The transistors are very similar. The SOA curves indicate that the MJL3281/1302 are more rugged. The SOA of the MJL0281/0302 is almost identical the the 5200/1943 devices.
I believe that the history of these ON transistors is that they are clones of the 5200/1943 devices. Pick whichever one is easier for you to source.
I believe that the history of these ON transistors is that they are clones of the 5200/1943 devices. Pick whichever one is easier for you to source.
This Cob comparison is apples to oranges, and not really an issue. An examination of the Cob curves shows that these devices are very similar. If the MJE parts are slightly higher, bear in mind that each of them is probably equivalent to at least two of the 2SC5200/A19432 in parallel.
Right now, I'm spinning up an amplifier, and I'd like to use MJE0281/0302 because they are quite a lot cheaper than the 3281/1302 parts. I designed an SOA limiter that lets me use the lesser part (hopefully).
Right now, I'm spinning up an amplifier, and I'd like to use MJE0281/0302 because they are quite a lot cheaper than the 3281/1302 parts. I designed an SOA limiter that lets me use the lesser part (hopefully).
The On Semi 3281is a bigger chip version of the Toshiba 3281. The 2SC5200 was Toshiba’s latest “version” of the 3281-type part when it came out. The TTC5200 is a shrunk-die version. It still “meets the SOA spec” . Chances are the PEAK junction temp is about the same. It will average a bit lower on the larger die, but is the hottest spot that determines reliability. If you want to see for yourself the relative die sizes, crack them open and see. I have.
Smaller dies have better second breakdown (high voltage SOA) performance per square millimeter than larger ones do. Larger ones have more of a tendency to hot-spot and current hog. This has always been the case. Manufacturers have found they can get “equivalent” SOA out of a smaller die. And you do get lower capacitance.
On Semi’s 0281 is a shrunk die version, and pretty comparable to the 2SC5200. I believe the TTC 5200 is smaller still than the 0281.
The On Semi 4281 is a Higher voltage version. It gives up some performance to get there - quasi-saturation is worse. That doesn’t matter much in an amp with 100 volt rails and 10 devices in parallel, which is when you resort to using these. I cringe when I see someone using 4281’s in an amp with 35 volt rails, when clearly they could do better with the 0281. I don’t know about the relative die size of the 4281 compared to the others as I have never ruined a perfectly good $6 4281 and the only ones I’ve ever bought were from Mouser. My guess is that it’s a bit smaller than On’s 3281 since high voltage SOA is better.
Smaller dies have better second breakdown (high voltage SOA) performance per square millimeter than larger ones do. Larger ones have more of a tendency to hot-spot and current hog. This has always been the case. Manufacturers have found they can get “equivalent” SOA out of a smaller die. And you do get lower capacitance.
On Semi’s 0281 is a shrunk die version, and pretty comparable to the 2SC5200. I believe the TTC 5200 is smaller still than the 0281.
The On Semi 4281 is a Higher voltage version. It gives up some performance to get there - quasi-saturation is worse. That doesn’t matter much in an amp with 100 volt rails and 10 devices in parallel, which is when you resort to using these. I cringe when I see someone using 4281’s in an amp with 35 volt rails, when clearly they could do better with the 0281. I don’t know about the relative die size of the 4281 compared to the others as I have never ruined a perfectly good $6 4281 and the only ones I’ve ever bought were from Mouser. My guess is that it’s a bit smaller than On’s 3281 since high voltage SOA is better.
wg-ski is correct. Hot spots limit the SOA of a larger die, without some kind of internal ballasting arrangement. Of course, that also generates heat and reduces SOA.
I'd like to say that this is why I'm trying to use MJL0281/0302 instead of their big brothers, but really, it's because I'm being cheap.
I'd like to say that this is why I'm trying to use MJL0281/0302 instead of their big brothers, but really, it's because I'm being cheap.
It applies to Vertical MOSFETs that are used as amplifier transistors, but not when they are used as switches. Vertical MOS still have a sharp positive tempco at typical bias voltages, and can do thermal runaway.
Vertical MOSFETs are better than BJTs about thermal runaway, because at very high currents, the tempco becomes negative. They can still suffer from runaway and hot-spots, but there is at least some correction mechanism built in. Often not enough to save the device, however.
Laterals have their famous zero tempco at a reasonable bias current, which is what makes them so popular.
Vertical MOSFETs are better than BJTs about thermal runaway, because at very high currents, the tempco becomes negative. They can still suffer from runaway and hot-spots, but there is at least some correction mechanism built in. Often not enough to save the device, however.
Laterals have their famous zero tempco at a reasonable bias current, which is what makes them so popular.
Older Hexfet types were pretty good about SOA, even the 240/9240. But with newer processing they’ve sped them up - getting lower and lower Qgs. The resulting “switching” mosfets are absolutely terrible in terms of second breakdown performance. They call it something else with Mosfets, but that’s essentially what it is. Some of these types may even be “500 watt” rated, but have almost zero DC SOA. Don’t use these modern types in an amplifier. You will blow it. Many tube amp builders make the mistake of using them for linear voltage regulators - at several hundred volts. They blow up. To be successful, one need to choose one with “poorer” switching performance - higher Rds iwith a relatively low current rating will get you high enough DC SOA, even if it’s only rated at a fraction of the Pdiss of a “better” one.