Power GaN FETS now at DigiKey...

[scolino]

There have been many high density packages released lately, and with demand for increased system density, these packages will become more difficult to use. EPC has recently produced an assembly application note, http://epc-co.com/epc/documents/prod...aNassembly.pdf that discusses bench assembly.

[theAnonymous1]

Originally Posted by zhoufang

I'm not sure if I missed something.
The figure 2 (Id vs Vgs) of the document shows great linearity.
If that graph is truly plotted using test data instead of just being an illustration skatch, I think this MOSFet might be great for linear operation too, I.e. Class A/AB amps

A discussion was started in the Pass Labs forum about linear use......

This is some FET !!!

[bwaslo]

Originally Posted by scolino

EPC has recently produced an assembly application notethat discusses bench assembly.

Thanks for the link! I hadn't noticed that. They refer to a way of soldering with flux but without solder paste. Might be possible to do it with a homebrew heating setup. Boards would have to be made with a solder mask, but that's not too expensive to have done these days. Tempting... I'll have to read a lot more about class D design before jumping, though.

[zhoufang]

Originally Posted by theAnonymous1

A discussion was started in the Pass Labs forum about linear use......

This is some FET !!!

Thank you for the note!

[bwaslo]

Originally Posted by zhoufang

I'm not sure if I missed something.
The figure 2 (Id vs Vgs) of the document shows great linearity.
If that graph is truly plotted using test data instead of just being an illustration skatch, I think this MOSFet might be great for linear operation too, I.e. Class A/AB amps

Yeah, but the values in the left column are quite a few Amperes where the graph gets linear (maybe 10 or so). I don't see one of these dissipating a lot of watts in that teeny package like it would have to do for Class A.

[iand]

Originally Posted by ChocoHolic

Well, for class D in the range of some single kW and switching frequencies up to 1 MHz, I could survive with TO-220 and would love TO-252 (D-pack). The resulting inductances can be handled without to much headache.
But the body diodes of Si-MosFets are a torture for my naive soul, and placing a series schottky + additional fast freewheeling is also not charming at all.

Some of the Fairchild 500-600V MOSFETS intended for LCD TVs have trr down to about 50ns.

You need quite a few in parallel but since they cost less than a dollar each this isn't really an issue, and you do get much lower inductance as a bonus.

Ian

[jackinnj]

Since EPC has a model for the EPC1011 I did a comparison with the venerable IRFP240:

[alvaius]

Pretty cool part. I can certainly see these in my future.

I find the low Vgs pretty exciting especially when coupled with the low gate charge. That makes for an easy and fast logic level driver.

In terms of soldering something like this it is not that hard. You will need a proper PCB of course. Put down a bit of solder paste, put on the part, then sit it on a heat plate. It is not 100% reliable, but with practice you can get fairly reliable results.

Any interest in some TO-220 adapters?

Alvaius

[Calvin]

Hi,

Jackinnj Distortion sims don´t show the distortion values for equal currents.
Taking Jacks Vgs-values of 6Vdc the idle currents are 560mA for the EPC1011 and 220mA for the IRFP240.
If the dc-part of the IRFP240s Vgs is corrected for same drain currents, distortion figures read 0.186% for the EPC and 0.132% for the IRFP240 (V6 changed to 8.9Vdc). If the Vgs of the EPC1011 is lowered to 2.1V to give 220mA idle current its distortion values rise to 1.265%.
So I don´t see an improvement here and the EPCs seem to need considerable idle currents to show low distortion behaviour.

jauu
Calvin