Low Rds(on) tempco
in [Design]
|
Prev: Novel? Guitar amp input stage, single 12AX7, footswitchableusing "folded" feedback loop.
Next: ucontrollers (AVR/ATMega) and radio transceiver modules
From: legg on 24 Jan 2010 20:32 On Sun, 24 Jan 2010 12:23:13 -0800, John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >On Sun, 24 Jan 2010 13:01:28 -0600, "Tim Williams" ><tmoranwms(a)charter.net> wrote: > >>A little while ago, someone was asking if anyone makes MOSFETs with a low >>Rds(on) tempco. Fairchild is claiming an usually low ratio of ~1.63 = >>Rds(on)(175C) / Rds(on)(25C) for some low voltage FETs: >>http://www.fairchildsemi.com/ds/FD%2FFDP8860.pdf >> >>But they're also claiming 254W dissipation in a TO-220. Tempco of just the >>leads, in air, at rated current, is more than Fig.3 I'm pretty sure. >> >>Tim > >Insane. 254 watts times 0.59 K/w is a 150K rise, and that's just the >silicon to the case. The DC SOAR curve sure looks like 160 watts max >to me. > Fig 11 shows 1ADC @30V limit, with Tc @25degC, so obviously something is not quite up to snuff, where data sheet accuracy is concerned here. RL
From: Hammy on 24 Jan 2010 21:07 On Sun, 24 Jan 2010 15:22:48 -0600, "Tim Williams" <tmoranwms(a)charter.net> wrote: >"Hammy" <spam(a)spam.com> wrote in message >news:bjbpl51m0fv2p63c8dnt6v5v1sbd5cv4lh(a)4ax.com... >> Even Lower from TI their new CICLON line. >> http://focus.ti.com/lit/ds/symlink/csd25301w1015.pdf > >That is quite flat. > >'Spose it helps not having leads, too. Average metal has an awful tempco. > >> They sent me an email about them. They are nice for LV only 25VDS max; >> easy drive requirements as well. > >I spotted this phenomenon while exploring a low volt, high amp synchronous >supply. I'm looking at things a bit harder to drive though (~200nC gate >charge). :-) Yea that Fairchild device is quite the brute. Have you seen the IRL3713S same ratings only half the gate charge at the same price it's a DPAK. http://canada.newark.com/international-rectifier/irl3713spbf/mosfet/dp/63J7773 NXP has some cheap LFPAKS with similar ratings as well; with a quarter the gate charge http://canada.newark.com/nxp/psmn4r0-30yl/transistor/dp/04R7692 The biggest FET I have gate charge wise is the FQA24N50, TO-3P, 500VDS,24A,0.20ohm RDSON, 90nC. Newark had them on sale for $1.14 each so I grabbed 25 of them. http://canada.newark.com/fairchild-semiconductor/fqa24n50/mosfet/dp/58K1512 Sub 0.25 ohm RDSON fets with 500 or 600VDS blocking capability usually cost an arm and a leg. Infineons IPB60R099CP is a prime example $14. http://canada.newark.com/infineon/ipb60r099cp/mosfet-n-to-263/dp/33P7131 >> It's mainly the lower voltage ones that have the best tempco's. When >> you get up in higher voltage they are all usually about 1.8 to 2 x >> 25C. I'm sure I used to now why this is but forgot. > >I would guess channel length. Something about short channel effects, or the >way they have to build them when Vds(max) > Vgs(max)? > >I wonder, does anyone know what the typical feature size is for power >MOSFETs? Still micrometers as usual? Anything gained from teensy sizes? > >Reminds me of an idle thought I had the other day... just imagine, if you >took all the transistors on one of those big pigfucker FPGAs and put them >all in parallel. You'd only get maybe 1.5Vds(max), but think of the >amperage and switching speed. ;-) > >Tim
From: John Larkin on 24 Jan 2010 21:19
On Sun, 24 Jan 2010 20:32:41 -0500, legg <legg(a)nospam.magma.ca> wrote: >On Sun, 24 Jan 2010 12:23:13 -0800, John Larkin ><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > >>On Sun, 24 Jan 2010 13:01:28 -0600, "Tim Williams" >><tmoranwms(a)charter.net> wrote: >> >>>A little while ago, someone was asking if anyone makes MOSFETs with a low >>>Rds(on) tempco. Fairchild is claiming an usually low ratio of ~1.63 = >>>Rds(on)(175C) / Rds(on)(25C) for some low voltage FETs: >>>http://www.fairchildsemi.com/ds/FD%2FFDP8860.pdf >>> >>>But they're also claiming 254W dissipation in a TO-220. Tempco of just the >>>leads, in air, at rated current, is more than Fig.3 I'm pretty sure. >>> >>>Tim >> >>Insane. 254 watts times 0.59 K/w is a 150K rise, and that's just the >>silicon to the case. The DC SOAR curve sure looks like 160 watts max >>to me. >> >Fig 11 shows 1ADC @30V limit, with Tc @25degC, so obviously something >is not quite up to snuff, where data sheet accuracy is concerned here. > >RL Switching-type fets are happiests at low drain voltages, and tend to detonate if they dissipate a lot of power at higher voltages. The corner of the DC SOAR curve is about 160 watts (2 volts, 80 amps) but you're right, that tapers off to 30 watts on the other end. So it's not a constant-power curve. John |