Component tester issue with testing JFETs

[kmc212]

I had recently bought a cheap component tester with an ATMega328p chip. One of the issues I have been having with the meter, is testing JFETs. The meter will identify JFETs as 2-Diodes when they are genuine JFETs. Is this an issue with the meter that can be fixed?

This is the meter I am using:

Aideepen Mega328 LCR-T4 Transistor Tester
https://smile.amazon.com/gp/product/B07RJLK7MC/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1

[exe]

What jfet model are you testing?

[madires]

Have you done the self-adjustment?

[duak]

N-channel JFETs and depletion mode MOSFETs require a negative voltage to be applied to their gates to shut them off.  I'll bet this tester does not generate that voltage.  Could you post a schematic or a picture of the parts on the circuit board showing the part numbers so we could see if it does?

[Kleinstein]

These cheap testers have a  limited 5 V voltage. So they have a principle problem with parts that need a high control voltage like some JFETs or MOSFETs with a high threshold. Another Problem are thyristors or similar which requite a significant trigger current.
AFAIK some of the tests work Ok for low threshold JFETs. So 2N4393 (rel. low threshold) has a good chance to work - 2N4391 (4-10 V threshold) would be near impossible with only 5 V supply.

[kmc212]

I've done the self test several times with the capacitor. I will try again. The readings are off.

[kmc212]

I am having trouble with J111  and J105 that are Fairchild.

[kmc212]

I have successfully tested Fairchild J112 that I bought on Amazon. The difference with the J112 that test good is they have the arched legs with the Three Color Bands. The J111 and J105 I have received have straight legs legs and not color bands. 

[kmc212]

these are photos of the  tester

The product link is here

https://smile.amazon.com/gp/product/B07RJLK7MC/ref=ppx_yo_dt_b_asin_title_o04_s00?ie=UTF8&psc=1

[kmc212]

I have noticed MOSFET voltage regulators don't test properly either. What's a good tester?

[Kleinstein]

I have successfully tested Fairchild J112 that I bought on Amazon. The difference with the J112 that test good is they have the arched legs with the Three Color Bands. The J111 and J105 I have received have straight legs legs and not color bands.

J111 is the higher threshold version (3-10 V), while the J112 is lower threshold (1-5 V). The tester probably can't detect the JFETs if the threshold is higher than some 4.5-4.8 V.

Similar problem with J105 - it is a high threshold one.

[kmc212]

That could be it. But why identify the JFET as 2 Diodes? That also happens if two legs are touching in the same port.

[Kleinstein]

For a high threshold JFET drain and source are more or less connected with a relatively low resistance (some 20 Ohms for the J111). In addition there is the normal diode function of the gate-drain/source junction. Well below the threshold the N JFET behaves like 2 diodes with common anode (gate) and a resistor between the cathodes (drain+source).
The tester may not have a separate entry for this configuration, as it normally does not make much sense - from this perspective it could as well assume a high threshold JFET.

So there can be some room for an improved software in this detail.

In theory the tester could even identify the JFET even more, as even below the threshold there is some change in the drain-source  resistance when the gate voltage is changed. However the change may not be large enough to be well measured with the limited means.

[exe]

That could be it. But why identify the JFET as 2 Diodes?

Afaik, jfet is a diode-like structure that is reverse biased. So, base to source has a diode structure. Small signal jfets a usually symmetrical*, source and drain are interchangeable. So, no wonder it also measures as a diode from base to drain too. Look at the picture here: https://en.wikipedia.org/wiki/JFET .

*I don't know how non-symmetrical jfets behave.

[indman]

Here is the method from the Karl-Heinz manual:
" 3.8 Measurement of E-MOS transistors and IGBTs
You should know for enhancement MOS transistors (P-E-MOS or N-E-MOS), that the measurement
of the gate threshold voltage (Vth) is more difficult with little gate capacity values. You can get a
better voltage value, if you connect a capacitor with a value of some nF parallel to the gate /source.
The gate threshold voltage will be find out with a drain current of about 3.5mA for a P-E-MOS and
about 4mA for a N-E-MOS. The RDS or better R DSon of E-MOS transistors is measured with a gate
- source voltage of nearly 5V , which is probably not the lowest value. In addition, the RDS resistance
is determined at a low drain current, which limits the resolution of the resistance value. Often in
the case of IGBTs and sometimes also withn enhancement MOS transistors, the available 5V of the
tester is not sufficient to drive the transistor across the gate. In this case, a battery with about 3V
will help to make a detection and measurements with the tester possible. The battery is connected
to the gate of the transistor with one pole and the other pole of the battery is then connected to a
test port (TP) of the tester instead of the transistor gate. When the battery is correctly polarized,
the battery voltage is added to the control voltage of the tester and the detection of the transistor
succeeds. The battery voltage must then be added to the indicated gate threshold voltage of course,
in order to get the correct threshold voltage for this component."

I tested it successfully on the Proteus simulator. Works great on the 2N4391 transistor 
In the 1st screenshot, the 2N4391 is connected directly to the test contacts

[kmc212]

Thank you for your helpful responses. I am not certain if the tester I have can be updated in any way.