Does anybody know of a SOT-23 n-channel enhancement-mode mosfet good for 300 volts or so? Current will be low. I can't find any. Lots of bipolars exist... Zetex is fearless, going to 500 volts in a SOT-23.
John
I wonder how that stacks up regarding recommended pcb trace clearances @
500V ?Graham
Hole in the middle with radial slots.
--
Boris Mohar
Hello John,
If SOT223 is acceptable check out the Infineon BSP324 or for more oomph the BSP298.
But it will be tough to get them and they'll probably make you buy a reel just to try them out. I found that obtaining samples from Infineon is a real pain which is why I don't use them much anymore.
The highest voltage SOT23 n-channel I know of is the Infineon BSS131 at
240V, around $0.10 in qties and this one is available at Digikey:Regards, Joerg
Supertex TN5335
-- -- kensmith@rahul.net forging knowledge
Excellent. Exponential thanks.
John
There was the Vishay VN50300, a 500V 240-ohm high-voltage FET.
Supertex made lots of parts in a TO-243 SMD package, like their 450V 20-ohm DN3545, 400V 25-ohm DN2540 and 350V 6-ohm TN2435.
There's the Sanyo 2sk3072, a 450V 210-ohm part, and the similar 2sk2867. Then there's Supertex' TN5335K1, a 350V 15-ohm MOSFET in TO-236 = sot-23. Mouser has 1,443 in stock at $0.43 qty 100.
There's Infineon's 250V 30-ohm bss139, special order from Newark, in stock at Farnell, and second-sourced by Fenghua. :-)
Hey, why not consider my favorite depletion-mode n-channel MOSFET, the 500V Supertex LND150 (TO-92) and LND250 (sot-23) 1k-ohm parts, or one of their bigger brothers? There's even Supertex' new 700V 42-ohm 500mA dn2470 in the TO-252 D-PAK, amazing! I may deploy those in my 10kV 1us pulse project to reset things between shots. Hah, maybe a stack of them can provide the 0.5A output-current!
--
Thanks,
- Win
Potting it would help.
Ooh, good stuff. Thanks.
We have the LNDs in stock, in sot-23 - they're even on this board already, in another place - but we need enhancement mode here. It's a programmable precision current sink, very low current, and all we have is one +5 volt supply, nothing available to pull the gate negative. It's got to sink 100 na to 100 ua with anywhere from +1.5 to +300 at the drain.
John
"John Larkin" a écrit dans le message de news: snipped-for-privacy@4ax.com...
Sure you have. Didn't you say you had +5V? You could pull the source positive. Probably not the better solution though.
If you already have a 7002 or something on your board, what's wrong with cascoding it with the LND? The additional 7002 may be better than just another PN on the board.
-- Thanks, Fred.
In article , John Larkin wrote: [...]
If you want to use a part you already stock: How about hooking the source to +5V?
..or cascode it with another MOSFET
-- -- kensmith@rahul.net forging knowledge
Maybe not. The closest point of the pads in under the body of the part. It is likely that the potting material wouldn't get completely under it. Even if it did, the knit line would still be the weak spot.
You could submerge the whole thing in Fluorinert. This would help with keeping it cool too.
-- -- kensmith@rahul.net forging knowledge
Two problems: not enough voltage headroom, and not enough pcb area.
Isink | | a k diode |
+0.5 ref----------- + | d opamp out----------g enh mosfet s +--- - | | | +------------------------+ | | r1 | | gndR1 is switchable to set the current. This must accurately sink current when the voltage at Isink is in the range +2 to +3, and must tolerate
+-300 volts without damage.An LND won't work for this fet because the gate can't go negative enough. The min drain voltage is +1.4, so I can't see a way to cascode this.
This circuit is replacing a much simpler thing that seemed to be a good idea at the time.
John
"John Larkin" a écrit dans le message de news: snipped-for-privacy@4ax.com...
le
Hook the LND gate to +5V and put your circuit with a 7002 under the LND. Try it, that'll work. When the LND has not enough bias the lower fet does the job.
-- Thanks, Fred.
Egad, that works! But if we can't get a nice HV sot-23 enhancement fet, we'll do that. This circuit is really crowded, so we'll try the simpler way if we can get the parts.
Clever!
| | d +5 --------g LND s
John
Hello Ken,
One method frequently used in med where much higher iso voltages are required is to round off all pad and trace edges that point into the critical direction. I'd certainly do that in this case. Costs nothing in production, just a few minutes of work on the CAD library.
Regards, Joerg
[...]
It did generate a phone call. "Oval pads on your SOT23s that can't be right".
-- -- kensmith@rahul.net forging knowledge
With good PCB material, and with a clean surface, 10 mil spacing easily stands off 1KV with very low leakage.
Potting IS NOT a "cure-all". In fact,one can have inferior results due to trapped moisture and ions.
Rather than potting do a search on "conformal coating". Widely used in avionics for environmental protection and improvement of dielectric strength. Applying it properly can be involved and expensive though.
-- Tony Williams.
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.