MOSFET musing

[bodgy]

I'm thinking of using a N-channel Mosfet as a free wheeling diode across a small motor or pump that is itself switched by a Mosfet in single quadrant formation.<p>Now of course I can just connect gate to source and then Drain to +V, Source to Drain of switching FET, but I get the feeling I will need to delay the turning on of the switching FET otherwise if the switch switches on at free wheel time I'm going so send an almighty amount juice through the top FET and bye bye FET.<p>Has anyone used this formation and had any problems?<p>I'm only toying with this because 1) you don't see this concept in hobby equipment all that often and 2) I can't be bothered to drive 20 miles to pick up a fast 10A Shottky at exorbitant prices.<p>Colin

[Chris Smith]

It sounds like what we call a chopper circuit, or "light dimmer" circuit with a Diac? Is that what your looking for? The gate in the light dimmer is set to a varaible delay but there are other mods you can add.

[russlk]

The inductive spike from the motor will occur just as the drive FET turns off. If you delay the turn on of the clamp FET, the spike is already gone. If you don't delay the clamp FET, both may be on at the same time causing heat and possibly smoke. I don't see any alternative to the schotkey diode.

[bodgy]

Looks like I'll have to go driving.<p>I remember seeing in the dim distant past an early application note circa 1967 that used a Mosfet in this fashion along with using a P-Channel with a N one in place of rectifiers and triacs. I have a circuit where one is used in this fashion, but it's primary use is as a motor brake and the flyback protection is one of those spin offs, used in upmarket golf carts and robots.<p>however I don't want the added complexity that this particular circuit uses.<p>I suppose I could always kill a few fets in the interests of science!<p>I would have used a Pro-Fet or Topfet, but they are very expensive out here, and I'm not sure how easy they are to get in the US.<p>Thanks for your help.<p>Colin

[Bob Scott]

bodgy,<p>I don't quite understand your hookup information ie:<p>"Now of course I can just connect gate to source and then Drain to +V, Source to Drain of switching FET, but I get the feeling I will need to delay the turning on of the switching FET otherwise if the switch switches on at free wheel time I'm going so send an almighty amount juice through the top FET and bye bye FET."<p>Is this a power MOSFET? If you connect the gate to the source, the FET part will always be off but the intrinsic internal substrate diode will catch the flyback energy just like a Schottky.<p>The only other way I've seen to use a FET with the gate shorted to the source is with a depletion mode JFET. It then becomes a constant current source ~20mA.<p>This is quite a coincidence. I have been working on a simple circuit to use MOSFETs as super low voltage drop rectifiers. I have been pondering the synchronization of the gate for years and have just found the answer. I plan to get the schematic notarised this week and apply for a patent (It will be my second patent).<p>Have you seen anything published about using MOSFETs as low loss rectifiers?<p>Bob <p>

[bodgy]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by Bob Scott:
bodgy,<p>
Have you seen anything published about using MOSFETs as low loss rectifiers?<p>Bob <p> <hr></blockquote><p>Bob,<p>I'm all fetted out and can't remember half the sites I've been to.<p>I think I've seen references to rectifiers, but I think they were as IC's rather than discretes.<p>I'd suggest searching at the IRF site and then look at Vishay and the like.<p>I found ST to be the worst site for information. There is an engineers info site, which I know I visited yesterday, but it isn't in my history so I can't give you the URL.<p>Colin<p>[ May 29, 2003: Message edited by: bodgy ]</p>

[rshayes]

FETs can be used as high efficiency rectifiers. It is usually referred to as synchronous rectification. With normal output voltages and currents, it usually isn't worth the trouble. The exception is low voltages, such as the 3.3 or lower supply used by some microprocessors, at high currents.<p>The drawbacks are that the gates of the FETs have to be driven with accurately timed drive signals, and that low resistance FETs are usually needed. A .05 ohm FET carrying 10 amps will drop .5 volts, at which point you might as well use a schottky rectifier. If you really need the efficiency, the cost of large FETs and the drive circuits may be worth it.<p>Unitrode gave design seminars on switching power supply design for about ten years before they were sold to Texas Instruments. The seminar notes are available on the Texas Instruments web site. Other manufacturers may also have ap notes on switching supply design that would mention synchronous rectifiers.

[L. Daniel Rosa]

The MOSFET has a reverse doide already in it (perhaps an artifact of it's construction). Using one as a free wheeling diode in a DC chopper circuit would defeat the purpose. If you could get one without the source to drain diode, then that would work, but you'd have to switch it externally and run it in parallel with a standard diode because your timing won't be perfect. Even so, the efficiency would improve.