[Fwd: LF: testing FETs]

[Andre Kesteloot]

Alan Melia wrote:

> Hi Peter, I think you said when we chatted that you couldnt tell the
> difference with a test meter (ohm-meter) The reason ofr this is the battery
> in an Ohm-meter is usually 1.5 volts. This is sufficient to turn-on a BJT
> but not sufficient to turn on a power FET of the type used in the Decca. The
> Audio FETs in the B&K may be different as they are used in linear audio
> applications.....they probably start to conduct at much lower gate source
> voltages. I suggest a little test rig with an ohm-meter across Drain-source
> (I dont think it matters which way round as the channel is resistive). Then
> use an external battery (PP3) or a PSU to put +9v through a 10k0
> (series....i.e.current limit) resistor on the gate. Most of these power
> switching FETs take about 6 volts gate-source to switch on....hence no
> reading either way with an 'AVO'.  The meter in d-s should be high
> resistance and swing to very low resistance when the 9v is applied to the
> gate..my IRF640s read about 2.6ohms on an Altai h/h DMM when on and
> fullscale when off.
>
>  I think that any failure would mean that this sequence could not be
> repeated. I am not sure how they fail, but the aftermath of the initial
> fault could well be a ball of melted and recryatalised silicon. The high
> resistor in the gate lead should ensure that if the gate has broken down
> there will not be enough volts to switch the device on. A quick check could
> be made directly across g-s to show that all the 9v was being dropped across
> the 10k0 resistor. If the FET has suffered from over-current it will show
> resistance between d-s with no gate drive I think. Maybe those who have
> fryed fets will confirm.....I have not lost a single FET to my Class E
> experiments !!
>
> Cheers de Alan G3NYK
> alan.melia@btinternet.com
>
> I hope this helps