M.O.S.F.E.T. diode clamp

[MNFsoft]

If you look at the vishay datasheet for the M.O.S.F.E.T. some times you would discover a diode clamped on the drain-source semiconductor junction while another time you would discover a zener diode in its place I mean a semiconductor manufacturing company company; how could it be so careless about their datasheets where I have noticed more than once with datasheets!

 

[duke37]

All diodes have a limited voltage. Some may be specified closely and some may have the energy specified which they can absorb.

 

[hevans1944]

There is always a diode (one of two body diodes) between the drain and source of every mosfet. There is another body diode between the source and the substrate (body), but it is normally short-circuited by an internal mosfet connection between the source and the substrate, a connection that is necessary to properly bias the conduction channel with the gate-to-source potential.

So, all that said, every diode has a reverse breakdown characteristic that falls somewhere between zener conduction and avalanche conduction, depending on diode construction. It is not uncommon to represent this fact by showing the body diode as a zener diode, perhaps to emphasize that there is a specified breakdown voltage limit on the drain-to-source voltage, V_DS, but the important thing to realize is that this diode is never intended to operate in a reverse breakdown condition. Most, if not all, datasheets fail to specify the conditions under which it is safe to operate the body diode as a zener diode, although some will specify how much (pulsed) energy it can dissipate without failure, as mentioned by @duke37 in his post #2. Best to avoid pushing device limits if you want long and continued operation from your mosfet.

 

[(*steve*)]

Marking the body diode as a zener may indicate that the breakdown between drain and source is due to the body diode rather than the channel.

 

[MNFsoft]

There is always a diode (one of two body diodes) between the drain and source of every mosfet. There is another body diode between the source and the substrate (body), but it is normally short-circuited by an internal mosfet connection between the source and the substrate, a connection that is necessary to properly bias the conduction channel with the gate-to-source potential.

So, all that said, every diode has a reverse breakdown characteristic that falls somewhere between zener conduction and avalanche conduction, depending on diode construction. It is not uncommon to represent this fact by showing the body diode as a zener diode, perhaps to emphasize that there is a specified breakdown voltage limit on the drain-to-source voltage, V_DS, but the important thing to realize is that this diode is never intended to operate in a reverse breakdown condition. Most, if not all, datasheets fail to specify the conditions under which it is safe to operate the body diode as a zener diode, although some will specify how much (pulsed) energy it can dissipate without failure, as mentioned by @duke37 in his post #2. Best to avoid pushing device limits if you want long and continued operation from your mosfet.

Thx a lot; but I am concerned that if you study the material side (construction) how does this affect its max. values??

 

[hevans1944]

I am concerned that if you study the material side (construction) how does this affect its max. values?

Are you designing mosfet wafer dies? There is a huge body of published information, dating back at least fifty years, that pretty much defines all there is know about how to do that. What is your educational background that would prepare you to "study the material side" and what is your goal in doing so? Do you have a background in solid state physics? Electronics Point is not a teaching forum, but if you ask the right questions someone here might be able to point you in the right direction where you can look for answers.

 

[Harald Kapp]

how could it be so careless about their datasheets

To err is human - even if you are a semiconductor manufacturer. It really looks careless showing two different symbol for the same transistor within a single datasheet. As mentioned before this is not really an issue as one never should exceed the max. ratings and in that case the breakdown properties of the body diode are not relevant.

Any information in a datasheet should be taken with a grain of salt and checked thoroughly if relevant for your application. I've seen a totally wrong application circuit being propagated through the datasheets of the LM723 voltage regulator from many manufacturers. They obviously copied the same circuit without verifying the correct operation (instead of current limiting the regulator blew up). This circuit is no longer shown in the current datasheets.

 

[duke37]

To err is human, to forgive divine.

 

[hevans1944]

In God we Trust, all others pays Cash.

 

[Audioguru]

Gary in this forum or on another forum is using an N-channel Mosfet to drive a car ignition coil to electrify a fence to keep (kill?) farm animals from running away. Most circuits in Google show an added diode to prevent the drain of the Mosfet from going negative and nothing to prevent the drain from going to a high positive voltage.
1) When the Mosfet turns on causing its drain to apply current to the coil, does the coil transformer stepup the 12V to produce maybe only 2kV?
2) When the Mosfet turns off then doesn't its reactance cause the drain of the Mosfet to go to +60V or more and the transformer produces 10kV or more?
3) Does the drain voltage ever go negative? If it did then wouldn't the Mosfet body diode clamp it?
4) What prevents the positive drain voltage from going too high? Wouldn't the Mosfet channel be zapped since the Zener voltage of the body diode is not specified?
5) If a reverse biased diode is parallel with the coil like a relay coil then the drain is clamped from going higher than about 13V and the high voltage output is reduced. Is a 60V Zener diode to ground used instead to prevent the drain from being zapped?

 

[duke37]

The electric fence transformer will have ratio of about 20 so 12V in will give about 200V on switch on.
On switch off, the energy will be fed to the fence which will have a considerable capacitance, this will determine the fence voltage (about 5kV).and about 250V on the primary. Fencers are often specified on the length of fence they can supply.
The energy will be about 1J (within a factor of 10) which has to be dealt with about once a second, 1W. I have not looked up the zener specfications to see if a fet can absorb this with no fence connected.