• Not so fast, cowboy

    RoGeorge10/23/2016 at 16:55 0 comments

    Six hours ago, the measured Vt for the altered BUZ11A was lower:

    Factory Vt = 3.84 V, C = 2.69 nF
    Modified Vt = 1.57 V, C = 2.99 nF
    After 6 hours of rest, Vt = 1.64 V

    I would like to say it's because of some temperature variations, but it looks like the transistor is slowly recovering to it's original higher Vt.
    [DeepSoic] have similar results: https://hackaday.io/project/16384-enhancement-to-depletion/log/48018-testing-tolerance-to-switching-drive

    So far, the self recovery seems to happen the same, no matter if the transistor is powered or disconnected.

    Later edit:
    2016-10-24 (one day later) Vt = 1.66 V; C = 2.93 nF
    2016-10-25 Vt = 1.64 V; C = 2.92 nF; T = 27.4 *C
    2016-10-26 Vt = 1.64 V; C = 2.90 nF
    2016-10-27 Vt = 1.62 V; C = 2.92 nF; T = 26.5 *C
    2016-10-28 Vt = 1.63 V; C = 2.89 nF; T = 26.1 *C
    2016-10-29 Vt = 1.62 V; C = 2.91 nF; T = 26.8 *C

    Not sure any more if any self recovery happens at all. Vt was stationary for the last 5 days. During all this time the transistor was not connected.

  • Instant gratification

    RoGeorge10/23/2016 at 05:47 1 comment

    I was trying to reproduce the [DeepSOIC] results from https://hackaday.io/project/16384-enhancement-to-depletion/log/47876-depletion-mode-achieved, but with other MOSFET models. Before starting to torture the transistors with negative Vgs voltage, I wanted first to trace the original Id versus Vgs characteristics of the MOSFETs.

    First DUT was a BUZ11A from ST. In the datasheet, the absolute maximum rating for the Vgs of a BUZ11A is +/- 20 V. By mistake, for a couple of seconds I applied +32V for Vgs. During this time the Vds was at 5V, and Ids was limited by the power supply to 3.2A only.

    To my surprise, instead of a permanently damaged transistor, the result was exactly what I was looking for in the first place: a MOSFET with Vgs(th) low enough to be controlled by a 3.3V pin. Please note that a positive over voltage was applied to the transistor's gate, not a negative over voltage like in the original experiment.

    So, no schematic, no setup, no measurements, no long hours of careful voltage adjustments and testing, no hard work, none of these.

    Just a dumb mistake and that's it. Mission accomplished!


    This is the brand new BUZ11A:


    and this is the same transistor after applying a Vgs of +32V for a few seconds:


    The change from 3.84V to 1.57V for the threshold parameter seems to be a permanent change, but I need to experiment more before jumping to conclusion.