I chimed in on the thread for the Linear Tech LTC6090 high voltage CMOS opamp. When going through the datasheet I noticed that they used the Ixys IXTH50N20 and IXTH24P20 Mosfets for a demo of the LTC6090 and LT1166 as drivers. I ordered some of them to see whether we might not have a solution to the disappearance of the high power Fairchild Semi devices, and the peculiarities of the IRFP240/9240 which aren't really complements.
There aren't many of these devices in the US at the moment.
There aren't many of these devices in the US at the moment.
Very cool. Pretty big Ciss, but high YFs to counteract a little bit. Pretty beefy as well.
Last edited:
Hi Guys
Most of the IXYS devices are meant for switching apps. The ones for linear are part of their L-series and have guaranteed dissipation ratings listed on the spec sheet. These also have lower Pd ratings overall than some of the extreme models.
Generally, for linear use, the largest standard package is good for no more than 300W. Those 1kW things have a thermal resistance rating that when looked at closely relies on liquid cooling.
The P-channels are inherently optimised for linear operation if you read the articles written by the IXYS engineers. However, as is usual for North American mosfet design since the 1990s, there are no designed or intended complimentary pairs.
Note also when looking at the SOA curves for most of the IXYS mosfets - and anything with a huge power rating from all manufacturers - that there is distinctly no DC curve, and rarely a 100ms curve. Extrapolating downwards from whatever the longest time curve is provided, you find the linear power rating to be very uninspiring. The high gm loses its appeal and very high cg has to be dealt with by using a low-z driver.
For class-D, you would probably want to go to devices with lower gate charge requirements.
Have fun
Kevin O'Connor
Most of the IXYS devices are meant for switching apps. The ones for linear are part of their L-series and have guaranteed dissipation ratings listed on the spec sheet. These also have lower Pd ratings overall than some of the extreme models.
Generally, for linear use, the largest standard package is good for no more than 300W. Those 1kW things have a thermal resistance rating that when looked at closely relies on liquid cooling.
The P-channels are inherently optimised for linear operation if you read the articles written by the IXYS engineers. However, as is usual for North American mosfet design since the 1990s, there are no designed or intended complimentary pairs.
Note also when looking at the SOA curves for most of the IXYS mosfets - and anything with a huge power rating from all manufacturers - that there is distinctly no DC curve, and rarely a 100ms curve. Extrapolating downwards from whatever the longest time curve is provided, you find the linear power rating to be very uninspiring. The high gm loses its appeal and very high cg has to be dealt with by using a low-z driver.
For class-D, you would probably want to go to devices with lower gate charge requirements.
Have fun
Kevin O'Connor
- Status
- Not open for further replies.