At those values the power dissipated by the part is "only" 73W. Actually the part is "package limited" to 195 A because bonding wires can't take it.
I am not sure how they come up with 375W but here is a simple calculation: The package's thermal resistance is 40 C/W when mounted on a 1 inchsq 1 ounce copper pad (I usually don't even have that much area but multi-layer helps) so if you limit junction to 150 C, the max one can do is 3.75W which comes out to ~38 Amps.
-- Muzaffer Kal DSPIA INC. ASIC/FPGA Design Services http://www.dspia.com
375/3.75 = 100. So it's practical to use this part at 1% of its specified maximum power dissipation!
Maybe they just slipped two decimal points on the datasheet power dissipation, and one decimal point on the max current in the flyer.
John
-- Really? You ship finished goods without testing them?
-- Oh, well, it's your loss... 60Vdss, Rds(on)max 2.5milliohms, 375 watts @25C Tc, it's a nice part. JF
No, no, no. Experience is the worst teacher. Thinking and math rule; experience only reminds you of things that shouldn't be ignored for analysis, not of actual expected results. Experience must be constantly questioned because things keep changing.
I know it will explode at 340 amps, and even at 190, unless the connections and cooling are done under some insane laboratory fixturing that is absurd for a DPAK part on a real board.
No, I know the literature is absurd and deliberately deceptive.
There are penty of perfectly good, and cheaper, mosfets around, from people who are not blatantly dishonest. I'll buy them.
Rant all you want. If you want to use DPAKS at 195 amps and 375 watts, go for it.
John
think this has part of the long explanation as to how the come up with the numbers
-Lasse
-- Well, since the spec reads 375W @Tc = 25C, if you're running it in an ambient of 25C, there's _no_ way. However, what's not practical for you might be eminently practical for someone else who's willing to use, say, chilled water. Then, of course, there's always derating and using multiple devices. JF
Chilled water to cool a tiny surface-mount package?
This uses TO247 packages, clamped to a gold-plated, water-cooled copper block, at those sorts of power levels. But the lead current is a lot lower than 340 amps.
John
John Larkin: Pomposity at its glorious extreme. ...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
The only thing bipartisan in this country is hypocrisy
-- And you learned that, how?
Quote "To actually utilize our parts in an application that targeted currents at the level of the ultimate current would likely be costly and impractical. Nucleated boiling is an expensive and tricky proposition."
... and ...
Quote "The heritage fleet of power MOSFET products has been rated via the classic method as described by equation 1. Those ratings fall on a very conservative engineering side. Newer, advanced products like the expanding suite of IR Trench products will begin carrying the new ultimate current limits, if appropriate."
Why does this remind me of the real estate market?
I want heritage datasheets :-)
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
It seems "normal" these days. Take a look at this datasheet as an example, just came up in Raveninghorde's thread:
Here you have a driver that is touted as 2A, 3A peak. The figures in the datasheet are of little use, not output loading curves and the table data has a wee remark that the device sits at VDD/2 when sinking 2.5A, sourcing is even worse. Easily overlooked. A young engineer might exclaim "hooray!" and design it in, only to have it almost unsolder itself or going *PHUT*. Then after a few experiences like that he might think that analog is a frustrating business to be in and move on to other fields :-(
"Modern" datasheets can often only be used when reading between the lines and applying a healthy marketing hype correction factor.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
-- Pussy! Here's high current: news:huc7q55cdj7nodallcs33mon2p03bnv2p8@4ax.com JF
The IR testing context isn't 100% clear to me, but using boiling liquid may be the way they get the DPAK source lead to not melt at their specified currents. Ditto not melting their solder joints to a pc board.
John
Sanity, at last!
John
...in that case, i have some #30 wire rated at 1,000 amps and will guarantee it or double your money back!
You mean going out of business like Signetics?
John
PC board?? Tack 24AWG wires onto it and sink the whole thing into the refrigerant. PC boards are excellent insulators in comparison.
Tim
-- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
Anybody here remember the Signetics WOM? As I recall it was intro'd back in 1972 about this time of year. Maybe IR is following suit? Art
Nope. Intro'ing a product like the WOM. Art
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.