DIY Electric Car Forums banner

1 - 20 of 160 Posts

·
Registered
Joined
·
3,219 Posts
Discussion Starter #1 (Edited)
About 2 years ago there was a thread discussing Evnetics possibly developing a DC-DC converter. The project was shelved mostly because I couldn't get the manufactured cost below $200, and then the locomotive drive system project came along and derailed (ahem) all other projects.

Well, we finally shipped off the prototype locomotive drive system to the customer so I've revisited this gaping hole in the market, except with the intent of making the DC-DC myself, because, again, it still wouldn't be possible for Evnetics to make it profitably - not alongside the DIY EV and locomotive drives, anyway.

And as explained in that thread from 2 years ago, I have already worked out most of the details for a 14V/50A DC-DC converter that can run on a very wide input voltage range of 100-400V.

So given that Evnetics can't make this product profitably, I've been contemplating funding the development and/or production of it via a Kickstarter campaign. This thread, then, is to discuss the technical details of the converter as well as the prospective Kickstarter campaign. For example, it seems to me that most of the products which have been successfully funded on Kickstarter were open source. I don't have a problem making this open source, but I'd rather not offer a kit for people to build this themselves because, a) it will be mostly surface mount and, b) I don't want to deal with the people who really shouldn't be building something like this buying the kit and attempting to build it anyway just to save $100-$200. And besides, I know from experience that it takes just as long to pack a kit as it does to simply populate the board with the parts.

However, that does leave a pretty big spread between the obligatory altruistic $10 pledge to give your support with nothing much to show for it, and the (estimated) $350 pledge for the assembled and tested product. Any ideas for intermediate rewards would be welcome, as this is where my imagination tends to fail me.

Finally, I used to run a small electronics design & manufacturing business, so I know from both that experience and from Evnetics that a lot more people will say they want your proposed gadget than will actually buy it. Funding the development - or the production of the first batch, at least - via Kickstarter would spread out the entrepreneurial risk rather than me shoulder it all by myself. Also, if this experience with Kickstarter turns out well then I (or Evnetics) will be much more inclined to use it to fund even more ambitious products like, say, an AC inverter or a drive for switched reluctance motors. Baby steps first, though.
 

·
Registered
Joined
·
296 Posts
However, that does leave a pretty big spread between the obligatory altruistic $10 pledge to give your support with nothing much to show for it, and the (estimated) $350 pledge for the assembled and tested product. Any ideas for intermediate rewards would be welcome, as this is where my imagination tends to fail me.
Ok you can count me in, I have only seen the option between
1) cheap and dangerous (due to low isolation standard)
2) integrated with controller (siemens)
3) very expensive ( I was going for that option)

So if you think you are able of doing a good DC/DC converter for voltage up to 400V and less than 500 USD I would very much like that
 

·
Registered
Joined
·
600 Posts
I don't know anything about kickstarter, but I'm still interested in a quality dc-dc converter for $350-$400.
 

·
Registered
Joined
·
4,051 Posts
Sounds like a great idea actually. There's also a gap at lower currents and 100-400V for Motorcycles, which may only need 20A or so. Not sure if that's easily done or not, but keep it in mind.
 

·
Registered
Joined
·
288 Posts
Unfortunately I'm all set for DC/DC converters so I can't say I'd be in for pledging to buy one with Kickstarter. I could probably do one of the altuistic pledges for nothing in return though. I would really like to see something like this succeed. Disclaimer: I haven't found your previous discussion about this so if I've gone over anything already said I apologize.

For tiers on kickstarter this is my initial brainstorming: You could have batches and make the first batch more money and each subsequent batch a bit less. This lets people who are eager early adopters get theirs first and people who aren't in as big of a hurry can hold off for a second or third batch. I have seen this done on kickstarter. Another idea might be to include wiring with more expensive pledges. If the unit would be externally fused maybe have that as an option as well. I'd imagine most converters maintain a fusebox in the car but if you could source something like that for people who need it then that might be a potential option as well.

Another idea (maybe dumb): Maybe include an output that supports limited amperage but never turns off. You know, something that could support an ignition switch and maybe keep people's radio presets and such. This eliminates any reason to keep a small 12V battery for bootstrapping purposes.
 

·
Registered
Joined
·
296 Posts
Interesting idea Tess. Good helping hand at the DIY market. Thanks.

But what would be the main gain over currently available product? Let say Mean Well SP-750 at 200$ (127-370VDC in, 13.5-16.5v out, 50A).
The only option is SP-750-12V
These small details....
I need at least 400 V input,
I need at least 14V out (13,5 is not enough, my not exceed 15V)
I would prefer better isolation than 100 MOhm per 500V
I would need the developer to have extensive experience with EV and it is a bonus to be able to bug him personally:)
 

·
Registered
Joined
·
3,219 Posts
Discussion Starter #9 (Edited)
Sounds like a great idea actually. There's also a gap at lower currents and 100-400V for Motorcycles, which may only need 20A or so. Not sure if that's easily done or not, but keep it in mind.
Yes, easily doable, though I would think it would be more useful to cover a lower voltage range rather than 100-400V. Are you trying to accommodate the RMS drive system by any chance?


...Disclaimer: I haven't found your previous discussion about this so if I've gone over anything already said I apologize.
I put a link to it in the opening post. It drones on for many pages with me often waxing acerbic and cantankerous, as is my wont.

For tiers on kickstarter this is my initial brainstorming: You could have batches and make the first batch more money and each subsequent batch a bit less.
Interesting suggestion. Definitely not something I would have thought of. Can you provide some example projects which used this kind of tier system as I would very much like to take a closer look at it.

Another idea might be to include wiring with more expensive pledges.
This ain't the GEVCU with 30 something wires - it's just going to have 5 or 6 terminals - HVIN+, HVIN-, Chassis Ground, LVOUT+, LVOUT- (optional), IGNIN (optional - enables the converter, fed by 12V from ignition switch).

If the unit would be externally fused maybe have that as an option as well.
Not a bad suggestion, but a surprisingly expensive one relative to the overall cost of the product. Still, I'll look into it as it could be for a premium tier?

Another idea (maybe dumb): Maybe include an output that supports limited amperage but never turns off. You know, something that could support an ignition switch and maybe keep people's radio presets and such. This eliminates any reason to keep a small 12V battery for bootstrapping purposes.
Nope, this one is a bad idea on two counts: 1) without a 12V battery to buffer peak currents the output voltage will simply collapse every time, say, an electric power steering pump starts pulling 65A; 2) the DC-DC will draw current from the traction pack continuously to supply those "parasitic" 12V loads and to prevent overdrawing the pack would require adding intelligence to what is otherwise a dumb black box can be told what the battery pack voltage is so it can shutdown if it drops too low.

Believe me, simpler is better in this case - just hook the thing up to HV and 12V and let'er rip.
 

·
Registered
Joined
·
4,051 Posts
Tesserect,
There's a big gap for the Electric Motorcycles with DC-DC converters. People do use some Vicor, but they're just the module, without terminals or filtering. Also, they basically stop at 96V for most of the 400W or smaller models, and then its off to TC Charger to get something else. Rinehart, Sevcon, Curtis, Soliton, are all over 100V ready controllers, but when you put them in a motorcycle, not sure we need 50-60A worth of 12V is all.
 

·
Registered
Joined
·
3,219 Posts
Discussion Starter #11
But what would be the main gain over currently available product? Let say Mean Well SP-750 at 200$ (127-370VDC in, 13.5-16.5v out, 50A).
This is the central question, and the answer is simply that my converter will be designed for use in an EV. Either that is immediately worth the $150 premium to you, or you can effectively pay ~$550 for it later on... after having to replace the $200 Mean Well that, well, meant well, but couldn't quite handle the often-hostile EV environment.

1. The topology I've chosen (buck current-fed full bridge) is uniquely suited to handling wild swings in pack voltage. The typical commercial power supply is ill-equipped to handle this because it either uses a voltage-selector switch and so can only run over the range of around 275-375VDC, or it uses a PFC stage to cover the entire AC mains range but then suffers from extremely slow transient response because the loop bandwidth of the PFC stage needs to be well under the line frequency to meet the IEC power factor and distortion specs.

2. It won't be bothered by ripple from the controller.

3. It will be in a closed box and passively cooled - no fan, no perforations - and designed for life outside at 50C, rather than in an air-conditioned office at 20-25C.

4. It will totally disconnect from the HV pack when shutdown, and will not place a parasitic drain on the main 12V output, either.
 

·
Registered
Joined
·
600 Posts
This is the central question, and the answer is simply that my converter will be designed for use in an EV. Either that is immediately worth the $150 premium to you, or you can effectively pay ~$550 for it later on... after having to replace the $200 Mean Well that, well, meant well, but couldn't quite handle the often-hostile EV environment.

1. The topology I've chosen (buck current-fed full bridge) is uniquely suited to handling wild swings in pack voltage. The typical commercial power supply is ill-equipped to handle this because it either uses a voltage-selector switch and so can only run over the range of around 275-375VDC, or it uses a PFC stage to cover the entire AC mains range but then suffers from extremely slow transient response because the loop bandwidth of the PFC stage needs to be well under the line frequency to meet the IEC power factor and distortion specs.

2. It won't be bothered by ripple from the controller.

3. It will be in a closed box and passively cooled - no fan, no perforations - and designed for life outside at 50C, rather than in an air-conditioned office at 20-25C.

4. It will totally disconnect from the HV pack when shutdown, and will not place a parasitic drain on the main 12V output, either.
+1 I like it!
 

·
Registered
Joined
·
3,219 Posts
Discussion Starter #13
...Rinehart, Sevcon, Curtis, Soliton, are all over 100V ready controllers, but when you put them in a motorcycle, not sure we need 50-60A worth of 12V is all.
No problem on making a smaller version that only puts out 20A, and while it will be less expensive, the price won't be directly proportional to output current. Off the cuff I'd say that halving the current would drop the price from $350 to $250. This is because while the basic design will be the same, most of the components will change: a different case, PCB, power transformer, MOSFETs, diodes, heck, even the values of the resistors and capacitors for loop compensation will change! So I don't know if that would be terribly attractive, pricewise, but it could make for an interesting stretch goal.

Oh, and a more general comment - I want honest feedback, not people blowing smoke up my *$$. [edit - this is not a jab at palmer_md :D ]

Also, I'm pretty sure I can hit a price point of $350 for a 50A output unit. Near as I can tell, that means I would actually be competitive with the Chinese "made for EV" stuff like the Chennic.
 

·
Registered
Joined
·
296 Posts
Oh, and a more general comment - I want honest feedback, not people blowing smoke up my *$$. [edit - this is not a jab at palmer_md :D ]

Also, I'm pretty sure I can hit a price point of $350 for a 50A output unit. Near as I can tell, that means I would actually be competitive with the Chinese "made for EV" stuff like the Chennic.
I am full in.
The model I know for kickstart-like-project is based on subscription
I would like to start with USD 100 / month for my part.
Meaning after 6 month I would have put in 600 USD and hopefully get one (prototype) converter.
First would not need to be certified, that is ok with the second (bought) one.
The rule who pays most gets first would apply of course..:)
 

·
Registered
Joined
·
336 Posts
I'm very happy to see you put this back on the table. If you can get close to the $350 price point I'm definitely in; that's a bargain in my books.

With regards to the tiered supporter question, perhaps you could offer different case options: a cheap stamped steel option for a contribution of $350 and a nice aluminum case version for $500?

Keep up the good work.
 

·
Registered
Joined
·
3,219 Posts
Discussion Starter #16
...
The model I know for kickstart-like-project is based on subscription
Another interesting suggestion, though I'm not quite sure how that would be implemented; it seems to reward me for procrastinating, which doesn't sound too good to prospective pledgers!? So I suspect I am missing something here, which is the real point of this thread: I am far more familiar with the design & manufacturing aspects of this project than I am with the Kickstarter part.

First would not need to be certified, that is ok with the second (bought) one.
Ah, yes... certification. I am designing this with eventual certification in mind - and can do some basic "pre-compliance" testing myself - but understand that the cost to have a product certified for the EU can be *very expensive* and whether or not I pursue that will depend a lot on the success of the Kickstarter project and the volume of subsequent sales. So you European fellows are very much depending on your North American compatriots to make a good showing.


...
With regards to the tiered supporter question, perhaps you could offer different case options: a cheap stamped steel option for a contribution of $350 and a nice aluminum case version for $500?...
That is something I have considered, but if I had the freedom to use any case shape/design I wanted I probably would design the board differently! Anyway, here is the case I have tentatively selected:

YGKT enclosure

I chose it because of the following reasons:

1. I estimate it is large enough to contain all the major components (won't know that for sure until I start to design the board);
2. it's black so it won't be too noticeable;
3. it has mounting flanges;
4. it is split, so I will be able to bolt the power semis to the sidewall, maybe with a copper heat spreader (I don't expect losses to be high enough to warrant that, though);
5. and, finally, it is cheap.

And interesting to see you join the fray, as you were the one that started the previous thread 2 years ago :D
 

·
Registered
Joined
·
600 Posts

·
Registered
Joined
·
3,219 Posts
Discussion Starter #18 (Edited)
Its a little different than the BUD you had in mind earlier, but I guess the case is not as important as the function..
Correct. Also, that Bud case wasn't all that attractive, really, and was rather expensive - $56, IIRC. If you all think the $40 premium in price for the Bud box is worth it, well, now is the time to cast your vote.
 

·
Registered
Joined
·
600 Posts
Correct. Also, that Bud case wasn't all that attractive, really, and was rather expensive - $56, IIRC. If you all think the $40 premium in price for the Bud box is worth it, well, now is the time to cast your vote.
I vote no for cost, but I really don't know the other considerations such as the shielding, sealing and such that the Bud seems to offer. I'd say its an engineering decision and its outside my expertise. As far as looks, you are correct, the less expensive box probably looks better.
 

·
Registered
Joined
·
296 Posts
Another interesting suggestion, though I'm not quite sure how that would be implemented; it seems to reward me for procrastinating, which doesn't sound too good to prospective pledgers!? So I suspect I am missing something here, which is the real point of this thread: I am far more familiar with the design & manufacturing aspects of this project than I am with the Kickstarter part.
Ok the same here I would be the worst salesman ever I think, so I have to get advice from some friends on this part. Here is one of the site used in Iceland similar to kickstarter , the karolinafund
http://www.karolinafund.com/project/index.
It is the same type of projects which are successfully funded
1) Their funding target is 1500-3000 EURO
2) They have 50-100 backers
3) Most common backup is about 20 EURO
4) The outcome is a product which could cost 20 Euro each or a bit more.
5) Funding was raised in less than 30 days
6) Product was in many cases delivered 30 days after the complete funding
7) Successful fundraisers had at least one successful product before.

So it is same for 300.000 (Iceland) as for 300.000.000 market (USA).
(only difference USD instead of EURO :) )

It is a bit hard to read into the backers mind but we tried...
1) Backers are not investors, they are potential buyers
2) (Early) Backers are ready to loose the money invested
3) Later backers are the result of herding, they don't want to miss the opportunity to join the successful buying
4) Very few are ready to lose more than 20 bucks
5) No interest after 6 months (much less after 1 month)

So what to do if you have a project which funding might be 5 to 10 times the size of most common success regarding funds and time?
Points 4) and 5) are a obvious problem,
Also that backers can only afford a certain amount and they do not if they are not a bit sure of the funder. So this typical funding websites may not be the answer but rather working in a known community.
The model of subscription has the same target (buy a product) but offers less risk for backer and also fundraiser.
1) Expected product price will be devided in affordable parts, monthly payment for example
2) There is always a possibility to back out if necessary (could be a 30% repayment rule or another rule up to certain amount).
3) Benefit of (early?)high backing would be first to receive (the Tesla effect)
4) Deadline for early funding, goals must be clear. this could be realized for example with paypal.

1. Step startup
To start the product 50% of the total funding goal must be reached at day X (less than 30 days for example). This could be paid on a paypal-account with a payback-option until day X. If goal is not reached, then the product will not be started. Full repayment is easy.

2. Step work start
After day X the (conceptual) work will begin. At this time both fundraiser and backup have committed. the fund in step 1. will not be repaid. Next payment will result when next goal has been reached. Meaning design and cost of product will to be have fixed. This is the ugly way of the compromise (other will be to expensive)

3. Step production phase.
Before this begins, Second payment must be maid, again there is a option to back out before it results. Here backing out would still leave an (time limited) option of turning back later at the cost of getting later delivery. The payment from 1. could act like a future price-off for the product. The risk is still that there will never be a product.
If there is a lot of backout the fundraiser has two options
a) To delay the product until further raising has been achieved
b) To stop the product and compensate the ones that paid the second payment, either with repayment or with an open source contribution.

4. Delivery prototype phase
Only to backers which have backed up at least the prototype price.
The higher the backup, the earlier the deliverment.
Other than backers can ofcourse buy but at higher price

5. Complete.
Some backers have their product, other backers which have paid less or stepped out earlier have a sort of price off of the product based on their backing (could be considered as prepaid up to a certain amount).

Just ideas...
 
1 - 20 of 160 Posts
Top