need experts in biasing!!!

[ξεναγος νεκροπολης]

hi guys!
i just drew a vero for a request that MaheshSK did.
mr l3ct3r pre amp
http://guitar-fx-layouts.42897.x6.nabble.com/mr-l3ct3r-pre-amp-tp14828.html

i'm just not sure what we should read (and where...yeah i'm still newbie on this...) to bias the j-fet. there are 2 trimmers and 4 j-fet though, so i would appreciate if someone could tell what the other 2 stand for.
i don't want to just draw...i want to learn too

also if someone has time, maybe could take a look if i put the right labels on the pots numbers.
lug2 and lug 3 is always a nightmare for me when there are no numbers on the schem...!!!

thanks a lot guys!!!

[Silver Blues]

Presuming I'm understanding your question correctly, what you want to read to bias a JFET is the voltage on the drain, which is generally ideal at about half the supply voltage. (Depending on the circuit topology, the effect may sound better biased a little high or low, however, which is up to the builder to decide.) There are four JFETs and two MOSFETs in that layout - the MOSFETs have no trimmers (a choice, I guess, I've seen lots of MOSFET-based builds with drain trimmers) and the two JFETs without trimmers are just buffers, thus they need no biasing and no trimmers.

[Frank_NH]

I think there is a mistake on that schematic.  That one comes from this archive:

Buzzefall

Compare it to this one:

Krankenstein

There are some threads related to this on DIYStompboxes.com as well.

My take:
(1) The first J201 can be biased with a 50K trimmer, no problem.
(2) The second J201 has a 27K source resistor.  That seems WAY too high.  They make have meant 2.7K.  With that value it will bias with a 50K trimmer.
(3) Alternatively, they may have meant to make that a MOSFET stage with a 27K source resistor (like the Krankenstein schematic).

I any case, I would do some research before putting a lot of effort into this.  It would be a bummer to develop a large vero and have it not work at all.  Good luck!

[ξεναγος νεκροπολης]

thanks a lot guys!

silver blues
yes that was my question...my English are not so good...i thought too that the other 2 j-fet were baffers but i just wanted to be sure.i 've never saw or hear mosfet-trimmers though!!learning every day...

and frank
you're right..i could have read first...i usually do, but i don't know what happened with this...!!!anyway...first thing tomorrow!
and i think this could be a nice practice to see how some things work, so i'll definitely build this.

[Frank_NH]

If you build the Buzzefall, definitely change R17 to 2.7K and use a 50K trimmers for R3 and R15.  As far as I can see the rest should work as is, assuming the J201 JFETs bias OK.

I'm not a big fan of really high gain distortion, but if this can get a crunchy Marshall tone I'd be very interested.   I'm also curious about the dual JFET-->MOSFET stages - many of the Marshall-style drives are either multiple JFET stages, Mu-amp stages, or a combination of the two.  Haven't seen a lot of MOSFET stages being used like this...

[MaheshSK]

One more thing i'll be using bf245C in place of j201. i can't find the normal substitutes either.

[Frank_NH]

Keep us posted on this project, as I'm quite interested to know if it will work.  At some point I'm going to attempt the infamous JCM800 Emulator which is a similar multistage JFET-based distortion.  I think I have all the bias numbers for that one and the J201s to do it.

[motterpaul]

I am no expert in anything here, but I can tell you biasing has been easy for me - as long as you start by using your ears, and then dial it in. At that point if you can make it work by ear you can try a meter, but if you are substituting parts, you should try by ear first to make sure you have a trimmer that is going to be able to bias that tranny.

The one thing that tripped me up is that takes some time to understand that substituting JFets is far from a sure deal. You get vastly different results from different pieces, but especially when substituting parts.

A J201 may work where they call for a 2n5484, but they will bias much different and you may even have to change trimpots (by my experience). Even a 2n5485 will bias differently from a 5484 - and they are the direct replacement for the original.

Fets are very interesting for sure, and aside from oscillation, noise and being tricky to work with - they are the most effective and versatile trannys in my opinion.

[Silver Blues]

My advice, especially if you're going to be swapping FETs - build a tester and use Frank's spreadsheet. Takes all the ass-pain out of JFET circuits. Seriously.

[ξεναγος νεκροπολης]

i've build my board.
it's working.
but not as the real amp at all. it's more like this

https://www.youtube.com/watch?v=__7R3oZPjnM

the dry signal in the demo, was not so dry, so i've tried it with a light boost in front of it to get the demo's tones and it's quite good.

i've used 2n5458. unfortunately the only j-fet i can find here and i believe that's why i can't take out of it what i was expecting. i know i should order lots of j-fets but there are much more i'm in for now to order...anyway all pots are working as they should, definitely use a 250k for gain and maybe 300k too.
if someone else could build this so we can have a second opinion, maybe we could tag it verified.

[Frank_NH]

Excellent.  Definitely get some J201s if you can.  What did you use for R17.  27K??  Thanks.

[ξεναγος νεκροπολης]

thanks man.
miro send a link with 100pieces at 25euros i think, so i'll get them as soon as possible.

yes 27k. i've used a 300k trim to bias q3. it was really easy to bias the hole board. now thinking to add a sweep pot it place of switch, cause it would definitely be useful.
also i wasn't sure about the orientation of 2.2uf cause there was not a + at the schematic, but as i thought, it was negative leg to ground.
last thing...it's too treble for me. do you have any idea how we could make it darker?the caps near output maybe?

[MaheshSK]

great man... bit occupied for the week . will try to build it next week.... do try tpost some pics... would be great to see..

[Frank_NH]

" i've used a 300k trim to bias q3"

Sounds about right.  My JFET spreadsheet calculates the J201 bias for 4.5V at the drain requires a 180K resistor if the source resistor is 27K.  You get a pretty healthy gain for that too.   If you use a 2N5458, the drain resistance drops to 63K, with about a 33% drop in gain (perhaps explaining the behavior your seeing with 2N5458s).  Let us know how the J201s work out.  

"it's too treble for me. do you have any idea how we could make it darker?"

Try replacing C8 and C14 with 1 uF or even 10 uF caps.  Those 47 nF caps are accentuating the treble in the gain stages.

[ξεναγος νεκροπολης]

HUGE THANKS Frank!

the info about the gain of the j-fets is really helpful. unfortunately my j201s will be here by christmas...santa will bring them to me...!!!
and for the treble.i'll replace those caps and report back. i wouldn't even think that those were the caps that do the treble thing. still too newbie on those things. i need to read more theory!!

thanks again man

[Frank_NH]

Here's some further explanation of the treble boost effect with JFET stages.  Here's the typical JFET stage schematic:



Rs is the source resistor and Cs is the source bypass capacitor.  If you remove this capacitor, the gain for the stage drops by around one half (maybe a little more).  I won't go into the reasons why, but basically the capacitor is passing AC signals to ground, while allowing the stage to maintain the DC bias set by Rs and Rd.

Now think of the source resistor and capacitor together as a high pass impedance for AC.   The corner frequency above which the AC signals are NOT attenuated is calculated by the formula (you've seen before):

fc = 1/(2*PI*Rs*Cs)

where Rs is the source resistance in ohms and Cs is the source capacitor value in farads.   So, as the value of Cs is reduced, only higher frequency signals are passed to ground and thus are boosted by the stage.  If Cs is too small, the stage will act like there is no bypass capacitor and will have lower gain.  Conversely, if you increase Cs to 1 uF - 47 uF, then eventually you boost all frequencies that matter (above 50 Hz or so).

I used this principle in my modded Subdecay Liquid Sunshine overdrive.  I substituted a dpdt on-on-on switch for the second gain pot second to allow selection of 100 nF, 1uF, and 10 uF values for Cs.   The difference in gain and treble content is quite noticeable, going from thinner treble boost to a higher gain, full frequency sound.

In the present case, your 47 nF caps are removing some bass content from the gain stage, so increasing those to 1 uF should bring back the bass.  Of course, there are other places in the circuit to modify caps and resistors to tailor the bass and treble content to you liking.  Let us know what you come up with.