Author Topic: Plexi-Drive JFET misery  (Read 6700 times)

lars-musik

Plexi-Drive JFET misery
« on: October 01, 2016, 04:59:50 PM »
This question comes up from time to time and now it's my turn. I am still on to a Wampler Plexi-Drive clone. I know that Brian's given some information about appropriate JFETs on the "other forum" stating that the drain resistor of 15K is essential for the sound: "You MUST sort through jfets to get a bias as close to 4.5 - 5v max as possible. This will allow to keep the drain resistors the same."

I will go with a trimmer at the drain anyway but in my bag of 150 J201 (I tested about 40 in my test rig  - drain 15K to 9V, gate 1M to ground, source 1K to ground -) none hat a drain voltage of over 2V. Most ranged between 1.2 and 1.8V which is nowhere close to 4.5V.
The assortment comes from 3 different suppliers, so it's unlikely they are all one "bad" charge.

Given that Brian states you'll get the real Plexi-Drive sound only by a 15K resistor to the drain of a JFET that results in a drain voltage of approx. 4.5V would it be correct to assume that if I'd charge pump the input voltage of the whole circuit to a value that allows for the JFETs to reach that voltage I'd be good?  Would that  be essentially the same as lowering the drain resistance to a point where the drain voltage is biased correctly? 

I can't really understand how biasing a JFET wouldn't work but as my JFET test voltages are so off I wonder if I could get them to sound good anyway. The first build sound too low in volume and gain for a start.

PRR

Re: Plexi-Drive JFET misery
« Reply #1 on: October 01, 2016, 11:07:38 PM »
The simple logical thing is to adjust the Source resistor to get the Drain where you want it.

If this Source resistor is some Magic Value, then you "DO" have to find and select JFETs for a suitable combination of Idss and Vto. (Which IMHO makes it a "bad design" for mass production, and maybe for casual DIY construction.)

Gus

Re: Plexi-Drive JFET misery
« Reply #2 on: October 02, 2016, 06:39:01 AM »
Have you checked the IDSS etc.?
A search of this forum or the web should find easy to make jfet testers.

lars-musik

Re: Plexi-Drive JFET misery
« Reply #3 on: October 02, 2016, 07:59:23 AM »
I built a Phase 90 and built a Jfet Tester for that. I was under the impression that for the Wampler Jfet builds the drain voltage was the value that counts. My knowledge about transistors is less than sketchy. How do I translate the measured values into usability for a Plexi Drive?

Gesendet von meinem GT-I9505G mit Tapatalk


GibsonGM

  • more
  • Poster2
  • *****
  • Posts: 5466
  • Total likes: 671
  • Mike Parker (aka. Guitar Mike, Parks, etc)
Re: Plexi-Drive JFET misery
« Reply #4 on: October 02, 2016, 08:38:15 AM »
Would would such a matching process be required, rather than adjusting source and drain R values to get you to the bias point you want which will cause the type of operation desired?    Anyone know this circuit well enough to hazard an answer?  I have assumed you match/measure for consistency between 2 or more devices, or to know where to bias your device...

I've always just tweeked them for optimal headroom, and have been very pleased with the results...but there COULD be a real reason for this matchy matchy thing, I dunno...
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

Gus

Re: Plexi-Drive JFET misery
« Reply #5 on: October 02, 2016, 10:39:52 AM »
If the 15k drain resistance is needed you could use a trim potentiometer in the source leg to set the drain voltage. then remove the trim measure it and install a fixed resistor(s) JFET microphones often at set this way
However the source voltage might/will be different than the real circuit and the gain will be wrong some of the gain is 15k / 1k along with the open loop gain of the jfet
I think you are in luck you should bias up with >1k source you posted you bias around 2VDC so you need to make the gate more negative to the source. Then you can add a resistor and series cap in parallel to equal about 1K at AC frequencies. You will need to size the cap for the needed resistor

The IDSS of a J201 should be .2ma to 1ma
This is why I asked if you measured it as a check if they are real J201s and not counterfeit
Measure the ones you have and sort them
Install the highest IDSS with the stock resistor values
Install the lowest IDSS with the stock resistor values
Note the way the drain voltage changes

anotherjim

Re: Plexi-Drive JFET misery
« Reply #6 on: October 02, 2016, 02:08:38 PM »

Is this it?

You have 1k source resistors. Raise R3 experimentally. Sound may be ok if drain reaches at least 3V. Use that value to replace the others. Some loss of gain. New R3 can have a bypass cap 47u (but 22u probably ok) WITH a resistor in series  - this resistor should make the parallel value with R3 nearabouts 1k. Ditto with R10 if you want.
R6 already is bypassed capped & looks like it's meant to have most gain - so no series resistor in that cap.

"Designing with PLLs in production circuits is reserved to those with the courage of the thoroughly uninformed."
Nuts & Volts
"Turtles have the short legs, not for the walking."
-Kenji "Damo" Suzuki

Pojo

Re: Plexi-Drive JFET misery
« Reply #7 on: October 02, 2016, 04:35:35 PM »
Where did you get your j201s? Having built a dozen or so pedals using a design based on the plexidrive structure, I can say that the 15k drain and 1k source resistors seem to be good for most j201s in my inventory to get the drain voltage between 4.5v and 5v. If most of yours are that far off I would suspect you may have gotten fakes or a relabeled batch. Which unfortunately enough is common among Chinese eBay vendors and some other sources.

Adjusting either the source or drain resistances to set the bias point will change the gain and response of the stage and is not what you want to if you want as close to the intended sound as possible. The only way is to find another source of j201s. If you get a couple dozen or so from a reputable dealer (like Steve from small bear) you should have no problem finding matches.

As wisely mentioned by others, it's also good to get acquainted with measuring the Idss and Vp of jfets. These values determine how much current the drains conduct and how much signal the gate will take before clipping. And you should find that all the j201s that work in the plexidrive fall within a certain range of Idss.

J0K3RX

Re: Plexi-Drive JFET misery
« Reply #8 on: October 02, 2016, 05:42:08 PM »


I will go with a trimmer at the drain anyway but in my bag of 150 J201 (I tested about 40 in my test rig  - drain 15K to 9V, gate 1M to ground, source 1K to ground -) none hat a drain voltage of over 2V. Most ranged between 1.2 and 1.8V which is nowhere close to 4.5V.
The assortment comes from 3 different suppliers, so it's unlikely they are all one "bad" charge.

Given that Brian states you'll get the real Plexi-Drive sound only by a 15K resistor to the drain of a JFET that results in a drain voltage of approx. 4.5V would it be correct to assume that if I'd charge pump the input voltage of the whole circuit to a value that allows for the JFETs to reach that voltage I'd be good?  Would that  be essentially the same as lowering the drain resistance to a point where the drain voltage is biased correctly? 

I can't really understand how biasing a JFET wouldn't work but as my JFET test voltages are so off I wonder if I could get them to sound good anyway. The first build sound too low in volume and gain for a start.

Are you saying that when running the plexi drive circuit at 9v using the correct resistor values in the schematic for drain, source and gate you get a drain voltage of 1.2v to 1.8v? What is your "test rig"? That really doesn't make sense, to me anyway..? Do you mean that you had to drop the drain voltage down to 1.2v to 1.8v using a trim pot for them to sound as they should?

Running the circuit at higher voltage will not help.. If they will not bias with the given values at 9v then they will not bias with those values a higher voltage either. They are biased at 1/2 the supply voltage...
« Last Edit: October 02, 2016, 05:45:19 PM by J0K3RX »
Doesn't matter what you did to get it... If it sounds good, then it is good!

lars-musik

Re: Plexi-Drive JFET misery
« Reply #9 on: October 02, 2016, 06:16:54 PM »

Are you saying that when running the plexi drive circuit at 9v using the correct resistor values in the schematic for drain, source and gate you get a drain voltage of 1.2v to 1.8v? What is your "test rig"?

My test rig is:
(I tested about 40 in my test rig  - drain 15K to 9V, gate 1M to ground, source 1K to ground -)

Is this completey off? The schematic posted by anotherjim is essentially what I used.

J0K3RX

Re: Plexi-Drive JFET misery
« Reply #10 on: October 02, 2016, 09:51:57 PM »
If your J201's are that far off then I would tend to believe (as Gus already mentioned) that they may be counterfeit :icon_sad: Kinda hard to believe that 3 different suppliers sent you the same jacked up jfet's but perhaps they all got them from the same supplier.. who knows?

If they are counterfeit then you will just be beating your head on a brick wall trying to get them to comply to this design...

Get some authentic ones from a supplier that is legit and keep some for comparison.. If you don't have a real problem with SMD you might consider getting some MMBFJ201 and then some of the small SOT-23 adapter boards. I have found that the MMBFJ201 are nearly all identical  and right on the money with the spec sheet.. 
Doesn't matter what you did to get it... If it sounds good, then it is good!

induction

Re: Plexi-Drive JFET misery
« Reply #11 on: October 03, 2016, 02:47:30 PM »
Having built this and several other Wampler pedals, I wouldn't say that your J201's are counterfeit just because they don't bias correctly in this circuit. Note that any J201 that biases correctly in the Plexi-Drive circuit will not bias correctly in the Black '65, the Tweed '57, etc. The tolerances on the specs for these jfets is notoriously wide, and Brian builds his circuits around certain fairly tight spec bins. Yes, you generally have to buy thousands of them and then sort them for the different circuits (and then use 1/2 to 2/3 of them for something else entirely). I don't know that using SMD J201's will help, since your J201's can be perfectly within spec and still not work in this circuit. I personally bought 200 or so J201's from several different suppliers. All of them tested within spec, but even so, I just barely managed to scrape together enough usable J201's to build one of each of the above three circuits.

You can certainly adjust the bias via the resistors, or you can choose a different bias voltage, but both of these things will alter the sound, pretty noticeably in my experience. The most obvious result will be that the gain of the individual stages will be wrong, thus the distortion and EQ characteristics of the circuit as a whole will change. Bottom line, if you are willing to tinker and happen to like what you end up with, you're golden, but if you want a clone of the commercial pedals, you'll have to sort jfets.

I appreciate Paul's comment that this makes it a 'bad design' by standard EE criteria, but such is the history of guitar amplification. I also agree that this J201-sorting business makes many of the Wampler circuits somewhat DIY-unfriendly. I wonder whether that works to Wampler's advantage or not. On the one hand, it may convince more people to buy them rather than build them (I've seen enough clones and read enough "my Wampler circuit won't bias" threads to be skeptical). On the other hand, if a potential buyer hears a bad-sounding clone, that may negatively influence his decision to buy. I'd guess that the impact of DIY on Wampler's sales is negligible, but that's just conjecture. I don't have any statistics one way or the other.

lars-musik

Re: Plexi-Drive JFET misery
« Reply #12 on: October 03, 2016, 03:39:41 PM »
Thank you all for your input.

After all what I have learned here I wonder how all the ROG designs are working around the variabilty of J201 or jfets in general. I recall they mostly implement bias potentiometers but if this heavily impacts the sound how's there "a" Dr. Boogie" for example?

Anyway I just saw that I can get 250 mmbfj201 at Farnell's for .17 EUR each. Maybe I'll give them a try.







induction

Re: Plexi-Drive JFET misery
« Reply #13 on: October 03, 2016, 03:55:40 PM »
I don't think you can generalize too much from the Wampler pedals. Brian designs his pedals to be highly sensitive to J201 specs, as he obsessively aims for a very specific sound from each pedal. His J201 designs seem to require the jfets to sit at the extreme edge of their tolerances. Not all jfet circuits are like that. In fact, as Paul mentioned, good circuit design tends to minimize the effect of variation between different devices, so usually you can just plug and play. The Wampler J201 circuits (as well as the Fuzz Face family of circuits) are pretty unusual in that regard.

Furthermore, ROG, like this forum, is a firmly DIY endeavor. Not only are they focused on reproducible DIY designs, they don't make commercial pedals, as far as I know. So you can't compare the sound of your circuit to a stock unit and see whether it's different. Wampler, on the other hand, is focused on selling commercial pedals that capture a specific tone, not on DIY-friendliness. Few other pedal companies that sell in large quantities want to spend the extra money to sort through tons of jfets or fine-tune each individual pedal. It's cheaper for them to design for plug-and-play as well.
« Last Edit: October 03, 2016, 04:02:13 PM by induction »

Gus

Re: Plexi-Drive JFET misery
« Reply #14 on: October 04, 2016, 05:14:40 PM »
Try what I posted and anotherjim posted

Bias with the source resistor then calculate what value resistor in parallel will equal about 1K.
Use a large enough value cap in series so the RC is 10 hertz or lower
Add the resistor in series with the cap in parallel with the source resistor that gives the correct drain voltage

Not exactly the same but it should be close enough.
For Q3 you can adjust the voltage divider (R7 and R8) for close to the same input sensitivity.

lars-musik

Re: Plexi-Drive JFET misery
« Reply #15 on: October 07, 2016, 06:17:32 AM »
I can't leave this alone. So I ordered 250 MMBFJ201 SOT-23 transistors and until they arrive I'll build myself this (hopefully working) Wampler-Jfet-Matcher. Per 3PDT (there might have been more elegant ways to switch this but as I am no EE I was afraid of interference if both circuits were connected...) I can choose between RG's Improved JFET Matcher (DMM2) and the Wampler Bias Finder (DMM1). The rotary switch dials in the drain resistors: Black '65: 5.6K; Cranked AC: 10K; Plexi-Drive and Cranked AC : 15K; Tweed '57: 16K; Pinnacle: 22K.
Let's see if this works out.



maoriente

Re: Plexi-Drive JFET misery
« Reply #16 on: October 12, 2016, 09:00:07 AM »
I too built a Wampler Bias Checker and had sorted through the rest of the 100 smd J201s I had purchased from Mouser.

Not a single one would read between 4.5-5v with anything less than 22k drain resistor. Although I did have quite a few fall in range with the 22k.

I was looking over the Tweed 57 schematic and noticed the 1st J201 is biased with 16k and then the next 2 parallel pairs have a single 22k drain resistor.

Question: How does paralleling the J201s with a single drain resistor affect the gain/tone etc vs a single J201? Could this be a viable workaround for the Plexi? Higher drain resistor with parallel J201s?

lars-musik

Re: Plexi-Drive JFET misery
« Reply #17 on: October 12, 2016, 10:01:08 AM »
Welcome to the jfet mourners club!

I am sitting in my basement at the moment (again) with a magnifier and sorted through my (re-)stock. All (ALL!) of my TO92 J201 do not work. None biases over 1.5 V although I bought new ones (20 each) from two trusted sources here in Germany. So my old stock (don't know when and where I got them) about 30pcs, 40 from trusted sources and 150 cheapos from a possibly dodgy chinese ebay seller behave basically the same. With the highest consistency in the chinese stock as far as they all bias the lowest.

Parallel, I ordered 250 MMBFJ201 at Farnell. They ALL bias between 3.9 and 5.4 V. Enough to make the pedal - finally.

So I can only repeat J0K3RX's advice: Take the SMD version!!!!!

gght

Re: Plexi-Drive JFET misery
« Reply #18 on: October 13, 2016, 12:07:34 AM »
I wonder if there is a simple explanation for the SMD jfets being good, yet so many thru hole are turning up bad?  Were there bins of rejects
that weren't disposed of? And are the SMDs manufactured now? And if so, why are they successful, or are the SMDs old stock, and just more plentiful?  So many questions. 

Gus

Re: Plexi-Drive JFET misery
« Reply #19 on: October 13, 2016, 04:34:58 AM »
 lars-musik
Did you test the IDSS of the leaded JFETs?  The 201 should be 1ma or lower.
https://www.fairchildsemi.com/datasheets/J2/J201.pdf