JFET Shootout

That will work. Put your voltage DMM on the other side of the current DMM because otherwise the voltage drop in the current DMM will affect the voltage reading. You also need a resistor in series with the power source to limit the diode current to a safe value.

Your DMMs need to resolve better than 100μA and 10mV. 20μA and 20mV is better. We want to make multiple measurements between 100μA and 10mA. Most pedals run the diodes at 1mA max for soft clippers, 4mA max for hard clippers. Doing it manually takes a lot of time.
 
Yes, it should be directly across the diode, not the diode + the ammeter. Just like real world voltmeters do not have infinite input impedance, real world ammeters do not have zero voltage drop.
 
Chuck D. Bones said:
Yes, it should be directly across the diode, not the diode + the ammeter. Just like real world voltmeters do not have infinite input impedance, real world ammeters do not have zero voltage drop.
Click to expand...
Ok, I see what you mean. I'm not sure I'll do it, but if I do, I'll report back here
 
If you do make measurements, try this to get the most useful data from the least number of measurements. Measure Vf at 5 currents:
100uA 300uA 1mA 2mA 4mA
That should give a pretty good indication as to which diodes & JFETs are similar.

It is so much easier with a curve tracer. Here's some data I collected. The squiggles are caused by the resolution limit of the scope. JFETs were measured with D & S shorted together.

1708210861480.png
 
lrgaraujo said:
Could you identify the fake 2n3819? Are they jfets, at least?
Click to expand...
I realized it is a fake 2n3819 after measuring Vp and Idss. They're jfets. Anyway... when I bought them on Aliexpress, I didn't expect anything else. I bought them rather as an experiment.
Here's a comparison. Can you guess which is which?

1715719972339.png
 
Left one is a dud...


Take a look at this photo - left transistor. There's still some kind of a residue visible. Is it a leftover from a laser engraver?
As for the measured parameters - Vp around -0.4V, Idss - around 2.4mA,
1715721699486.png
 
I checked one of mine, the difference is subtle but it's a little different from yours (letters and numbers cover a larger area). And I can't see that residue
 
I guessed left was bogus simply because I've never seen a fake whatever split up the lettering/numbering as per the one on the right.

I'm not saying there aren't fakes with the letter-number pattern:

xxx
#x#
###

...just that I've never seen them.
 
The MMBF5458 is generally unavailable, but the Mouser website offers suggestions for similar transistors.

For example, the 2SK932 (datasheet)
1764674209678.png
or MMBF4117 (datasheet)
1764674280739.png
Any experience with them?
There is also TF412S but the SOT-833 package rather discourages its use.
 
temol said:
The MMBF5458 is generally unavailable, but the Mouser website offers suggestions for similar transistors.

For example, the 2SK932 (datasheet)
View attachment 107069
or MMBF4117 (datasheet)
View attachment 107070
Any experience with them?
There is also TF412S but the SOT-833 package rather discourages its use.
Click to expand...
Hard to find a perfect match, especially given the large range of specs.
I tend to go for 2sk208-GR when I have to replace mmbf5458
 
The 4117's Idss is extremely low, <0.1mA. This means the biasing will have to change when subbing it for 2N5458. It also means that the gain (gm) will be low and the channel resistance will be high.
 
You must log in or register to reply here.
Back
Top