Do jfets sound different lately???

[BigLloyde]

For the last year or so...more or less.......seems like pedals I have built using jfets (mostly j201s) have been sounding different.  Seems like they are more distorted and brighter with less mids.  I use Fairchild brand mostly, so I don't know if this is brand specific or not.  Also, I first noticed this on a fresh batch that I bought about the time Mouser was starting to switch over to RoHS compliant components, but I'm not sure these were, but the latest batch sounds about the same.  Not sure if that's at issue or not.  Mostly just wondered if anyone else had noticed this and if they had, are there any good suggestions to "bring them back" maybe by changing the resistor and/or capacitor values off the source pin if that makes sense.

[Phorhas]

Man, don't go there!

Next thing we know, people will be raving about NOS JFETs - which only they posses the full and dynamic "tube sound" !!!



In what circuits are you having these problems?

[BigLloyde]

LOL......no kidding.

Just some variations of amplifier type circuits with jfets replacing the tubes, similar to some of the ROG stuff, but I don't have a link to a specific schematic.  I kinda thought/hoped it was just my imagination until I built one for this guy that already had one that was built a while back and he mentioned that it sounded different.

[km-r]

Man, don't go there!

Next thing we know, people will be raving about NOS JFETs - which only they posses the full and dynamic "tube sound" !!!



In what circuits are you having these problems?

hahaha, use NOS JFETS to emulate NON-NOS TUBES...
anyway, i'll be buying loads of J201 just in case theyd become rare... 

[DDD]

Our ears are prone to change (i.e. get older, to our regrets) 100 times more rapidly than jfets. Maybe it's a reason...

[brett]

One batch might change relative to the next.

It would be unusual for JFETs to affect tone very much.  Their main game is gain. 
Having said that, low gain/low volume often sounds "thin", "brittle" and "tinny" because at low volume the human ear mainly picks up the midrange. 
Some of the older ones amoung us remember the "loudness" button on 1970s hi-fi gear.  The idea was to boost bass a lot and treble a bit when using low volume.  Of course, most of us wanted to rock the neighbouhood, so we just left the loudness button "on" all the time.
cheers

[db]

Brett beat me to it, but it's true that there are differences between individual devices, different batches and manufacturers.  The spec for JFETs is notoriously wide.  For a J201, IDSS varies between 0.2 and 1mA, VGS(OFF) varies between -0.3 and -1.5V, gm is only specified as a minimum value!  Any device within these limits could be called a J201.

I recently measured IDSS for a batch of 10 J201s and this is what I got:
0.42 mA, 0.60 mA, 0.83 mA, 0.86 mA, 0.48 mA, 0.60 mA, 0.85 mA, 0.61 mA, 0.83 mA, 0.53 mA

The big question is: how does your design cope with this variation and how would it affect sound.  Many circuits use a drain trim resistor to adjust for these variations, but this has the effect of altering the gain of the stage so an individual device would probably sound different in that circuit.  Having said that, you would expect this to vary between individual devices as much as between batches so it may be that the J201s you have bought have had a tighter spread than the ones I got.

[MartyMart]

Also, don't forget that the tolerances on all the other components will be "slightly" different.
You could, for instance build 2 X the same circuit on the same layout and they would be a
little bit different.
Perhaps if you used ALL the exact same parts and IC on a pair of TS-9's they would get
close, but with Jfets, you would have to match all the "Q" positions from a batch to get
real close, say only use the 0.60mA and 0.83mA spec fets for same "Q's" on both .....
then set the trim bias "exactly"  - that would be a tough one

MM.

[markm]

I honestly have not noticed a difference in the device per-se but, what Marty says certainly is true.

[BigLloyde]

What you all say makes perfect sense, I guess I knew most if it all already, but it has made me think.  Maybe the design is pushing one of the jfet sections too hard, or trying to get too much gain out of one of the jfet sections, but I got away with it because of the specs of that bunch of jfets.  Thanks for the help guys.......back to the breadboard for me....

[JimRayden]

Yargh, those RoHS-bastards are killin our tone!

---------
Jimbo

[brett]

Oh no... it's a choice between "save our tone" or "save the environment".

Not really.

[Paul Perry (Frostwave)]

It wouldn't suprise me at all if the semiconductor fabrication process was changed.
It's happened before, that components were replaced by ones with suprisingly different characteristics. A classic was the old 2N3055 power transistor. Which suddenly had its gain and bandwidth "improve" alarmingly.
I certainly wouldn't dismiss it out of hand... the J102 is one of those componebnts that is so well known, that it is impossible to kill.
Siliconix, the old J102 maker, is now owned by Vishay - I bet the original plant is long gone.

[km-r]

whos this RoHS guy??

lol, lets murder him!

[BigLloyde]

I never wanted it to come down to this................MURDER

In my particular application, I am having some success increasing the size of the j201 source resistor from 100ohms to 470ohm, then, re-biasing.  Seems to restore the sweetness pretty well  IMHO, without losing too much gain.  Thanks all.

[JimRayden]

whos this RoHS guy??

lol, lets murder him!

Let's not be so lame. First we build a RoHS-Ruby with all sorts of RoHS FX up front, then tape headphones on the evil RoHS-guy's head and let him enjoy that... all RoHS'y sound.

Now that'll wipe any environment-friendly thoughts offa his mind. Long live non-RoHS pirates! Yarr!

If you excuse me, I'm off to play "RoHSy Lady" now.

--------
Jimbo