# MosFet transistor ratings?

#### neorules

Joined Nov 30, 2007
12
I am looking for two types of Mosfet or BJT transistors.
When I search for mosfets in the 100 amp plus range I find these IRF7769L2TR1PBF
The ratings are confusing since they are rated at 350 amps and 100v.
But they are only rated at 3.3w......
So it is obvious by the surface mount and case design they could never handle 100 amps at say 48 volts. That would be 4800 watts right?

I guess I should search by watt rating but that gets me some 2000v 40amp transistors...

Is there anyway to navigate the ratings game easily?

I am trying to control a BLDC motor in the order of 1000 amps max with normal draw in the 100 amp range.

I don't like the idea of paying $2000 for an IGBT since I need a bunch. Thanks JC P.S. if I use a dielectric oil can I submerge the whole controller for cooling purposes? or is that just a stupid question. #### Papabravo Joined Feb 24, 2006 16,168 I am looking for two types of Mosfet or BJT transistors. When I search for mosfets in the 100 amp plus range I find these IRF7769L2TR1PBF The ratings are confusing since they are rated at 350 amps and 100v. But they are only rated at 3.3w...... So it is obvious by the surface mount and case design they could never handle 100 amps at say 48 volts. That would be 4800 watts right? I guess I should search by watt rating but that gets me some 2000v 40amp transistors... Is there anyway to navigate the ratings game easily? I am trying to control a BLDC motor in the order of 1000 amps max with normal draw in the 100 amp range. I don't like the idea of paying$2000 for an IGBT since I need a bunch.

Thanks JC

P.S. if I use a dielectric oil can I submerge the whole controller for cooling purposes? or is that just a stupid question.
The answer to your question is no. You have to know what you are doing and what you are looking at. Each of the different numbers in the data sheet represents different conditions of temperature and heat sinking. You are correct in your assessment that a surface mount part is unlikely to be suitable for your application. You are also working under dangerous conditions and IMHO this thread needs to be closed as it is a violation of the terms and conditions. You'll shoot yer eye out kid!!

Joined Dec 26, 2010
2,148
No, these ratings do not work in the simple way that you imagine. The maximum voltage rating can only be applied continuously when the device is either non-conducting or passing such a small current that the wattage rating is not exceeded.

The full current rating applies only when the device is turned on so that the voltage across it is small - and note that the full ratings may not be available continuously with a feasible heatsink temperature.

Liquid cooling may be possible, but it's fraught with difficulty. All that draining down or heaving out when something goes wrong, worries about lethal breakdown products, or toxicity full stop, bubbles in the oil, leaks, compatibility problems turning polymers into something like snail slime...

I hope one day that my nightmares about drowning in the dielectric fluid tank will go away. After all, that stuff is dense enough: why don't I float?

#### neorules

Joined Nov 30, 2007
12
The answer to your question is no. You have to know what you are doing and what you are looking at. Each of the different numbers in the data sheet represents different conditions of temperature and heat sinking. You are correct in your assessment that a surface mount part is unlikely to be suitable for your application. You are also working under dangerous conditions and IMHO this thread needs to be closed as it is a violation of the terms and conditions. You'll shoot yer eye out kid!!
OK... I didn't know I violated anything...
Could you please clarify? I have tried to look up the Terms and Conditions but I cant find the link.

Thanks JC

#### neorules

Joined Nov 30, 2007
12
http://mcmanis.com/chuck/robotics/projects/esc2/FET-power.html
Thanks I bookmarked that page back in 2004 ish... it is a great link.
I was wondering if anyone had come up with a more easy to use cheat sheet or numbering scheme.
If not I will just pour through all the data sheets of instock items.

I have built about a half dozen or so (open source) controllers for brushed applications. Much of what I have learned about that I got from chucks page and 4Q- page and OSM- page (i didn't list names because I don't know if that violates the terms (see previous post))

Thanks much JC

#### neorules

Joined Nov 30, 2007
12
Liquid cooling may be possible, but it's fraught with difficulty. All that draining down or heaving out when something goes wrong, worries about lethal breakdown products, or toxicity full stop, bubbles in the oil, leaks, compatibility problems turning polymers into something like snail slime...

I hope one day that my nightmares about drowning in the dielectric fluid tank will go away. After all, that stuff is dense enough: why don't I float?
LOL sweet dreams.....

The motor I am drying to control is oil cooled and all my previous controllers are forced air cooled. I could stick with the forced air but since I am already pumping oils around I was wondering why oil cooling wasn't done on the board itself.

I guess being compatible with all the different plastics is the problem.

I already have a water based heat exchanger on one of my controllers (again forced air over the tubes that are indirectly placed) and it works great. I even thought about modifying CPU liquid cooling but that causes so many other issues like physical contact with the heat sink on the different parts of the H bridge causing shorts.

hmmmm... probably stick with forced air and heat exchanger idea....

Thanks JC

#### SgtWookie

Joined Jul 17, 2007
22,221
One of the local colleges around here took a novel approach to cooling a computer; they removed the motherboard & supply from the box, and put it in a fish tank, then filled the fish tank with mineral oil. The cooling fans run quite slowly, but the computer doesn't overheat.

#### neorules

Joined Nov 30, 2007
12
One of the local colleges around here took a novel approach to cooling a computer; they removed the motherboard & supply from the box, and put it in a fish tank, then filled the fish tank with mineral oil. The cooling fans run quite slowly, but the computer doesn't overheat.
Awesome LOL....
That is exactly what I was thinking.... instead of having 16 or 32 mosfets standing out in the air why not "sink" them all at once without the fear of shorting out...
Back in 76 or 77 I built a cantenna for a dummy load and that was the basis for my question.
I used it for years until amateur radio lost it's charm with me.

I might just take one of my brushless controllers and submerge it in transformer oil or mineral oil and see what happens.

The biggest problem with my OSM- controllers is they do get hot and the fan by itself isn't the greatest.

Thanks JC

Joined Dec 26, 2010
2,148
You might try looking at fluids specifically designed for the job, such as the Fluorinert family of fluids manufactured by the 3M corporation. http://www.3m.com/product/information/Fluorinert-Electronic-Liquid.html

NB I have no affiliation with 3M, and of course other manufacturers of such materials may exist. I do not know either whether they can even be sold legally to amateurs, nor what the cost might be.

There are certainly some serious downsides to these chemicals: under adverse conditions fluorinated liquids can generate extremely poisonous breakdown products, and may be very aggressive to some materials, as I recall particularly silicones (not silicon itself).

#### colinb

Joined Jun 15, 2011
351
The 3M products are something like 100 times more expensive than mineral oil. What are the disadvantages of mineral oil as a bath that make the Fluorinert worth it?