OT?: LiFePo4 battery
plasmadust
lithium iron phosphate bats that are starting to get mass produced.
This technology has been around for more than 15 years, but it looks
like it is finally going to be available to all.
http://www.gebattery.com.cn/product/lf.asp
I am getting some in for evaluation, will let you know how it goes.
DC
Joe
I thought this was going to be about life post 40.
Best,
Joe (mobile)
Jeff at EV Bones
On Mar 1, 9:26 am, "plasmadust" <dcanfi...@evtech.us> wrote:
> I thought I would put a link to the write-up from the mfg. about the
> lithium iron phosphate bats that are starting to get mass produced.
> This technology has been around for more than 15 years, but it looks
> like it is finally going to be available to all.http://www.gebattery.com.cn/product/lf.asp
> I am getting some in for evaluation, will let you know how it goes.
> DC
Hi Doug.
I've been testing some chinese made 26650 3Ahr cells for a few months
now.
Charging is at 1C, discharge at 4/C, and they're holding up well, with
no capacity or impedance issues at 230 cycles to 80% DOD.Temp has been
maintained between 24 and 30 degrees C throughout the testing.
The US rep claims better than 1k cycles to 100 % DOD, so we'll see.
I also began "spanking" paralleled clusters of A123 M1 cells. Though
only rated to 2.3aH each, the 30 cell clusters have been constant
current discharged at 300a to 85-90% DOD (~2.25v). Amazing batteries.
The biggest challenge has been to devise a volume-production cell
module interconnect structure to withstand an automotive application
while keeping the weight down, and not restrict module ventilation.
Regards, Jeff
Devin
on the road
deerfencer1
only rated to 2.3aH each, the 30 cell clusters have been constant
current discharged at 300a to 85-90% DOD (~2.25v). Amazing batteries.
>>
Holy cow, Jeff--amazing is the word! What kind of amperage have you
been recharging them at? Time to recharge? And what kind of device do
you use to draw them down at such high power--some kind of super-duty
load tester?
Larry
deerfencer1
looks
like it is finally going to be available to all. >>
Doug,
Why has this battery tech been held hostage so long? Not another big
oil conspiracy I hope! I know nothing about these batts, so appreciate
the link.
Thanks,
Larry
Jeff at EV Bones
On Mar 1, 3:15 pm, "deerfencer1" <deerfenc...@aol.com> wrote:
> Holy cow, Jeff--amazing is the word! What kind of amperage have you
> been recharging them at? Time to recharge? And what kind of device do
> you use to draw them down at such high power--some kind of super-duty
> load tester?
>
> Larry
Larry, as the A123 cells are paralleled in groups of 30, they are
constant current charged at 50 amps to 80% capacity, then constant
voltage to full. Full is achieving 3.7v at 500ma/cell finish.
I've been told these 26650 size cells can can withstand 70A discharge
for a few seconds. Wow. I'd never consider that kind of abuse. I have
the cell cluster interconnect straps sized to fail at 50A if an
individual cell is damaged and goes to short, as it's parallel wired
kin can push 1500A into it.
Full charge requires about an hour and a half per cycle. Discharge is
less than 20 minutes.
The discharge system is limited to 300 amps max. I'd like it to be
double that, but they are just too expensive.
The discharge tool is a Dynaload electronic load. Basically a buttload
of parallel wired TO-3 transistors in a big rack mount air-cooled
system. It was later modified for operation down to 3v. Model
RPL-50-300-3000
You can see an image of one here:
http://www.testmart.com/search/asp.cfm/ELECLO/DYNALO/RPL-50-300-3000.html
The charger is just a monster size Hewlett Packard lab supply with
programmable output control.
Both the charger and discharge load are controlled by a DAQ board from
National Instruments.
The DAQ board is peecee controlled through an application program
called Lab View. Assembled recipes can control external devices
through a graphical user interface. It also monitors a half dozen
temperature sensors, along with a shunt to track current and calculate
the energy I/O from the cells.
A similar system is also testing chinese made 2.8aH 26650 LFP cells
wired in clusters.
And two big ABC chemical fire extinguishers. Hope I'll never need
them.
Regards, jeff
deerfencer1
Thanks for all the details. Have to admit I probably don't understand
90% of it, but I'm sure the engineers and more technically savvy here
really appreciate it. Due to your ongoing exemplary efforts and
excellent contributions here, I hereby nominate you as Honorary Mad
Battery Scientist of the TF Google Boards, with all the perks that
that honor confers--i.e. unlimited complimentary access to Uma's 900#
Hotline, a free lifetime supply of fire extinguishers, and a roundtrip
ticket to China for "patent research"--as long as you bring back the
goods <g>.
Best,
Larry
BTW, what are your plans for the A123s if they test out--anything you
can share? And how are the Cinese cells looking?
dermot dobson
did some 5A charge / 20A discharge tests then did a 1hr charge to
full and a brief ride on my Heinzmann 500/750W Trek6045 - just a
couple of miles.
So far, so good! Eventual plan is to have a 36V pack of parallel
cells (depending on price) of 3 - 6 cells, so 6 - 12AH.
What are people using to charge these things?
If there isn't a good and easily available charger - that helps to
ensure long pack life - I'll have to brew my own.
Current thoughts would be to use an old PC 350W or 400W power supply
(or available new dirt cheap), and rewind the output side of the
transformer (plenty of info on the net for this - ham radio operators
are always looking for 13.8V at 30A PSUs for instance).
I'd wind 10 individual secondary windings, each would have a simple
CCCV (ConstandCurrentConstantVoltage) circuit to suit the cell's 3.6V
charge regime. An LED on each charge circuit would give me some idea
of the pack's health - or at least I'd soon spot if one cell finished
its charge cycle early or late.
That way, I'd avoid any concerns about charge balancing a series
stack, as each cell would only 'see' its own individual charger.
Haven't yet decided on a suitable 11 pin connector. Alternatively, I
could just build that a switched mode circuit with those multiple
outputs into the pack itself and run in a suitable DC supply from a
dumb mains psu. 12V would bean interesting choice, the e-bike packs
could easily be charged from a car battery(*).
I've kept the electronics block from the DeWalt original pack - might
hack it open to see what's inside.
dermot
(*) I do already have one of these in bits somewhere: a 150W switched
mode device, 12VDC input with a 3A max output controlled by a single
smart charger chip for NiMH or NiCd. Mostly done, but never quite
got round to finishing it off.
Far better route than a 12V to 240V inverter then running a standard
charger off that.
JJ
I've been running these packs(Dewalt 36V) set up as two packs in
series for a 72V Xlyte on my homebrew since September of 2006. They
are good. I have left them in the original packaging and still use the
Dewalt charger daily. There are about 500 kilometers on them at this
time. When I first started using them I was new to the e-bike world so
everything was experimental to me. The two packs will take me about
six km at 30km/hour. That's without pedaling. If I pedal in my highest
gear I can usually keep up with the motor and maybe add a little extra
power to the system.
Normally I keep the speed limited to around 30kph for a couple
reasons. 1) I don't want silly run-ins with the law. and 2) Much more
that that and I'm pushing the packs hard enough that they are close to
100% DOD when I get home. That said I have a fun little story about
the other day.
I live in central California so I rarely get wet. But I got caught
Tuesday after a beautiful sunny ride to work in the morning. It
clouded up and rained hard most of the afternoon. I believe it rained
hardest on my ride home. I was able to make some preparation and cover
my electronics and batteries to protect them. Due to this covering and
the rain I was unable to monitor my speed or remaining charge on the
way home. I reached about a quarter mile from the house and the bike
just flat ran out of juice. While that is normal for the A123
batteries when fully discharged from everything I've read, it had
never happened to me before. So of course I finished the ride home
under pedal power only. Here is the fun part. My previous max speed
while pedalling was about 47kph. My Drainbrain in this case said my
max speed was 61kph. I had to do a bit of a doubletake and am still
not sure I believe the reading myself. I have occasionally run the
bike at 47kph for short distances without pedalling just to see what
it would do. Never expected to approach 60kph though.
And the even better part. Early on I ran the batteries pretty hard
just to see what they could withstand. In a couple of cases the packs
would not recharge when I was done with the ride. Fortunately I
discovered the reason why and was able to repair the problem. One cell
in each of the two packs I had go bad was down to 1V. Apparently the
Dewalt charger is not happy with this reading and tells you that the
pack is bad. Guess what? If you charge the one cell individually until
it reaches a level close to the other cells and then reassemble the
pack, the charger is happy again and the entire pack recharges. This
time neither pack caused a problem and both charge normally even
though I ran them completely empty.
Other good things. I limited myself to a 20A controller so I wouldn't
get too out of hand with speed. I have run up to 29A(short periods)
fairly regularly without damaging the controller(might have something
to do with the aluminum plate it is mounted to for heat sinking) or
the packs and not toasted its electronics. On a normal quick run to
the local grocer I will often run the 750W motor at 900 to 1100W for a
few minutes at a time with no noticeable adverse effects. Voltage drop
at these levels puts me down to around 52V which is about 18-20A.
Again no noticeable problems at this time. When the weather turns
warmer this summer that could change.
One last item. I have mentioned this at another site. the Dewalt packs
are not true 36V packs. Each cell is rated at 3.3V but charged at
3.6V. If I were to build a 36V pack, there would be 11 cells in the
pack. I verified this in a couple of ways. I wrote to Dewalt and they
told me that their algorithm is set not to charge to a full 36V
normally. Also I meter or check the final voltage daily and normally
see a max voltage of around 34.5V.
Sorry about the dissertation
JJ
Richard Papa
Fascinating stuff guys and I am soooo excited for you both. Seems these
cells are simply amazing!
I too like the idea of running an extra cell to give you 37v as that would
keep the performance up :) I wonder if we can have these in out TF bikes?
Please keep us posted you two pioneers you.
Take care,
Rich
dermot dobson
>I too like the idea of running an extra cell to give you 37v as that would
>keep the performance up :) I wonder if we can have these in out TF bikes?
There is an alternative solution here, that might be of interest for
other battery chemistries where the voltage droops a bit with discharge:
I have a small PCB designed for mobile ham radio use, it produces a
steady 13.8V from the varying output of a (nominally) 12V car battery.
Here's the idea: when the engine on a car is running, there is always
enough volts for the regulator inside a ham radio transmitter to run
at full power, but when the engine is stopped, battery volts soon
drop to 12V or lower - especially when you pull, say, 20A for the transmitter.
This device (designed by a friend of mine) uses a 150W high frequency
switched-mode power supply that adds up to several volts on top of
12V to achieve a regulated 13.8V. Since the transformer/psu only
handles the 'top up' voltage, the 150W one is capable of,
say, providing up to 5V 'top up' at 30A, giving for a maximum of 1KW
on a nominal regulated 36V supply.
I've modified it to operate on 36V+ input and it seems to work quite
well. Only bench tested so far, using a big resistive load, needs
quite a bit of tidying up before it could go on the road.
The transformer runs a bit warmer than I would like, but I had to use
multiple parallel thinner wires for the windings - better efforts
there should fix that problem.
In 'real' operation, the circuit would do nothing when the battery
volts are high, then gradually cut in with increasing assistance as
the voltage falls.
I think you'd need a DrainBrain or similar, or you'd have no way of
knowing that the batteries are approaching flat.
I think NiCd has the flattest discharge curve, probably conventional
Li or SLA is the worst. I do notice a bit of voltage drop-off with my
13AH NiMh pack, so would even help there.
dermot
Richard Papa
It's the sag in the NIMH that is the problem as Lithium seems to hold on
better. I see your talking about adding enough power to stay true 36v and
this is something all the packs should try and do. Love the idea of a true
36v pack and you have a way of doing it (that was what I was trying to say
with 37v).
Cheers my friend but you're way ahead of me/all :)
Rich
dermot dobson
>It's the sag in the NIMH that is the problem as Lithium seems to hold on
>better. I see your talking about adding enough power to stay true 36v and
>this is something all the packs should try and do. Love the idea of a true
>36v pack and you have a way of doing it (that was what I was trying to say
>with 37v).
That would certainly work, but 2x5 packs of cells physically fit
together so neatly, in a way that 6+5 might not...
11 cells is probably the most sensible idea, from an electrical
standpoint, unless the higher fresh-from-charge unloaded voltage
might be enough to pop some electronics - or breach the law on power
output, or course (grin!).
A friend of mine has an old 36V 150W Powabyke - normally powered by
3x 12V 12AH SLAs to form a monstrously heavy box.
Just one tiny, lightweight pack of A123's should run that for 30
minutes - somewhere, I have an old battery box, so might give it a try.
dermot
deerfencer1
are not true 36V packs. Each cell is rated at 3.3V but charged at
3.6V. If I were to build a 36V pack, there would be 11 cells in the
pack.>>
JJ,
Thanks for that great report.
Re adding cells due to A123's conservative voltage range, I'm almost
certain you could go to 12 cells per pack and our TFs would run like a
dream. Evidence? Many of us riding the Point 1 lipos regularly see
close to 42V on our voltmeters off a fresh charge. 42-34.5 = 7.5V
leeway here, very close to the additional voltage you'd get by adding
two more A123 cells.
Larry
JJ
I think one could build a pack of almost any size. I agree with Dermot
about the physical sizing and have been checking into methods to
create a clean installation for the odd packaging. Here is a link to
my favorite idea straight from A123.
http://www.a123systems.com/html/products/packSys.html
Click on the ckt. board for an expanded view. In their photo notice
that the cells are only on one side of the board. I would place cells
on both sides. Obviously have to pay attention to heat gain but the
way the Dewalt packs are fully enclosed I don't believe heat would be
too large an obstacle in a small system. The next part of the design
would be to place the charging/balancing circuitry directly on the
board for the individual cells. I like this circuit at the following
link.
http://www.metricmind.com/ac_honda/main2.htm
This guy is seriously wired and definitely appears to know what he is
doing. To see the actual schematic and pictures of his board click on
his links toward the bottom of the page.
As I see things, if his circuit works as he says, all a person would
need would be a power supply adjusted for the particular number of
cells on the circuit board(i.e 3.6V/cell). So if twelve cells were
installed one would need a supply capable of 43.2V. Then the only real
concern would be determining the max rate of charging from the supply.
A123 recommends no more than 10A to extend the cycle life. I know
there are others trying higher currents for quicker charge times.
Very good information is coming from Everydayflyer over at the RC
group. He has a data acquisition system connected to his planes and
cells and regularly updates his information.
I am really not familiar with how the TF battery packs are laid out
other than the fact that they are located in the front wheel. Anyone
want to give me a quick tutorial? How are they internally mounted? And
what is their original chemistry? Are they difficult to replace?
Also Dermot,
The voltage regulation circuit you described would probably work ok
but it might be simpler to find a dc-dc convertor. There are many on
the market made specifically for electric vehicles and some are
capable of handling high current outputs. Can't give you a specific
model right now because I haven't researched them that much. I've just
seen references to them in reading.
JJ
Richard Papa
As you can see they are welded. Series in a spiral pattern. Each cell
having individual charging/monitoring in a truly engineered design. Only
drawback is the weight. Near double that of the cells but you need a
shockproof housing to be a hub. Plenty of space in there though,
Rich
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com] On
Behalf Of JJ
Sent: 04 March 2007 18:08
To: TidalForce Forum
Subject: [TF] Re: A123 cells
dermot dobson
>Also Dermot,
>The voltage regulation circuit you described would probably work ok
>but it might be simpler to find a dc-dc convertor. There are many on
>the market made specifically for electric vehicles and some are
>capable of handling high current outputs. Can't give you a specific
>model right now because I haven't researched them that much. I've just
>seen references to them in reading.
>JJ
What I am describing is essentially a DC-DC converter but with the
difference that it only handles the 'top-up' part of the load, so it
can be much smaller/cheaper than a conventional implementation. The
major difference is that the feedback senses the *total* output
voltage of the stack yet regulates just the additional voltage from
the converter to make it stable.
With a bit of hunting, it can be built for maybe $35 or so - far
cheaper, I suspect, than a commercial one.
And, of course, much more instructive and fun!
I probably wouldn't use it to replace extra cells, but to turn a real
world 'soggy' battery into an essentially 'perfect' regulated one
makes a bit more sense. Most of the single chip controllers
incorporate an undervolt cutout, and for those simpler e-biks that do
not have that function in the controller, it'd make sense to include that too.
On the topic of DC-DC converters, I've been meaning to get/build one
to take in 36V and output 3V for a couple of the Luxeon 3W LEDs. A
couple of Amps to the LEDs will only represent ~175mA from the main
pack - not a significant drain compared to the motor.
dermot
Nimbuzz
two threads are together. What are these pics of -- TF front hub? Are
we talking about putting A123 cells in a TF hub? If so, I just bought
a dead TF hub for that purpose.
Please clarify,
Unc A
Devin
in the hub, they used 3 parallel strings of (10) A123 Lithium for
triple the range... and since the cells are so high powered, they
could up the amperage draw of the motor to 300+ amps =) thus giving a
~10kw power system, 70 mph top speed.
deerfencer1
It's a Google glitch triggered, I believe, when someone e-mails a new
post and topic in reply to an existing post within an older folder.
Google gets confused and renames the older folder rather than starting
a new one, thus creating a two-headed beast that will ping and pong
between subject headings depending on which post gets replied to. Rich
sez Google is aware of this bug and attempting a fix. In the meantime
there's no easy way I can see of cleaning up the "contaminated"
folder. Now aren't you glad you asked <g>?
As to your other query, yes, we're talking about the A123 cells that
several members are playing with now--JJ, Dermot, and Jeff for three.
In any case, Dermot and Jeff are the most technically experienced here
by far and can likely help us pikers along as we come to terms with
these wonderful batteries.
Best of luck on your project--exciting stuff!
Cheers,
Larry
deerfencer1
hint Bob <g>) put together an A123 pack complete with BMS for use on
our bikes (or Clytes for that matter) as an auxilliary plug-and-play
pack? Bob could build the packs to individual specs/wants/needs at
whatever voltage or Ah the customer calls for.
Initial cost would obviously be higher than the Chinese packs we've
been playing with, but I'm guessing the A123's might be close to
lifetime packs if cared for, so in the end a better deal.
Larry
Steve Mooring
Would it be possible to send us a couple of photos of how you have your
"original packaging" A123 cells setup to mount on your bike?
Thanks,
Steve
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com]
On Behalf Of JJ
Sent: Saturday, March 03, 2007 9:00 PM
To: TidalForce Forum
Subject: [TF] Re: A123 cells
Hi Dermot,
One last item. I have mentioned this at another site. the Dewalt packs
are not true 36V packs. Each cell is rated at 3.3V but charged at
3.6V. If I were to build a 36V pack, there would be 11 cells in the
JJ
Very cool. What a neat trick these guys did. Is this your shop with
all the battery packs? Anyone out there try using these with a
Crystalyte motor?
Thanks for the pix...gives me something to ponder.
JJ
On Mar 4, 10:23 am, "Richard Papa" <papap...@btconnect.com> wrote:
> I wonder if A123 will do an F type cell?
>
> Motor%20assembly%208%20-%20Battery%20packs%20are%20assembeld%20the%20same%20way.jpg
> 174KViewDownload
>
> Motor%20assembly%209%20-%20Battery%20minus%20shell.jpg
> 178KViewDownload
> > Larry- Hide quoted text -
>
> - Show quoted text -
JJ
I think Rich posted them here several months back. I'm not sure how to
add a photo here yet but I have them posted at the following link.
http://www.rcuniverse.com/forum/m_4803317/mpage_2/key_/tm.htm
JJ
> > charger off that.- Hide quoted text -
JJ
six miles.
JJ
deerfencer1
lion packs, in rereading an earlier post of Jeff's (EVBones) about
using the Dewalt/A123 packs as is, it appears you need 5-6 in parallel
for an "unstoppable" ride on our TFX bikes due to DW's conservative
BMS and the high current demands of our X bikes. I personally think
most of us would want that kind of range offered by six of these packs
anyway (18 miles or so based on JJ's results to date on his double),
and of course the more the merrier as far as speading the load out
over the individual cells.
Here's Jeff's original post:
http://groups.google.com/group/Tidalforce/msg/35cb9c2ed9c22b30?hl=en&
Larry
mace
been an inspiration for several months now.
Due to lack of electronics knowledge, I'm limited to simple means of
charging (with off the shelf items). Stuff I use with RC electric
models. The Astroflight 112D charger, meant to be used for up to 40
NiCd or NiMH cells (using a peak detection means to stop a charge) can
be used with A123 cells using the Dapter from slkelectronics.com.
I've used this setup on 2S and 4S A123 packs, but not yet with the 10S
pack. Mostly now I use a 60W photovoltaic panel and the Dapter to do
the job on the 2s and 4S packs. Works like a charm. 3 PV panels in
series should do a 10S pack in about half an hour, with the Dapter to
cut things off. The Dapter senses the number of cells and terminates
charge at 3.7 V x number of cells. The PV panels put out up to 4
amps. (I have 6 of these PV panels not dedicated to anything else
right now.)
mace
Nimbuzz
#16 post? If they are TF hubs why are they in this thread -- I find no
explanation of their presence. If they are being discussed for adding
A13 cells -- well let's discuss it!
Uncle man
Richard Papa
Someone? Changed the Subject of this A123 thread, no idea who?
Rich
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com] On
Behalf Of Nimbuzz
Sent: 05 March 2007 05:48
To: TidalForce Forum
Richard Papa
Rich
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com] On
Behalf Of Devin
Sent: 04 March 2007 21:33
To: TidalForce Forum
dermot dobson
I limited my test on a single pack to 7A, to be a bit closer to
realworld usage.
First high current cycle (only done a couple of charge/discharge at
1A up to now):
Charged to full, then left overnight (which matches my normal use).
offload voltage, 33.5V, dropped within a few seconds to 31.5 at 7A.
Voltage stayed pretty stable until close to end of test, when it
started to drop sharply, so I disconnected the load before it hit the
minimum voltage per cell.
Approximate capacity achieved - 2AH - could probably have squeezed a
bit more out, but aiming for pack longevity.
Measurements taken by hand with a DMM and a clock - so doubtless some
errors there.
I'll wait until I get the other packs stripped and paralleled up
before doing too much more, I don't want to mix cells that have seen
significant differences in their usage.
I've been taking some pics as I go along and will post them -
currently looking at how best to physically arrange the cells in
series/parallel configuration.
dermot
Richard Papa
There were some guys on another group who used aluminum tubing (Natural heat
sink) to house their non BMS regulated cells. I like the idea of using a
manual BMS (Drain Brain) to keep an eye on things.
Keep it up my friend.
Cheers,
Rich
Btw, We also just found out WHO is recycling the messages with new topics
(It's you) :)
Sorry my friend but Google has a bug in the software that does not recycle
new topics very well and I will try and explain.
When you receive a message via e-mail, it has a Topic title in the "Subject"
cell. If you change this and start a new topic with an older message then
the website will change the original topic title in the original topic.
So when starting a new topic please send a new fresh e-mail to
Tidalforce@@@@@googlegroups.com (@@@@@ = @) with the new topic title in the
"subjects" Cells. Sorry my friend, you did nothing wrong and I should
explain this to all new members.
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com] On
dermot dobson
this is old news for many, but just documenting stuff as I go).
The NiMH pack contains 10 cells, 12Volts of 13AH = 156 WHr at a
weight of 10x250gram = 2.5Kg.
The A123 pack contains 10 cells, 33.1Volts of 2.3AH = 76 WHr at a
weight of 10x70gram = 0.7Kg.
So, assuming other matters are equal, A123 is around 57% of the
weight of NiMH for a comparable energy density.
And around 30% of the weight of a comparable pack of NiCd cells.
These are nominal, rather than measured Ah values - in practice, due
to the (claimed) superior ability of the A123 cells to approach 100%
DoD, you can probably argue that A123 is a bit better still, since
you wouldn't want to do that routinely with NiMh.
My original 36v 13AH NiMH homebuilt pack of GP Batteries cells is
still going strong after 3 years of daily use, just rather heavy! I
doubt that I've discharged it much more than 50% - and then on only
a handful of occasions.
This shows my current setup -
http://knoxie.ath.cx/new%20site/electro/dermot/HTML/index.htm -
except that the 6 sticks of 13AH cells were mounted on the bike
shortly after these photographs were taken.
dermot
deerfencer1
configuration. The whole conversion looks very professional and pretty
darm sleuth. Thanks for the pics!
Regards,
Larry
dermot dobson
controller for A123 cells today - using a handy digital X-Ray set...
Here's the result. A bit difficult to see very much detail; I had to
use a moderately high Kv to get through the Aluminium backplate, so
lower attenuation parts like the semiconductors are not very well
shown. I'm guessing that large the white (white=higher density, in
x-ray images) objects at upper right and centre left are
ferrite-cored inductors.
The two large rectangles to the left and again at lower right are the
connectors for the wires that went off to the individual cells. At
top are connectors where it connects to both the drill and the charger.
I've reduced this image in size quite a bit, the original was around
5Megapixels, but that would take a long time to load.
dermot
(I tried to create a new topic in the approved googlegroups fashion,
but strangely, it didn't seem to work, so tried again here)
Steve Mooring
Steve
-----Original Message-----
From: Tidal...@googlegroups.com [mailto:Tidal...@googlegroups.com]
On Behalf Of dermot dobson
Sent: Wednesday, March 07, 2007 1:41 PM
To: Tidal...@googlegroups.com
Subject: [TF] Re: A123 cells
Jeff at EV Bones
On Mar 7, 2:40 pm, dermot dobson <dermotbdob...@gmail.com> wrote:
> Ever-inquisitive, I thought I'd have a quick look 'inside' a DeWalt
> controller for A123 cells today - using a handy digital X-Ray set...
> dewalt-a123-x-ray.jpg
> 75KViewDownload
Cool! Well if you want to know what it is doing exactly:
MCU monitors all ten cells individually.
MCU can control individual cell charging rate by means of analog
switch IC.
MCU tracks total energy I/O, time, and one temperature sensor.
Shunt is employed to report energy I/O using MCU ADC channel.
I see the analog switch in the xray (the quad flat pack at lower
right, but the MCU does not appear in the image. Perhaps since it's on
the other side, it might not be easily visible. BTW, the shunt is at
upper left, next to the LDO series regulator.
Really, it's nice bit of engineering, and it cost them (B&D) a lot to
produce it. I'm guessing a bit over $20 in tens of thousands quantity.
How'd I know what was inside the potting?
Heatgun and a dentist's tool set, followed by generous agitation in a
50/50 mix of Methylene Chloride and Formic acid. Also known as "secret
sauce" for when you need to look into a competitors epoxy packaged
semiconductor IC. And it worked afterwards.
Me so bad.
Regards, Jeff
dermot dobson
>Cool! Well if you want to know what it is doing exactly:
Nice description!
Do you know what the maximum current capability might be?
I wonder if it would work to get, say, five cells paralleled onto
each point where the controller expects to find one, then reduce the
value of the shunt by a factor of five so that the controller doesn't
cut off too early.
That way, you'd end up with an 11AH pack but still using the original
charge controller.
But I suspect that this is an overly simplistic idea...If nothing
else, I'd guess that the charger has a watchdog timer as a back up to
the smart electronics.
dermot
Jeff at EV Bones
>
> Do you know what the maximum current capability might be?
Dermot,
The internal mosfet will open when the shunt detects greater than a
~15 amp load peak average, or 25 amp "short".
The gating mosfet is the item directly left of the quad flat pack
analog switch.
>
> I wonder if it would work to get, say, five cells paralleled onto
> each point where the controller expects to find one, then reduce the
> value of the shunt by a factor of five so that the controller doesn't
> cut off too early.
>
A good idea, but the mosfet in series with the output cannot
accommodate those large currents.
The module is basically a paperweight outside of the intended
application.
But, that was a really good idea.
Regards, Jeff