'New' Jfet low-gain overdrive idea...

[liquids]

Good question and reminder.  I could do it right now so maybe I will. But in the past I found it of no consequence as I have two BSIABs that sound about identical enough to me.  Maybe I'd notice it more with the lower gain.  I'll also try some 2n5457s when I do, for fun...

[aziltz]

you designed it for your neck pickup, but described the things i like to do with my tele bridge pickup and fingers.

I'm thinking you will really dig this!  If/when you build it (or anyone), contact me or post here if you want help on tweaks for sounds you can't seem to get.  I really think that if you're looking for that picking-strength regulated clean to gritty type overdrive, it will be here.

sure will!  it'll be a while though, I have more circuits than drilled boxes right now!

[liquids]

With regards to JDs comments, did you find much audible variance when taste testing other mpf102s at the front end?
I may draw this up on the weekend if I have time, perhaps with an optional output buffer.

A quick test of 4 MFP102s and a pair of 2N5457s yielded interesting results...

I wanted the MPF102s to not matter one hand for easy reproduction, and on the other hand I would feel silly saying any MFP102s sound the same because of the way they are said to vary so much....I got a surpising combo of two.

My limited experiments showed me this: I had a 'magic'--really a 'odd man out' MPF102 in there originally, it seems.  Whatever it's stats are, it was yielding noticably lower gain.  It was the one I originally had as the upper of the two MPF102s. Other Jfets proved noticably louder and hence clipped a bit more, all else being equal.  OF note, this exact MPF102 yielded much less, if any, noticable difference when it was the lower JFET.  The other three JFETs all sounded close enough to the same to be negligable to me given the gain control and 20% tolerance pot that follows that stage.

2N5457s clipped way more.  With my hot bridge humbucker pickup, they had a bit of unwanted 'input' clip with that first spike of a chord, etc, which is one reason why I've used MPF102s in my personal BSIAB pedals. along with extremely improved guitar volume knob control.

I was, however, able to tweak the source resistors to get desirable and similar gain levels. With 2N5457s, I needed to go as far as about 33k (!) on the source to tone down the gain of the first stage to get the response I was looking for.  To be nitpicky, It did still have a 'not as smooth' clipping with my bridge pickup, but that may not be as much of an issue if you aren't using such hot pickups, YMMV.

Using any of the other MPF102s, if I scaled the source resistor back to 10K, it was about right. I'll probably go with this now, as using that arrangement was not only more consistent with the other MPF102s, but maybe a little clearer than using the 'dud' MPF102 and a 4.7k...?

So, if your trying this, swap a few different JFETs around for yourself. But the moral of the story is, thankfully, it's not something that can't be compensated for.

[Toney]

 But it sounded the best with that 'misbehaving' Jfet, right.

Serendipity = happy accident.
So now, I guess, it would be good to find the characteristics of that mpf102 with this...

http://www.geofex.com/Article_Folders/fetmatch/fetmatch.gif

and perhaps Johans idea from 2006:

"a trick i found out a long time ago...just use you DMM to open and close the gate..first you measure the resistance between drain and source..is it open or closed? 
<a few "k" or >1M ? 
Holding one of the probes at either the drain or source and touching the gate with the other, either opens or closes a functioning FET( measure resistance drain/source again), depending on polarity..swapping the probes will do the opposite.
I realize my description might be a little cryptic, but once you sit down and try, you will see what I mean..."

Once you have an idea of the characteristics, perhaps there will be another device which replicates them whilst 'in spec'.

[liquids]

But it sounded the best with that 'misbehaving' Jfet, right.

Nah, like I said, I'm not going with the dud/misbehaving MPF102.  The other 3 MPF102s I tried were pretty similar in gain etc. While they were louder than the one, all else being equal, I liked them with a 10k source to ground compared to the dud with a 4.7k--which while yielding about the same level in volume, the more consistent ones sounded a bit 'clearer.'  Given that one MPF102 was the odd man out of my bunch, statistically it makes more sense to design around that than the misbehaving MPF102.   Or use 2N5457s with a 25-33k resistor, and it's 'close enough' to not go crazy nitpicking over...just tweak to your liking.

I realize many will say that MPF102s are just too inconsistent.  To me, so are pots.  It's not a production pedal circuit, so it's infinitely tweakable anyhow.  It's a starting point for people to try, but yeah, anyone's results are going to vary a bit.

I'll eventually get to testing some JFETs.  I have all I need to do it, I just haven't done it. Once I do, I can try and post stats for those interested in getting close.  To me, from the start I was shooting for a sound and tweaked until I got it with the JFETs I had...so with different JFETs, I do the same and use my ears.    Plenty of knobs on the thing to boot... 

MOre so, from my findings, If you can isolate the the first mu-amp stage with whatever random pair of MPF102s you have (with sockets for sure, and on a breadboard ideally) and can then swap source resistors and/or JFETs until the gain is slightly above unity with the bypass signal (though it will have a distinct EQ curve), than you are in the ballpark, I'd say, for the right response and gain range for the first stage.  

[Toney]

So now the magic one is a dud? 

Anyhow, I will have a crack at this on the weekend.
Looks like it could be fun.

[liquids]

I really can't wrap my minds around how to wire up the gain pot in real life(!)...I always get stuck on this. 

The pot is supposed to act as a variable (log-type) resistance of 500 k to let varying levels of the first stage signal to ground.  There should always be at least a 1K resistance, so some signal, however little it may be, goes to the next stage.  a 470pf 'bright' cap should be involved so that the high end is preserved at all gain levels. I'd like to keep the cap on the pot and off the board, to boot.   Can anyone do a picture?  I seem to be able to do it in theory better than reality.  Following the BSIAB schematic on GGG (http://www.generalguitargadgets.com/pdf/ggg_bsiab2_sc.pdf), it would be akin to R19 and C6....      ....

[waltk]

I used your schematic (1.1?) to make a layout and etch a board.  But that was before I had seen the second page of posts, so I might have to make some adjustments before wiring up some pots and trying it.  I guess I could just socket the source resistors to swap values.  Fortunately, the JFETs are all socketed.

Do you have any other suggestions before I actually hook it up?

Here's the layout


...here's the top side (I made one change after the picture - when I realized the cap at the top-right should be 1uF instead of .1 uF)


...and here's the bottom side

[liquids]

That is awesome!   Scary that you haven't breadboarded it and have done this anyway. I really hope you like it...

As per previous posts and findings which are definitely worth reading, change the source resistor going to ground to 10K.   I had a certain MPF102 that made for extremely low gain combined with the 4.7K, was originally expirimenting based around that.  MPF102s are said to be 'all over the map,' and that one happened to be on the low gain part of the map.       I've swapped it a few times with others, and between the other three it was the ugly ducking (or, magic, depending on how you see it), with the others being negligably different one ro another, but all were much higher in gain.

With other factors compensating for the increase in gain (changing the 4.7K to a 10K in that stage), I do pefer the 'other three' and it should make other's results...uh, less inconsistent. So it needed an update.   I'd socket that resistor while you are there unsoldering it, even, since you haven't breadboarded it.   As mentioned in that post, if you can audioprobe out of that first JFET stage .1uF cap---what you are mostly looking for between the MPF102s and the source resistor combination you use, is it should basically match your clean signal volume, and the tone should be clean and clear, rather than muffled.  It turns out its sort of an oddball buffer stage (near unity gain if nothing else) with a non-flat frequency curve, followed by a gain control after it, which for all practical purposes dials down the signal hitting the next stage from between unity gain to zero, without sacrificing huge losses in clarity...that's how I see it at least.   

Also, on the tone stack I went to .1uF on the bass side...it's a subtle difference if any at all, but it may be worth socketing, again, if you haven't breadboarded it. 

I am away on vacation until tomorrow morning, and I really want a boxed up build of it for a gig I have in just over two weeks!   My first vero layout attempt was a bust in size for the enclosure, so that reminds me that I need to get going on the second version.   Looking at yours is simultaneously encouraging and unsettling!

Anyhow, read through the posts that you may find interesting because I may be missing something...keep us posted and don't be afraid to say you don't care for it, if that be the case!  It's a starting point, if nothing else.

[liquids]

Also, seeing your layout helped me wrap my mind around how to wire the 'Mid' (it's really just a gain control) combination set resistor and pot. I was more so leaning in the direction that I needed to use the 470R between the third leg and ground, but its easier the way you have it, and good to know it will work that way.

[liquids]

Oh no...a closer look at your layout (and even the v1.2 schematic I have right now) shows a horrible oversight that the second stage, the input of the 2N5457 is only seeing what's coming out of what the 470pf filter lets through!  It's supposed to be wired a gain pot to ground with a 470pf 'bright' cap on it. My schematic drawing skills are seriously lacking...my pleas for someone who is good with schemas and has a good program to draw it met silence, so your stuck with my bad drawings.    I've got to fix and update the schem if and when at all possible , which is a bit misleading since I know what I have but can't flesh it out...      Maybe I should pull it for the time being?  I'm very sorry about this.

Help!

[waltk]

I'm very sorry about this.

Don't sweat it... I think you've made a fine start with this.  I never do breadboarding, and usually do experimental layouts like this when I have some extra space on a PCB I'm etching.  Also, I consider the soldering part to be fun.  So it's no big deal if this particular one doesn't work out.

the input of the 2N5457 is only seeing what's coming out of what the 470pf filter lets through

I see what you mean about the 470pF cap into the 2N5457... oops.  I was just blindly translating your schematic into a layout without thinking about it much.  It may be possible to work around this with a little board surgery.  If not, I'll give it another try with an updated version next time I do some etching (and I don't mind making an extra one for you).

I've got to fix and update the schem if and when at all possible

Please do continue tweaking your design, and post an updated schematic when you think the issues are worked out.  I don't know what you used to draw it, but it actually looks very nice.  I use DipTrace, and the results aren't any nicer looking - and when you're done it's very easy to create the layout (auto-routing and all).

[liquids]

Thanks for the kind words and offer for a board, good to know it was just extra PCB space. 

I use the DIY layout creator for schematics, but I find it a bit tedious to work with as it's not so friendly to error or bad mouse drags.  After I wrestle with it, I transfer it as an image to MS Paint to chop it up, and load it on google 'pages' (sites?) as an image.   

Maybe you can solder a jumper somewhere between the gain pot and the 2N5457 input to 'fix' the current board, as I'd love to see what you think of the sound of it from that 1.1 version even, if you can do that.   The tweaks I've made since then are subtle, and add some additional control, and are kind of 'personal' for my rig--so there is nothing to say that the version you build off of (if corrected) isn't better, depending on your tastes.  I'd definitely socket a few of the caps and resistors particularly in the tone stack, for easy modding, if you like the core sound and later solder up a corrected layout.     

I would update the schematic right now even, but I'd have to start from scratch, because the file for the working schematic is about 1,000 miles away until the weekend.   

[Ben N]

I am interested, just don't have a lot to say. Except that this is not the first attempt at a mu-amp based low-gain overdrive in these parts. Long ago, Jake Nady put out something he called "Clean Machine", IIRC. I didn't build one, and don't recall how well it faired, and don't know where the schematic might be, but I'm pretty sure it had two mu-amp stages, and maybe a buffer.

[liquids]

Old link should work the same...here's a cleaner shortcut of the updated schematic, which should be....well, at least closer to 'correct' 

http://sites.google.com/site/teaandfiction/Home/overdrive-1/Liquidrive1_3.JPG

[liquids]

I've put this circuit (and a few others) to the side recently, not for lack of want, but due to business prepping for a gig.  I wanted to box this up for the gig, and spent most of a previous Saturday attempting to do so but had no luck debugging, so I abandoned it on vero for now as a later debug or rebuild. 
Hooking up the breadboarded version again and tweaking even today, I've noticed some parts and values weren't as previously posted, and tweaked some others.  Also, I retract my comments which say if you signal probe out of the first mu-amp stage out should be unity gain -- it's not.  I'll eventually do some JFET testing especially for the MPF102s.

I've updated the schematic to reflect what I have on the breadboard, and simplified some of the odd values, so to speak.  Whenever I update the schematic it is reflected on this page: http://sites.google.com/site/teaandfiction/Home/overdrive

Some things I will say; the 'bass' control should be considered subtle, and if more range is desired, upping it to 1M option might help, though I personally find it overkill.  Also, the 'treble' control was designed to be run as full tilt as possible, where it is then basically functioning like a .1uF bass filter, with, again, a very subtle treble roll off (and some bass increased) as you turn it back or if your guitar is too bright, as well as useful for settings with more gain and/or bridge pickups.  The subtly of the EQ controls kind of speak to liking the core of the sound and the comparatively low and 'narrow' gain range this is designed for, hence not wanting or needing drastic controls here, though the BMP tone stack is essential, mostly for draining a (controllable) amount of gain between 2nd and 3rd the stages. 

On this circuit, and my modded BSIAB which are at the core the same circuit, I don't like the sounds yielded by low setting of the "Drive" knob, which are overly bass-shy due to the cap across it.  I recommend and personally run it at least 1/3 of the way up, no matter what. Removing the cap affects the sound too much and is a necessary ingredient. 

That being said, even with the Drive knob 1/3 up, if you set the mid/gain pot to minimum, even with the volume maxed you'll be very clean on gain, and just over unity gain on output, which is all I ever need volume wise.  So that means that for the most part, there's no reason to run the Drive below  1/3, and why the "Mid" pot mostly acts as really a gain pot, with the 'drive' being your fine tuning within that gain range. 

It's an odd beast and the more I use it the more I see it is, unlike the BSIAB, designed for a very narrow range of use and my personal tastes, so if it is not well-understood internally, it would be very odd knob-turner for the average guitar player!  That being said, I'm still looking forward to hearing from anyone who has messed with or is messing with it--again, critiques are very welcome.   

[aron]

dynamic in the way that a clean tube amp set on the edge and sound clean when picking single notes normally, but just starts breaking up with accented single notes if you dig in, or play multiple notes/intervals, etc.  Absolutely no Mid hump, because besides for having humbuckers, I need it to mesh well with my Fender's clean sound, and it needs to 'cut' just right etc.

Sorry to hijack, but I have had this for years now. It's a variant of my Shaka HV. Hopefully I will have something to show in a reasonable time.

Aron

[liquids]

I tried to get this up and running weeks and weeks back, vero'd and soldered it up, but could not get my layout debugged, time after time, for weeks.  I was frozen at the first stage, sadly--I was getting a signal, but it was extremely quiet no matter what I did.  I couldn't figure it out, evening after random evening of re-looking the layout, swapping parts, continuity checking, testing with the meter, and audio probing.  Nothing came to me, and it was depressing.

I'm back at it, so I ditched that board and started from scratch.  At some point this week while I was re-doing my layout, I realized...while I thought I had read that I could run all the 'upper' jfets in mu-amp circuits off the same pair of 1Ms between power and ground (Like a Vr), on the breadboard I had always given each stage it's own pair of 1M resistors.

So I re-breadboarded it trying to run them all off the same pair...and--same problem as I had on vero.  I'm assuming that was the main issue, and I won't do that again either way.  I tweaked it while it was on the breadboard since I've learned a thing or two from other builds since, and hence have  updated the schematic to reflect what I'm now building. 

150pf does more of what I want over the volume pot and the gain control between stage 1 and 2 to preserve high end in a more natural sounding way than the 470pf cap did.  AS mentioned before, 470pf didn't sound as good at low gain settings--which sounded unnaturally bass shy rather than than simply maintaining clarity on the high end. The 150pf is much more natural.  In fact, I may go back and re-breadboard/mod my BSIAB-type pedals I use a 150pf in that spot, now that I understand the difference.   

I really get a lot of use out of the rather subtle but useful tone controls, though it would be easy to make them more drastic.  I'm seeing that sticking with the LTD Silver (or Aron's Shaka HV) sure would have been easier in some ways, but hey, what fun is there in that.   
But I find there is a wide (and now wider) range and more shades of gain via the interaction between that pot and the 'gain' control on the BMP-derivative tone stack (now called 'cleanliness' on the schematic) that wasn't available with the LTD, so it's worth it to my ears. 

So the schematic that is up now is the version I'm re-building.  I hope to have up and running soon, and then will be testing out in real-life situations on gigs and in a band mix with high hopes. I'm getting two potential useful 'core' tones out of it, so I'm even considering building a second depending on how much I end up using it and/or knob twiddling between songs.