1. Home
  2. Electronics Design
  3. Backfed MOSFET current limit

Backfed MOSFET current limit

Got a circuit.

I'm using a high-side P-channel MOSFET to prevent damage when the battery is put in backwards. The FET is put in "backwards" (battery to drain, circuit to source) to point the body diode in the correct direction.

When the battery is inserted, a capacitor on the circuit side of the FET charges up, sucking a considerable amount of current very briefly. The current is briefly conducted through the body diode, then when the voltage rises high enough the MOSFET (whose gate is connected to ground) turns on.

I can calculate the number of Joules that this dumps into the MOSFET, and the data sheet has a transient thermal response curve. So I could calculate whether it's all safe -- except that I don't know if those curves would apply when the thing is being back-fed, or during the time that the current is flowing through the body diode junction.

Any one have any mileage with this?

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com
Reply to
Tim Wescott
Loading thread data ...

What difference would it make whether it gets heated up by the channel or the body diode? Something in your circuitry should make sure that it will not draw power until the FET has reached a guaranteed low Rdson level.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

most appnotes talk about recovery time etc. when used for free-wheeling

most datasheets list similar current limits etc. for the body diode and FET and it is the same piece of silicon so I'd expect the specs for the device is similar in both directions

-Lasse

Reply to
Lasse Langwadt Christensen

The PMOS supply reversal circuit is so pretty, though. I use it all the time, with maybe a resistor in the gate when using gate-protected parts.

Usually there's a polyfuse in series somewhere, whose 20 C resistance helps a lot in limiting the inrush.

0-----------* *--*---------------*----0 Vraw | A | | --------- | ------* | | | | CCC R R CCC R | R | | | | | GND GND

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

Depending on expected voltages there should be a zener across GS. Also a bidirectional TVS at the input because it's really hi-Z for negative going pulses. The usual ... carpet .... sneakers ... tzzt ... phut.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

I think to not draw power would need back to back p-fets. Using your logic, nobody would ever build a soft start circuit.

Personally I don't see anything wrong with the plan. Your failure mechanism will be electromigration, and it shouldn't matter as to the source of the current flow.

Reply to
miso

It certainly is possible to blow out the gate if the part isn't gate-protected.

I usually put a few microfarad ceramic cap upstream, which pretty well gets rid of the carpet shock problem. The old rule of never having silicon connected directly to the outside world....

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

Test the body diode to destruction! See how much margin you have.

Just a few minutes ago, I destroyed a 1000 amp, 1000 watt mosfet. I charged a 10,000 uF cap with a little current-limited supply and pulsed the gate drive of the fet to fully discharge the cap in 1 millisecond, once every 10 seconds. Then I cranked up the voltage until it died. It was around 45 volts, 300 amps, 10 joules, about 10x the SOAR rating.

Smaller mosfets might die at a joule or less. I'm guessing that the body diode can eat more joules than the normal mosfet mode. Mosfets used to be bragged about as not having a secondary breakdown failure mode, but now they do.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation
Reply to
John Larkin

I like a unidirectional transzorb after a polyfuse. That handles reverse, overvoltage, and downstream shorts. Mostly.

Incidentally, the surface-mount polyfuses are awful. We always used radial leaded parts.

--

John Larkin         Highland Technology, Inc 

jlarkin at highlandtechnology dot com 
http://www.highlandtechnology.com 

Precision electronic instrumentation 
Picosecond-resolution Digital Delay and Pulse generators 
Custom laser drivers and controllers 
Photonics and fiberoptic TTL data links 
VME thermocouple, LVDT, synchro   acquisition and simulation
Reply to
John Larkin

For stand-alone boxes, I agree entirely. For boards that go into somebody's system, I sort of lean towards the open-circuit (PMOS) vs. the short-circuit (Transzorb) approach to protection, because it makes it easier to debug the system by localizing the problem.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

BTW have you tried the Tyco PolyZen devices? They're little SMT baby boards that use the Transzorb to heat the polyswitch, which is a cute idea. They've been extending the voltage ranges lately.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 USA 
+1 845 480 2058 

hobbs at electrooptical dot net 
http://electrooptical.net
Reply to
Phil Hobbs

I don't see any reason why the diode wouldn't be as robust as the channel -- it has higher current density (minority carriers!) and I think it should have lower voltage drop (parasitic resistance should be much less than Rds(on), which will dominate under transient conditions).

If it's avalanche rated, that would occur in the diode junction as well; the energy rating (for very short transients; note that Larkin's "electrolytic zapper" has a longer time constant than conduction through silicon) will apply, at least for transients of the same shape and duration. The traditional inductive test gives a triangular waveform, a sufficiently crude approximation of a capacitor charging, I suppose.

You could always ugly-ifiy your circuit with a precharge resistor and bypass PMOS (the other direction). :-)

Tim

-- Deep Friar: a very philosophical monk. Website:

formatting link

Reply to
Tim Williams

Only enough to know that manufacturers like International rectifier full spec out transient and steady state capability of the body diode. It will be shown as section labeled source-drain diode. E.g.:

formatting link
Looks like it's junction temperature limited for smallish currents and bond wire limited for parts with really current specs.

Reply to
bloggs.fredbloggs.fred

Huh? It has nothing to do with soft-start.

There needs to be a proper UVLO on the other side. Else an erratic "brown-out style" input behavior can fry the FET.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Not a reference to a contact type switch.

"Soft start" is a power supply design term.

Power supplies can be poorly designed and have a "hard start' issue, where the main oscillator fails to start.

a "soft start" circuit insures that the inrush on start-up does not cause this "latching" effect, and insures that the oscillator will start-up properly.

Reply to
DecadentLinuxUserNumeroUno

I know. The problem I pointed out is not related to soft-start (which every good piece of electronics should have). What I mean is the fact that when you have a brown-out, loose contact or undervoltage situation and the electronics already starts consuming a lot at a few volts it can fry the FET.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Use two FET's back-to-back, as in this previously posted example...

formatting link

For your purposes, ground both gates, and shunt MP2 with a resistor of value appropriate for your surge-limiting needs. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

That's a good long-term solution. At the moment I'm trying to help out someone who gave me a last-minute call: "everything's ready for production, is this going to be a disaster?"

I _think_ that while what he's got isn't optimum, he's still in the ballpark of "working".

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com
Reply to
Tim Wescott

A single P-channel is usually fine for reverse-battery protection. But keep the brown-out scenario in mind, that could really rain on someone's parade. He should make sure that there is sufficient UVLO protection downstream.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

But doesn't limit initial surge.

Just put it under the gates of the P-FET's... minimum voltage or ho pass.

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.