Friday afternoon weird mosfet behaviour ZVN2106A reverse Vgs conduction

[fourtytwo42]

Has to be Friday, Just bought some new mosfets from a reputable distributor and they are behaving VERY strangely.
I am using an unusual circuit that reverse biases the gate/source junction during part of the operating cycle, the applied voltages all being <=5V BUT I find the gate is sourcing/leaking current, in this case 170uA with a Vgs held by the device at ~800mV. This smells diode to me but the specification claims there are none!

Here is the product overview sheet claiming no diodes https://www.diodes.com/products/discrete/mosfets/mosfet-master-table/part/ZVN2106A
Here is the data sheet https://www.diodes.com/products/discrete/mosfets/mosfet-master-table/part/ZVN2106A

And attached is my test circuit, current flows in R5 with the actual device on a breadboard

I just downloaded and incorperated there own spice model and of course it doesnt show this problem!

My guess is the devices (I have several the same) are out of spec, any comments ?

[fourtytwo42]

I also tested a small SOT23 mosfet (BSS123) and suffered a similar problem with a leakage of ~300uA
BUT a power mosfet (RFB12P08) had no measurable leakage in either bias direction as expected.
So is this phenomenon related to small mosfets I wonder ?

[T3sl4co1l]

That shouldn't be.

First of all, try to rule out any mistake in your setup, see if you get positive and negative results for dummy components (like an actual diode, zener, BJT, JFET, other MOSFETs..).

Is the response resistive?  The gate may be blown out.

You may simply have bad parts, or they have been handled badly.  Small parts are much more ESD sensitive than power parts, after all.

Tim

[fourtytwo42]

Thank you so much for your reply Tim, I thought I had lost the last marble!

The humidity has been very low here recently and I have nothing but extremely rudimentary static precautions
One of the batch failed with classic punchthrough symptoms of 60R bi-directional from gate to source.
I don't often use tiny mosfets, I built several circuits years ago with ZVN2110A's but that circuit was not sensitive to reverse gate leakage (not many are).

Better get myself some protection 

Rog

[fourtytwo42]

I think I may have found one culprit, my ancient soldering iron was no longer earthed so any leakage current from the mains powered heater was just developng a nice voltage on the tip. A weld had broken internally between the earth lead and the element shroud (see pic).

I measured the leakage using a Fluke 87 (10Meg impedance) and got ~40Vrms or 4uA quite enough to kill these devices Of course no effect on the majority of the stuff I handle so not sure how long this fault has been waiting to bite me!
Next time to be doubly sure I will push the leads through a bit of anti-static foam (without touching them) and leave it in place till after soldering then pick the remains off with tweezers!

I did an exhaustive search for protected TO92 or similar mosfets and found very few, and they are very hard to buy!
STQ1NK60ZR   TO-92
2N7000K   TO-92
STN1NK80Z   SOT-223
STD1LNK60Z-1   IPAK
BSS138BK   SOT-23

[T3sl4co1l]

Aha!  A very circuitous diagnosis, but a fortunate discovery!  That would definitely do it.

Zener protected gates seem relatively common on smaller parts, but keep in mind they use the smallest diode possible, so as to minimize added capacitance and leakage.  I don't recall seeing any rated for ESD > 2kV (HBM), if even that.

Tim

[fourtytwo42]

And there we have it, handling and fitting the mosfet with an anti-static foam tutu works 
Fairly easy to remove afterwards with tweezers.
Of course having an EARTHED soldering iron helps too 

Cross posted in renewables GTI blog, sorry!