Has anyone done a detailed time line analysis of what happens internally to a MOSFET (or at least given it some educated thought) when it fails in these Class E situations ? Voltage sequencing ? High SWR ? Break-Before-Make ? Inductive circuit discharge spike ? TVS ?, ESD ? We all know the result is local overheating which cascades to device failure.?

The frog did not go deaf when you cut his legs off and he will no longer jump will you tell him to.

73 Kees K5BCQ

I'd be surprised if that ever happens.

They cost $0.34 each.? Unlikely anybody will undertake much Failure Mode and Effects Analysis work.? To learn anything you have to get down to the die and the interconnects.? On a plastic encapsulated part that means mechanical removal (micro sandblast) or chemical removal.? It's not like a custom chip where you can remove a metal cap and see everything exposed after only a couple minutes of work.

If you can see inside you can tell the difference between overcurrent and electrostatic discharge damage.? They look very different under the microscope.

Pretty likely the depth of analysis will be "Darn.? It failed ... replace it".

I used to work as a Reliability Analyst.? If you want high reliability you have to derate the parts, meaning you run them at a fraction of their rated limits of current, voltage, temperature, vibration.

With hobby radio equipment we are surely not running at a conservative fraction of rated limits.? Some of us run these puppies hard and hot.

If you have repeated failures you should be thinking about the Why? yourself, and removing the likely causes until your circuit survives for a long time.

If it failed before you started operating it may have been caused by electrostatic discharge during handling.
If it failed during operation it's most likely VSWR or excessive voltage.

Mouser has them in stock.? Go gettem!

Failure mode is no different to every tube, tranny or resistor since day one.
Exceed it's maximum power dissipation and it fails ;-)

- 73 de Andy -

Kees

As others are conveying, transistors are zapped with too much instantaneous voltage, too much current or getting too hot while operating. Placing an imperfect load on it can greatly raise voltage. These longer duration steady on waveforms make things warmer. And those humans who want every last watt of radiation get into the mix.?

Curt

I'll add another thought to my earlier comment.? I've had no personal experience with it, but I've heard that an antenna can gradually pick up a static charge which can zap your finals at the instant you connect your antenna.? This is probably more likely with a long wire antenna, with some wind blowing.? To avoid it you only have to briefly short the center pin of your coax connector to the threaded ring / shield.? The hard part is to remember to do it.

Antenna static would be just another way to subject your final amplifier transistors to electrostatic discharge.? The thing is, the transistor might look perfect from the outside, and leave you puzzling as to why and how it failed.

It's easy to guess What Went Wrong if your component is visibly damaged, maybe with a smokin' hole punctured, or worse, burn marks.? Its those cases where it still looks intact that make you wonder.

Buy a few spares.? I've got 60 BS170s on hand and haven't needed to replace any yet.? "Carry an umbrella to prevent rain" kind of thinking!

If this is likely, what value bleed resistor at the coax socket would remember better than we might ?

Michael 2E0IHW

Adding a megohm bleeder resister does seem like a good idea.? Wouldn't it need to go on the antenna somewhere, to keep any static dissipated before you go to connect the antenna connector to the rig's connector?? And it's not at all a sure thing that the antenna's coax shield would represent a good ground anyway, to dissipate charge away.

We need Allison or someone similarly with deep knowledge.? Deeper than mine, anyway!? Where are all the RF engineers when you need them?

开云体育

For V/UHF antennas on military aircraft, we used a megohm or more to reduce the discharge. Those antennas used fiberglass radomes and not much could be used for grounding in the blades. ?We used special gaskets that provided contact with the aircraft skin. ?St Elmo’s fire is good example of discharge under certain conditions. ?Sometimes, the discharge is referred to as “P-Static” or precipitation static but never thought it had much to do with moisture. ?Mostly high speed through cold dry air, but the physics geeks didn’t always agree. ?Expect similar things occur on ham antennas subjected to wind. ?

On Sat, Dec 3, 2022 at 11:37 PM, Rich Brandt, AE8AA wrote:

Where are all the RF engineers when you need them?

Plenty on here, but pointless if no one listens.

That's why there's barely a week guess by without the same old questions being asked.

Gets boring after a while.
?
--
- 73 de Andy -

The only reason that the same questions get asked is maybe that they were not answered to where we understood the answer ? When someone answers that high SWR causes MOSFET failures, I need some more detail. Sure, if the?
antenna is not matched, power is reflected back into the LPF network and the PA for dissipation. This will heat up the PA.....maybe too much and the MOSFET goes into a heat, more heat, component value change, avalanche, more heat, burn out cycle. What about MOSFET Gate capacitance discharge? What about Gate voltage spike reduction by adding a small series Gate resistor. What about a reverse diode across the Gate resistor, What about a series Gate - Source resistor.

73 Kees K5BCQ

I'm here just a bit preoccupied...

The long and short of it is... It varies.

Over voltage or over current are likely SWR related.? Assuming nominal
constructions and tuning.

High operating voltage points to the over voltage or current fail case.
Pushing for one more watt!.

The other is ESD, lightening and not a direct hit but enough to induce
high voltages.? ?How high is too high 60V (breakdown voltage for BS170).

One condition all MOSFETs are susceptible to his a case where the
high voltage due to SWR does not exceed the break down voltage but
is in phase with the Gate drive and due to gate to drain capacitance
couples RF back exceeding the gate to source maximum (20V).
This failure can also take out the ACT08.

Generally milling the case will tell you little as often the device melts.

Hint if you blow devices one examine what you are doing.
If you do it more than one stop doing that.

Hi SWR is not alway s a problem but what the actual R,Jx
is more important as a reactive load that match that has
good SWR but not?50 ohms may induce havoc.

Also tuning up on the air using the radio (QCX/QDX) is a
very sketchy practice as invariably you will go through
very bad SWR to get a good one.? I call that
teasing the tail of the dragon.? It bites, and so does
replacing BS170s.

One last item in dry places ESD as in carpet static kills
BS170s as well.? caution when handling.


Allison
------------------
Post online only, please no email.

On Sat, Dec 3, 2022 at 07:03 PM, Mark Phillips wrote:

What about a pair of diodes or even a lamp of some sort?

My R-390(A) has a neon bulb across the antenna input, but of course it is only a receiver.? I have developed the habit of discharging any cable before connecting it to an instrument or radio; briefly shorting inner and outer conductors is sufficient.??
73, Don N2VGU

any of he gas tumes (even NE2) ignite at typically 67-75V.
Diodes are useless as its a TX and the voltage going out is?
going to turn them on and cause issues.


Allison
------------------
Post online only, please no email.

At my former shack, with multiple receive antennas and some long wires for 160, I often could hear little pop-pop noises coming from any non-protected SO-239's leading outward, generated by wind static.

I installed an NE-2 neon bulb at the antenna input of all of my receivers, to assist in clamping that spark to ground.

73! Mark K9TR

A 100k resistor will easily discharge the static?
without any averse effects.?

--
Allison
------------------
Post online only, please no email.

Kees,

the gate of mosfets can be protected with a zener.?

But it adds capacitance at that node, counter productive.?

p

--
Allison
------------------
Post online only, please no email.

In the tropics I got zapped by increasing static a few times when disconnecting
the ladderline? when thunder approached.? Connecting a resistor from each leg to earth,
made disconnecting less valorous - and probably saved my ATLAS 210x...
I can't remember the value - it was whatever I had two of in the frugal? junkbox :-)

Thanks for the confirmation, Allison.

Michael 2E0IHW