Post by thread:Power Mosfet

Ok I just went through my workbench and found this power mosfet from radio shack..... its a IFR510 N-channel 60 volt here is my question: are you ready:???? what the heck is this thing for? hehe dont yell at me please :) I realy dont know what its for.... is it some type of electronic switch? I have never used a Mosfet of any kind before,.... I havent used much of anything to tell you the truth. what would it be usefull for? sorry for the dumb question... just bored I guess.... regards, Ryan

Ryan Pogge wrote: > > Ok I just went through my workbench and found this power mosfet from radio > shack..... > its a IFR510 > N-channel 60 volt > > here is my question: are you ready:???? > > what the heck is this thing for? <snip> An IRF510 is a switch. The gate is the control, you charge it up (like a capacitor) to about +10 volts compared to the source, and the switch is on. On, it's resistance is about .6 Ohms. The source must be more negative than the drain, as the mosfet simply conducts, switched on or not, when hooked up the other way. Discharge the gate (short it to the source) to switch it off again. Thare's more, but you don't sound too terribly interested anyway, so I won't bother you with it. You can try experimenting with it, and learn more that way. International Rectifier also sell some MOSFET's that are fully on at 5 volts. Hmmmmm. Russell Hedges

Hi all, At 06:29 PM 8/25/98 -0700, you wrote: >on. On, it's resistance is about .6 Ohms. The source must be more >negative than the drain, as the mosfet simply conducts, switched on or >not, when hooked up the other way. I might be wrong, but I don't think that this is exactly correct. AFAIK, all fets are pretty symetrical when it comes to the source and drain. So, as long as you charge the gate more positive than EITHER drain or source by the VgsON threshold, it will turn on, and it doesn't matter which is more positive, drain or source. Usually, the actual source is kept more negative because the manufacturer optimizes the gate-source characteristics more than gate-drain characteristics (i.e., gate-source capacitance is kept lower in smal signal fets, I don't know if the same is true in power mosfets). Sean +--------------------------------+ | Sean Breheny | | Amateur Radio Callsign: KA3YXM | | Electrical Engineering Student | +--------------------------------+ Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 spam_OUTshb7TakeThisOuTcornell.edu Phone(USA): (607) 253-0315

Ryan Pogge wrote: > its a IFR510 > what the heck is this thing for? Russell gave a good answer, Ryan, but I'll add a bit from my perspective ... I've only just learned about these things in the past few months. They are sort of like a transistor, in that they switch current on and off, but they are most often used for solid on-off, sort of like how a relay works, rather than partial on-off like a transistor. How useful is it to you? Think of any case where you want to use a relay to switch DC, and try out the MOSFET in it's place. Quieter, more reliable, no moving parts, and much easier to destroy. -- James Cameron (.....james.cameronKILLspam@spam@digital.com) Digital Equipment Corporation (Australia) Pty. Ltd. A.C.N. 000 446 800

Sean Breheny wrote: > > Hi all, > > At 06:29 PM 8/25/98 -0700, you wrote: > >on. On, it's resistance is about .6 Ohms. The source must be more > >negative than the drain, as the mosfet simply conducts, switched on or > >not, when hooked up the other way. > > I might be wrong, but I don't think that this is exactly correct. AFAIK, > all fets are pretty symetrical when it comes to the source and drain. These MOSFET's have an inhearent reverse diode in them, so that they conduct when connected in reverse. As far as I know, all the big ones do. Sometimes the schematic symbol shows this diode, sometimes not. Keep in mind, though, that a MOSFET when on is a resistance, and not a junction drop. In this case, the on resistence is .6 OHMS, a number not at all related to a .6 VOLT junction drop. Russell Hedges

James Cameron wrote: > > Ryan Pogge wrote: > > its a IFR510 > > what the heck is this thing for? > > Russell gave a good answer, Ryan, but I'll add a bit from my perspective > ... I've only just learned about these things in the past few months. Aww shucks! Your turning my head. <snip> > How useful is it to you? Think of any case where you want to use a > relay to switch DC, and try out the MOSFET in it's place. Quieter, more > reliable, no moving parts, and much easier to destroy. > Not to mention, they take almost no drive at all. I have sorted batches of hundreds of MOSFET's by gate capacitance, big 14 Amp 400 Volt things, and the most gate capacitance I saw was about 35 pF. That's thirty five pico Farads. And they are usually cheaper than a relay. Russell Hedges

On Tue, 25 Aug 1998 22:03:00 -0400 Sean Breheny <shb7KILLspamCORNELL.EDU> writes: >Hi all, > >At 06:29 PM 8/25/98 -0700, you wrote: >>on. On, it's resistance is about .6 Ohms. The source must be more >>negative than the drain, as the mosfet simply conducts, switched on >or >>not, when hooked up the other way. > >I might be wrong, but I don't think that this is exactly correct. >AFAIK, >all fets are pretty symetrical when it comes to the source and drain. Power MOSFETs have a diode inherent in their structure so they will conduct if the drain is reverse-biased from the source. The gate-source voltage determines the amount of conduction, the gate-drain voltage has rather little effect. They are made with the gate and source close to each other on the top of the chip; the drain is the substrate of the chip. Should you need to operate the drain negative with respect to the source, P-channel power MOSFETs are made. But, there are optocouplers with a FET output that will switch signals of either polarity. Apparently the FET in these is more symmetrical, or at least it doesn't have the diode. I haven't seen just the FET part offered as a discrete device. >optimizes the gate-source characteristics >more >than gate-drain characteristics (i.e., gate-source capacitance is kept >lower in smal signal fets, I don't know if the same is true in power >mosfets). In a power FET, the gate-source capacitance is much more than the gate-drain capacitance, but the latter usually is more of a problem in switching circuits due to the Miller effect. _____________________________________________________________________ You don't need to buy Internet access to use free Internet e-mail. Get completely free e-mail from Juno at http://www.juno.com Or call Juno at (800) 654-JUNO [654-5866]

Hi Sean, You are thinking JFETs, not MOSFETs. -Bob At 10:03 PM 8/25/98 -0400, you wrote: >I might be wrong, but I don't think that this is exactly correct. AFAIK, >all fets are pretty symetrical when it comes to the source and drain. So, >as long as you charge the gate more positive than EITHER drain or source by >the VgsON threshold, it will turn on, and it doesn't matter which is more >positive, drain or source. Usually, the actual source is kept more negative >because the manufacturer optimizes the gate-source characteristics more >than gate-drain characteristics (i.e., gate-source capacitance is kept >lower in smal signal fets, I don't know if the same is true in power mosfets). > >Sean

... But, there are optocouplers with a FET >output that will switch signals of either polarity. Apparently the FET >in these is more symmetrical, or at least it doesn't have the diode. I >haven't seen just the FET part offered as a discrete device. Actually, these optocouplers generally have an N Power Fet and a P Power FET in series. Each has an intrinsic source/drain diode, but since they are pointing o pposite directions, no current can flow unless you turn on one or both of the F ETs. These devices are also made with IGBTs, in which case the intrinsic diodes are n ot present, they can handle much more current, but on/off time is slower. Ron Fial

Hi all, I've been toying arround with power mosfets for some time now,hanling some high voltage switching (300 volts,maybe 1 or two amps).N-channel mosfets require at least gate voltage to be 10-12 volts above source to get in full conduction,which means that if your high voltage rail is 300 volts,and you have two mosfets in a half bridge configuration,the upper mosfet will require 300 + 12 = 312 volts for turning on.There are drivers from IR to achieve this,but I'd like to do it without them. The approach is usually making the source voltage float to the source level, using a bootstrap diode and cappacitor to provide the overvoltage when the fet is to be turned on.This has the inherent limitation of not being able to keep the mosfet in conduction for a period longer than the charge of the cappacitor allows it to.Then, the lower mosfet must put the upper one's source to ground to allow the cappacitor to recharge from a 15 volt power source through the bootstrap diode. Please excuse the roughness of the exposition,but this is mainly the result of months of burning things on the protoboard with very little theoretical background. Can any of you show me how you would build the circuit to drive the upper side mosfet,with the bootstrap diode and cappacitor?

ahhhh cool man... so this thing is good for driveing DC motors and such...????? someone said somtin about an audio amp too.... so it would be just like a transistor based audio amp? I might try that. But what is the advantage of useing this for an audio amp versus a transistor. just more watts???? {Quote hidden}>Ryan Pogge wrote: >> its a IFR510 >> what the heck is this thing for? > >Russell gave a good answer, Ryan, but I'll add a bit from my perspective >... I've only just learned about these things in the past few months. > >They are sort of like a transistor, in that they switch current on and >off, but they are most often used for solid on-off, sort of like how a >relay works, rather than partial on-off like a transistor. > >How useful is it to you? Think of any case where you want to use a >relay to switch DC, and try out the MOSFET in it's place. Quieter, more >reliable, no moving parts, and much easier to destroy. > >-- >James Cameron (.....james.cameronKILLspam.....digital.com) >Digital Equipment Corporation (Australia) Pty. Ltd. A.C.N. 000 446 800 >

Regarding the comments on MOSFETs...Can anyone tell me if there is a MOSFET that can control say 100 Amps? I would like to replace a mechanical battery switch with a MOSFET one. For this I would like a minimum of on-resistance, say less than 0.1 ohms if possible. Of course, I realize that careful attention will have to be paid to the thermal design aspect. Does any such device exist? If so, please point me to its data sheet. TIA Jim Manzari

Hi all, Thanks for the correction. I guess I should read the WHOLE section on FETs in "The ART of Electronics" before engaging my mouth on this subject. Seriously, though, for my own edification, if I took a MOSFET that lacked a protection diode, wouldn't it be able to work backwards, albeit with different characteristics? Thanks, Sean At 11:35 PM 8/25/98 -0700, you wrote: {Quote hidden}>Hi Sean, > >You are thinking JFETs, not MOSFETs. > >-Bob > >At 10:03 PM 8/25/98 -0400, you wrote: >>I might be wrong, but I don't think that this is exactly correct. AFAIK, >>all fets are pretty symetrical when it comes to the source and drain. So, >>as long as you charge the gate more positive than EITHER drain or source by >>the VgsON threshold, it will turn on, and it doesn't matter which is more >>positive, drain or source. Usually, the actual source is kept more negative >>because the manufacturer optimizes the gate-source characteristics more >>than gate-drain characteristics (i.e., gate-source capacitance is kept >>lower in smal signal fets, I don't know if the same is true in power >mosfets). >> >>Sean > +--------------------------------+ | Sean Breheny | | Amateur Radio Callsign: KA3YXM | | Electrical Engineering Student | +--------------------------------+ Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 EraseMEshb7spam_OUTTakeThisOuTcornell.edu Phone(USA): (607) 253-0315

At 13:31 26.08.98 +0200, you wrote: >Regarding the comments on MOSFETs...Can anyone tell me if there is a >MOSFET that can control say 100 Amps? I would like to replace a Hi, instead of using MOSFETs, have a look at SCRs or IGBTs. Thomas -- ********************************************************** * Thomas Magin FON: ++49-761-4543-489 * * marquette-Hellige GmbH FAX: -507 * * E-FW / Emergency Systems email: maginspam_OUThellige.de * * Munzinger Str. 3 * * D-79111 Freiburg / Germany * **********************************************************

(given you don't want to use a chip designed for this purpose) How about an isolated power supply and a optical coupler. Mike {Quote hidden}> ---------- > From: Walter Markiw[SMTP:@spam@markiwKILLspamOVERNET.COM.AR] > Sent: Wednesday, August 26, 1998 7:40 AM > To: KILLspamPICLISTKILLspamMITVMA.MIT.EDU > Subject: Re: Power Mosfet > > Hi all, > I've been toying arround with power mosfets for some time > now,hanling > some high > voltage switching (300 volts,maybe 1 or two amps).N-channel mosfets > require > at least > gate voltage to be 10-12 volts above source to get in full > conduction,which > means that > if your high voltage rail is 300 volts,and you have two mosfets in a half > bridge > configuration,the upper mosfet will require 300 + 12 = 312 volts for > turning > on.There > are drivers from IR to achieve this,but I'd like to do it without them. > The approach is usually making the source voltage float to the > source > level, > using a bootstrap diode and cappacitor to provide the overvoltage when the > fet is to be > turned on.This has the inherent limitation of not being able to keep the > mosfet in conduction for a period longer than the charge of the cappacitor > allows it to.Then, > the lower mosfet must put the upper one's source to ground to allow the > cappacitor > to recharge from a 15 volt power source through the bootstrap diode. > Please excuse the roughness of the exposition,but this is mainly > the > result of months of burning things on the protoboard with very little > theoretical background. > Can any of you show me how you would build the circuit to drive the > upper side mosfet,with the bootstrap diode and cappacitor? >

> Regarding the comments on MOSFETs...Can anyone tell me if there is a > MOSFET that can control say 100 Amps? I would like to replace a > mechanical battery switch with a MOSFET one. For this I would like a > minimum of on-resistance, say less than 0.1 ohms if possible. Of > course, I realize that careful attention will have to be paid to the > thermal design aspect. Does any such device exist? If so, please > point me to its data sheet. > > TIA > > Jim Manzari > I think you'd want *much* less than 0.1 ohms P=I^2R =100*100*0.1 = 1000 Watts!!!! Thats a lot of heat to get rid of. Would this be 100 Amps continuous or pulsed? If the average current is lower, there are many devices readily available that can take 25 Amps or more peak current. Bearing in mind that the gate takes no current (unless changing state) you could by a load of these and parallel them up. This is the approach taken by the PWM motor controllers used for Radio controlled models. Mike Rigby-Jones RemoveMEmrjonesTakeThisOuTnortel.co.uk

Jim Manzari wrote: >Regarding the comments on MOSFETs...Can anyone tell me if there is a >MOSFET that can control say 100 Amps? I would like to replace a >mechanical battery switch with a MOSFET one. For this I would like a >minimum of on-resistance, say less than 0.1 ohms if possible. Of >course, I realize that careful attention will have to be paid to the >thermal design aspect. Does any such device exist? If so, please >point me to its data sheet. > >TIA > >Jim Manzari Yes, although I do not have any references near by. These typically are modules which mount several MOSFET die onto a metal back plate and then provide a sealed package. I have typically just use several parallel MOSFETs separately strapped to a heat sink. When doing this use separate gate drive resistors and do not attempt to use the parallel combination in a linear mode -- it will fail at a much lower than expected current as current is not shared well unless the parts are fully on. One project I did had an H-bridge which switch currents on the order of 400Amps at 24VDC at around 24KHz PWM. BTW it is very easy to release the factory installed smoke from these devices. Make sure the heat sink is adequate and make sure that the source to drain voltage does not get exceeded when you open the circuit. Typically this requires good bypass on the power supply and possibly external catch diodes (the parasitic MOSFET reverse diodes do fairly well in most cases) and typically double the source to drain voltage over the voltage you are switching. Chip Weller

At 20:41 25.8.1998 -0400, you wrote: >Ok I just went through my workbench and found this power mosfet from radio >shack..... >its a IFR510 >N-channel 60 volt > >here is my question: are you ready:???? > >what the heck is this thing for? >hehe dont yell at me please :) I realy dont know what its for.... is it some >type of electronic switch? :-) If you will find one more IRF510 and two more IRF9510 (P channel MOSFETS), you can build a nice H-bridge and drive motor with your PICs. Seriously, power MOSFETS are like switches, but you need to solve some specific problems when you want to use them. The whole problematic is probably a bit long for e-mail answer, but you can find very good explanation of these components in "Art of Electronic" or similar book. Also, International Rectifier (company which manufacture IRF510 and hundreds of other similar MOSFETS) have some PDF sheets about MOSFETS on their web page. PavelK ************************************************************************** * Pavel KorenskyÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * * DATOR3 LAN Services spol. s r.o.ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * * Styblova 13, 140 00, Prague 4, Czech Republic ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * *ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * * PGP Key fingerprint:Ê F3 E1 AE BC 34 18 CB A6Ê CC D0 DA 9E 79 03 41 D4 * *ÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * * SUMMA SCIENTIA - NIHIL SCIREÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊÊ * **************************************************************************

Rigby-Jones, Michael [PAI01:4837:EXCH] wrote: > > I think you'd want *much* less than 0.1 ohms > > P=I^2R =100*100*0.1 = 1000 Watts!!!! > > Thats a lot of heat to get rid of. Would this be 100 Amps continuous or Michael, How right you are! I meant to say 0.01 ohms. Now that people have directed me to some data sheets I see that 0.008 ohms is possible. Still a good deal of heat at 100A, but at least manageable. In actual practice I would expect continuous current to be less than 100A since the batteries can not be charged at greater than about C/10 to C/20. Thanks for pointing out my mistake. Regards, Jim Manzari

On Wed, 26 Aug 1998 13:31:05 +0200, you wrote: >Regarding the comments on MOSFETs...Can anyone tell me if there is a >MOSFET that can control say 100 Amps? I would like to replace a >mechanical battery switch with a MOSFET one. For this I would like a >minimum of on-resistance, say less than 0.1 ohms if possible. Of >course, I realize that careful attention will have to be paid to the >thermal design aspect. Does any such device exist? If so, please >point me to its data sheet. > >TIA > >Jim Manzari MOSFETs can be paralleled quite easily, so a few smaller ( much cheaper) devices could easily do it. You can get To-220 devices rated at over 50A, although in practice packaging and connection issues mean you wouldn't be able to get this performance in practice. Many of the big MOSFET makers (IR, MOT, ST, Semikron) do large module type devices with screw-terminal connections, with ratings to 200A but they are expensive. e.g. ST STE25N06, 250A, 60V, 0.004R Ron, UKP37 (US$60) 1-off you can get a lot of TO-220 or TO-3P devices for that money! ____ ____ _/ L_/ Mike Harrison / White Wing Logic / spamBeGonewwlspamBeGonenetcomuk.co.uk _/ L_/ _/ W_/ Hardware & Software design / PCB Design / Consultancy _/ W_/ /_W_/ Industrial / Computer Peripherals / Hazardous Area /_W_/

On Wed, 26 Aug 1998 13:31:05 +0200 Jim Manzari <TakeThisOuTmanzariEraseMEspam_OUTBLUEWIN.CH> writes: >Regarding the comments on MOSFETs...Can anyone tell me if there is a >MOSFET that can control say 100 Amps? I would like to replace a >mechanical battery switch with a MOSFET one. For this I would like a >minimum of on-resistance, say less than 0.1 ohms if possible. You need a much lower on-resistance than that. At 0.1 ohms, the FET would drop 10 V and dissipate 1000 W. There may not be a simgle FET with the ratings you need, but they are easy to combine in parallel. Five IRFZ40's at 0.018 ohms each may be enough, the combination would drop 0.36 V and dissipate 36 W (7.2 W per FET). A FET switch is not exactly the same as a relay since it will conduct through the diodes in the opposite direction. _____________________________________________________________________ You don't need to buy Internet access to use free Internet e-mail. Get completely free e-mail from Juno at http://www.juno.com Or call Juno at (800) 654-JUNO [654-5866]

At 17:27 26.08.98 +0200, you wrote: >Rigby-Jones, Michael [PAI01:4837:EXCH] wrote: >> >> I think you'd want *much* less than 0.1 ohms >> >> P=I^2R =100*100*0.1 = 1000 Watts!!!! >> >> Thats a lot of heat to get rid of. Would this be 100 Amps continuous or > >Michael, > >How right you are! I meant to say 0.01 ohms. Now that people have >directed me to some data sheets I see that 0.008 ohms is possible. Hi, IRF: IRFP064N, 98A DC, 390A pulsed, 0.008R Toshiba: 2SK2987, 70A DC, 280A pulsed, 0.004R (!) So long Thomas =8-) -- ********************************************************** * Thomas Magin FON: ++49-761-4543-489 * * marquette-Hellige GmbH FAX: -507 * * E-FW / Emergency Systems email: RemoveMEmaginTakeThisOuThellige.de * * Munzinger Str. 3 * * D-79111 Freiburg / Germany * **********************************************************

If you want to use the sort of circuit you have just described to provide high side drive you MUST occasionally effectively take the hi-side FETS gate drive power supply capacitor down to ground so that it can be charged (as you described). There are other schemes such as using a transformer and an oscillator to provide the high side drive or providing the drive directly via a pulse transformer. All of these incorporate an element of black magic. The very large and rather rapid voltage swings as the upper FET/IGBT switches tends to produce large dv/dt' in your driver circuit which can lead to false and/or no triggering etc. Stray capacitance (and any other stray effects you can possibly think may exist :-)) need to be accounted for. Try doing a web search for terms such as "hi side drive*" or "h bridge". There are many ICs available and if you are only building one off's it is probably much easier to use one. A major advantage of typical hi-side drivers is the provision of 'shoot through" protection (prevents both FETs being on at once, lock out if the hi-side power supply cap's voltage drops too low and lockout if the bottom driver supply voltage is too low. There is also often over-current sensing and possibly other protection. Doing all this yourself is possible but distracts you from the main application. If you want to build an oscillator type high side power supply the ubiquitous 555 timer is generally adequate to the task driving a small toroid with suitable turns. remember that the toroid needs to provide high side voltage isolation - an insulated core is advisable if you are using hundreds of volts. I haven't used a 555 for this specific application but there have been functionally identical units in hobby mags - some other PIC-listers may be able to provide actual values. {Original Message removed}

... >or providing the drive directly via a pulse transformer. All of these >incorporate an element of black magic. The very large and rather >rapid voltage swings as the upper FET/IGBT switches tends to produce >large dv/dt' in your driver circuit which can lead to false and/or no >triggering etc. Stray capacitance (and any other stray effects you >can possibly think may exist :-)) need to be accounted for. Try doing >a web search for terms such as "hi side drive*" or "h bridge". ... There is no black magic in using pulse transformer. We have one of the HP switched power supply in our lab and the drive in done on a half-bridge using separate pulse transformers. The design is very simple, and works perfectly ( pulse transformer + resistor + zener diode not to fry the gate ) I you are interested, please tell me, but only if it is for personnal design ( HP would kill me/us else :-) ------------------------------------------------------ N'oubliez jamais que nous voyons plus facilement la / paille dans l'oeil du voisin que la poutre dans le / notre. / --------------- / MERCY Jean-marc Maitre de conference / University Montpellier 2 / Sciences et techniques du languedoc / Place Eugene Bataillon / 34095 Montpellier cedex 5 / FRANCE / Tel : 33-4-67-14-48-20 / Fax : 33-4-67-14-37-60 / Email : mercyEraseME.....ges.univ-montp2.fr / ------------------------------------------------------