Electronics > Beginners
Operation in This Area is Limited by R DS(on)
Jwillis:
I purchased an inexpensive electronic load off EBay.They're not much different than the one Dave made other than the built in monitoring . It works fine and fairly accurate as long as it's doesn't get over loaded. I burned out the original MOSFET (IRFP260) but I had no exact replacement so I put in an IGBT (G60N100) and it worked great and could handle a bit more than the original. I also changed the heat sink to a much larger 6 pipe double fan. It doesn't seem to go much higher than 40oC with the heat sink.
I'm testing a 0-40V 0-20A power supply I'm building and made some changes to the biasing on the power transistors and wanted to do a retest . But I turned the control on the load maybe a bit to high and of course destroyed the IGBT. I thought I would order some MOSFETS that can handle the voltages and currents that would fall within the parameters that I need.
The IXFH50N60P3https://media.digikey.com/pdf/Data%20Sheets/IXYS%20PDFs/IXFx50N60P3.pdf
or The FDL100N50F https://www.mouser.com/datasheet/2/149/FDL100N50F-1007460.pdf
My question is in the Datasheets where the Maximum Safe Operating Area graph is theirs a note "Operation in This Area is Limited by R DS(on)" I've looked on the net for some explanation but I'm probably not searching right. Can anyone tell me what that means .Am I going to have problems turning these MOSFETs on at lower voltage and high current?
Ice-Tea:
The SOA maps the current in function of the Vds. However, the Vds can not be higher than the product of the current and Rds. Imagine a 1 ohm Rds and a current of 10A. Obviously, Vds can't be higher than 10V ;)
magic:
--- Quote from: Jwillis on December 31, 2019, 07:33:46 am ---Am I going to have problems turning these MOSFETs on at lower voltage and high current?
--- End quote ---
You will, because of their internal drain-source resistance, RDS(on) ;)
There are bigger problems with MOSFET SOA, though. Read the following thread and in particular this post by David Hess and the links he posted.
https://www.eevblog.com/forum/chat/darlingtons-used-as-audio-output-devices/msg1388118/#msg1388118
You have two MOSFET vendors who pretty much admit that their SOA ratings at DC are plain bullshit. International Rectifier even blew one of their parts at 16V / 4A, which was presumably well withing the part's documented dissipation and temperature limits.
And of course, the big "2.5kW power dissipation" spec applies only if you keep the heatspreader of the device at 25°C. Otherwise you need to multiply power dissipation by junction-to-case thermal resistance (from the datasheet) plus case-to-heatsink resistance, typical values scrapped from various datasheets below.
--- Code: ---TO264 .15 K/W
TO247 .21 K/W
TO220 .40 K/W
--- End code ---
IGBTs just suck for linear, don't use them :P
magic:
I apologize for misinformation, the device destroyed by IR actually has very limited SOA spec and is not supposed to withstand 16V / 4A operation. The appnote also shows that IRHM9260 (presumably related to IRF9260, the P-complement of IRF260) can withstand over 100W for 400ms. I observed a tendency for newer, low-RDS(on) parts to be rated much worse in terms of SOA than dinosaurs like IRF240/260/640/etc. BTW, these old HEXFETs are also widely abused for audio by hobbyists and even some commercial vendors and they seem to tolerate it.
Not sure what to think about Fairchild; on one hand AN-4161 clearly says that SOA curves are typically specified under assumption of uniform die temperature which may not always be true, on the other hand I cannot imagine how they would attempt to defend themselves if a part with guaranteed DC SOA rating of 10A/100V/Tc=25 turns out to blow up in such conditions.
I noticed the FDL100N50F myself and it certainly looks interesting. If you don't mind risking blowing it up I think we all would love to hear if it's any good ;D
Jwillis:
--- Quote from: magic on December 31, 2019, 09:04:12 am ---IGBTs just suck for linear, don't use them :P
--- End quote ---
I think I'm starting to understand the limitations when dealing with the RDS(on). Probably explains the voltage drops I was getting.
The IGBT G60N100 I used worked very well even over extended periods of time within its stated SOA parameters .The newest test is what killed it .At 40 V 20A it clearly went into second breakdown.https://www.mouser.com/datasheet/2/149/fairchild%20semiconductor_fgl60n100bntd-320282.pdf
I screwed up by not watching the load while I was checking the temperature of the power transistors on the power supply. The biggest problem was when the IGBT burned out it shorted closed. I panicked when copious amounts of smoke started pouring out under the heatsink. I could see the electronic load had reached the 20A limit I had set, but neglected to see what voltage on the other meter. IGBTs are fine to use but like BJTs you really have to watch that second breakdown.
--- Quote from: magic on December 31, 2019, 08:00:50 pm ---I noticed the FDL100N50F myself and it certainly looks interesting. If you don't mind risking blowing it up I think we all would love to hear if it's any good ;D
--- End quote ---
I ordered several of both IRFP260 and FDL100N50F. Destroying the IRFP260 could have been a power spike because I did find one in my scrap pile but its leads were broken off. I attempted to fix it by craving back the casing enough to weld leads on but the solder melts before I can get an accurate reading. That's really hot. So I think the IRFP260 should be able to take 40V at 20A but its going to be pushing it hard to the edge of the SOA. I just have to be more careful when approaching the top end. I have another electronic load coming so I can experiment with this one more. I think I should be able to parallel mosfets off board.
Navigation
[0] Message Index
[#] Next page
Go to full version