DIY Electric Car Forums banner

1 - 4 of 4 Posts

·
Registered
Joined
·
70 Posts
Discussion Starter #1
Re: [EVDL] Mosfet manufacturing request - Larger/Higher amp models

My view on this (and my experience in design and manufacturing
in general) is that it makes no sense for any manufacturer
to design and produce a products that will take many thousands
(likely around 100,000) dollars investment, then start up production
of a single batch, only to see that it takes years for that first
batch to be consumed by the few consumers.
Unless each consumer that buys a handful of them is willing to
spend a thousand dollars per part or so - then it can be very
lucrative, that is why very small scale production does happen
for special applications like nuclear plants, space vehicles
and so on. Spending only 4 figures on dedicated parts is
peanuts there.
Not so sure that any sensible controller designer would go down
this path, unless his design is expected to start running in
the tens of thousands of units, because at that moment the
NRE (Non Recurring Engineering) cost is only around $10 per
controller, so a dedicated development of parts becomes feasible.

Soooo - this is the old chicken and egg problem: until there
is a large market for EV controllers, there will be no dedicated
parts for EV controllers, which means that we will have to live
with what is out there in the market already.
Sometimes parts are present that are used for trains or other
high power EV applications, but most designs use low cost
transistors that are produced for other customers by the millions
and even buying several dozen of them is cheaper than a dedicated
part for an EV controller.

As soon as EVs take off big time (not a matter *if* they take off,
but only *when* they take off, in my mind) then you will see
plenty dedicated developments, just like you see plenty dedicated
developments for power tools or gensets.

Cor van de Water
Systems Architect
Proxim Wireless Corporation http://www.proxim.com
Email: [email protected] Private: http://www.cvandewater.com
Skype: cor_van_de_water IM: [email protected]
Tel: +1 408 542 5225 VoIP: +31 20 3987567 FWD# 25925
Fax: +1 408 731 3675 eFAX: +31-87-784-1130
Second Life: www.secondlife.com/?u=3b42cb3f4ae249319edb487991c30acb

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Michael Barkley
Sent: Sunday, September 09, 2007 2:13 PM
To: EVDL.ORG EVDL.org
Subject: [EVDL] Mosfet manufacturing request - Larger/Higher amp models

Wondering? - Since MOSFET's seem to be the choice component in the controllers, looks like we really need to approach the manufacturer of these devices to see if they can produce larger versions so that there isn't the need to parallel them up to make the amperage that is needed. Surely there are other applications that these HIGH AMPERAGE versions would be useful in other than EV's. This could motivate them to produce them, and then the controllers we need could be made with less components and maybe more reliable. Again, just my mind wondering into lala land again.




M. Barkley

www.texomaev.com

http://www.austinev.org/evalbum/1135


_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #2
Re: [EVDL] Mosfet manufacturing request - Larger/Higher amp models

From: Michael Barkley
>Wondering? - Since MOSFET's seem to be the choice
>component in the controllers

Actually, the fashionable part today is the IGBT. It keeps changing; MOSFETs were popular 10 years ago, darlington transistors around 20 years back, SCRs before them, and contactors before that.

> looks like we really need to approach the manufacturer of these
> devices to see if they can produce larger versions so that there
> isn't the need to parallel them up to make the amperage that is
> needed. Surely there are other applications that these HIGH
> AMPERAGE versions would be useful in other than EV's.

High-current IGBTs are already on the market. They weren't built for road-going EVs (because they use negligible numbers of parts), but instead for high power commercial and industrial systems market.

EV controller manufacturers are actually a bit slow in designing new products to use IGBTs. Cafe Electric's Zilla uses IGBTs; but they are the smaller, earlier-generation parts, so it needs a lot of them in parallel. No one (yet) is using one of the big IGBT modules to make a road-going DC motor controller (though a number of individuals and prototypes have been built).


--
"Excellence does not require perfection." -- Henry James
--
Lee A. Hart, 814 8th Ave N, Sartell MN 56377, leeahart-at-earthlink.net

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #3
Re: [EVDL] Mosfet manufacturing request - Larger/Higher amp models

earlier-generation parts, so it needs a lot of them in parallel. No one
(yet) is using one of the big IGBT modules to make a road-going DC motor
controller (though a number of individuals and prototypes have been built).

No AC vs DC debate, please... Most 'big' OEM's are likely to not use brushed
DC motors, but ACIM or BLDC motors... as Siemens, Mercedes, Honda, Toyota,
GM, Ford, Solectria, etc. all have done. For these power levels and motor
types, the 'standard' arrangement is three half-bridge IGBT modules which
have been around for a while. The parallel here is that industrial ACIM
controls in the 208V-240V range and roadgoing ACIM/BLDC controllers in the
300V-400V battery volt range are very similar and should be able to use the
same modules - 300A to 600A, high-speed 600V dual IGBT modules. A hobbyist
can get a six-pack worth of 300A, 600V IGBT (enough for about 60kW-80kW AC)
for around $260. One advantage of high-voltage (300-400V) AC systems is that
the maximum current levels drop to where a reasonably sized semiconductor
die can carry the required current.

Several markets are served well by these IGBT modules... welding, induction
heating, industrial ACIM control, large UPS systems, large power inverters,
and conveniently, AC EV's. IMHO, the semiconductor market in these power
levels are making the class of part you are looking for, but the target is
not only EV's, obviously. But when you combine all of the above markets,
it's substantial enough to justify making them. A high-current MOSFET for
the 96V-200V, 1000A+ range is not really that useful for nearly as many
industrial/commercial purposes. Powerex, by the way, offers 6-pack MOSFET
modules that could be paralleled, but I'd bet they are crazy expensive. You
can get a lot of TO-220 or TO-247 packaged MOSFET's, hand-select them, and
bus-bar them together for the price of one.

-Dale

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 

·
Registered
Joined
·
70 Posts
Discussion Starter #4
Re: [EVDL] Mosfet manufacturing request - Larger/Higher amp models

it's a fair thought although the zilla does use IGBTs. and quite large
IGBT modules exist. they are however more expensive than the equivalent
small ones. and looking at these cute little TO-247 three legged
liquorices I can see the appeal:
http://www.df-cad.dk/web/ev/TO-247-tube.jpg
http://www.df-cad.dk/web/ev/TO-247-front.jpg
http://www.df-cad.dk/web/ev/TO-247-back.jpg (this happens to be an
STMicro mosfet but same size as the ones used in Zillas)

each one capable of 80A if cooled to 25c or in practice perhaps 40-50A
watercooled. costing 2$ a piece you could have a very viable 300V 300A
rating with just 6 of these..

using small components you can treat the manufacturing more like regular
low power electronics where as with IGBT modules and large can caps it
begins to look more like mechanical engineering. looking at the zilla
for instance it looks much like any low power audio amplifier inside:
http://www.df-cad.dk/web/ev/Zilla2kdiscovery.jpg
Doesn't look like 2/3 megawatt but it is and not to belittle Oatmeal's
work :) it looks completely pro and works great but the real magic is
the power transistor technology. I wouldn't be surprised if you could
run a normal sized car perfectly well at high way speed using only a
single TO-247 IGBT. Indeed I would bet you it could be done.

Dan

Michael Barkley wrote:

>Wondering? - Since MOSFET's seem to be the choice
>component in the controllers, looks like we really
>need to approach the manufacturer of these devices to
>see if they can produce larger versions so that there
>isn't the need to parallel them up to make the
>amperage that is needed. Surely there are other
>applications that these HIGH AMPERAGE versions would
>be useful in other than EV's. This could motivate
>them to produce them, and then the controllers we need
>could be made with less components and maybe more
>reliable. Again, just my mind wondering into lala
>land again.
>
>
>
>
>M. Barkley
>
> www.texomaev.com
>
> http://www.austinev.org/evalbum/1135
>
>
>_______________________________________________
>For subscription options, see
>http://lists.sjsu.edu/mailman/listinfo/ev
>
>
>

_______________________________________________
For subscription options, see
http://lists.sjsu.edu/mailman/listinfo/ev
 
1 - 4 of 4 Posts
Top