as
past.
situations.
it
I hate the conventional mosfet symbol. I draw them as BJTs with insulated gates. Oh, I hate the usual IGBT symbol, too.
Hey, this is suitably goofy:
as
past.
situations.
it
I hate the conventional mosfet symbol. I draw them as BJTs with insulated gates. Oh, I hate the usual IGBT symbol, too.
Hey, this is suitably goofy:
also
We await your list of "lethal hazards". Why are you so afraid of actually making a true engineering claim? You're a fraud. ...Jim Thompson
Le Tue, 04 Jun 2013 12:29:36 -0400, Phil Hobbs a écrit:
There's that one and some others if you can stand the huge SO8 package and insanely low 10.6V BVDS abs max :
Maybe you can photocopy the datasheet and loose the annoying dots...
Available at Mouser, not Joergly priced, but still affordable.
It may be possible on a single substrate, i.e. single body diode. I'm pondering, and will let the group know. ...Jim Thompson
It would be a useful product with many applications.
Failing that, I think there are two dies in the digital transistors which come in SOT-23 and smaller packages.
However, making it on a single chip would allow a chip scale type of construction which would help in getting the heat out in a fault condition. IIRC, the SOT-23 packages are marginal power-wise for high fault voltages, and there are two of them, so they share with an AC fault.
as
past.
situations.
it
I like the old fashioned symbols too, because they remind me which way the body diode goes.
Cheers
Phil Hobbs
past.
situations.
Series opposing, right. In normal operation, neither of the body diodes turns on, so they just look like resistors. Once you get up near I_DSS, one body diode does start to conduct, but the other FET provides the current limit.
Cheers
Phil Hobbs
I'm done similar body-to-body structures, but in enhanced mode, to implement directional power switching. As in these previously posted links...
What catches my attention is the possibility to use such a gimmick as an HV DISCONNECT mechanism... rather than absorptive... ride it out.
I've done similar schemes in bipolar back in my alternator days... turn off the field device during a transient, and ride it out in BVcex mode. ...Jim Thompson
Body-to-body works nicely, though some non-linearity in the "resistance" versus voltage, then you reach IDSS in either direction.
So a monolithic structure IS possible. You just need to convince someone to build it ;-)
What interests me is the series disconnect connection in conjunction with the usual ESD diodes... perfect performance between rails, then limited current into the ESD structure. I have several monolithic processes available to me now that have depletion devices, so I get to play >:-} ...Jim Thompson
zener modulating Vgs of the FET. I'm not going to speculate about a bunch of what-if's as to what kind of environment the circuit has to tolerate, just too open ended. The original post said something about just 10mA load or so and a
48V supply, so that means he can easily power limit his source to something workable and eliminate all this crazy conjecture about spikes, blown junctions, unlimited rise times, and megalithic energy storage reservoirs all over the place ready to unleash destruction.Smoke is fine, it lets you know there is fire in the wire..
But really, how about a soft start on that cap? As you said, current limit is unknown, that could upset the customers supply if it is also operating other items?
Just thinking out load.
Jamie
I use Dep-fets on inputs of devices that only supply a resistor if you want to use the device in current mode. It's easy for some current transmitter to short or a stupid pulled.
I think JL was the one that through me the bone on that one some time back when I was looking for a cheap in line current limiter.
Jfets were also an option but not too many in the low R range.
Jamie
What a jerk you are.
S as
ts
he past.
ituations.
n kext -
rThanks, So it takes -1V (or so) to turn it off. (looking at the data sheet with 'new' eyes.) I guess the best thing is to just order a few.
For input protection I've only done the series (~1k) resistor and diodes to the power rails... which also has 'issues'.
George H.
t -
but also
I just browsed through the app note. $^&#($%($&^#^&)#^^& OMFH Every one of those circuits is a bad design. They look like a summer intern with no experience grade foolishness. Take figure 11, i don't see how to turn the FET off enough for it to work. It would require 4 volts across R2 at zero load current. A normal (enhancement) fet would work just fine.
?-)
Yep. Just make sure that U2-A,B have ESD diodes. ...Jim Thompson
.
That particular circuit originated with Advanced Linear Devices which produ ces depletion MOSFETs with very low thresholds like -0.2V and have fairly l ow currents at VGS=0, uA level. The choice of a depletion FET in a low vo ltage application buys you headroom you would lose using an enhancement FET .
also
Yeah, that Ixys appnote is horrible.
Not nearly as horrible as your "regulator" suggestion...
Have something to add? Share your thoughts — no account required.
Ask the community — no account required