ATF55143-incorrect DC current, works like a thyristor,I bought the fake conponents???

[Aprecede]

I would like to design a LNA and chosed the ATF-55143. But when I test their DC characteristics, they are extremely strange!!! I biased the circuit as the image showing, Rd=100Ω,Rg=10kΩ. When I set the VDD to 2.7V and alter the gate voltage(Vg), the current(Id) suddenly reaches 14mA when Vg is 0.25V, then the current(Id) keeps changeless although Vg increases. Then, even when I set Vg=0, Id keeps the same until VDD decreases to 1.8V or lower. I bought the ATF 55143 from two different place, the price is 2.2 U.S. dollar and 0.4 U.S. dollar, respectively. But their DC characteristics are the same. I doubt I have bought the fake conponents:-? ? Does anyone have some ideas about that???:smile:

 

[FvM]

You may be seeing circuit self-oscillations. Is your layout prepared for GHz frequencies? Avoiding resonant loads?

 

[Aprecede]

Thanks for your answer, but I was only testing its DC characteristics and they are very different from the design manual.

- - - Updated - - -

You may be seeing circuit self-oscillations. Is your layout prepared for GHz frequencies? Avoiding resonant loads?

 

[FvM]

I was only testing its DC characteristics

"Only testing DC" can turn out a bit difficult with GHZ transistors without well-considered test circuit design.

One point is that these components are rather sensitive to damage by ESD, so I won't perform unnecessary measurements before putting the component into the final circuit.
You can probably stop oscillations by damping resistors.

 

[Aprecede]

Thanks. I test some other ATF-55143, their DC characteristics are the same though different from the design manual. If the circuit oscillates, the DC current are aways jumpy, but here it keeps the same. I think it can't be a oscillator as the image shows.

"Only testing DC" can turn out a bit difficult with GHZ transistors without well-considered test circuit design.

One point is that these components are rather sensitive to damage by ESD, so I won't perform unnecessary measurements before putting the component into the final circuit.
You can probably stop oscillations by damping resistors.