
Li-ion the choice for an electric vehicle. In the past, to achieve such tremendous range for an
electric vehicle it would need to carry more than a thousand kilograms of nickel metal hydride
batteries. Physically large and heavy, such a car could never achieve the acceleration and
handling performance that the Tesla Roadster has achieved.

Due to their high energy density, Li-ion batteries have become the technology of choice for
laptops, cell phones, and many other portable applications. Precisely because they have all this
energy stored in a small space, Li-ion batteries can be dangerous if not handled properly. In fact,
there have been several cases of Li-ion batteries going into thermal runaway in laptop
applications leading to recalls by Dell, Apple, IBM, and other manufacturers. However, even
with this high energy density, the Li-ion batteries in the Tesla Roadster only store the energy
equivalent of about 8 liters of gasoline; a very small amount of energy for a typical vehicle. The
pack operates at a nominal 375 volts, stores about 53 kilowatt hours of electric energy, and
delivers up to 200 kilowatts of electric power.
The power and energy capabilities of the pack
make it essential that safety be considered a primary criterion in the pack’s design and
architecture.

(200/53= approx 4C, yes?)






2 of 5 Copyright © 2007 Tesla Motors Updated: December 19, 2007
Fundamentally, cells within the pack need to be protected from adverse situations that could be
electrical, mechanical, or thermal in nature. The entire design must also be fault tolerant to
reasonably expected manufacturing defects in the cells and in the pack itself.
In the body of the
paper that follows, we discuss aspects of the Tesla Roadster battery pack design that address
these concerns. However, this is not a complete summary of all the battery pack safety features,
since some aspects of our design and implementation are proprietary and/or patent pending.

Picking a Cell Design and Supplier
We started our design by purposely picking a small form factor battery cell. This cell is called
the 18650 because of its measurements of 18mm diameter by 65mm length (i.e., just a bit larger
than a AA battery). Due to its small size, the cell contains a limited amount of energy. If a failure
event occurs with this cell, the effect will be much less than that expected from a cell many times
larger. Billions of 18650 cells are made each year. Though the chance of a safety event in a
laptop is small, the number of safety incidents involving Li-ion batteries is rising each year
because there are so many more devices using small and powerful power sources.

The Tesla Roadster battery pack is comprised of about 6800 of these 18650 cells, and the entire
pack has a mass of about 450kg.

The engineers at Tesla Motors selected cells from reputable Fortune 500 battery suppliers that
have each produced billions of safe, reliable, Li-ion batteries. All the cell manufacturers that
Tesla Motors has considered invest a great deal of money and engineering resources to minimize
manufacturing defects within their cells. Overall, the selection criteria used by Tesla Motors
included multiple factors, confirmed by extensive internal and external testing, that directly
relate to the cell’s overall safety in the Tesla Roadster.

Design Safety Features: Cell Level
Since the 18650 cell is the fundamental building block of the battery pack, it is important that it
be fault tolerant. The cells used in the Tesla Roadster all have an internal positive temperature
coefficient (PTC) current limiting device. The primary role of this PTC is to limit short circuit
current on an individual cell level. It is important to note that this device is completely passive
and functions without any inputs from the rest of the battery pack systems.

A second level of protection is provided by the Current Interrupt Device (CID). Each battery cell
used in the Tesla Roadster has an internal CID. These devices serve to protect the cell from
excessive internal pressure. In such a case the CID will break and electrically disconnect the cell.
High internal pressure is generally caused by over-temperature or other failures that then result in
over-temperature.


The cells also incorporate numerous mechanical, thermal, and chemical factors that contribute to
their safety in the Tesla Roadster. For example, cells used in the Tesla Roadster battery pack are
all packaged in steel cans. This feature offers multiple safety benefits. From a mechanical
standpoint, the steel case of each cell provides structural rigidity and strength. This helps






3 of 5 Copyright © 2007 Tesla Motors Updated: December 19, 2007
dissipate extreme mechanical loading as well as providing protection against objects penetrating
or compressing a cell and thereby shorting it. From a thermal standpoint, the steel case also
offers good thermal conductivity. The dissipation of heat from a cell both extends battery life and
helps maintain the pack at an even temperature. From a chemical and materials standpoint, the
materials used in the cell’s construction can greatly impact the flammability and initiation
temperature of thermal runaway. Tesla Motors has chosen a very safe cell with great attention
paid to both these factors.

Design Safety Features: Battery Pack Level
Due to the size, weight, and cost of the Tesla Roadster battery pack, we have the opportunity to
add many more safety features than can be contained in a laptop battery pack. Overall, some of
these battery pack safety features are active and others are passive. Some are mechanical and
others are electrical. For example, the battery pack is controlled internally by several embedded
microprocessors that operate both when the battery pack is installed in the car, and when the
pack is being transported. An example of a passive safety feature is the selection of Aluminum
for our battery enclosure instead of plastic as in all laptop packs. The Aluminum provides greater
structural strength in case of mechanical abuse tolerance and does not easily melt or burn.
Collectively, the high levels of redundancy and layers of protection culminate in the safest large
battery seen by the experts in the field with whom we’ve consulted.

Architecturally, the battery pack is comprised of 11 battery modules (otherwise referred to as
“Sheets”), a main control and logic PCB (printed circuit board), and a 12V DC-DC power
supply. Each of the 11 modules carries a monitoring PCB (with its own microprocessor) that
communicates with the rest of the vehicle microcontrollers, broadcasting the voltage and
temperature measurements of its module over a standard CAN bus.


The method by which the cells are electrically connected together can have a huge impact
(positive or adverse) on the overall pack safety. In the Tesla Roadster battery pack, each of the
thousands of cells has two fuses (one each for the cell’s anode and cathode). This results in
tremendous safety benefits since a cell becomes electrically separated from the rest of the pack if
either of its fuses blow (generally by a short circuit). In addition to cell fuses, each of the 11
battery modules has its own main fuse that guards against a short circuit across the complete
module.

The picture below (Figure 1) shows the complete battery pack on a cart. Note the tubes and
manifold extending out of the battery pack at its lower long edge. These are used to circulate
cooling fluid (a 50/50 mix of water and glycol) throughout the pack via sealed fluid paths. This
enables us to keep the cells thermally balanced. This extends the life of the battery pack and also
has numerous safety benefits.








4 of 5 Copyright © 2007 Tesla Motors Updated: December 19, 2007


Figure 1 – Tesla Roadster battery pack

This cooling system design is especially effective because we have chosen to combine thousands
of small cells rather than several large ones to build an ESS, dramatically increasing the surface
to volume ratio. For example, with seven thousand 18650 cells the surface area is roughly 27
square meters. If there were an imaginary set of 20 much larger cube-shaped cells that enclosed
the same volume, the surface area would be only 3.5 square meters, more than seven times
smaller. Surface area is essential to cooling batteries since the surface is where heat is removed;
more is better. Also, because of their small size, each cell is able to quickly redistribute heat
within and shed heat to the ambient environment making it essentially isothermal. This cooling
architecture avoids “hot spots” which can lead to failures in large battery modules.


The multiple microprocessors within the ESS communicate via a CAN Bus, a robust automotive
communication protocol. During normal vehicle operation and storage, the battery logic board
communicates with the vehicle to initiate battery cooling, report state of charge, and signal
battery faults.
A fundamental element of the vehicle and battery pack safety design is the ability
to electrically disconnect the high voltage of the pack from the rest of the car (by controlling two
high voltage contactors) if any of a number of adverse conditions are detected.

The microprocessors, logic circuitry and sensors are continually monitoring voltages, currents
and temperatures within the pack.
These sensors also monitor inertia acceleration (e.g. to detect a
crash) and vehicle orientation to the ground (e.g. to detect a rollover). Our battery packs also
include smoke, humidity, and moisture sensors. If certain sensors exceed the specified range,
then the high voltage contactors will immediately (within milliseconds) disconnect the high
voltage of the battery pack from the car. In fact, the contactors are only closed (connected) when
commanded and energized to do so. Without the proper commands these contactors will open.

In more severe fault conditions such as a vehicle collision, active protection systems including
the logic board could fail due to damage. Therefore, the battery pack design incorporates an
array of passive safety features as well. The passive design improves the robustness of the
battery pack, particularly against mechanical damage and potential foreign object penetration of
the battery pack.






5 of 5 Copyright © 2007 Tesla Motors Updated: December 19, 2007

None of the Tesla Roadster’s high voltage systems are accessible to accidental contact outside
their protective enclosures and jacketed cables. Only with special tools can someone gain access
to any high-voltage components. Our high-voltage systems are enclosed, labeled, and color-
coded with markings that service technicians and emergency responders already understand.

Finally, the battery pack enclosure is designed to contain all the battery modules, fuses, bus bars,
and safety circuitry of the system. The enclosure is electrically isolated from the battery pack and
prevents users from directly accessing any high voltage connections. The enclosure is also
designed to withstand substantial abuse in the vehicle, including collision, while maintaining the
integrity of the battery modules and circuitry inside.

Testing
Upon completion of our design, we collaborated with an outside firm known for expertise in
lithium-ion batteries to perform hundreds of tests to validate the abuse tolerance and
effectiveness of our design.

We have performed further tests including SAE (Society of Automotive Engineers) shock and
vibration, crush, and vehicle collision testing. Additionally, the United Nations (UN) imposes
strict rules regarding the transport of lithium-ion batteries. Tesla Motors will not be able to sell
and deliver cars to its customers unless the production battery pack has met rigorous testing
standards set by the UN or substitute testing agreed to by the United States Department of
Transportation. Finally, we have passed all required tests from the Federal Motor Vehicle Safety
Standards (FMVSS). This involves crashing of complete cars with functional battery packs in
them.


Exhaustive, or what?


:mrgreen:


I can't imagine welding 6,800 cells into a pack. 10-14pcs sounds good to me for my little project. And as much as I hope to avoid the complexity of all the sensors, microprocessors, contactors et al, I'd like to "get the most out of" this pack in terms of longevity AND performance. 80% DOD sounds reasonable. A cell going bad before the rest sounds reasonably likely - hopefully when it goes it's easy to replace.

For my little application, I'm tempted to get an individual charger for each cell, providing I can recharge the pack overnight on a 15A residential branch receptacle.

I realize for others this is totally unrealistic, of course.

Thanks - and please by all means continue!

However, Jack, to be fair, if Luke, of all people, didn't know squat about batteries - I have a feeling he'd be trying to find out, and succeeding. :wink:
 
I agree Here about the great Tesla...

it's like the Tesla battery pack desing is THE EXACT OPPOSITE AS WHAT YOU ARE CLAIMING ABOUT BATTERY DESING...

When you say:
-Ri information is useless
-Multiple parallel-serie large number of cell group desing will end to alot of failure..
-Spotwelding cells is a big risk.....
-Monitoring HVC is useless

But.. from now.. The tesla battery have prooved that it is reliable and efficient!

This car is actually the best exemple of electric car performances in the world.

and... the most important part to ameliorate in an EV actually is the battery!


And.. You still think they are all wrong? :? ....

Now let's talk about Blue elephants... :roll:

Doc
 
Folks here are running 3+C, Jack

That's quite true. Right now I'm running off of a Headway pack (might upgrade to SkyE/CALB later) that is 50 AHr pack, 72.8 v pack (5p x 24s). It's not the final pack for the car but it's what I've got to get me on the road and for testing. Flooring it from a dead stop on level ground I can pull 300 amps out of my pack for a short period. That's a 6C discharge rate. I can spend several seconds at 200 amps - a 4C discharge rate. The Headway cells can take it - in fact highest voltage drops I have seen so far on the entire pack is from around 77 to 69.9 volts. In the next few weeks I'll have some instrumentation to get cell-by-cell voltage levels across the pack.

Headways are more expensive than TSky/SkyE for the energy, but they certainly appear to be able to handle the current draws without flinching.

Keep up the good debate, guys and let's try to keep the gloves above the belt. And nice videos, Jack. Thanks for sharing the cell dissection exercise - that was cool. I got to see the insides of a cylindrical cell through a much less coordinated method: mild explosion. :mrgreen: http://endless-sphere.com/forums/viewtopic.php?f=14&t=7235#p127184

Cheers,
--Adam
 
Doctorbass said:
I agree Here about the great Tesla...

it's like the Tesla battery pack desing is THE EXACT OPPOSITE AS WHAT YOU ARE CLAIMING ABOUT BATTERY DESING...

When you say:
-Ri information is useless
-Multiple parallel-serie large number of cell group desing will end to alot of failure..
-Spotwelding cells is a big risk.....
-Monitoring HVC is useless

But.. from now.. The tesla battery have prooved that it is reliable and efficient!

This car is actually the best exemple of electric car performances in the world.

and... the most important part to ameliorate in an EV actually is the battery!


And.. You still think they are all wrong? :? ....

Now let's talk about Blue elephants... :roll:

Doc


The question is...is it affordable for us as DIY's, I cant even imagine trying to put a pack like that together.

Roy
 
Well, if kitcars can be kitcars, we need not all aspire to build a pack like the Tesla Roadster's.

Just depends on what one is prepared to pay for, and live with.

The Tesla pack costs $36k to replace. You can order one in advance for $12k, though. Interesting economics.

Jack, how much do you have into the Clubman so far?
 
RoughRider said:
Who knows why TS changed to LiFeYPo4 cells with the Yttrium addition....???
What is the benefit of the Yttrium?
Here:
http://endless-sphere.com/forums/viewtopic.php?f=14&t=15227&p=226740#p226740
"The grain size of Y-doped LiFePO4 became smaller and grain morphology became more regular than that of undoped LiFePO4."

As pointed out elsewhere in this thread, makes surfaces a little more "nano"... larger surface area per unit of material (...so faster response times for electro-chemical reactions? more efficient? lower IR?)
Cheers
Lock
 
johnrobholmes said:
It is laughable that you are so anti-BMS. I have destroyed so many cells following 80% dod and 80% balance-charge cycles to know that a BMS will at least allow me to see the problem before it happens. I still don't run a BMS, and yet I know that on packs with more than 6 batteries in series it is in my best interest- especially with more volatile chemistries.


Let us pretend a pack has one cell that has decided to just give up the ghost after an accident or internal failure. Many times these cells will charge up just fine (voltage), but will not have the capacity upon discharge cycle. Even worse, when charging to lower voltages the damaged cell may not even be out of line with the rest at all. Voltage is fine under no load, but....When it gets run down before the rest it will get hot and maybe catch on fire depending on chemistry. If this cell was isolated it wouldn't be too bad, but generally packs are built with all the cells together. Just the heat generated from this one cell can damage surrounding cells, not even considering a fire. The only way to catch this bad cell is a BMS, because we don't always have the luxury of the "bad" cell showing itself until it is too late. It would be great if it did, but "accidents" are not called happy-fun-time for a reason.


I don't know what to tell you pard. I feel your pain. You do NOT understand battery chemistry or maintenance, and it costs money to watch them die. But there is no magic BMS pixie dust to make life simple for you. I wish there were. I am not anti-BMS at all. I just haven't found anything that will do what you want done, or even what I want done. As time goes on and we gain experience with these cells, it all starts to look more to me like instrumentation and control. If you want to call that a BMS I'm on board. But what I'm seeing out there looks like more blue elephant gun. "Bleeding" as a medical procedure went out 100 years ago. And LVC monitoring is about the same problem. I'm willing to talk about solutions, but I keep running up against some sort of malformed religion and I have a short fuse about that I have to tell you. Give me a circuit that will protect $13,000 worth of batteries and I'll buy it. But you have to "splain it" where I can understand it. I'm from Missouri and so far what you've SHOWN ME is that Endless Sphere is about as populated with engineering geniuses as American Idol.

Jack Rickard
 
AndyH said:
[[we interrupt this thread to bring a big-picture view...]]

jrickard said:
And I would take exception to every point you make here. They do NOT match anything I've learned and are totally wrong.

No disrespect intended, Jack, but of course what you're hearing here does not match anything you've learned - and thus you judge it to be 'totally wrong'. That's normal!

You've bought a lot of cells. You've built a beautiful speedster, and you're working thru the Clubman build. You've somewhat recently purchased a CBA and two amps and have very recently been publishing single cell bench test data.

From a macro, Joe-public behind the wheel, auto-manufacturer point of view what you're saying about management makes sense. To borrow your 'daughter mode' comments, most folks don't care that cell 37 is a bit lower than the other 43 (or the other 99).

But - and I think this is the important part - when one builds an EV and designs a pack to power it, they're not Joe-public -- they must 'play the part' of the pack engineer, the systems engineer, the Lab Rat that - on one hand - studies the cells and defines their operating envelope, while on the other hand defines the needs of the vehicle and owner. Last step is to find a way to configure and interface the pack to the vehicle so that the pack is happy, the vehicle is happy, and the driver is happy.

To become that engineer, one has to (at least) study the chemistry, test single cells,** test multi-cell packs, instrument the pack and log road data, reduce and analyze the data, and use that to expand the understanding of the engineer and also improve the pack.

[ ** I may be wrong, but I think you're here.]

Since I'm not aware that you've logged dynamic pack data on the road, I'm not surprised to find that you don't yet understand the benefit of Ri or understand the way a pack behaves in use. I think you absolutely will get there - and I suspect you and Brain will enjoy the journey! And - most important of all - you'll raise the sorely lacking knowledge level of the folks out there in DIY land.

For comparison - the electric motorcycles and scooters imported into the 'States from China are similar to your Porsche in terms of management. A pack of TS cells, a controller with a pack level LVC cut, and a bank charger set to about 3.62V per cell. I can guarantee you that the folks with 3+ years of experience with this minimal level of pack management are flocking to ANY BMS system that promises cell-level upper and lower voltage management - because they're tired of suffering thru reduced power, reduced range, and the expense and down-time waiting for replacement cells to arrive.

Carry on!

PS: I'm still looking forward to a 'speedster travelogue' to Lamberts - just a little clip of fresh rolls flying thru the air? :D Last time I was there I flew down for lunch from Scott AFB. When one flies in, Lamberts will pick you up at the airport and will 'sneak' you to a table thru the kitchen ahead of the line that wraps around the building. Fun!

I do understand the way a pack behaves in use. It's what I've been talking about. Monitor for WHAT and WHAT are you going to do with the information. I don't think you guys have been looking at "in use" data. Otherwise the shunting would have never come up, and now the LVC. It's a different world at 500 amps has been my mantra.

As to Lamberts, come on down. It's not widely known in EV land, but I have a hanger full of aircraft, including 2 DC-3s and a Huey. My last trip to Lamberts was in a UH-1H and yes, they picked me up at the airport. It's 30 miles from here. I've been so wrapped up in the EV thing, I haven't been to the hangar in four months. They are having a DC-3 fly in at Oshkosh, and at some point I have to break off and go get my birds in the air for the event. They're 70 years old and a bit crotchety on the startup.

Jack Rickard
http:///evtv.me

PS. DID you test the SE's manually? I actually have one in the BEDROOM thanks to you and an Arduino project I'm working on and I am NOT confused. Drops to 4.2 volt almost immediately (well faster than the TS cells).
 
AndyH said:
Jack, when I invited you visit, I pointed out that some folks have names with different colors. I promise you that's for a very good reason. :wink:

(no disrespect to my fellow monochromes. :) )


Andy. I'm sure it is. The problem is, I try to judge people not by the color of their signature, but the content of their post. It's kind of a real MLK day type of thing. If you want me to go away so you can preserve your little tin God pecking order and color scheme, I will gladly do so. There's nothing for me here so far, except the POSSIBIILITY of a mystery from YOU about cell settling levels, that is not really happening very well at the moment. But I AM ordering some LiFeYPo4 cells pronto, and if you have it wrong, I'll tell EVERYBODY. Beyond that, I really don't care squat what color you are. What can you tell me about Lithium cells, that I can then verify to be true. That's all I'm about, and I'll steal all of it from your wife, your daughter, your dog, and right out of the back of your pickup truck without blinking an eye. I want it all. From wherever. However. And from a mono if that's where it comes from.

But I have a short fuse on bullshit typed on the screen by people wanting to have a say about something. You've at least shown me data. I think you're equipment is dicked up. But at least it's data from SOMEWHERE. Maybe I can learn about your equipment.

Jack Rickard
http://EVTV.me
 
Doctorbass said:
I agree Here about the great Tesla...

it's like the Tesla battery pack desing is THE EXACT OPPOSITE AS WHAT YOU ARE CLAIMING ABOUT BATTERY DESING...

When you say:
-Ri information is useless
-Multiple parallel-serie large number of cell group desing will end to alot of failure..
-Spotwelding cells is a big risk.....
-Monitoring HVC is useless

But.. from now.. The tesla battery have prooved that it is reliable and efficient!

This car is actually the best exemple of electric car performances in the world.

and... the most important part to ameliorate in an EV actually is the battery!


And.. You still think they are all wrong? :? ....

Now let's talk about Blue elephants... :roll:

Doc

You know, I hadn't thought of it in precisely those terms. Thanks for pointing that out to me Doc. But in answer to your question, I have to say.... YES. I think they are ENTIRELY wrong. They do not use LiFePo4 cells. They use cells with a higher energy density, but that are more dangerous. They have a brazillion connections, each of which can spells disaster. There are heat problems. Basically I think it is ENTIRELY the wrong approach. I MUCH prefer the large prismatic cells. They are cooler. They do not have the negative temperature coefficient problems. The connections are more manageable. They are not proprietary - we can replace them with whatever is next at will and WITHOUT Tesla's permission or markup. And yes, while I hadn't thought of it in the terms you state, you are essentially CORRECT in that EVERYTHING I know and believe about in batteries flys directly in the face of your eggregiously childish appeal to authority, the Tesla design philosophy.

I WILL say I was impressed by the wreck in Denmark, where a much heavier car wound up parked ON TOP of the Tesla battery pack, with no disaster evident from it. But I very strongly believe that Tesla, AC Propulsion, and any idiot that tries to make a battery module out of 5000 itty bitty cells is actually a MORON in public. You've got me pegged to a T Mr. Doc.

Is ANYONE in this forum confused as to my position on the subject of little bitty batteries cobbled into a BIG battery pack? I can type MUCH more at length on this if you need/want it.


Jack Rickard
http:///evtv.me
 
AndyH said:
Nope - no magic. Just an upper and lower limit that many find it sensible to stay within.

While pushing 400A might seem potent and manly, you're doing that with 180Ah in the Porsche (according to your entry in EV Album) for a 2.2C peak discharge rate. Folks here are running 3+C, Jack - some significantly more than that. Please understand who you're typing to before you really step on it...

I'm starting to see where the problem is here. Lunging at nonsense is not engineering. Sure, I draw 400 amps on the bench. The Speedster, which you draw 70 inferences from inaccurately, routinely draws 540 amps at any freeway speed. And yes, I do have 180Ah cells. That's about 3C. Does that mean mine is just as long as yours is?

In any event, 540 is about what I've seen on the meter. No more in actual driving. But you rather glossed over the point. 400, or 540, isn't 15. So why are you preternaturally focused on what is happening while charging at the top end? The reason you are coming to the wrong conclusions is that you are looking for answers in the wrong freakin place.

And the carbon is all in the Anode. There isn't 5% carbon on the cathode of a TS or SE cell and they give a breakdown of what's in them on their website by percentage. THAT's why I'm tellling you that you DON'T KNOW SQUAT ABOUT WHAT YOUR TALKING ABOUT. And that'S IN TECHNICOLOR!!!

Jack Rickard
 
ahambone said:
[
Keep up the good debate, guys and let's try to keep the gloves above the belt. And nice videos, Jack. Thanks for sharing the cell dissection exercise - that was cool. I got to see the insides of a cylindrical cell through a much less coordinated method: mild explosion. :mrgreen: http://endless-sphere.com/forums/viewtopic.php?f=14&t=7235#p127184

Cheers,
--Adam

I DETEST explosions. They cause me to spill my whiskey, which is annoying. And actually, I like my batteries to be SILENT. ANY noise from them causes me to start looking them over carefully. It's never good. No good news from noisy batteries....

I am actually in AWE of these cells. The things I've done to them, a little venting, some gas. I've melted a wrench now and then. But largely, they are amazing....

I have little experience with Headways or little cells. I bought 512 of the A123 cylindricals and they just didn't make the cut after a couple of very small projects. They did put out HUGELY, but they weren't the girl I wanted to marry if you know what I mean. I bet I have 475 sitting here if anyone cares to make an offer.

BIg current. Little staying power. Kind of touchy on the charge.

Jack Rickard
 
northernmike said:
Well, if kitcars can be kitcars, we need not all aspire to build a pack like the Tesla Roadster's.

Just depends on what one is prepared to pay for, and live with.

The Tesla pack costs $36k to replace. You can order one in advance for $12k, though. Interesting economics.

Jack, how much do you have into the Clubman so far?

Criminy. I don't know. We're in the "dark hours" of the Clubman. Tesla was going for max miles, and they got it. I think we're looking at 125-150 real miles on the Clubman, not counting 10-12% regen and not counting 15-20% heat and air conditioning. The difference, is we have a simple battery system with virtually no temperature issues, no connection issues, and we can swap our our SE's for new TS Y cells if I want to. When I want to. Without a special "deal" from Tesla.

But we're getting slaughtered. The Quaife auto torque biasing differential I REALLY wanted in the car. Well, it turns out we got the transmission torn down, and the Quaife part doesnt even KIND of match the differential in the transmission. It has a different number of mounting bolts, it's a different size, and even the teeth for the drive shafts are a different size. We talked to Quaife, and they're familiar with "the problem." The problem is, BMW did a transmission change to an ENTIRELY different transmission, under the EXACT same BMW part number in mid-year. They referred to ours as the "large frame" and no, it won't fit. And they don't have one that will. And right now, nobody really wants one. The good news is that somehow our transmission is "better". The bad news is, that we don't get the Quaife in it.

We're still playing with the adapter. It was 3 months at VAC Motorsports, who didn't even start it, and now 3 months at Erlbachers, two blocks away. But it is mostly done. I'm hoping to film next week. Getting smart on the TIMS600 controller right now. I think we really did spend our way into something good with this MES-DEA drive train though. It's sealed, water cooled, looks like about 177 ft lbs at 300 amps, and while the BMW thing has turned out to be a pain in the ass (they're own telephone number is a state secret that the CEO isn't allowed to know) the car really is a twisted warped sort of engineering marvel. But the basic car was $32K and if we can do the conversion for $32K in parts and outside services, never mind us, I would consider it a miracle.

Is that a good value trade? I don't know. But we'll have about 40kWh, a 30kW motor that will really do about 120kW, and about 15 ft lbs MORE than the AC Propulsion systems """"200kW"""" and those quotes are in quotes, system. We'll have good cargo and good cooling. You'll be able to dial in EXACTLY how much regen you want any time you want it. ANd in all respects we expect to exceed BMW's effort albeit on a larger car. We just have more room for batteries - bottom line.

The heating system works quite well. Our 4KW water system puts out 165F water and about 105F at the vent. It draws about 10 amps, which is quite significant in our 100 AH pack. The air conditioning won't be much different. Environmental will be great, but costly in an energy sense. You turn on those heated seats and the heater and it will shorten your range, no doubt. The magic is HAVING heated seats and real heater.

The "secret" project is actually what we call Speedster part DUH. They shipped our new roller today. We are going to do a second Speedster with the HPGC AC system in what I hope is a much simpler, more manufacturable, somewhat more economic model with regenerative braking. That will go BACK to Special Editions Inc for evaluation and testing.

I'm hopeful it will result in a produced turnkey electric Beck Chamonix Speedster Electric that anyone can purchase and drive by the fall of this year. THAT should END the series of queries from those wanting to buy MY Speedster.

IN further goofy news, we've been doing quite a bit of talk with Neil Young on his LinkVolt - 1959 Lincoln Continental. I'm campaigning hard for him to DROP the hybrid concept and go entirely battery electric. He's going to film a documentary driving cross country in it. And he's stopping at EVTV Motor Verks along the way.

We're also TALKING about me doing an electric drive conversion of his 1938 JENSEN METEOR motor launch - a boat. I don't know WHEN I would do it, and not much of a clue as to HOW I would do it, but that's the conversation and it keeps coming up. Video of the METEOR, the LincVolt, the Speedster, the Mississippi River, and Neil Young all rather rolled into one unimaginable mess.

Meanwhile I'm in Arduino hell. I've fallen in love with this little beastie. It's an ATMEL microprocessor with a kind of "C++ for kids" interface that I've just fallen in love with. I'm measuring current, voltage, temperature, with it so far, and I have a $49 GPS sitting here trying to have sex with it that will track 51 satellites. If I can get them to mate in captivity.....

Jack Rickard
http://evtv.me

PS NO. I don't CARE if it can measure LV.
 
jrickard said:
I do understand the way a pack behaves in use. It's what I've been talking about. Monitor for WHAT and WHAT are you going to do with the information. I don't think you guys have been looking at "in use" data. Otherwise the shunting would have never come up, and now the LVC. It's a different world at 500 amps has been my mantra.

Jack,

I have seen a quick turn-around time from a talk on DIY or the TS yahoo group and either a video or a blog post. Based on that timeline, the fact that I've not seen sign of a desire to log data from a rolling pack, and from you telling me in no uncertain terms that you're not interested in dynamic logging, I have no reason to believe that you do understand the way a pack behaves in use.

Case in point. I have a previously mentioned Chinese scooter that is delivered free of pack management. The very first thing I did was pull the pack, wire it, and start logging data. I logged charges, discharges, cold rides, hot rides, partial charges and short rides, etc. I wanted to know exactly what was happening with the pack because, frankly, I think the manufacturer is crazy to put hundreds of packs on the street with no protection and a 2 year battery warranty. And it turns out that we're right -- the pack comes off the truck with cells in various states of charge. While it's true that the TS charger is set to a 3.62V average, nobody told the cells to be average -- and the low ones stayed below 3.3V for days while the more full ones marched their way up thru 4.5V. Strike one. Even balancing the pack didn't help for more than a few months. Cells were stying down longer, charging got more erratic, LVC events happened more frequently.

One year later, I'm working thru individual cell testing because I have a full six cells that don't rise above 3.5V after being on the bank charger for a full four days and one is very unhappy.

[edit...attachments didn't make it first time thru.] The PDF data is after the pack sat on the TS charger in it's 'pulse and glide' mode for more than 4 days - and still cells were low. After this data came a quick ride to the post office - with the pack-level LVC in place but no cell-level LVC connected. The blue line is 2.5V. Cell 4 is on a PakTrakr remote with 5 cells. Voltage dropped under 1.4 and that remote went off-line because it's powered by the cells it monitors...

View attachment 1

Because I know how the cells behaved when new, and have seen a number of uncommanded departures from controlled flight over the past year, I'm better able to recognize and deal with the strange stuff that's happening now.

Since you apparently don't log, and don't yet want to believe that the pack can do anything but fly in perfect formation 24/7, I predict that when it does happen to you - and just like gear-up landings - there are those that have those that will -- it'll be a larger shock to your system than is needs to be.

I'm still working thru a test sequence with the 100Ah cells. When its' complete I'll re-run the 24-hour bleed-down.

Andy
 

Attachments

  • Uncharged_Pack.pdf
    312.5 KB · Views: 75
Well, I'm sorry for my outburst. I apologize to Ggoodrum and Dogman.

Let me say that I've got 5 runs now on my battery pack without a BMS. I checked the cells today and they are perfectly fine. I'm still conditioning the battery so the most I've used so far is 3.5 AH. Checked the cells today after charging and in one 12v pack, cells were 3.46v and 3.65v. About .2v off. I got a small 3.2v charger to balance them so I can manually balance them very easily. All 12 cells takes less than an hour with 3 small chargers to top off. They still aren't reading all the same voltage but they are close enough. The 3.46v and 3.65v doesn't even bother me. I think that it'd have to be more than .6v off to even be close to a problem with overvoltage. And when I topped them all off with my small charger, each cell took at most 10 min to finish which means all the cells are near full.

Not bothering with a BMS. I agree with jrickard. I think Jack Rickard is trying to say that simplicity is the best way to go. My cells connect directly to a watt meter which connects to a controller. That's it. If you have a big BMS, you got how many circuits that could fail ? And then when it does fail, are you going to buy another $100 - $200 BMS? Or $50 BMS? And what if some of the cells don't want to be charged past 3.55 v? I find that some of my cells just seem to be full at 3.60v and some at 3.65v.
 
AndyH said:
Because I know how the cells behaved when new, and have seen a number of uncommanded departures from controlled flight over the past year, I'm better able to recognize and deal with the strange stuff that's happening now.

Since you apparently don't log, and don't yet want to believe that the pack can do anything but fly in perfect formation 24/7, I predict that when it does happen to you - and just like gear-up landings - there are those that have those that will -- it'll be a larger shock to your system than is needs to be.

I'm still working thru a test sequence with the 100Ah cells. When its' complete I'll re-run the 24-hour bleed-down.

Andy


Andy,

I've heard this so many times I can chant it myself. It goes kinda like this,

"Maybe so, but BOY are you going to be SORRY when the Blue Elephant DOES get here. You'll wish then you had bought one of our deluxe Blue Elephant guns then. Right guys? Hey, let's all vote! If enough of use all believe in Blue Elephants, maybe he won't matter....besides, internal resistance garble garble inbalance garble garble temperature garble garble creeping inbalance garble garble magic pixie dust yatta yatta yatta.."

Jack Rickard
 
Suppose Jack is right - and the 6,831 cell pack is a moronic idea.

Suppose 180Ah cells are the way forward. I'm OK with that!

But - what's so terrible about charging each one individually, especially if one has that many FEWER to charge, since they're so large?

AndyH, I was worried about EXACTLY what your graph shows - cells supplied at varying SOCs.

I have a bunch of SAFT NiCads that are a little large for my project and am considering replacing them with TS cells - but the (perceived?) complexity of "managing" them is what's keeping me away.

Jack - what would the ideal management circuit actually DO, in your opinion?
 
northernmike said:
AndyH, I was worried about EXACTLY what your graph shows - cells supplied at varying SOCs.

I have a bunch of SAFT NiCads that are a little large for my project and am considering replacing them with TS cells - but the (perceived?) complexity of "managing" them is what's keeping me away.

The best I can offer is to say that I wanted to see for myself how a pack behaves on the road and that's why I installed logging equipment before a BMS. My pack is my 'test bed' - my educational and development pack. I initially thought Jack was on to something with his bottom-balancing idea but based on the year's experience, I have to agree with Doc - it doensn't matter a bit. I cannot herd cats and cannot force cells to stay balanced with Jedi mind tricks. The best I can do is use a management strategy to keep cells within whatever parameters I set. That view seems to be shared by folks here that know more than I do, and by industry.

When Sanyo builds a laptop battery, they put a management system on it so that cells cannot be charged or discharged outside their limits. When Ford, GM, Hughes Aerospace, and Solectria built EVs in the 90s and early 2000s they installed plenty of computerized management. The Ford system - based around 8V lead acid batteries and 12V NiMH modules, monitors each battery module. The 'consumer level' voltage and SOC limits are narrower than the 'pack level' limits - and both are inside the ultimate limits for the pack. The owner cannot drive away unless the consumer level SOC is within limits, the upper and lower voltages are within limits, the module temperatures are within limits, and the modules are relatively close to each other in voltage and SOC. The truck will shut down and leave you on the side of the road to protect the pack.

I've attached two representative looks from other sources - Delphi and Valence.

I'm a dealer for xtreme scooters and have helped owners with pack trouble. I've built some of Gary and Richard's products and sent them to China for installation into scooters destined for Europe. The common thread here is that a bank charger set to 3.62V per cell and a pack-level LVC set to 2.5V per cell average isn't enough protection for a pack - and retailers are losing profits to replacing damaged cells under waranty.

I don't know enough to design a perfect system, and have plenty to learn about the cells, but I'm certain at this point that not managing the pack is not an option if one wants to see even half the service life these cells are capable of delivering.

I hope that helps,
Andy
 

Attachments

  • lithiumcontrol.pdf
    71.8 KB · Views: 120
  • Valence_ucharge_bms_user_guide.pdf
    478.8 KB · Views: 165
  • Uno-shield-Cell1.pdf
    16.7 KB · Views: 34
  • Uno-shield.pdf
    107.3 KB · Views: 31
Jack,

My idea of a nice "BMS" is very simple, let us call it instrumentation and control for your sake. I am from Missouri too, so I understand your attitude fully sir. I agree with statements made here- that simplicity is the way to go.


From here forward, "need" is determined by a users ability to monitor a pack, need for redundant safety measures, and need for an automated battery monitoring system (BMS). This is my basic "need" list with full automation of battery monitoring during run, fully automated charging, and maximal cell life being my goal. I really don't want to spend my time mulling over cell data and testing batteries if I don't absolutely have to.

1. Each cell "needs" low voltage monitoring. This prevents single cell damage. All lithium chemistries have a minimum voltage, whereupon cell damage occurs when it is crossed. This low voltage damage will increase RI and decrease capacity- a point proven thousands of times over by users of lithium batteries around the world.
2. The pack (as a whole) "needs" a low voltage cutoff point as well. This prevents total pack damage assuming that the pack is balanced. This feature is redundant with feature one implemented.

These can be done with an on-pack solution or a motor controller solution. I don't care how it is done, only that it is occurring.

3. Charging needs cell level high voltage monitoring. This is an absolute need, as series charging without this monitor will end with a damaged pack. Sure, a HV balanced pack won't need this every time, but who will guarantee me that my cells won't drift after 10 cycles?


To achieve a "balanced" pack we must either charge each cell individually or we must charge the pack in series with cell level balancing (via discharge or bypass circuits). To discuss low, mid, or high voltage balancing is a null point in my opinion. Series charging will allow balancing at any point. Parallel charging lends itself to HV balancing.

So my goals are minimal maintenance and maximal cell life, and I believe that any electric user would want these same goals. A lithium battery must be kept within HV and LV constraints during use if cell life is of concern, and these three points will ensure it is done.



I can't show you specific data or fancy graphs on my thoughts, but I can tell you my experience. When you tell me I don't know batteries, I would expect that you know my experience before passing judgment. I have been using, buying, selling, constructing, and destructing lithium batteries since 2002. My colleagues are electrical and chemical engineers that know a thing or two about electro-chemical storage. I am merely a penny pinching business man that wants to maximize utility and value on all battery packs. I have watched every pack that is perfectly factory matched on RI, voltage, and capacity drift over time. Cells that are cycled 10 or 100 times before matching into a pack will still drift over time. When it comes to having a long lasting pack, I know that a monitoring system is 1000x more diligent than I could ever be- I am but a human that makes mistakes. Where is the folly in this?
 
Can we all just agree on this:

1. LifePO4 cells are expensive things that
2. Have parameters during charge and discharge that should not be exceeded to maximise lifespan
3. The only thing we ever really do to these cells is charge or discharge at a given rate and to a given capacity
4. Therefore some regulation of these processes is required

But there is clearly disagreement on how best monitor these parameters, and manage the charge or discharge process. Particularly when these cells are grouped into batteries of varying complexity.

And here's where it get interesting, but sadly the good info is diluted by the interweb egos.

I have an interest in this tech as both a user and an investor, so this thread is of particular value to me. I am no engineer, but I am a reasonble judge of people, and I can say that the by the personal attacks smell like someone coming from a position of weakness.

This is particularly important given information is being posted on both sides, but to a casual observer, none is really backed by more authority than another.

Both 'sides' is an admitted oversimplification here.

Please can it, and stick to the real arguments.

This era is such an interesting echo of the turn of the last century, where a bunch of technically savvy engineers and hobbyists built a transportation revolution in their workshops and barns. Back then, I suppose there was just as much disagreement on the best way to manage the equipment. I bet they were far more gentelmanly to one another. Don't let the juvenile culture of the internet turn you into monkeys throwing poo at the zoo. Many of you are clearly talented engineers, let's not waste this opportunity to have an interesting debate.

Thanks, particularly to those of you in this thread that never stooped to this level. You know who you are.
 
I don't know how I missed this thread, it's a good read :)

Hi Jack,

you really should try some of the cells I'm selling. I think you'll be presently surprised at the current they can provide with minimal voltage sag. All the power of the A123 26650s but with higher energy density than any other LiFePO4 currently available and in a form that lends iteself to a pretty starightforward pack assembly. If a single 20Ah cell could be made to work in tandem with the big cells you are using, you should be able to reduce the sag so you no longer see less than 2V per cell. The way I see it, is that if a cell is going to less than 2V even momentarily under load, it's being pushed beyond it's normal limits. A pack built entirely of these cells would offer power levels far above the present favourite for high power EV which is the Kokams.

There is no comparison between using a 2-3C pack at anywhere near it's limits and then using a 20-30C pack of similar voltage/capacity at the same discharge level. Like night and day.
 
Back
Top