Monday, April 16, 2012
Throttle Bodies and Bottom Balancing
This week we do some final testing on our dual Soliton1's and dual 11 inch motors before breaking down the test bench to install the motors into the Cadillac Escalade EXT.
One of the issues with this vehicle is where to get the signal to serve as the "throttle" input to the Soliton1's.
Of course, we have two such signals routed from the accelerator itself to the Engine Control Unit or ECU. We could simply steal one and use that.
But the Cadillac is complicated by a lot of systems. The ECU is the heart of the beast and it takes a lot of factors into consideration before finally sending a drive signal to the throttle body. The engine of course has fuel injection for the gasoline, but the actual engine speed and power is regulated by controlling the flow of oxygen containing air into the intake manifold. This is done with a throttle body - a round hole with a round plate we used to call the butterfly valve in the carburetor days of the 1960's.
As it runs out, the throttle body has two potentiometers each provided with 5v from the ECU. They move with the throttle plate. They actually operate in different directions and at different scales. And so one varies from about 3.8v down to something less than a volt, while the other ranges from 1.2 to up over 4volts. The ECU uses these two signals with a function in software to determine throttle position, so if either one is "off" by any appreciable amount, the mathematical relationship between the two signals fails and the ECU will shut down the system.
So we have to be careful not to "load" one signal down by feeding it into the Soliton. I'll probably build a little opamp buffer for the signal. Something else to fail. But it will avoid throwing the ECU into a data storm.
We've had some interest in our little cobble up to bottom balance cells. Understand that this was cobbled together in an hour out of stuff laying in the pile on my bench. And it was mostly a test case for the little voltmeters we found. It actually works pretty well. Sufficiently so that we had some viewer interest in purchasing them or at least a parts list so they could make their own and indeed two have made their own.
The problem with this is that in many ways, your mission and mine are forever different. The bottom balancer I built has large components that are easy to identify, and it makes it easy for me to describe on camera what they do and why we want to do that.
To ACTUALLY do that, there are some better choices. They don't show particularly well on camera, and they don't illustrate the principles quite as well. But they are less expensive, do more and are easier to operate once you get past the small learning curve.
I've reviewed four or five of these, and I think the winner hands down is the Revolectric 1344 watt device. It both charges and discharges and will even do "cycles" of charge and discharge. You can interface it to a Windoze PC and at that point, the sky is the limit on cell testing, bottom balancing, cycle life testing, cell matching, and more.
Economically, by the time you buy a power supply, a meter, a $55 bleed resistor, and a contactor, you're talking the same money with far more capability in the Revo device. http://www.revolectrix.com
This week we also add to our online store with David Kerzel's J1772 kit for DIY conversions.
David sells these things on eBay. He has a web site that isn't very well done frankly, so he sells on eBay quite successfully. But the plastic inlet ports from China are now down to about $51. He can mark those up and sell them on eBay all day long and MOST EV builders prefer this.
I don't. He started out making his own J1772 ports with a work of art in 6061 billet aluminum. The problem is, the aluminum billet costs more than the finished Chinese port BEFORE he machines it. So he had dropped the product from his eBay store.
The problem is, I have fallen in love with it. Yes, thrift is always a virtue. But we build nice electric cars and do so for the PURPOSE of attracting attention so we can demonstrate the cars and hopefully convert a few more PEOPLE than we do cars. And I've noticed they are ALWAYS curious about the charge port.
We used to use a Marinco NEMA 5-15 recessed male plug. This allowed us to use any ordinary extension cord and plug into any ordinary 120v outlet. In practice, we almost never charge on 120v. We always charge on 240v at home. But the problem was that the NEMA 5-15's are just not meant for daily use. The act of plugging in and unplugging wears them out over the course of a few months or a year.
I've fallen in love with the J1772 standard not for the safety features, but for the heavy pistol like plug and sturdy inlet connector that it is. This is designed for 50,000 insertions. It is just a much more physically substantial system.
On the eCobra, we actually put David's billet aluminum device inside a billet aluminum flip over gas cap. David liked it so well he's working on a version of this on his own. But we also used it on the Swallow and we had one mounted behind the normal fuel door on the Escalade. The door and interior are of course glossy black and the shiny aluminum billet just looks like a piece of jewelry there. As one of the first things anyone is going to want to see is the charge port, we've got a nice one to view.
So I asked David to make me up 10 so I would have them. We added them to the store, and the first three were sold out within four hours of the release of this weeks video. Should have seven more next week. As he makes these by hand, I'm not sure we can keep up. Even though they are QUITE a bit more expensive than the plastic ones.
He's also updated his little circuit board that actually properly does the copilot signal AND the proximity switch. This is what actually lets you use J1772 in your car and triggers the available J1772 EVSE, such as the GE Wattstation, to put out power.
It is a very small device and quite inexpensive. His update includes packaging it in a plastic case that can be easily mounted. The terminal strip is still readily exposed and clearly marked.
So with these two components, it is very EASY to add J1772 functionality to your build. You can STILL have a NEMA 5-15 in parallel for opportunity charging at the Walmart parking lot light pole if necessary. But with this kit, you can also make use of the many public charge stations going up, and install a proper J1772 EVSE in your garage.
In any event, we've gone from one thing to another in the EVTV store. First it was braided straps, then the JLD 404 meter. This J1772 inlet is a welcome addition. I don't know where all this leads. I hadn't really pictured EVTV as a component retailer. But we have received eight Soliton 1's and two Soliton Jr's in stock as well and I'll be adding those to the store this week. We kind of intend to pursue it to see where this all leads.
This takes me back to the early days of the BBS. We originally thought BBS would advertise in our little newsletter. Instead they became readers and we wound up running an eight line BBS ourselves. Our advertisers turned out to be Cisco and Sun, and US Robotics and the companies making tools for online communications.
Similarly, I thought all these little dealers of EV components would be a market for advertising on EVTV. They are all VIEWERS of EVTV, but not a single online retailer showed up to advertise. Just like the BBS/Internet thing, we'll wind up with the component manufacturers, but the guys just running online shops are simply not sufficiently sophisticated in the art of selling and running a buiness to be a market for advertising.
I would predict these SAME guys will grouse that we have an unfair advantage with the EVTV show. And they will never put two and two together there or pick up on the irony of it all.
Nature of the beast is that we aren't much of a threat. The average EV builder is if nothing else, thrifty. And our focus is entirely going to be on the very best components we can find. My belief is that if you build a rocket entirely of components provided by the lowest bidder, always include the cost of hiring a test pilot to fly it. You don't want to actually be IN the rocket yourself at any given time.
If I'm building a car, I want to make careful choices, but generally go for the better device. These choices are cumulative, and will result in a car that is either the sum of the cheapest things I could find, or the sum of some inspired, but sometimes pricey choices.
Our entire mission is to build ATTRACTIVE and DESIRABLE cars people will WANT. And of course I like fine rolling stock myself. So it's all part of God's plan.
If you just want some crate to creak you along the four miles to work and back without buying gasoline, let me strongly note that that's an ENTIRELY VALID MISSION. I don't share it. But it is entirely valid.
And our A123 work is a nod in that direction. I think you'll STILL find a small lithium pack more durable and more practical than Pb chemistry battery cells. And SOMEONE is going to figure out a way to package them in a small 60Ah or 80Ah pack that will do 30-40 miles at a competitive price without 200 hours of labor.
But let me reiterate, in building a car for ME, these cells would never make the cut. The CALB prismatics are just better in all respects and make a better, more trouble free car to last a lifetime. I like having 100 mile range even if I'm never going to use it.
The ongoing soap opera with the company A123 is just beyond belief. I am at this point actually embarrassed for them and about the entire topic. I cannot conceive of a product this good manhandled so badly and on such a grand global scale involving hundreds of millions of dollars and at least three continents. Along with taxpayer investment in an effort to make us a world "leader" in battery development. It has reached the point of public obscenity and defies the willing suspension of disbelief necessary to follow the story.
But did I mention the 3p28S module we have in shipment from China as we speak?