Pushing OnSemi transistors to the limit

[wg_ski]

I'm not sure what you mean by "does not count". Are you saying that current limiting is pointless with mosfets because they do not have the same SOA as bipolar transistors?

Instantaneous IV limiting isn't really needed. MOSFETs aren't subject to any secondary breakdown mechanism. You can overdissipate them at full VDS by a factor of two, and you may very well melt the epoxy case before the die catastrophically fails. In a conventional audio amp with MOSFETs, you may only need a mains fuse for protection.

 

[Kevin_Murray]

I don't mean to hijack this thread so this is my last post regarding output protection.

Instantaneous IV limiting isn't really needed. MOSFETs aren't subject to any secondary breakdown mechanism. You can overdissipate them at full VDS by a factor of two, and you may very well melt the epoxy case before the die catastrophically fails. In a conventional audio amp with MOSFETs, you may only need a mains fuse for protection.

I would argue this because I have had mosfets destroyed several times from over driving them. This was under lab conditions where the output was monitored. Vertical fets do not self-protect in the manner of lateral mosfets. Furthermore, absense of secondary breakdown is no reason to dismiss the benefits of output protection. Mosfet failure modes are more than fast enough to protect a fuse.

 

[ilimzn]

I'm not sure what you mean by "does not count". Are you saying that current limiting is pointless with mosfets because they do not have the same SOA as bipolar transistors?

No, what I meant was that pushing MOSFET specs does not count in comparison to pushing BJT specs in this context because MOSFETs don't have secondary breakdown, whereas with a typical BJT this tends top be the problem when pushing it as far as it will go.
Like you, i have found that some form of protection is needed, with MOSFETs so far a simple current limit and appropriately dimensioned fuse has proven itself adequate. Vgs limiting is of course always mandatory (AFAIK only laterals have built-in zeners).

 

[EWorkshop1708]

I think that Onsemi's BTJ are very "conservative" rated...

They usually can take much more than rated - especially the 4281 and 21194 with their opposite

The only "Onsemi's" that I have blow up in the last 2 years are the counterfeits

I've had good luck with those transistors..........almost driving a complete short at nearly full output swing!
I'll tell you a story

I built a single mono amp a while back with 5 pairs of the MJL 4281/4302 transistors to drive 350W @ 4 ohms into a subwoofer. I imagine only 2 or 3 pairs were actually needed. 50-0-50V toroid transformer. +/- 75V rails unloaded.

Since I built the amp, it's blown two different 12 inch subs! 1st sub was a generic, 2nd sub was Sony Xplod P5W 1200W MAX, 350W RMS.

When I blew the 2nd sub, I was driving full volume into it playing some rock music, when a bright flash of light flickered through the subwoofer's cone, and the 6.3A mains fuse on the amp blew. After that, the Sub measured only 0.9 ohms and the sub smelled burnt to a crisp. After replacing the fuse, amp worked as normal - although I needed a new Sub.