Irf540 Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi Lars,

That is almost true if the silicon die in the TO247 is as small as the die
in the TO220. The idea is nice though, cause a weak point is the insulation washer.

You may mant to have a look at IXYS isoplus packages with internal
AlO3 insulation laminated with the dies copper base. Nice !

/ Mattias
 
ekaerin: Do you mean one of these babies? :D
 

Attachments

  • ixys.jpg
    ixys.jpg
    50.8 KB · Views: 695
To continue where last where som days ago, the high capacitance in the mosfet's is hard to overcome without some extra drivers.

But typically it is very close related to the RDS on.

When this go down, the capacitance goes up with the increasing density in the mosfet's...

If we can overcome the capacitance problem with a driverstage, with ... lets say smaller mosfet's.
There is some posebillity to get low loss types with very low Qg...

I have found some Generation 7 mosfets with:

Qg=43nC@vdd = 250V,id=22A

td(on) 8nS,tr = 6nS
th(off) 18nS,Tf = 2nS

PD = 265W for a TO247/D3PAK house... unfortunatly the are only available in N-channel and they are expensive..
 
Lars,
that guy is big! Like motor drives and other use where speed is not
that important, dare I say audio in this context......

The Isoplus feature goes down at least to TO220.
I've used the PLUS247 package with this feature and it's great.

But as Sonnya pointed out, most development surrounds N ch FETs.
The main target is switchmode power where P ch FETs are seldom
used.

/ Mattias
 
Thanks
(( You can overcome the small-package-problem by mounting it directly on a solid copper block of 2 by 2 by 1 cm, without insulator between TO-220 and block (only compound), and instead insulation between block and heatsink. This way you can dissipate as much power in a TO-220 as a TO-247. It is not a very cost effective solution tho.....))

actually I do the same with Irfp240 & Irfp9240 , these devices due to their good speed can sink high currents faster than bipolar transistors , hence the get heated faster too
 
Hey Dj BASS AMP

I came across this cct today. It is an old elektor design. This design will definately swing rail to rail.

As it stands with the rails shown it will only deliver about 20 watts into 8 ohm (40w into 4 etc) but if you use a higher voltage op amp you can raise the rails and probably achieve around 200 watts into 4 ohms.

Of course all of the other components will need to be changed to suit the higher rails, but not a major problem.

So if your not worried about Hi-Fi and really want to use IRF540 / 9540 then this could be it.

Cheers
 

Attachments

  • simple mos amp.gif
    simple mos amp.gif
    14.7 KB · Views: 730
Then you are lucky with your thermal situation :)

I think you ought to modify your components list or at least point out the pinout differences:

T7 = BUZ900DP, IRFP240, 2SJ201, 2SJ162
T8 = BUZ905DP, IRFP9240, 2SK1530, 2SK1058

These devices have different pinouts and the uninitiated builder will not know what to do. The result will be destruction of the devices upon power up. It's not clear on your layout which pinout the design is for.
 
Commentable Experienced Thoughts

richie00boy said:
Hi Ampman, I like your designs, looking forward to seeing more of them.

RITCHIE TELL ME ABOUT UR SELF?
Which design do u like most and why??

Look at the pic of my younger brother when he was just 1 month old, but now he is 16.

ampman
 

Attachments

  • ritchie.jpg
    ritchie.jpg
    11.8 KB · Views: 403
ampman,

I just enjoy reading and learning from your often animated posts :) A lot of them contain new and intriguing concepts.

I have a degree in electronic engineering and have been into DIY audio for about 10 years. At first I was into speakers, now I like amplifiers as well. I think class-d is the future for >100W, but it's a challenge to DIY.

Nice picture of your brother, I bet he is pleased you posted that on here ;)
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.