Electronics-Related.com
Forums

"Jellybean" HV low power MOSFET

Started by Ralph Barone February 26, 2015
Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole
MOSFET, preferably one that's designed more for analog use than for sub-ns
switching. Thanks.  I'm looking to squeeze a cathode follower into an
existing vacuum tube amp and adding one MOSFET is easier than adding half a
12AT7.
Ralph Barone <address_is@invalid.invalid> Wrote in message:
> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole > MOSFET, preferably one that's designed more for analog use than for sub-ns > switching. Thanks. I'm looking to squeeze a cathode follower into an > existing vacuum tube amp and adding one MOSFET is easier than adding half a > 12AT7. >
BUZ74 --
On Thursday, February 26, 2015 at 12:55:36 AM UTC-5, Ralph Barone wrote:
> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole > MOSFET, preferably one that's designed more for analog use than for sub-ns > switching. Thanks. I'm looking to squeeze a cathode follower into an > existing vacuum tube amp and adding one MOSFET is easier than adding half a > 12AT7.
IRF820? (in my parts box) Digikey has a nice search engine. George H.
On 26/02/2015 16:55, Ralph Barone wrote:
> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole > MOSFET, preferably one that's designed more for analog use than for sub-ns > switching. Thanks. I'm looking to squeeze a cathode follower into an > existing vacuum tube amp and adding one MOSFET is easier than adding half a > 12AT7. >
Be a little careful if you are running a MOSFET with high VDS and significant power dissipation (some drain current) at the same time. Modern MOSFETs often have a positive coefficient of drain current with temperature, for a given gate-source voltage. Even different parts of the same die have that property, so that if one side of the chip gets hotter, it will dissipate a greater fraction of the power than the cooler side, and so it will get hotter still. If things work out badly, a hot spot can form leading to failure. Some MOSFETs are specifically characterised for linear operation at high voltage, and have DC safe operating area graphs. For example I am using one of these: http://ixdev.ixys.com/DataSheet/DS100086%28IXTN60N50L2%29.pdf Perhaps they also make something smaller that would suit your application. Perhaps you are even running the device at low enough power that an ordinary switching MOSFET would be fine. Chris
George Herold <gherold@teachspin.com> wrote:
> On Thursday, February 26, 2015 at 12:55:36 AM UTC-5, Ralph Barone wrote: >> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole >> MOSFET, preferably one that's designed more for analog use than for sub-ns >> switching. Thanks. I'm looking to squeeze a cathode follower into an >> existing vacuum tube amp and adding one MOSFET is easier than adding half a >> 12AT7. > > IRF820? (in my parts box) Digikey has a nice search engine. > > George H.
Thanks (and thanks to bitrex for the BUZ74 recommendation). Your mention of parts box reminded me to look in mine where I found an IRFD110, which didn't fit the bill, but helped me find IRFD310, which also does. Now back to design before I check price and availability.
Chris Jones <lugnut808@spam.yahoo.com> wrote:
> On 26/02/2015 16:55, Ralph Barone wrote: >> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole >> MOSFET, preferably one that's designed more for analog use than for sub-ns >> switching. Thanks. I'm looking to squeeze a cathode follower into an >> existing vacuum tube amp and adding one MOSFET is easier than adding half a >> 12AT7. >> > > Be a little careful if you are running a MOSFET with high VDS and > significant power dissipation (some drain current) at the same time. > > Modern MOSFETs often have a positive coefficient of drain current with > temperature, for a given gate-source voltage. Even different parts of the > same die have that property, so that if one side of the chip gets hotter, > it will dissipate a greater fraction of the power than the cooler side, > and so it will get hotter still. If things work out badly, a hot spot can > form leading to failure. Some MOSFETs are specifically characterised for > linear operation at high voltage, and have DC safe operating area graphs. > For example I am using one of these: > http://ixdev.ixys.com/DataSheet/DS100086%28IXTN60N50L2%29.pdf > Perhaps they also make something smaller that would suit your > application. Perhaps you are even running the device at low enough power > that an ordinary switching MOSFET would be fine. > > Chris
I haven't done all the calcs (hell, I'm not even sure WHAT I'm going to do yet), but I suspect drain current will be under 2mA with a supply voltage of 250 V, so a device rating of a couple Watts should be ballpark OK.
Ralph Barone wrote...
> >Chris Jones <lugnut808@spam.yahoo.com> wrote: >> On 26/02/2015 16:55, Ralph Barone wrote: >>> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole >>> MOSFET, preferably one that's designed more for analog use than for sub-ns >>> switching. Thanks. I'm looking to squeeze a cathode follower into an >>> existing vacuum tube amp and adding one MOSFET is easier than adding >>> half a 12AT7. >>> >> >> Be a little careful if you are running a MOSFET with high VDS and >> significant power dissipation (some drain current) at the same time. >> >> Modern MOSFETs often have a positive coefficient of drain current with >> temperature, for a given gate-source voltage. Even different parts of the >> same die have that property, so that if one side of the chip gets hotter, >> it will dissipate a greater fraction of the power than the cooler side, >> and so it will get hotter still. If things work out badly, a hot spot can >> form leading to failure. Some MOSFETs are specifically characterised for >> linear operation at high voltage, and have DC safe operating area graphs. >> For example I am using one of these: >> http://ixdev.ixys.com/DataSheet/DS100086%28IXTN60N50L2%29.pdf >> Perhaps they also make something smaller that would suit your >> application. Perhaps you are even running the device at low enough power >> that an ordinary switching MOSFET would be fine. >> >> Chris > > I haven't done all the calcs (hell, I'm not even sure WHAT I'm going > to do yet), but I suspect drain current will be under 2mA with a supply > voltage of 250 V, so a device rating of a couple Watts should be > ballpark OK.
Your IRFD310 is rated at 1W, but that's for a junction-temp rise of 125 degrees, and they assume you manage to keep the case at 25 degrees, which you can't do if the junction is at 150 degrees. Bottom line your 2mA and 250V will probably make the junction VERY hot. A larger part with higher Pd rating would be better for your power dissipation issue. The two suggested TO-220 parts would have no trouble. Like vacuum tubes, MOSFETs have high Zin, but unlike tubes they have very high input and output capacitances. IRFD310 has Ciss = 170pF, which is about 100x higher than a 12AT7. The IRS820 is only 2x higher, 360pF, but it's rated at 50W. -- Thanks, - Win
"Winfield Hill" <hill@rowland.harvard.edu> wrote in message 
news:mcpu4601og@drn.newsguy.com...
> Like vacuum tubes, MOSFETs have high Zin, but unlike tubes > they have very high input and output capacitances. IRFD310 > has Ciss = 170pF, which is about 100x higher than a 12AT7. > The IRS820 is only 2x higher, 360pF, but it's rated at 50W.
It's funny that Cin and Gm are substantially higher, so that the coefficient of performance in a circuit of limited Zo can end up much worse. Well, that's just to say the impedances aren't matched, so not really saying much. If they made RF transistors with the ratings of a 12AT7, it would be much better. Sadly there's no market for a 10mA, 400V transistor that's capable of 500MHz, so the only thing you'll see is couple-ampere switching types. It irritates me more that, although they could make switching transistors with specs similar to LDMOS RF parts, they simply *don't*. The R_g and Ciss are always so bad. So if you ever wanted a 1-10MHz switching part for a couple bucks... nevermind. I suppose GaN is moving in these days, but it's still a tiny market segment. Tim -- Seven Transistor Labs Electrical Engineering Consultation Website: http://seventransistorlabs.com
Winfield Hill <hill@rowland.harvard.edu> wrote:
> Ralph Barone wrote... >> >> Chris Jones <lugnut808@spam.yahoo.com> wrote: >>> On 26/02/2015 16:55, Ralph Barone wrote: >>>> Can anybody suggest a cheap, easily obtainable >= 400 V, < 2A thru hole >>>> MOSFET, preferably one that's designed more for analog use than for sub-ns >>>> switching. Thanks. I'm looking to squeeze a cathode follower into an >>>> existing vacuum tube amp and adding one MOSFET is easier than adding >>>> half a 12AT7. >>>> >>> >>> Be a little careful if you are running a MOSFET with high VDS and >>> significant power dissipation (some drain current) at the same time. >>> >>> Modern MOSFETs often have a positive coefficient of drain current with >>> temperature, for a given gate-source voltage. Even different parts of the >>> same die have that property, so that if one side of the chip gets hotter, >>> it will dissipate a greater fraction of the power than the cooler side, >>> and so it will get hotter still. If things work out badly, a hot spot can >>> form leading to failure. Some MOSFETs are specifically characterised for >>> linear operation at high voltage, and have DC safe operating area graphs. >>> For example I am using one of these: >>> http://ixdev.ixys.com/DataSheet/DS100086%28IXTN60N50L2%29.pdf >>> Perhaps they also make something smaller that would suit your >>> application. Perhaps you are even running the device at low enough power >>> that an ordinary switching MOSFET would be fine. >>> >>> Chris >> >> I haven't done all the calcs (hell, I'm not even sure WHAT I'm going >> to do yet), but I suspect drain current will be under 2mA with a supply >> voltage of 250 V, so a device rating of a couple Watts should be >> ballpark OK. > > Your IRFD310 is rated at 1W, but that's for a junction-temp > rise of 125 degrees, and they assume you manage to keep the > case at 25 degrees, which you can't do if the junction is > at 150 degrees. Bottom line your 2mA and 250V will probably > make the junction VERY hot. A larger part with higher Pd > rating would be better for your power dissipation issue. > The two suggested TO-220 parts would have no trouble. > > Like vacuum tubes, MOSFETs have high Zin, but unlike tubes > they have very high input and output capacitances. IRFD310 > has Ciss = 170pF, which is about 100x higher than a 12AT7. > The IRS820 is only 2x higher, 360pF, but it's rated at 50W. >
Point taken. The 250 V I had quoted was total supply voltage, of which the device should only see around half, but I suppose a TO-220 isn't any harder to use that a TO-92 or DIP. PS: Thank you for AoE. In a previous lifetime I designed a firing angle meter for an HVDC converter using many hints from your book.