I hooked up a 2N7000, 9-volt battery+led connected to source and drain.
You can do the usual fun floating-gate 3-state tricks.
OK, the real point was to connect a power supply to the gate through a
100K resistor and crank it up. It died at -74 volts on the gate, at which point my LED turned on dimly and I lost gate control. The gate now measures roughly 5K to the source or drain.John
What was the connection? BATT to drain to source to BATT and power supply minus volts from where to where?
You can regain gate control but only for one more action: Connect a
470uF cap to your supply, let in run to 100V, connect to gate ... *BAM* ... there, LED goes out :-)Disclaimer: Professional driver on close course, do not do this in real life.
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
Batt +9, resistor, LED, drain. Battery negative to source. Floating gate initially. If you touch the source and gate, the LED goes off. Touch gate and +9, bright. Touch gate and drain and the LED goes on half-bright, where Vg = Vd ~~ Vgs threshold voltage. The gate will float in any of those states for a long time.
The blowout voltage was applied to the gate against the source. As I cranked up the supply, the LED stayed off until I blew out the gate at
-74. Most people rate them at +-20 Vgs max.
Maybe I'll try some power fets. I've heard that they typically stand very high gate voltages.
John
I like to test parts to destruction, but I'm not outright cruel.
John
I use them as a 'touch switch'...... 3M3 gate to Vcc, transorb gate to source ( cos I found that some got 'blown'...) touch button- 10K- gate, other touch button to source. Vcc 5V Unlimited life(almost ) :)
Those VGS,max are for sustained voltages- never bothered to find out why, but a 20V device can take something 40V for a limited time like
20ms without damage and it doesn't seem to matter whether the offending source can supply any current or not.
This is fun:
+9----1K--LED-+-----------------+ | | | | | | | | | | | 1M | | d \ | g-----o \o-----+ s | | 1 uF | | gnd gndIf you hold the switch closed, the fet settles to Vd = Vgs(th) and the LED is dim.
Now release the switch and then tap it every couple of seconds. Initially nothing happens, but after some number of taps the LED brightness starts to seesaw, more and more, and eventually it becomes an alternate-action switch. Press and hold to return to the dim state.
It's also fun to play with the floating gate. You can move blobs of charge around with a small insulated screwdriver or some such. Or induced voltage from charged objects.
Gate leakage seems to be on the order of thousands of electrons per second.
John
Then put them out of their misery quickly - make a Cockroft-Walton multiplier with the rectifier/reservoir blocks from a half-dozen SMPSUs.
Other testing might be useful: in addition to catastrophic failure, high gate bias can cause the threshold to shift. That is JUST as destructive, in terms of calibration shift and eventually ON resistance and logic threshold.
And the transconductance can suffer, too, I suppose. That's rarely a limiting factor, one doesn't choose a high-current MOSFET for its input voltage small-signal sensitivity...
How do I bill you for the time I lost playing with this the first time you mentioned it?
How do I pay you for the fun I had?
Ed
I think the charges cancel.
John
Any remaining balances can be leveled out at Zeitgeist :-)
-- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
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.