Posts
Here are some further detailed photos.
This is the slotted heat sink for the Curtis 1238. This should provide excellent cooling for this controller.
The Spyder filler cap is most interesting. In this case, it allows external connection of the power cord.
This is a shot of the front compartment
This shot shows the rear compartment. Click on any of these for a larger view.
We continue to work on Speedster Part duh. Truth to tell, we did our first test drive this week. It will be featured in next weeks show.
We put some further thought and experimentation into the daughter box and valet function. Essentially, we eliminated the box entirely. We're down to a couple of resistors and a switch wired to the Xantrex LinkPro. Here's the Linkpro connections:
It simply applies a current limit resistor of 100 ohms and a 5k 10-turn pot across the 0-5v throttle output to throttle return lines. Either the valet switch or the Xantrex can do this. We measured throttle inputs TO the controller using teh Curtis 1311 Programmer and it works fine.
We also found that the Curtis would NOT drive our two series Gigavacs. These comprise a 24 volt coil together, but the Curtis did not agree. So we replaced them with a Kilovac LEV200A5NAA contactor and the Curtis is now quite happy.
Jack Rickard
Nice Conversion Duane, hoping to build one soon myself.....won't be letting the Wife do the camera work tho! lol
ReplyDeleteJack - Is there gonna be some 3C Charge/Discharge videos soon, or were you just teasing?? I love watching paint dry....watching grass grow is pretty cool to if it means I learn something! :)
AWESOME SHOWS!!! Love them!! Have over 55GB of you videos saved! (All of them)
To Jack: Congratulations on your webpublishing. You are one the very, very, few people who (in my opinion) know what you're talking about.
ReplyDeleteOn another thread: A North Vancouver EV owner has an Opel sportscar EV with a CVT transmission that he has demonstrated to be Mountain Highway (i.e. very hilly) capable. The CVT is from a California outfit that is building them for EV's.
BTW, an analog voltmeter is useful for LIFEPO4 EV's in detecting when the batteries have "run out of breath" (exceeding the discharge rate) and thus need to rest.
Rob Matthies
a Vancouver Gadgeteer
co-builder, Revived Battery Electric Pickup Truck
99.9% cheaper to run than a Tesla Electric Roadster
matthiesr -at- yahoo -dot- ca
Gents:
ReplyDeleteThat's an elegant build, Duane & Scott, full of nifty little details that make a lot of sense, and every little thing is finished beautifully. Congratulations.
I'm really interested in those battery boxes, and how they are restraining the cells from swelling. These questions are as much for Jack as for you guys, as you seem to both be doing something very similar wth your battery boxes.
I can see from these pictures that at least the rear boxes are the cell restraints, and that there aren't any end plates or strapping holding them together. I'm wondering whether there is some preferred compression scheme of the cells in the boxes, and whether you have shims or some other way beyond getting the right dimension to achieve and adjust that clamping force. Maybe the batteries just don't care that much. At any rate, it will be interesting to see over time whether the end panels of the cases bow out with swelling cells. I would watch the larger boxes very carefully, because those end plates are unreinforced aluminum sheet, and if the cells do swell even a little in there and bow the end plates, it would be a royal PITA to have to recompress the boxes to get the cells out and then stiffen the end walls of the boxes before reassembly. How you would then get swollen cells out of a reinforced box is another good question...
I can't see the front box in the pictures, but appears to have at least one rod running the length of the cells, probably to restrain them. That indicates maybe this set has plates and same strapping setup, because there are a lot of them in a row and they no doubt will exert a lot more swelling pressure, but I'm asking how was that done and what are the considerations?
Same thing for you, Jack. What are you doing about swelling in the RES? (I think for the sake of history, and moving forward, the car needs a simple consisent name, like Rickard Electric Speedster, so that 30 years from now people will just call them "Rickards" at auctions and car shows...) The way your cells fit into the composite boxes doesn't seem (from the available images) to have any clamping system or adjustment for compression of the cells. Is that a conscious decision based on your experience with what these cells require? Are you watching the cells in the RES to see if the design of your boxes is permitting any of the cells to swell?
I'm ignorant of the actual forces involved here, but from the factory-supplied plates it seems there's a fair amount of swelling force that needs to be contained in these cells. In my application, the cells will be in wells that are built into the composite body of my trike. If they slide in there with no further ado, that's great, but I'm not sure if making the wells the precise size the cells need is such a great idea. If they swell in there and become impossible to remove, I'll have a problem.
What are you learning about this?
Tom
PS, sorry to hear the heat exchanger is too small, but I never liked it hanging under the rear axle anyway. You might try the radiator/fan set from a Honda Helix 250 scooter. Look up ebay item 130392808277 to see one. The dimensions and capacity seem about right for you.
OK, here's a better idea for the radiator:
ReplyDeletehttp://www.mishimoto.com/honda-crf150r-aluminum-dirt-bike-radiator-07.html
This would just replace your current fill tank. With a small fan, you'd be all set. Note that you can get it with the connectors on either side (left or right orientation.)
Pretty slick...
Tom
Jack,
ReplyDeleteI now have some time on my Swift at higher ambient temperatures. The heat sink is working fine. Yesterday at 92F outside temp and 60 mph on the highway the controller remained below 45C. I'm also getting much better range with the warmer temperatures. Drove about 63 miles yesterday and was at 46% soc after. Cells in the insulated boxes upon return were at 105 F in rear box inside the car, 90 F in front box which gets airflow around it.
tomofreno
Tom:
ReplyDeleteWe've done quite a bit of plate clamping and so forth in the Mini. It's proven unnecessary. We understand a little better now how the swelling occurs and what causes it. It's almost solely a function of overcharging. In fact, I would say at this point if you experience swelling your can take that as a SIGN that you're overcharging the cells. We're just not experiencing the swelling in the cars.
We're now using a little radiator and fan that came as the oil cooler on the original Speedster. I just received a very nice unit from Summit, but it is too big. What it DOES have is -6AN male fittings already. So I'm still shopping.
Meanwhile the small one actually does a good job. And to my relief, the entire cooling system is indeed very necessary. I ran the car a bit today with it totally shut off. The controller pretty quickly reached 87C despite having a good bit of aluminum under it in the form of the chill plate. It goes into current limit at 85C so I was noticeably power limited. Stopped the car, turned on the pump and fan, and it dropped to 57C almost immediately. I could coax it up to about 63C by firing up a hill full power in the wrong gear. But that's as high as it would go.
My temperature controller is a little flakey. Serves me right. I found it for $35. You get what you pay for.
Jack Rickard
My controller hit 82C yesterday after a long uphill at the end of a 20 mile trip in 85F ambient, no active cooling. I'll try a small fan on the bottom of the aluminum mounting plate to start, and get more aggressive from there if I need to.
ReplyDeleteJRP3
Jack:
ReplyDeleteVery cool...
I don't know if I ever showed you this directly, but there's a store-bought alternative to your cooling plate:
http://cgi.ebay.com/Cold-Plate-single-pass-/380183803497
Keep in mind I have NOT bought this, and I'm typing myself smart, but its cheap and even if it couldn't be sanded flat at home and required a milling operation to produce a flat mounting surface for mating to the controller, it would still be cheaper and (probably) more consistent as a fabbed item than your custom plate set at $300 or so.
Just an idea, and thanks for the tip on the batteries; I will make my boxes slightly long in the stacked cell dimension, with a closed-cell foam sheet to take up the slack, and a couple of shims behind that so I can feel how tight the fit is getting from time to time. Easier than I thought it was going to be. Lighter, too...
Regards,
Tom
We were at about 96F ambient I think. I drove it for about twenty minutes I would guess on winding blacktop country road at 45-55 mph. Pulled up to a stop sign and it was reading 87C. I turned on the pump and fan to the liquid cooling system and it dropped it into the fifties within 10 seconds.
ReplyDeleteRunning from there, I could get it to about 63C when I pushed it.
I was a bit concerned. I had concocted this entire theory on heat for this controller. Largely, it was from observing people in the past:
1. Burning up Curtis controllers. The old 1231s would char themselves. I always kind of thought this happened mostly to people who mounted in fairly enclosed places with no heat sink.
2. Current limit. The controllers generally have gotten better about protecting themselves. I've found a half dozen guys who's performance was almost instantly improved by moving the low voltage cutoff or adding heat sink. The LVC would make sense statically, but the cells sag to that level on acceleration. The current limit works so well, you just think you don't have much power. They basically hold you there at the LVC. Heat works the same way. As the controller heated past 85C, I found my power diminishing. Without the temp view on the Curtis 840, how would I know?
So I concocted about a $1000 set of parts adding much in the way of complexity and possible failure items to make this liquid cooling system, and while it's a great theory, it could just as easily have been unneeded with the controller remaining cool in all regimes.
Fortunately or Unfortunately, it actually does get hot and needs the cooling system quite badly.
Of course, fins and fan are much less expensive and much less complexity. But it takes a lot of fin, or a lot of fan, to do that. Where I have the controller, which is where I REALLY want the controller in this car, there just isn't much vertical room to spare. With the chill plate, I can reach all the connections nicely. With an air heat sink, I don't even know how I would make the connections.
So to some degree, I'm pleased it wasn't a wasted effort. On the other hand, I would not have been sad to remove the whole rig from the car to simplify it.
Jack Rickard
Tom:
ReplyDeleteThanks for the link on the chill plate. Actually that was EXACTLY what I was looking for originally. I knew they used them in ice cream shops and salad bars and so forth. I couldn't find one on eBay at the time.
But we're not paying $300. I think Cape Precision Machine is making it for us for $195. It's sized perfectly for the Curtis 1238, it has quite a serpentine route instead of single pass, and it is a smooth finish which mates well thermally with the bottom of the Curtis.
The thing works like a champ moving heat from the controller to the water. But when it's 96F, you don't have much of a temperature differential. This isn't an air conditioner. It just moves the heat to a heat exchanger. We had to lose our little finned work of art and go to a pretty ugly little radiator and fan. I've bought a nicer one, but it is also larger and I don't know that Brain can get it fit into the car.
As the controllers grow in power, they are increasingly going to water cooling. The Netgain controller, the Solitron, the Zilla all are water cooled at this point. But the little weeny radiators and pumps scare me. So we're using Mezier pumps, AN fittings, etc. And it's pretty standard "car stuff" so replacement parts won't be a problem.
Jack Rickard
I don't know that single pass in this instant refers to the path in the plate- but it rather means that there is a single pathway for a single soda spigot. Multi-pass units have additional in and out fittings.
ReplyDeleteA bigger consideration with this unit is drilling holes into or through it. You'd hate to perforate the cooling lines by accident, so at the end of the day your custom plate with the right holes in the right places, the right size and a flat mating surface is actually a better deal at $200 than starting at half of that and having to create all the details...
Tom
What do you think about the scalable 'platform' Trexa has out ?
ReplyDeleteBuild your own electric car with Trexa
http://www.gizmag.com/trexa-vehicle-development-platform/14023/
It looks like any kind of classic car could be built...
I think the Trexa is not really "out" and we can't buy one.
ReplyDeleteIf they complete their plans and it does become "out" I would be interested.
Jack Rickard
I was wondering whether I'd find anything on that car. Thanks for the ride Duane!
ReplyDelete