Best MOSFET for output stages

Status
Not open for further replies.
Jump to Latest
M

millwood

Disabled Account
Joined 2003
#1
I know, I know, there is no generic, universal best. But I am trying to get your favorite.

Mine goes like this:

1) IRF540/9540: they are cheap, rugged, and impossible to destroy. and I got tons of them, 🙂
2) 2sk1058/2sj162: impossible to destroy, low gate capacitance (CFP is possible with this but hard with the IRFs). But they are hard to find and are pricey.
3) IRFP240/9240: I just got some of them and haven't tried. Nelson seems to like them in SE applications.

your favorite?
 
#9
millwood said:

2) 2sk1058/2sj162: impossible to destroy, low gate capacitance (CFP is possible with this but hard with the IRFs). But they are hard to find and are pricey.
Click to expand...

Have you considered the two British clones made by Exicon
and Magnatec? They are cheaper and the Exicons even have
slightly better specs. Magnatec also has twochip devices in
one package for higher power handling.

You can buy Exicon devices directly from www.profusionplc.com/hitachi.htm
for prices that are at least much better than what I have to
pay for the Hitachi devices here in Sweden. The nice thing is
that there is no extra charge for shipping and handling etc.
as far as I understand. I will probably place an order myself
from them some day. The only reason I haven't done it yet is
that I am somewhat worried about matching, since I expect
one is likely to get devices from the same batch, and so may
end up with N and P devices from batches that don't match
well.

Edit: I just checked out that site again and they have dropped
the prices for Exicons quite a bit since last time I checked, so
they are only around 4 Euro now (+VAT, I think).
 
#12
peranders said:

It's smart to have some anyway because this is a nice theory, not very realistic.
Click to expand...

Why? Because of bias drift and its influence on the sound?

Assuming you have at least some heatsinking they won't
need it for pretection due to their temp. coeff. behaviour.
I have seen a lot of schematics using lateral MOSFETs and
I can't remember any of them having any temp. compensation.
 
#15
Drafance007 said:
millwood said:

1) IRF540/9540: they are cheap, rugged, and impossible to destroy. and I got tons of them,

Is there any good project on forum for this devices....I also have about 40 of them...very chep realy...?
Click to expand...


you can always try Nelson's Citation 12 mod: a little bit dated but still good. I also dreamed up one, with a T-driver stage posted somewhere here a while back.
 
#18
5th element said:
Ive used the exicons from profusion as direct replacements for the sk,sj's, no problems at all and MUCH less expensive for me in the UK. Farnell stock the sk and sj's but at like three- four times the price!!, thankgod for profusion!😀
Click to expand...

Did you have any problem with matching of the Exicons? Since
profusion seem to have direct connections with Exicon I suspect
one gets devices fromt the same batch and not some average
mix from the shelf, so I am worried about ending up with
perfectly matching N devices and perfectly matching P devices,
but the Ns and Ps not matchin each other?

BTW, they were very friednly and helpful when I mailed them
asking for Spice models once.
 
#19
2sk1058/2sj162 have a nice low Cfb (10pF) but their other characteristica seem to be inferior to say IRFP240/9240.
Especially the RDSon is very poor, and therefore losses are high.
RDSon for 2SK1058 is 1 Ohm at 5 A, and max ID is 7 Amp.
For IRFP240 these data are: 0.3 Ohm (at 150 C) and max ID is 12A at 100 C. Capacitance though is higher 130 pF, but i see no real problem in that.

I will have to agree with Nelson Pass, if he likes the IRF devices more, and i don't really see the point in spending extra money for 2sk1058/2sj162.
 
#20
RE:
"i don't really see the point in spending extra money for 2sk1058/2sj162"

It may be significant to some designs that the 2sk1058/2sj162 changes from posative to negative thermal coefficient at a very low current (less than 1 Amp). THe IRF devices switch over at something like 10 amps. Thus the Toshibas can be configures in a SF output section requiring no thermal tracking. While the IRFs may be self-protected from thermal runaway they may require some kind of thermal tracking .when operated below the 10 amp point.

There are probably ways to design around the imperfections of either device, but tossing one or the other out of consideration without regard to tradeoff limits ones choices.
 
Status
Not open for further replies.
Top Bottom