IRS2092 2x500W/1000W BTL

[b_force]

I'm orienting on some designs for a 2 x 500W (@4ohm) / 1000W BTL (@8 ohm, full bridge) design with around ±80-90Vdc power supply voltage.

I really like the "simplicity" of the IRS2092

However, I was wondering if this driver can be used to load around 2 x 500W/1000W by default, or do I need an extra MOSFET driver, like the TC4420 (or a bjt pair)?

 

[sweetperfume]

Yes to increase the pair of fets you will need an extra driver....this will increase the speed of switching....and you will have to drive it on 75volts to achieve that output if you use two pair of irfb4227 at 4ohms load per channel...means 500watts non clipping

 

[b_force]

Ok, so I understand that it is not possible to drive 500W with one pair of fets?
You need to parallel them.

Is there a way to determine when an extra pair is needed?
(or could that be find in the application notes of IRF?)

 

[sweetperfume]

The irs2092 can only draw 1amp current....for more power you will require to increase the fets....without buffer the switching speed decreases causing a load on the driver and resulting in a good smoke out of it.....

 

[b_force]

The irs2092 can only draw 1amp current....for more power you will require to increase the fets....without buffer the switching speed decreases causing a load on the driver and resulting in a good smoke out of it.....

I understand.

But how much driving current (from the IRS2092) is needed for a given output power?

 

[sweetperfume]

The fets and the voltage rails decide the output power with reference to the impedance you use on the output...and the output from the chip is just related with how many fets are you using with it.....normally if anyone wants to drive more than one pair they use buffers

 

[nigelwright7557]

The irs2092 can only draw 1amp current....for more power you will require to increase the fets....without buffer the switching speed decreases causing a load on the driver and resulting in a good smoke out of it.....

The 2092 can supply 1a5.
If you start driving more mosfets the rise time becomes too slow and the 2092 thinks it is having an overcurrent event and resets the 2092.

 

[b_force]

Ok, but a pair of IRFP250 can be used to drive 500W or 1000W BTL if I'm correct.
So no extra buffer is needed?

 

[luka]

for the love of god, never ever use those old fets, when you have 10x better ones out there

 

[b_force]

for the love of god, never ever use those old fets, when you have 10x better ones out there

You are right, it was just an example.

Are there other fets that can drive 500W (or a little bit more) as a pair?
Or is it better to use four fets for better cooling (more cooling surface)

 

[luka]

You are right, it was just an example.

Are there other fets that can drive 500W (or a little bit more) as a pair?
Or is it better to use four fets for better cooling (more cooling surface)

yes, irfb4227 is good start... there are others too, this one is just good to start with, far better then 20y old fets

 

[sweetperfume]

using more fets is a better option but only wen u use a buffer

 

[b_force]

@luka
Good point.
Old products, or (even worse) absolute products is not an option.

I am trying to find out if I need a buffer.

Is there a real benefit to use the TC4420 instead of twp NPN/PNP pairs?

According to AN-1070, I need a MOSFEt with a BVdss of at least 200V.
Also a low Rds(on) is needed, so the IRFP4668 would be a good choice.
However, the IRFB4127 is cheaper and has a lower Qg