Zen Mod said:
did you tried good one ?
( I presume that peeked one wasn't good for trying ......)
No, as I saw no point in doing so, due to the rather hefty input capacitance.
It was actually a perfectly good mosfet I opened, as I wondered if it was one big mosfet, or many smaller dies in parallel.
If you take a look at the data sheet, you will find that the input capacitance VS. wattage, is not too far off being the same as a bunch of IRFP xxxx in parallel.
Now, go take a look at "home".
Magura 🙂
Tyimo said:
Check out advanced power technologies power mosfet modules. 750w in a SOT 224/227 (recalling from memory) package (more for bus bar use). They are almost certainly what the devices are.
I was referring to much bigger mosfets like 50A-200A continous types for motor use.
PS. Can't seem to find this company anymore! Has it been taken over recently? Or was it Applied power technologies ? US company.
Kevin
Of course it is clear that the devices are not being run any where
near their ratings. If you want to play with this circuit, there is no
need to chase down those particular parts, and it is clear that they
are parallel devices in a package anyway.
Interestingly, because they have been designed for switching, you
do not have an assurance that those chips are actually matched
for linear operation.
😎
near their ratings. If you want to play with this circuit, there is no
need to chase down those particular parts, and it is clear that they
are parallel devices in a package anyway.
Interestingly, because they have been designed for switching, you
do not have an assurance that those chips are actually matched
for linear operation.
😎
I'd suggest going for a single die device instead of a random SOT227 part.
The biggest one you can get will disappate upwards of 800W from a single die, and then you don't have the potential oscillation problem associated with using the random twin die part, in the linear region.
The biggest one you can get will disappate upwards of 800W from a single die, and then you don't have the potential oscillation problem associated with using the random twin die part, in the linear region.
Lars Clausen said:
This is wrong, being made on the same case/fabrication means the capacitances/characterisitics align very closely otherwise you will get hotspots etc and it wouldn't be a 750 watt part with manufacturer's warranty or SOA! .
Regards
Kevin
Fanuc said:
I don't doubt that the chips are from the same lot code, but that is no
assurance that the Vgs is matched. If the part is designed for switching,
the manufacturer is unlikely to care about that and is unlikely to go to the
trouble of matching dies. If they did, I expect that the data sheets would
be bragging about it.
😎
Steve at apexjr has some large mosfet devices for a song if anyones interested. I can post the data tomorrow at work.
Szervusz, Tyimo!
I found this for you suddenly. I don't know, this is good for an amplifier?
But only a thousand HUF.
http://www.lomex.hu/pdf/irfp9n2d.pdf
Gyuri
I found this for you suddenly. I don't know, this is good for an amplifier?
But only a thousand HUF.
http://www.lomex.hu/pdf/irfp9n2d.pdf
Gyuri
Sorry i found some devices with 800W dissipation in a plastic housing, but i'm not 100% they are single die.
http://ixdev.ixys.com/DataSheet/99194.pdf
This one contains the biggest Coolmos single die available:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SPW47N60C3IN-ND
But only 415 W.
http://ixdev.ixys.com/DataSheet/99194.pdf
This one contains the biggest Coolmos single die available:
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SPW47N60C3IN-ND
But only 415 W.
Lars Clausen said:
ONLY 415W!?!😀
I'm trying these with the Mini-A no doubt. Would the front end bias need to be increased a bit to properly drive the Ciss of 6800pF? I' m not too good with the graphs on the datasheet, but what would be the optimum voltage and current to operate these devices?
A diff pair biased at about 50 mA would be about right. The original
Mosfet input circuits would do the job with some heat sinking.
😎
Mosfet input circuits would do the job with some heat sinking.
😎
That Coolmos is in a stock TO-247 package, so you can load it at best with about 30 W - just as the original parts.
Well, but the 40 S transconductance could help 😀
A real switcher, not even a transconductance plot in the datasheet.
Have fun, Hannes
Well, but the 40 S transconductance could help 😀
A real switcher, not even a transconductance plot in the datasheet.
Have fun, Hannes
h_a said:
I was thinking the same thing. How is it possible for 415W, in a TO-247 pkg? I would think that a TO-264 would be capable of something closer to that.
That's a theoretical limit for high (but short) pulses. Certainly not a continous rating. With Silicon as semiconductor no way to dissipate such power in a single package.
Have fun, Hannes
As long as you're spending big money on output devices, try the
Semi-South SJEP120R063 at 273 watts for $60.
😎
Semi-South SJEP120R063 at 273 watts for $60.
😎
Hannes: That depends on how you mount the device onto the heat sink, and what your end heat sink temperature is.
I've been running TO247 devices on 200W+ dissapation for weeks without problems, running thermal tests.
A TO264 package can potentially contain more than one die.
I've been running TO247 devices on 200W+ dissapation for weeks without problems, running thermal tests.
A TO264 package can potentially contain more than one die.
Nelson Pass said:
Oooh...and its a Jfet too!! Sample request has been sent. Dare I ask what the minimum order is?
- Status
- Not open for further replies.