First off, I would like to say that I love what the wikispeed project is doing. I think there is a lot of potential there, and would love to get directly involved, specifically with designing an electric drive module (email sent last week
). I found out about Wikispeed recently from someone I met at a convention, and while as far as I can tell from publicly available information, I get the feeling it's not quite as far along as the promo videos seem to suggest, like I said, I think there is a lot of potential there, and I am excited to see what it can grow into.
I figured I would open with that, since I know a lot of what I am about to say is very critical and may rub someone the wrong way. That being said, if I am reading anything that was said incorrectly or incompletely, I am always happy to be proven wrong. To paraphrase a wise saying, when we fail/are proven wrong; that is when we learn.
Having not heard back from the team yet (only sent the email on the 10th, so I'm not complaining, just keeping the brain working
), I figured I would search and see if I could find anything on the 'two different electric drive module designs' that were mentioned in a few of the wikispeed videos...thus, how I found this thread.
Reading through this thread, frankly, I was a bit disappointed. I completely understand the budget problems, but for a few different reasons, I think what seems to be the current approach is flawed. But some of that may be the idealist in me.
First and foremost, there is a question you have to ask yourself. 'Am I designing this as a product or am I just trying to minimize costs?' Assuming that this is following the same line as the rest of the wikispeed project, I am going to say that this is intended as a product. Not now, but at some point down the road. I understand the budget concerns, I really do...but when the ICE module was being developed, did they just wander down to the local scrap yard and say "Oh, hey, look, that's an ICE. It's a 5HP 70cc dirt bike, but that'll work."? I'm guessing not. I'm guessing they found a commonly available, reasonably priced, appropriately powered, ICE engine, that they could purchase new, reliably and repeatably. That being said, scavenged parts are fine for prototyping, but in designing the module you must always keep in mind that the parts that will go into the end product will be very different than what you are using now..which ties into another point, but I'll go into that later. I would hate to see the result of this be a module that makes the car behave like an overgrown golf cart...if it is to succeed, IMO, you need to match or exceed what the ICE module is doing....which is not as hard as you might think.
And that leads me to my other initial point.
I understand the reason for the $5,000 price point. I really do. You want to match the cost of the ICE module. But that is not realistic just yet...not if you want something reproducible that will actually behave like a car. Final price will really depend on what range you want to achieve...but there will be a certain, minimum, possible, price....and it will be a good bit above $5,000. Double that, and you might be able to make it. For example, an AC-50 with a 7602 and 38 CALB CA-60's, charger, and associated contactors/fuses/etc, would give you a nice, peppy, little, roadster. Wouldn't give you the same performance as the 130HP ICE, but it would still act like a
car. You
might be able to do that module for less than $10,000, and it would probably only give you about a 40-50 mile range. ..also, run the numbers before assuming that cooling will take up too much room...if you're used to lead batteries, it may shock you how small even the large-format LiFe batteries are (check out the chart in my sig for figures). Looking at your layout diagram, it looks like you're dedicating a full 1/2 of the module to batteries. You should be able to fit a good 30kWh in that space...enough storage to give the car probably about a 200-300 mile range (I'm guessing ~100-150Wh/mi, based on the mpg numbers). Granted, that would also be quite expensive, and larger than (imo) necessary, but it also means that with a smaller pack you would have a lot more space for other things.
Ok, for more specific points.
Regarding where to put the 12V aux battery, that should be determined by the rest of the vehicle design. If it is in the ICE engine module, it should be in the electric drive module. If it is not in the ICE module, it should not be in the electric drive module. If it is external, you will want to have a DC-DC converter in the electric drive module, with leads feeding out of the module, that will connect to the external 12V aux battery. If the aux battery is external, I would assume there are already such hookups for feeding off the alternator on the ICE module.
The last thing that jumped out at me is the 'BJB'. I understand the idea behind it; you are trying to modularize the module itself, to make it agnostic to what controller/charger/etc you are using. This is admirable, but not, imo, practical at this point. Things are not yet sufficiently standardized in the EV world that this could be made to actually work well. For example, some controllers will accept two switching signals (ie, key/enable), some only accept one. Some accept a charger disable, some do not (I would just tie this one into the throttle feed, personally). Some have built-in pre-charge, some do not. Some have built-in contactors, some do not. If you are feeding the motor control lines through the BJB, you need to keep in mind the possibility for differing numbers of lines (series-wound DC or PMDC == 2, 3-phase AC or PMAC/BLDC == 3, sep-ex DC == 4). If you choose to use a BMS, you will have to account for it and its communication with the various parts.
In my opinion, there is just too much variation between different component combinations at this time to make a BJB work. You could have everything from half a dozen different components, all needing to be interconnected, all the way to a single unit that handles everything, and any point on the spectrum in-between. As such, I would stop the universalization at the parent module level, and create a separate design for each electric drive module component combination. There can be some shared sub-modules, but everything will not be shared. For example, you could have battery modules of different capacities that are constant across all electric drive modules, and you could have the control components (controller, charger, etc) in one sub-module that is the same size/construction for all electric drive modules, but I think that level of universal modularization is as far as it is logical to go at this point in the industry.
Lastly, I would encourage you to post this design log on other forums as well. EV's just aren't the primary focus of this site, so there won't be as much EV expertise as somewhere that does focus on them. My 2 cents, post it on as many relevant boards as you care to pay attention to; tap as wide a knowledge base as possible. diyelectriccar.com is a good one. I've heard good things about EVDL.org. elmoto.net is focused on motorcycles, but a large amount of knowledge overlaps.