05-14-2009, 03:40 PM
|
#1251 (permalink)
|
EcoModding Lurker
Join Date: May 2009
Location: Bucharest,RO and Copenhagen,DK
Posts: 42
Thanks: 0
Thanked 1 Time in 1 Post
|
end caps: sheet metal bended in a Z shape with square angles, bolted to the same heatsink and the box. Strips of rubber (from a tire tube) glued to the sides will seal the box perfectly.
|
|
|
Today
|
|
|
Other popular topics in this forum...
|
|
|
05-14-2009, 11:44 PM
|
#1252 (permalink)
|
EcoModding Lurker
Join Date: Oct 2008
Location: USA
Posts: 31
Thanks: 0
Thanked 1 Time in 1 Post
|
It seems like you would want a Lexan box anyways, for safety. No matter how well it's designed, there's no way to prevent a metal shaving getting in there and zapping the whole outside. The last thing you need is someone sueing you into oblivion because they were lazy and careless.
You don't need a very thick layer to trap EMI - heck, you could do it with aluminim foil or a cannibalized mesh garbage can.
I love seeing this project 'go pro,' but remember the little guys here!
EDIT if you're worried about heat dissipation, what prevents you from wiring up a little 10W 12V fan to the top of the heatsink? You can get them dirt cheap for computer cases.
|
|
|
05-14-2009, 11:56 PM
|
#1253 (permalink)
|
PaulH
Join Date: Feb 2008
Location: Maricopa, AZ (sort of. Actually outside of town)
Posts: 3,832
Thanks: 1,362
Thanked 1,202 Times in 765 Posts
|
I'm just worried about the ambient temperature of the box to keep going up, because most of the walls are insulators. I really don't know how it would perform with hard long driving with only the heat sink and a fan, surrounded by lexan. Ben did some pretty hard driving up big hills, and the thermal shutdown never came on, so maybe it's not really a problem??? I'm not sure. Man, I like lexan as much as the next guy! Maybe people can make their own box without too much trouble, and they can pick what they like best?
|
|
|
05-15-2009, 12:30 AM
|
#1254 (permalink)
|
EcoModding Lurker
Join Date: Oct 2008
Location: USA
Posts: 31
Thanks: 0
Thanked 1 Time in 1 Post
|
What are the approximate (ballpark!) dimensions of the box and thickness?
|
|
|
05-15-2009, 12:30 AM
|
#1255 (permalink)
|
EcoModding Lurker
Join Date: Mar 2009
Location: Florida
Posts: 37
Thanks: 0
Thanked 2 Times in 2 Posts
|
I have been lurking and following this thread closely. Fantastic work by everyone involved.
I just wanted to comment about the enclosure. A certain commercial brand of controller simply uses an aluminum extrusion with plastic end caps. All they do is adhesively bond on the end caps.
The idea about a cooling fan would work nice but I think you will want to keep the controller sealed to prevent any unwanted small metal parts from entering. It sounds like you have all of items connected to a heat sink which would be perfect for the bottom of the enclosure. My recomendation would be to then mount the controller to another larger aluminum plate which will act as an external heat sink and help pull the heat out of the controller. A thin film of heat transfer paste will greatly help the thermal conductance between the two plates. Just remember you want to use as little as possible. Also, aluminum is much better at transferring heat than steel so my vote would be for an aluminum enclosure.
Keep up the great work....
|
|
|
05-15-2009, 12:33 AM
|
#1256 (permalink)
|
EcoModding Lurker
Join Date: Jan 2009
Location: canada
Posts: 7
Thanks: 0
Thanked 0 Times in 0 Posts
|
case ventilation
Quote:
Originally Posted by MazdaMatt
Wouldn't lexan end caps defeat the EMI purpose of the metal box?
|
How about using copper screen on the ends? It's a decent shield, it could even be wrapped around a lexan case with the ends open for ventilation.
Here's a C&P on copper screen.
Choosing a copper RFI shielding mesh
Radio frequency shielding and Faraday construction can be accomplished with three popular copper meshes. 16 mesh .011inch wire diameter has a high airflow but the lowest shielding effectiveness. 22 mesh .015 inch wire diameter is specified by the U S government Tempest program. It is a very sturdy mesh which still has good airflow but than 16 mesh. 100 mesh .0045 inch wire diameter is the finest practical copper shielding mesh. It has restricted airflow but is quite effective into the higher frequencies.
Last edited by jimini; 05-15-2009 at 01:02 AM..
|
|
|
05-15-2009, 12:42 AM
|
#1257 (permalink)
|
PaulH
Join Date: Feb 2008
Location: Maricopa, AZ (sort of. Actually outside of town)
Posts: 3,832
Thanks: 1,362
Thanked 1,202 Times in 765 Posts
|
I like the idea of ends open for ventilation, but no one else would I bet. I think it would be cool to have a fan blowing along where the mosfets are. That would really keep things cool. But there's probably the risk of crap going inside it. Mine is mounted inside my car, so that's not really an issue, but most people have it in the motor compartment, exposed to the outside world.
I like Steiner's idea too. I've been thinking about that today. Maybe it would be best to just bolt the internal copper heat spreader to an aluminum plate (sort of like the curtis), and allow people to connect that plate to whatever they want, like an external heat sink if necessary? It would have to be thick enough so that you could embed the bolts that attach the plate to the heat spreader.
Hold on a second! Is that what those little rubber circles are at the bottom of the Curtis??? Just covering where the bolts are? Sneaky little buggers! Maybe we should just use an aluminum plate instead of a heat sink? Let people then bolt the plate to where ever they want???
|
|
|
05-15-2009, 01:38 AM
|
#1258 (permalink)
|
EcoModding Lurker
Join Date: Oct 2008
Location: USA
Posts: 31
Thanks: 0
Thanked 1 Time in 1 Post
|
So I ran the numbers. assuming a box 10" x 8" x 4", 1/4 inch Lexan would dissipate 150 watts at a 25 C temperature difference. So the thermal shutdown starts at 75, measured at [close to] the MOSFETs. Assuming it's 30 C out (86 degrees)...
MOSFETs are 167 F (75 C, when the throttling begins)
box air temperature is 131 F (55 C, or 36 F cooler than the mosfets; anyone with a case and CPU thermometer want to sanity check this?)
outside temperature is 86 F (30 C)
So even with that rather large temperature difference between the MOSFETs and the inside air, you're dissipating quite a lot of heat. Most of the heat is going to go out the heatsink, of course. If the inside temperature was higher, still more heat would be dissipated through the Lexan. If the box was bigger, same thing. Ditto if the lexan is thinner.
Are these assumptions good? Is that too high an operating temperature?
EDIT: also, whichever material you choose remember to paint it matte black! Silver or white wouldn't radiate away heat as well.
Last edited by MissileStick; 05-15-2009 at 01:44 AM..
|
|
|
05-15-2009, 01:47 AM
|
#1259 (permalink)
|
PaulH
Join Date: Feb 2008
Location: Maricopa, AZ (sort of. Actually outside of town)
Posts: 3,832
Thanks: 1,362
Thanked 1,202 Times in 765 Posts
|
Oops! I forgot to write the dimensions! Sorry Missile!
Ya, maybe 11x7x3.5 is a good guess.
Maybe the Lexan is fine?
|
|
|
05-15-2009, 09:00 AM
|
#1260 (permalink)
|
Master EcoModder
Join Date: Jun 2008
Location: London, Ontario
Posts: 1,096
Thanks: 0
Thanked 17 Times in 14 Posts
|
Call me crazy, but in this business of electronics I have NEVER seen somebody custom design a box just to hold a couple circuit boards. It may be too late, but i think it would have been a good idea to pick an off-the-shelf case and put screw holes on the circuit boards to match it. Personally, i'd shoot for one of those standard cases that car amplifiers come in. They are already designed for heat.
On the other hand, despite Missile's worries, bending sheet metal is EASY stuff. I did some 90* angles to create ducting for my race car using a steel table, a couple C clamps, a 2x4 and a hammer..
|
|
|
|