[EVDL List]

regarding the inductance build-up behind the transistors during hard
switch off where the curtis does soft switching to avoid it, do you
think that the typically 1200v rated IGBTs would be able to take that
buildup from maybe 300v operation of hard switching so as to avoid the
need for parallel protective caps?

Dan

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

 

[EVDL List]

[No message]

 

[EVDL List]

Morgan LaMoore wrote:
> (I bought a bunch of 1000V IGBT snubber
> capacitors for under $2 each.)
>
what brand and model?

> p.s. Did you just buy a bunch of 1200V 600A IGBT's from ebay? The week
> DoctorHD mentioned the ebay IGBT's, there was a bidding frenzy. (I took part
> in it, too.) Now, there's another batch for sale, and most are unclaimed. I
> feel kind of stupid for bidding so much.
>
>
hehe no I bought a couple several months ago. got them for 25$ each.
they seem to be in regular supply so don't pay too much. not sure who
have used that many but somebody is getting rid of a lot of them. while
I have them now I'm also getting some small ones like the curtis and
zilla approach for experimentation. it has the advantage of being
smaller, lighter and in dependable supply such that an open source
design can be guaranteed makeable for everyone and not rely on special
finds on ebay. while I got the modules for 25$ the new price is 450$ for
a single one. compared to the 80$ cost of all the power electronics in a
typical curtis.

similarly with caps it seems several small ones work better than fewer
big ones. Panasonic EU-EE series is the best I've found from what I
understand. or TS-EE depending on compromise. I'm not sure yet but I
think the zilla uses those too

Dan

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

 

[EVDL List]

hehe no I bought a couple several months ago. got them for 25$ each.
they seem to be in regular supply so don't pay too much. not sure who
>have used that many but somebody is getting rid of a lot of them. while

Dan, nice to see you digging in.
In the 300A region, I've used the Fairchild FMG2G300US60E with good
success in my BLDC controller, it's a bit low for a brushed motor
but in the 250V - 380V pack range with BLDC or ACIM it should be
reasonable. A six-pack of them (three modules) goes for about $250
which is quite reasonable for new parts, I think. That's new parts,
not used, and it has a good temp coefficient so you can use the
voltage drop across it to estimate current for current limiting and
control. With enough voltage, 60 to 80kW should be reasonable and
those are pretty cheap. I looked at the Ebay ones and decided that
I wanted something that was serviceable and repeatable, so when I
found these, I used them. Since I was repairing an already (partially)
debugged controller I felt the risk of using a non-self-protected
IGBT module was pretty low so I went with those.

I would think that the lower voltage drop of 600V modules, generally
lower cost of 600V modules, coupled with better snubbing capacitance
may be a more cost-effective choice, especially if some builders are
going to buy new parts for their controllers.

I don't know if it's of interest, but I'm working on a new BLDC
control board and I can post schematics and gerbers and source code,
but I don't have the time for tech support. I don't know how many
people out there are using BLDC's, but I am.

-Dale

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

 

[EVDL List]

Dan Frederiksen wrote:
>
> regarding the inductance build-up behind the transistors during hard
> switch off where the curtis does soft switching to avoid it, do you
> think that the typically 1200v rated IGBTs would be able to take that
> buildup from maybe 300v operation of hard switching so as to avoid the
> need for parallel protective caps?

If you want to minimize/avoid the capacitors, then you should use a
different topology converter. The buck converter normally used
intrinsically has a huge input ripple current; battery current switches
from 0% to 100% to 0% to 100% of motor current, etc. One way or another,
either the batteries or the capacitors have to carry it.

One possible topology is the Cuk' converter. It uses the same 4 parts
(transistor, diode, inductor, capacitor) but wired a bit differently.
Done right, it has *zero* battery current ripple. The Cuk' is also
interesting because the motor voltage can be either lower or higher than
the pack voltage.

The drawback of the Cuk' is that the stress on the transistor and diode
are higher -- i.e. you need higher voltage semiconductors. This is
getting easier though, since 600v and 1200v IGBTs are common. The
capacitor in a Cuk' converter is also vastly smaller than in a buck
converter; but it needs a very high ripple current rating. The end
result is that a Cuk' converter costs more.

--
Ring the bells that still can ring
Forget the perfect offering
There is a crack in everything
That's how the light gets in -- Leonard Cohen
--
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