What can i do with fake J201 and 2N5457's?

[Rob Strand]

I took your data and processed some statistics

I also computed Rds0, which represents the channel resistance of the JFET,

   gm0 = yfs0 = 2*Idss/Vp
   Rds0 = 1/gm0

Typically J201's are around 600 or so.
2N5457's and many other JFETS are around 200.

Your JFETs are around 200ohms, more like 2N5457's with low Vp's.

Device   Vp [mV]  Idss [mA]   Rds0 [ohm]
avg            540.2    1.34          207.4
            
min           365.0      1.01        100.3
max           765.0      1.82        369.8
sd             114.2      0.22         57.3
sd%           21.1       16.7        27.6
avg-3sd      197.5     0.67         35.4
avg+3sd     882.9     2.01         379.4

[abakuzam]

I took you data and processed some statistics

I also computed Rds0, which represents the channel resistance of the JFET,

   gm0 = yfs0 = 2*Idss/Vp
   Rds0 = 1/gm0

Typically J201's are around 600 or so.
2N5457's and many other JFETS are around 200.

Your JFETs are around 200ohms, more like 2N5457's with low Vp's.

Device   Vp [mV]  Idss [mA]   Rds0 [ohm]
avg            540.2    1.34          207.4
            
min           365.0      1.01        100.3
max           765.0      1.82        369.8
sd             114.2      0.22         57.3
sd%           21.1       16.7        27.6
avg-3sd      197.5     0.67         35.4
avg+3sd     882.9     2.01         379.4

Thanks a lot! , i guess they will be useful in some way.

i don't know much about calculations and jfets, but that's a perfect excuse to learn more stuff about them . I should measure my to92 2n5457's and do the same calculations as you did and see where they belong. Thank you very much!

[Rob Strand]

i don't know much about calculations and jfets, but that's a perfect excuse to learn more stuff about them

A good start is the Fetzer calculator that rankot mentioned a few posts back.
It's at the bottom of this page,

http://runoffgroove.com/fetzervalve.html

JFETs are tricky things to deal with because they have wide tolerances.  That's why you will see a lot of designs with Trimpots on the drain resistors.   I put-up a lot of results a year or so ago.  The main aim was to show how the designs and performance changes when you consider the JFET tolerances.    I include the Fetzer method and it gives quite useable results.

[abakuzam]

i don't know much about calculations and jfets, but that's a perfect excuse to learn more stuff about them

A good start is the Fetzer calculator that rankot mentioned a few posts back.
It's at the bottom of this page,

http://runoffgroove.com/fetzervalve.html

JFETs are tricky things to deal with because they have wide tolerances.  That's why you will see a lot of designs with Trimpots on the drain resistors.   I put-up a lot of results a year or so ago.  The main aim was to show how the designs and performance changes when you consider the JFET tolerances.    I include the Fetzer method and it gives quite useable results.

I understand, my knowledge on jfets are , correct me if I'm wrong
- in production they make jfets, a lot of them, then they sort them out in some specs because manufacturer can't be sure about how they will end up, like CPU manufacturing, they just produce them then sort them depending their performance. That's why they have wide tolerance hence in some designs , designer needs to put trimpots etc to compansate that tolerances.
- they are similar to tubes without pilot lights, I have a little bit knowledge of basics about tubes. Which I can relate to your calculations

I'm so grateful for your time to answer my question and helping me out to understand electronics better

[Rob Strand]

I understand, my knowledge on jfets are , correct me if I'm wrong

That's a good start.

The only thing I can add is you will see them used as voltage-controlled-resistors (VCR) in Phasers, Compressors, Tremolo's.  When the voltage across the drain (D) and the source (S) is small they behave a little differently.   

You will also see them used as constant-current sources but this isn't used so much in effects pedals.

[abakuzam]

I understand, my knowledge on jfets are , correct me if I'm wrong

That's a good start.

The only thing I can add is you will see them used as voltage-controlled-resistors (VCR) in Phasers, Compressors, Tremolo's.  When the voltage across the drain (D) and the source (S) is small they behave a little differently.   

You will also see them used as constant-current sources but this isn't used so much in effects pedals.

Now,for me , it makes more sense, thank you again!

[PRR]

Those all look like fine amplifiers, in a stage working around a half a mA, source resistor 400-500 Ohms, drain resistor 10K.

[Eb7+9]

JFETs are tricky things to deal with because they have wide tolerances. 

jFET's are only tricky when one doesn't bother measuring Idss and Vgs(off) per device
and instead rely on the vague idea of manufacturing tolerances ... with their arbitrary widths, etc.

all small signal jFET's are more or less the same in function
parametrically different maybe in the parasitic capacitance part and operating ranges
the later usually not affecting use much in our 9v world

otherwise, with the same input/output mapping shape, stretched or scaled by Idss and Vgs(off) "limit" values

...

maybe not a bad idea to characterize jFET's (fake, mis-labelled, or otherwise) and catalogue them by their Vgs(off) value

this at least will give us the opportunity to pair or quad them together for VCR applications (eg., phasors) if we want to go the "matched" route, and in that application Vgs(off) can then also tell us what Zener bias voltage be required to bias them properly, etc ...

nothing tricky or weird about jFET's ... if you "know" what you're working with

[diffeq]

I took your data and processed some statistics

I also computed Rds0, which represents the channel resistance of the JFET,

   gm0 = yfs0 = 2*Idss/Vp
   Rds0 = 1/gm0

Typically J201's are around 600 or so.
2N5457's and many other JFETS are around 200.

Your JFETs are around 200ohms, more like 2N5457's with low Vp's.

Device   Vp [mV]  Idss [mA]   Rds0 [ohm]
avg            540.2    1.34          207.4
            
min           365.0      1.01        100.3
max           765.0      1.82        369.8
sd             114.2      0.22         57.3
sd%           21.1       16.7        27.6
avg-3sd      197.5     0.67         35.4
avg+3sd     882.9     2.01         379.4

Thanks for the equations! How would one calculate the gate voltage required for a particular resistance (say, 10k)?

[diydave]

I measured 36 of the J201 jfets, can't find the rest of them , probably, under my bed ahah

i measured vgsoff with this : https://www.rullywow.com/jfet-matcher-project-vgsoff/

Idss with this : http://stompville.co.uk/?p=112

I used same 9v regulated power supply , these are my results:

Note: i didn't bother to put -0,xxx type of measurement for vgsoff, you can put -0, at the begining of the numbers.

Vgsoff values seems to in spec but Idss values are waay out of specsi checked my smd j201's (mmbfj201) they all seem to be in specs.

VGS(off)
Volts   Idss mA
488   1,54
522   1,15
668   1,23
424   1,23
442   1,73
460   1,42
620   1,19
698   1,51
696   1,52
618   1,24
448   1,58
406   1,3
600   1,65
452   1,45
610   1,12
625   1,31
450   1,41
485   1,16
460   1,15
436   1,64
747   1,01
637   1,76
765   1,13
424   1,17
443   1,02
589   1,07
374   1,11
585   1,51
512   1,19
451   1,18
753   1,48
365   1,82
526   1,14
490   1,3
638   1,47

Year's ago I've made a jfet calculator: http://www.diydave.be/tools/fetcalc/index_eng.html
Since you have Vp and Idss for every fet, it's a "walk in the park" to get Rd and Rs to get your fet working.
E.g, the last one: if you put in the numbers, and want a mid-bias for Vgs and Vd, you get Rs = 870 ohm and Rd = 11k.

[Rob Strand]

Thanks for the equations! How would one calculate the gate voltage required for a particular resistance (say, 10k)?

You should be able to find all the JFET equations on the web and also many pages about JFET biasing.  Some will make sense to you and some won't.  Just dig around until you find one you like.

Equation 1 from here is the basic JFET equation,  most results follow on from that,
www.kennethkuhn.com/students/ee351/jfet_basics.pdf

You have to watch out for the signs of VP and VGS so I'm writing it like this,

   ID = IDSS * [1 + (VGS/ |VP|) ]^2       (eqn 1)

where VGS is negative.

We can re-write this as,

   VGS = |VP| (sqrt(ID/IDSS) -1)     (eqn 2)

The above equations are just for the JFET.

For biasing you might typically Bias the drain voltage at 4.5V for a 9V supply.  That mean the voltage across the drain resistor is (9-4.5) = 4.5V and the current through the 10k drain resistor is ID = 4.5V / 10k = 0.45mA.   So you take that value and your JFET parameters |VP| = 0.5V and IDSS=1mA  and plug then into (eqn 2)

VGS       =   0.5 * (sqrt(0.45/1.0) - 1) = -0.16V

FYI, the Fetzer calculator *starts* with VGS/|VP| = -0.35 and then calculates RD.

[Rob Strand]

jFET's are only tricky when one doesn't bother measuring Idss and Vgs(off) per device
and instead rely on the vague idea of manufacturing tolerances ... with their arbitrary widths, etc.

OK for home but not for production.    For phasers you don't have much choice but to measure since you already know the device tolerances are too wide to build a circuit that will work - even with a variable bias pot.

[Marcos - Munky]

Couple of years back, I had 2 fake 201's. At least I think they were fake. I'd tried one in a Mayqueen and they didn't work until I use a 20x trimmer to bias them (by ear) and even then the sweet spot was within about 1/4 turn. I ended up with a great sounding pedal with so much bass that only some amps can handle it. I posted about it here at the time.

Earlier this week I started to build for a friend an 18V ep booster, but it wasn't sounding right. I used an 2N5457. Drain resistor is 8k2 on the original. Today I adjusted the bias to half the supply. Used a 20K pot, then a 50K pot, a 100K pot and finally a 250K until I got half the supply (167K resistance). Now I'm hoping the mriend don't like the sound of this booster, so I can keep it

[stallik]

Quick! Build him another

[italianguy63]

As a related aside... I just got fake 2N5458's.  Same issue, very low gain.

MC

[Eb7+9]

OK for home but not for production.

DIY is about doing at home what THEY won't do at the factory ...


if that's what's needed to give your home-made product an edge, why not ... ?!

[rankot]

As a related aside... I just got fake 2N5458's.  Same issue, very low gain.

MC

Could anyone suspecting fake parts post average Idss/Vgs(off) values, so we can have a better idea about a quality of parts we have? What is the source of your fake parts, too, so we know what to avoid?

[reddesert]

Here is a plot of Idss and Vp that I made from several batches of TO-92 JFETs that I have. Everything here is genuine except for 10 putative J201s. These were bought from a US ebay seller before J201s completely dried up, I knew they would likely have some issues, but felt they might not be complete duds. This turned out to be the case: they are real JFETs but out of spec, which is fine as long as I know it.  The values of Idss and Vp were measured with the simple test circuit on the Runoffgroove Fetzer Valve webpage.



The way that a Fetzer-type gain stage is biased, if you substitute a JFET with larger -Vp, then to bias it you need a smaller drain resistor and the gain is reduced. Thus if you substitute a 2N5457 for a J201, the gain winds up lower and you get a different-sounding pedal.

[rankot]

Those are my values:

[duck_arse]

reddesert - your off j201's look quite a bit like on j202's.