It’s cold here in Cape Girardeau for July. Actually, not. It has been over 100 degrees Fahrenheit every day and most of our shop is devoid of air conditioning, making EVTV a pretty miserable place to be, much less pinned to for two hours writing down battery measurements every minute on the minute.
But we are kind of on a mission here. Our enthusiasm for CALB’s new CA series continues to grow. As we look at these cells in more detail, we are finding ourselves quite taken with this latest offering of Chinese LiFePo4 technology. It is enabling for DIY car conversions.
[jwplayer file=”news070612-iPhone.m4v” hd.file=”news070612-1280.mov” image=”http://media3.ev-tv.me/news070612.jpg” streamer=”rtmp://s3einxnpkaij93.cloudfront.net/cfx/st/” provider=”rtmp” html5_file=”http://media3.ev-tv.me/news070612-1280.mov”]
In this episode, we undertook our first cold temperature testing ever. We’ve talked about doing this for years, but really never have actually put a number to the decreased performance we have noticed in winter months. In the event of the Escalade Project, we had decided to heat not only the passenger compartment with circulated ethylene glycol, but also the battery boxes. This so we could safely charge the vehicle in low temperatures but in the hopes of a more seasonal performance as well. Without really defining what happens to these cells in the cold, it is kind of a blind solution to an undefined problem.
Indeed, we still haven’t’ tested a 400Ah Winston battery in these conditions and I’m a little hesitant to undertake that on a battery that size. But one of CALB’s claims for the new CALB CA series is improved low temperature performance.
We rather failed to reproduce there results, but we didn’t use exactly the same methodology. Ours varied in five important respects:
1. Temperature. Implied in the CALB graph is a steady -20C or -4F. We cold soaked our cells to 0F, but during the test, held the cell in ice in a cooler. The temperature rose over the test for both cells from 0 to about 25.5 degrees F.
2. Discharge level. CALB discharged their cells to 2.0 v. We end the test at 2.5v. First, 2.5v is supposed to be 100%DOD. But second, we wanted to compare directly discharge tests we did last week at 88F ambient. Those tests were done to 2.50v.
3. Discharge rate. The CALB test was performed at 0.3C which in our case would be 60amps. We did our testing at a uniform 100A constant discharge which is a little over 0.5C by spec, and about 0.5C for the actual tested capacities.
4. We don’t know WHAT they meant by RETENTION by percent. Retained capacity in Watt hours? In amp hours? In physical weight? In geographic location?
5. We used the larger 180Ah cells rather than the unequal sizes of 60Ah for SE and 40Ah for CA. I find larger cells give longer, slower, larger results less affected by testing errors and noise.
Our first tests were of the familiar SE180AHA cell. This was a new cell which we had tested the previous week at 88F.
The actual decrease in amp hour capacity was about 10Ah or 5%. Not too shabby. But the voltage of the cell throughout the test was DRAMATICALLY lower than the voltage at 88F.
This would obviously affect our performance on the road and the feel of the car. It would not be as snappy or accelerate as well and this very much matches our observations during winter driving. So we really have two results here, diminished capacity and diminished performance.
Actually we don’t. They are one and the same but not apparent in this graph. We measure amp hours because the cells are rated in amp hours and that is what we have good tools to measure with. When your only tool is a hammer, all the problems begin to resemble nails.
Your car using electrical POWER. This is actually more properly measured in watts. As most are familiar, a watt is the product of one amp of current flow at one volt of pressure. And if that current flow is maintained at one amp by one volt for an hour, we can term it a Watt-hour of energy. A level of electrical energy expended over time. We often DO talk about our cars requiring 200 Wh per mile for example.
But a watt at 3/4 of a volt isn’t’ a watt at all. If we compare the voltage and current levels of our test in a minute by minute fashion, we find a cumulative test result of 621 Watt-hours from the SE180Ah cell as shown in the graph below.
As you can see, 535 watts at the lower temperature indicates 86.15% of capacity, a more substantial loss of very nearly 14% in cold weather. This is a very believable figure for me.
The CA series voltage comparison is quite similar:
This is an improvement to over 97% of amp hour capacity, but again, if we look at total Watt-hours it is probably a more accurate representation of “retention”.
This represents a decrease in total energy available of more like 7.75%. This is about 56% of the decrease we see with the SE series and so a pretty serious improvement over the SE series. If you live in an area with frequent temperatures below freezing, this represents a significant advantage to the newer CA series cells. If you live in Florida, Arizona, or California, the advantages are dubious here. But here in Missouri we would pay attention to that pretty smartly.
Raymond Harp joins us on camera here at EVTV with Brian Noto in beginning a small, and hopefully both easy and inexpensive project to convert a John Deere zero turn lawn mower to LiFepo4 power as well. We expect a good outcome here as it should dramatically reduce the noise signature of this heinous device and ideally the torque of electric drive will make it quite a grass thrower as well. We are using a used John Deere, have rebuilt what was actually a locked up junk motor discarded from a previous build, no controller, and probably a series of 10 or 12 survivor cells from my earlier test ministrations. Hopefully this little side project will not distract us too much from our main Escalade project.
The Escalade is not well. Yes, we can start and idle it and even rev it up impressively in neutral. But if we drop it into gear, it moves into gear quite smoothly, but then the ECU slams the throttle plate closed in quite authoritative fashion. It would appear not to like one of our sensor inputs. I’m guessing manifold pressure, which I’ve always thought should be ore tied to current than rpm. Back to the drawing board. I don’t see much reason to spend a lot of time on pumps, heaters and accoutrements when we don’t have it driving yet.
Royce Wood operates an automotive repair shop called K&R Automotive up in Ohio. He’s got a lot of GM experience and most notably with ECU issues. He’s offered to drive down this week with his GENESIS scanner to do an analysis on the Escalade and ascertain for certain WHICH ECU sensor signal is likely causing the issue. That won’t fix it. But it will point us more accurately to where we should be working.
We include a video update on Royce’s electric Mercury Cougar project. He has already signed up to bring the Cougar to EVCCON 2012.
We continue to enjoy kind of phenomenal response to our online components, and continue to work on additions. Our most serious omission is an ongoing problem in the EV community, and affordable and flexible charger.
We did quite a bit of discussion with Elcon. This is a very reliable and low cost charger, but curiously inflexible. You have to set the voltage and current levels at the factory and live with them forever more.
But there is a chink in the resistance this problem has presented. They do offer a CANbus controlled version of the charger. WE are looking at the feasibility of coupling a CANbus equipped Arduino to this otherwise excellent charger, to produce a configurable version of this charger. Any Arduino gurus who wish to assist are welcome.
I personally think this weeks video is about as good as we’ve done.
Enjoy.
Jack Rickard
I personally think this weeks video is about as good as we’ve done.
I don’t know Jack, I think all of them have been excellent. As always thanks for another great show.
Mark
Excellent show Jack, What is Brians expected run time and max current draw for the electrified John Deere? Even though the motor you are pulling off is 20 hp, I would be very surprised if it is actually using more than 8-12 on average, with an occasional burst in heavy grass.
My guesstimate would be about 35-40 minutes with 12 100ah cells.
Brian:
We of the “Great Unwashed” who have converted the more common $800-$1,300 worn-out garden tractors have learned that the more amp hours you can stack into the battery space the better since more run time is very desirable. I suggest 48volts at 180amp hours. Also, if you want better control of the tractor, i.e. the ability to limit the current drawn, an easy ramp-up, etc. control of the motor rpm, software programming, I suggest a small motor controller:
http://www.alltraxinc.com/Products_SPM.html ………….Full disclosure: I have nothing to do with this company!
I found with my basic conversion that had one switch (on/off) and a single speed, that you can exhaust the batteries rather quickly with only one speed available and controlling the speed with the transmission lever. It’s was fun at first because the tractor would take off faster than a snail moves through a goose’s gut.
Jack:
On Apple; How about if Apple can take over A123’s IP and bring the products to market. I really like the specs on the batteries; but, as you infer, it’s difficult to trust A123, especially since they seem to have chosen to do business on the open market with Chinese Traders. Seems to me it takes a lot of money to make ’em in the U.S., ship ’em to China and then ship ’em back. But, hey! what do I know, I just a poor voter, floating in a sea of confusing politic rhetoric.
Jack,
My Sunday and Monday is a pain eagerly awaiting an episode from EVTV
And when sometimes an episode is a day or two late, I have a sense of loss when I’m unable to watch it
Trully, you give a blow by blow of everything that is EV
Having an e-Store from your site is a real service to all the builders out there.
I’ve been missing your road tests as you drive along the places in your town.
Hey why not shoot a behind the scenes video while you do your prep work, do your e-Store orders processing, Bloopers videos, design and ideation video where you draft your circuit design work and do iterations.
I bet that would be fun to watch as well a interesting.
Cheers,
Marl
Holy smokes, that’s quite an improvement!
For us guys up in the Great White North, cold weather performance is of considerable importance. Here in Toronto, winters hover around 0C, say 20 to 40F on average. We get a cold snap down to 0-10 about two weeks a year in total. So these test results are pretty much perfect for practical application for me.
The fact that the CA’s lose only 8% capacity at the *coldest* weather we’ll see is nothing short of amazing. I normally work with lead-acid, and we generally use a 50 to 60% loss of capacity for that same temperature! Even the SE is fantastic in comparison.
Maury:
Some days I feel like Paul Revere, except I’m not sure anybody is listening. THE CHINESE ARE COMING THE CHINESE ARE COMING, and they bring BATTERIES.
Yes, Pb cells lose 50-60% in the cold. SE loses 14%. CA loses 8% And it just keeps getting better. These batteries, nothing else. No other technical innovation. THESE BATTERIES, take the electric car across a great divide, the threshold of viability, from being a novelty and an interesting experiment, to being useful AS A CAR. We’ve waited 100 years. They arrive, and everyone is intent on still waiting…..
I bring you tidings of great joy. You can now build a successful electric car that is SUCCESFUL, as a CAR. It isn’t’ cheap. It isn’t’ free. In fact it is QUITE COSTLY – as things of value often are.
As in many cases, it is not how very elegantly WELL the bear dances. The trick is, can a bear be taught to DANCE AT ALL?
Jack Rickard
Great show..
I am leaning more and more toward the ELCON charger for my project. (At least to Start…) The Manzita just seems too fragile and the manual method to set up the proper charging voltage level is insane. I do a lot of work with bus protocols including CanBUS. I would be happy to give it a try, but looking at the specs, I am not sure you can change the charging curves using CanBUS. Also when you buy the CanBUS model, it looks like it will NOT shut off the charger until it get the signal from the CanBUS. I hope that is not true, but that is how I read their specs…
I know that the little PLC I am using can easily act as a charge controller (It would help justify the cost a little). I picked this little PLC because it has two 20Khz chopping PWM outputs which can drive a buck/boost charger. It is one reason why I got the very Accurate 0-150V to 0-10V transducer. It should allow me to accurate measure the pack voltage (Using a 16bit signed analog input) down to 0.01 Volts. It can be programmed so that the desired charging level and profile can be set through variables on a custom configuration screen. I can easily write the software in the PLC and built an isolated digital interface to the charger. My only problem is that I am a little weak on the necessary analog circuits. (Been about 15 year since I worked on anything with this much current)
The best thing I have found is the kit from Electric Motor Werks… http://www.emotorwerks.com
Does anyone have any experience with this kit?
If you have any other ideas, you can contact me directly at: Jeff@cuttingedgesoftware.com
P.S. Jack, I can program the little PLC to spoof just about any signal. It can probably do it more accurately than any analog circuit. If you can get enough empirical data, I can put it in a table and interpolate the values out of the table on the fly. This is more or less how the ECU in cars do this anyway. I suspect in addition to manifold pressure, you are going to have to re-work the “Torque Vectoring” table in the transmission ECU. Since the torque curve of the ICE and Electrc Drive are so vastly different, it may be the Trans ECU that is shutting you down. I had this same issue on my 2005 Mustang GT that I was stupid enough to put a 10psi Saleen Super charger on. Until I completely reworked the Torque Vectoring table, it would slam the throttle plate shut on hard accelerations. I am sure that GM calls it something else, but it’s operation is most likely similar….
We are looking at the Elcon charger and trying to find away around the madness of the charge curve policies they have been following. They are a little disinterested in the problem and I’m beginning to see their point on this. They are going in a different direction and ultimately CANbus is of course how everything will be done. We will be stocking Elcon as soon as we get this worked out.
They have two mutually exclusive control modes, you can have either but not both.
The one you will no doubt be interested in allows an external device to control the charge current with a 2 to 5 volt signal on the enable pin. Output current is scaled to this analog input. SO your PLC monitors the voltage, holding 5 on the pin until it gets to CC/CV, and then brings the voltage gradually down reducing the current to maintain pack voltage at the CV point until it reaches some nominal value. Voltage below 2.0v shut off the charger.
The other, and the one I’m leaning toward, is the CANbus interface. Actually there are only two messages. The charger sends a message to the controller stating the current pack voltage and current applied. The controller sends the other message – target voltage and commanded current.
The controller is supposed to send its message AT LEAST once per second. If the charger does not receive one for five seconds, it automatically shuts down. Good medicine.
The advantage to this over the other, is that the charger provides the pack voltage and apparently does this fairly accurately. So the only connection the controller really needs is the CANbus connection to the charger and 12v power. Unfortunately ,the 12v output of the charger is limited to 50ma, and so most likely cannot be used for powering the controller.
I have had the charger from Electric Motor Werks. IT is an arduino running kind of a bad boy switching power supply. Simple coil and a single IGBT in buck mode. It works. And it does actually fairly high power. But it is not very mature. The key deb ounce on the Arduino for example is a little hosed – with ready code floating all over the Arduino space. The code is very ugly in this thing. It has some entirely unnecessary calibration procedures. And it is kind of ugly. But it does CC/CV fairly accurately. And it does put out a good amount of power. Promising, but immature.
We may take you up on your offer to spoof a signal. I’m going to try to capture the Manifold Pressure signal on my wife’s ESV to make sure we are looking at the right thing. And Royce Wood is going to come examine our ECU with a scanner to determine that that IS in fact the signal. I don’t think its the transmission.
Yesterday, I went back to a current output from the SOliton with a larger cap and a pot so I could adjust the level. Kinda/sorta worked. IF I held my foot on the brake, and then applied accelerator, it actually drove like a car, a car with very hot and smelly brakes. That tells me we do not have a strong enough signal from the Soliton’s current output at low RPMS. I’m going to do some further stuff just with a pot and 5v and a meter to see if I can manually spoof it. But some translation form the Soliton signals to the ECU is going to be necessary. A linear voltage is not going to do the trick.
Jack RIckard
Jack,
Regarding spoofing the MAP signal… I think it might help a bit to start with a clear visualization of the source of manifold vacuum. It really is a COMBINATION of throttle position and RPM. It’s all about intake flow restriction fighting engine pumping. An idling engine would have a low pressure (lots of vacuum) despite low revs because even though the engine is pumping very little air, the throttle is closed and restricting even that low flow rate.
Think of the ambient air as an input voltage (atmospheric pressure), the throttle restriction as a pot, the air flow as current, and the MAP as the output voltage of the pot, resulting from the voltage drop across the pot. In fluid mechanics you talk about pressure drops through restrictions in a similar way as electronics guys talk about voltage drops across resistors.
Stomp on the go pedal in gear at a low RPM and the MAP goes to atmospheric pressure because the throttle is wide open and the engine isn’t pumping much. Other scenarios:
o Part throttle and high RPM (maybe going down a hill and using some engine braking) could have a low pressure because the engine is pumping too much air to readily flow through a partly opened throttle.
o Same part throttle position and lower RPM (maybe in a higher gear on level ground) could have a much higher pressure since less air has to flow through the same restriction.
The ECU may be using the MAP to decide to downshift when you open the throttle going up a hill, or upshift when you lift off the accelerator a bit on a down hill.
But I have no idea how to actually spoof it!
-Paul
Hi Jack !
Nice to see that you finlly found an interest in the Can-bus option of the Elcon chargers. I suggested that to you on the blog or in an email about two years ago. But you were not to found of adding controller (almost BMS) stuff to a perfectly working charger at that time. But guess what times change….. I don’t blame you for not looking into it, I could have done it myself. You have been busy with other stuff. I have to. I don’t even own an Elcon charger. I totally love your work!! But I do feel sorry for you sometimes. You have all these nice gadgets that you don’t use. Doesn’t the PL8 have a logging funtion that you could use when doing these discharges, or do you really enjoy watching paint dry? Let me guess it has some nice features, butit runs on windows?
Kind Regards
/Per
Elcon charger tuning mods.. Prius owners to the technical rescue:-
http://priuschat.com/threads/elcon-charger-end-voltage-and-other-mods.104914/
And lot’s of chatter about it here:-
http://priuschat.com/threads/elcon-charger-end-voltage-and-other-mods.104914/
Hoping its good and valid.
R..i..i…i..i..g..g..g..h..h.hht.
Andy do you ever LOOK at any of the links you post here. This one goes to a discussion that references a PDF with a malformed URL. No information at all And the Prius owners are clearly clueless as to batteries and chargers in their entirety. They’re having a discussion about A123 cells whether or not they can do 3C.????
Jack
Jack,
I haven’t watched the show, but just read the blog. Did Elcon say anything about the 520V curve to you? It uses the 10 curve presets to set voltages in increments of 1Volt I believe. I think a couple people have doe it so far and I am toying with doing that as well. Can you think of any downsides to it?
Dave:
I’ve done some pretty serious begging on just that topic and they did not mention the 520 curve. They turned down a sale for 20 units doing it my way loading 10 presets and said it is “uneconomic” for them to do this. Where do I find out more about the 520 curve.?
Jack Rickard
hmm, maybe they did it for a few people and now it has opened a can of worms. Someone on DIYEC started doing it as far as I can tell. On the Evolve site, you can go to the Elcon chargers and click on the 500 curve. On the far right is the 520V. Its V4 is set at 1V. I guess they then specify a different number of cells for each of the 10 presets. ie. 151 cells representing 151 Volts up to 160 cells representing 160 Volts.
I will forward you the sheet someone sent me. I am still not sure if it will do what they say it will, but take a look at it and tell me what you think.
Jack,
I had asked you quite some time ago and you had mentioned doing a special curve and I followed your lead. I talked to the guys at Elcon in Sacramento and they built my program like you mentioned to me.
Here is the link to the 520 Curve and the 10 different voltage settings they gave me. It is also set up to charge at a low 3 amps until it reaches at minim of 50 volts then jumps to full current until the set voltage range is set. It terminates at 2 amps. Works great.
I have an Elcon 3000 240 volt charger with this curve and voltage settings.
http://greenev.zapto.org/images/McWadeElcon.jpg
You might want to copy this jpg.
Jack Rickard
Jack:
I am a little concerned about the CA series cells in that the charge curve seems so steep at the end of charging. If you have a long string of cells and one is lower capacity than the rest I would worry about overcharging that one cell. With the SE series it seems like there would be more overlap with other cells because the curve is not as steep and less risk of overcharging a single cell. Other than matching cells is there good way around this? I know you dont like top balancing, but it seem like a potential partial solution. Or is this not as big a problem as I think it is?
Also, do you ever worry about a cell going bad (buss bar) during charging and causing overcharging of the remaining cells?
Thanks,
David Seabury
It is quite steep, but do not be concerned. A flatter charge curve is actually a benefit. But yes, we may back off to 3.45 v per cell rather than the 3.50 we’ve been using.
Top balancing is an excellent way to trash cells left and right. The curve is also sharp on discharge. And that’s where the current is, not at 15-20 amps while charging.
Jack Rickard
Jack
I’m surprised your MAP spoof works as well as it does. From what I could glean on your prior posts, you’re using RPM to drive a linear MAP reading. Not sure if high voltage is high vacuum or vice versa. Either way, ICE just don’t behave that way. On my truck, no load (regardless of RPM) gives high vacuum (around 20 in Hg). Dropping it into gear lowers the vacuum to about 15 in Hg at idle. Under load, I can drop the vacuum to about 3 in Hg at any RPM by simply mashing the accelerator. Alternately, letting off the accelerator gives about 23 in Hg regardless of RPM (until you coast to a stop and the motor loads up again). Lastly, cruising at a constant speed gives between 10-15 in Hg depending on terrain, with a small correlation to RPM.
You can roughly look at ICE performance with three variables: MAP, RPM, and engine load (assuming stoichiometric fuel ratio). Given two, you can figure the third. My guess is the Escalade ECU is seeing RPM and engine load (data from the transmission gear and vehicle speed) and figuring out there is a serious disconnect somewhere and engages limp mode. Assuming that low voltage on your spoof equals low vacuum, the ECU may be figuring this as a massive vacuum leak or sensor failure.
Long way of saying you might be better served by hooking your MAP signal to something relating to the accelerator position rather than RPM.
Stabbing in the dark here, and I’m not the one doing the work, so take it as you will.
I am using an Elcon 4000W charger with the 520v curve. I purchased this charger from EV West. These guys seem to have a contact at Elcon that is actually helpful if you can believe that.
10 options 1volt increments. The only difference I have is they set my battery capacity at 300Ah so the I3 value = Ah/50 = final end of charge current is 6 amps.
Jack you might want to give Matt Hauber a call on this subject.
On another note, I really prefer the 4000W charger over the 3000W Elcon charger. The 4000 has fans which keep the charger much cooler. The 3000W charger would get very hot. The 4000W charger barely gets warm and pushes 31.1 amps into my 38 200Ah cell pack. It’s 50% more $$$ but for me worth it.
Richard
I use an ElCon as well. After the first month of use I sent it back for reprogramming. Wasn’t at all happy with the low cutoff current setting I originally had. One thing I learned was that yes you can have them (at least the non CAN models) programmed however you want subject to the constraints of their model. You can pick the voltage steps between the 10 programmed charge curves, the CC/CV transition point, and the CV cutoff current to be whatever you want. Either ElCon or you can back calculate what that equates to as far as cell voltage, number of cells, and cell capacity for the purposes of programming the charger.
For me right now this is enough to keep me on the road. It would be nice to have a more programmable charger but for the price I am thrilled with my little ElCon. I can still add or remove 2 or 3 cells from my pack without any issues. Just need to select a new charger program. As long as I’m not making radical changes to my car the ElCon should have enough flexibility for me and it can be sent back for more extensive reprogramming if needed.
Hi Richard, What is the front to rear weight ratio in your 914? Just building mine now. Mark
Hi Jack,
Seeing the difference in the cold temperature data I am thinking of laying down a few paths of the pipe tape used to keep water pipes from freezing in the bottom of the battery boxes. Not a very elegant solution but fairly inexpensive and easy and it has a thermostat in the system already
x
Hi Jack,
Looking forward to your power density testing when you receive your order (even if you’re not …).
Now I sure you aware that within the SE range the power density changes between each cell type. But just in case …
The old calb spec sheets rate the 180ah at 5C, the 100ah at 8C and the 60 and 40ah at 10C.
From your earlier testing (and you are practically the only person to have done this, thankyou) we saw the Calb 180ah sag 22% at 5.5C. I have collected one other data point and that is the Calb 100ah also sagging around 21 to 23% at 8C.
This is from two reports from car users so unaware of their testing accuracy. Both of these tests are “real world” tests and include losses through cables and connection hardware like your previous dyno testing.
I’m not aware of any tests on the SE 40ah (or the 60ah) cell. So I gather you’d need to do this in tandem with the testing on the 40ah CA cell.
Best regards,
David.
David:
I think you’re in a little too deep there for me. I would be astounded to learn there were any particular differences in power density between the various cells.
Power Density and Energy Density represent the classic tradeoff in any Lithium ion battery design and certainly LiFePo4. Both are almost entirely a function of the thickness of the active material deposited on the cathode of the cell.
If you apply a thicker coating of active material to the aluminum cathode, you increase the percentage of material of the entire cell that is cathode material. This increases the energy density – the total amount of power storage. But thicker coatings tend to increase the diffusion delays and make it more difficult for individual ions to penetrate to the interior layers and intercalate with the LiFePo4 crystalline lattice. This diffusion delay limits the ability of the cell to deliver a particular amount of current in a particular period.
If you want greater power output, generally you have to decrease the thickness of the cathode material layer. This makes transit of ions into this material much less difficult and so the cell can more quickly produce current and thus power at any particular voltage.
But it does so at the cost of a decreased energy density.
And so you can have more energy storage capacity, or more power delivery capacity, but not both at the same time.
There ARE some tricks. To increase the conductivity of the cathode materials, and so decrease diffusion delays, they generally mix a bit of carbon in the cathode slurry. The degree of this carbon doping of the crystal lattice is part of the magic sauce each battery manufacturer guards as a state secret. Simliarly the inclusion of very trace amounts of rare earth materials such as vanadium, yttrium, etc.
Finally, they can also do minor improvements in the granulation of the cathode powder itself – smaller particles lead to more surface area – energy density, but smaller “tunnels” for ions – increasing diffusion delay again. This leads to complex 3D granulations where the cathode is made up of particles of one size, which in turn are made up of sub particles of a much smaller size.
And so this is the art. CALB is never going to tell me what they did to up the power density of the CA series. In fact, 90% of the people working for CALB are never going to be allowed access to this information.
When actually constructing the batteries, they are not going to use different formulations, and it is likewise pretty nearly absurd that they would have different settings on the equipment that “prints” this cathodic ink, made of cathode powder mixture and some binders, onto the aluminum foil. In fact, getting that thickness just right, at very very small thicknesses in all cases, is probably THE central problem in automating battery manufacture. It is also the central cause of variation in capacity from one cell to the next. Picture the beat of a butterly’s wings as sufficient to throw this value off for a week’s production.
So given what little I know of battery manufacture, and the small difference really between 8C and 10C, and I would say you are peering at your meters a bit too hard, and perhaps reading messages from God in cloud formations. I could be wrong and they perhaps have done this on purpose, but what is the purpose?
As to the specification sheets. I’m sorry. I have looked at too many of them. I’ve found over time I can actually often get them to change by doing an episode on EVTV. They issue them constantly, update them constantly and the values change each time, but in now way consistently or particularly accurately. We actually have NO specification sheet for the CA series at all that even makes a passing lick of sense. I got an internal document with some test results that I’m working from – badly translated and quite confusing.
They have also published a little booklet on the car of these cells that seems to have complied every myth every coined with regards to these cells into one single document. I was amused to see our technique of sanding off the terminals with a piece of sand paper when installing the connections included. PLease do NOT use sand paper by the way. You wind up with sand down in the terminal threads and it is hard to blow out and damaging to the threads. We use a 3M foam sanding block that sheds a very fine dust and much less of it than sandpaper.
Jack Rickard
Don’t forget that there is more than just the cell to look at. In the case of the 180Ah and 100Ah cells the terminals are the same size. There is a similar voltage drop across these terminals in each case, 900A and 800A, voltage drop as a fraction of Ah size is quite different. Naturally the smaller cell will appear to have a higher power density when in reality is likely doesn’t as Jack described in his answer.
David D. Nelson
Hi Jack,
Great show this week (as usual like the last 3 years that I listen to ..). How ironic that I was watching you comment on Apple and Android as I was watching the show for the first time on my LCD TV using my Android Motorolla XT860 wich has an HDMI port wich I recently acquired BEACAUSE it has that HDMI port with the intention of watching the EVTV show in streaming on the LCD TV….. IS it that what is call syncronicity in english?
I’m also glad to see that you carry cables in the store. I do live in the province of Quebec in Canada and the provincial authorities that provides license plate and allow (or not) conversion to be legal on the raod are VERY picky. Not only they do not allow the use of welding cable ; the cable HAS to be automotive grade, orange and UL or CSA or ULC (etc..) compliant. So you can imagine that it is pretty hard to find a company that will accept to provide such kind of cable for a single purchase of 50 feet….
Speaking about charger. I don’t know if anybody could help on that but I have about the same problem here with charger than cable. The only one (canadian) UL certified charger for EV are the Delta-Q. My conversion is a 1997 Audi A4 and I have a Siemens 1PV5133 motor with simovert (short) controller. So tha only way I can see to provide charging for a 300+V system is to hook up 4 84v volt chargers in series for 336 volts (105 lifePO4 cells in series). So I will end up WITH a fixed voltage charger at the same price than a Brusa….. At that point I am wandering about how much will it cost for Brusa or Elcon to get UL certified…and pay them to do so in exchange of a ”free” charger and some share in their buissness… Unless Delta-Q are going to do a programable high voltage charger. Anyways, just my two cents about chargers.
Xavier
X,
For isolated high voltage chargers…http://www.metricmind.com/category/ev-battery-chargers/
These chargers are sold all around the world.
Mark Yormark
Mark:
Had you been paying attention, the first charger we used was a BRUSA. Yes, they are sold all over the world, but if you buy one in the U.S> Victor has a monopoly on them – and they’re about $4000 his way.
Then too, if you blow one up, Victor doesn’t know anything about them, and there is no warranty or support here. Don’t ask me how I know.
So $4000 for a 3kw charger with no support doesn’t really work for me. I don’t’ think it will with our viewers either.
Jack Rickard
Jack,
Give me some slack Jack, I tuned into your blog after the build of the Mini Cooper.
My reply was not intended as a plug for Victor, but merely a source for high voltage chargers no matter the conditions of price and service.
So do you have a source for high voltage isolated chargers over 450 vdc?
I already know you do not like Current Ways.
Regards,
Mark Yormark
We did indeed use the BRUSA on the Mini Cooper Mark. But no, our FIRST build was the burgundy Speedster, which we have now redone and refer to as Redux. It was our very first build and used a BRUSA charger. Speedster Duh uses an Elcon/Chinnic PFC3000 at 1/4 the price which has served flawlessly for several years.
I have commissioned a gent to develop an Arduino/CANbus with display to serve as a controller for CANbus equipped Elcon chargers. We will offer charger and controller as a package you can easily configure for less than $2000 in that 3kw range where the equivalent BRUSA is about twice that price.
RechargeCar is actually developing an Arduino Mega sized board with CAN on the board itself. I think this will be a huge addition to the EV community and enable a lot of things like this.
Jack
I have the 520v algorithm on my Elcon 2000W charger that I purchased from Chenic. I took it to Mark at Zivan/ Electric Conversions, 515 N. 10th St, Sacramento, CA 95811, p 916-441-4161 to reprogram it. He was very helpful. It took about 1.5 hours. I went out to lunch and when I got back it was done. The 520v algorithm gives me 10 algorithms, in one volt increments from a charger shut off from 154V to 163V so I can tune the shutoff point as required by my pack (45 SE180). All this after Chenic told me that the charger would only support one charge algorithm for lithium.
There is also that brusa state that their NLG513 charger ”Complies with all applicable specifications (EMC, harmonics suppression)”: http://www.brusa.biz/index.php?id=206&L=1
If they did the same thing with this model than with the previous that Victor is now selling at ”rebate”, It does mean that they have been designed to meet the standard but are not certified by any testing lab at the moment. Correct me if I am wrong (I would really do want somebody to correct me!)
I received a reponse from Delta-Q this week : ”Currently Delta-Q is only working with lithium BMS/integrators for supply of lithium chargers, we are not supporting lithium charging directly via distributors.”
So to all of you who already have an Elcon or any other lithium charger : I’m really jalous of you! ALL!
Where the hay is the the goldanged show this week?? Just kidding, I was there. Thanks for the weekend, Jack. I wish the magic man could have done more for you. Time is always helpful in these situations, but it wasn’t to be this time around. Anyway, we did take 146 to 57 to 70. It seemed faster once you saw a big map. And here’s news: Indiana still blows no matter which Interstate you’re on. Looking forward to returning to the diamond in the rough that is Cape Girardeau, so we’ll see you next time. K
I’m actually still editing this morning. Should have it up by this evening I would think.
Jack
After pouring over the protocols. I’m warming to the idea of a CANbus equipped Elcon; if the Arduino has a separate circuit to watch its heartbeat independent of the serial port and it is metal cased on a fully isolating DC supply. …. We need to completely cover our bottoms.
3KW is my size, so for many car drivers with space and desire to charge out on the road why not consider the 8KW charger because either Elcon type can be current limited depending on the connected port?
Actually, the Elcon has a very endearing trait, if it fails to receive a message from the controller for 5 seconds over the CANbus, it shuts down entirely.
EMP.ENU shielding is kind of important, and sort of at cross purposes to having an LCD touch display.
Hi Jack,
In all you years,successes, and mishaps of testing these batteries have you seen any evidence of temperature being an indicator of eminent cell swelling? In other words instead of a BMS would a thermal warning system that is not electrically connected to the cell terminals actually be much better?
There is of course a temperature correlation. But I would personally doubt it would be either sensitive enough or sufficiently in advance of the fact to be useful as a Battery Management System.
If you bottom balance your cells, don’t overcharge, and don’t’ over discharge, you really do not have trouble with these cells. There is of course the scenario of one cell going bad and we have no adequate defense of that. Lee Hart’s battbridge would probably be sufficiently sensitive to alert you to that if you were both observant and knowledgeable in that event.
Beyond that, the expense and complication of a BMS is simply not warranted, needed, or worth the expense and the dangers posed to life and limb.
Jack Rickard
I’m not at all promoting any BMS, I’ve always felt any electrical device has the possibility of failure and the worst place to have a short is between two poles of a battery. I’m mainly wondering if swelling, puffing of A123 cells occur or can be warned against by some sort of temperature set point.
Saw this on youtube today I thought you might like it.
Chevy trucks like the Volt featuring Bob Lutz
http://www.youtube.com/watch?v=p5STy6HRZEQ&feature=g-u-u
I enjoy your show every week.
Thanks so much.
“And the same thing could be done with an Escalade” – Bob Lutz
Interesting VIA video.
So much for Bob’s retirement… Can we guess the price of jet fuel, for his 2 personal owned business jets, his fully restored Russian Jet trainer, And his personal jet helicopter, has thinned the Air under his Golden Parachute? Either way, it’s good to see Bob out pitching the battery design (100 mile) that GM should have started with, for the Volt..
I can only say, It’s too bad Toyota wouldn’t supply VIA with the roller. A good paint job and reliable electrical systems, are still worth their weight in gold. Mainly because, it take a little gold and other precious minerals to to make a quality contacts, and paints. Something which still eludes GM and their rust belt associates.
More times than not, an EV owner has a real good understanding, when they expect their desired travel plans too exceed the range of their charge… Using a simple COTS $800 10 hp gas power generator mounted and towed in a motor-bike trailer, could easily provide a 10-15 KWh source, for the average long distance cruising current. Leaving the existing battery charge for acceleration and the odd hill current boost. We just need Evnetitics and Curtis to wake up and add an Auxilaiy AC inlet, with full wave rectifier and the IGBT/Diode switching, into their existing controller designs.Thus, a Gen-set being towed down the long straight freeways to see lost relatives, makes Range Anxiety drop slight below Bladder Anxiety on double doses of Toviaz and Xanax. The nice part.. You can leave the Gen-set parked the other 95% percent of the time you drive to store and back.
Maybe we can even talk U-hual into to setting the standard for the power connector, and corner the Gen-set market. I would love nothing more, than to pull my EV into a U-Haul and connect up to one, for the long drive to EVCON.. And after, maybe drive on up to Bob’s house in Michagan and show him, what an EV Auxiliary Range Extention, really looks like. Maybe try to understand his thinking of why you need a V6 mounted under the hood with a tank of Sour Gas, if not just to add another $30K to the sticker.
And in doing so all of sudden, all this talk of Elcons, and Brusas charger issues becomes mute… As you would then have both the battery charging and motor controller circuitry housed in one unit… Elon Musk didn’t buy into AC Propulsion’s patents, just to bury Alan Cocconi ideas and design philosophy, in typical Microsoft competitive fashion.
If VIA had any imagination they would have a least used Ethanol for fuel, so Jack could get the VIA Escalade version. And throw a few oak chip in the tank, so draining the tank by the glass full on a regualr basis would be at least worth the effort. And hay, maybe a 20 year old Escalade, would finally have some value, at least by the quart.
On another note…
While I fully agree that an active BMS is truly inherently risky to both to Batteries and the World in general. A battery monitoring setup is still (as Martha would say) “A Good Thing” . But like all good things, It should be done in Moderation!. By moderation, meaning you don’t need to monitor each and every cell (KISS is a factor). Simply monitoring 3 of 4 points along the string, and taking an average (Vdrop / n-cells) would easily indicate a section imbalance between each group. When a large imbalance is indicated either during charging or discharging, the actual cause always requires in-depth examination, both electrically and mechanically, Which can’t be done, looking at a PC monitor, speeding down the road at 60+ MPH. We have enough idiots busy texting, and touching-up their makeup, at high speed, as it is…
“…There is of course the scenario of one cell going bad and we have no adequate defense of that…” Jack you talked a while back about maybe splitting the pack into two equal halves and comparing voltages off charge to detect shorted cells. Did you get any further with that? I guess a cell that goes open circuit will make its problem known by stopping the car from working….
We did a whole segment in one show showing Lee Hart’s “batt-bridge” concept in comparing one half of the pack voltage to the other rather closely. If they aren’t within a volt or two, you have a problem Houston.
Jack
Thanks Jack – yes I recall it being discussed. Is it something you are using actively – in the Escalade for example?
John,
Thanks for your hard work. I look forward to meeting you and EVCCON….
Many thanks Jeff – I enjoy it. Yes it will be good to meet many at EVCCON who are just names at present
John,
I did a variation on Lee Hart’s Batt-Bridge and it is a very simple device to make. I wrote about what I did on my blog at http://2003gizmo.blogspot.com/2011/07/battery-pack-balance-monitor.html.
David D. Nelson
Many thanks David – I’ll take a look
Jack,
When did Minute Maid start making brew? 🙂
Jack,
One thing you may or may not have considered is that the ECU on the Escalade is most likely running a PID loop to maintain RPM ( at least at idle). That sounds great except that the tuning parameters for PID loop driving an ICE vs. an electric motor are quite different and if not tuned correctly can easily lead to oscillation. In particular the “I” and “D” coefficients are very sensitive. I was playing with PID parameters for a robotics project and if you set the “I” or “D” too high it damn near self destructed due to violent oscillations. This could explain why you saw the RPM oscillate from idle to 1000 then back to idle and then 2000 etc. The problem is you do not have access to those in the ECU (at least I don’t think you do). In addition you have two PIDs fighting each other (the one in the ECU and the one in the Soliton).
If this is the problem the only solution would be to drive the Solitons direct from the pedal, but then you have the transmission issues. You could try using the ECU to idle the Solitons, but I would be very careful and be ready to shut it down quickly.
Just some thoughts.
David Seabury
Jack,
This is a total shot in the dark but I was wondering if the noise Royce mentioned he was seeing might be eliminated by temporarily running the 12V system entirely off of a 12V battery just to see what would happen. I could be wrong but if this eliminated the noise you could then determine if it had anything to do with the strange behavior you are seeing.
David D. Nelson
Have you seen the new blogpost about this weeks show?
http://blog.evtv.me/2012/07/elescalade-maf-and-map-sensor-drama/
Since you are discussing the show in this thread, I guess not. For some reason it is not featured as the top blogpost.
Carsten, The blog post wasn’t up when I posted that. If I had waited a short while it would have been. That is why the earlier posts are here and not on the new blog post.
Here is a recent article about the possible merger….
http://blogs.marketwatch.com/thetell/2014/02/16/apple-held-secret-meeting-with-tesla-san-francisco-chronicle/