Wampler Plexi Drive issues

[percyhornickel]

Hi, last night I built the Wampler Plexi Drive using 2N5458 instead J201 Jfets. Drein Bias Q1 4.5B, Q2 and Q3 tried from 4.5V to 6.5V

I couldn´t make it to sound good, no much gain and the effect is not present in clean channel. Just when I connect the amp gain in min is when the "plexi sound" is noted a little.


I know 2N5458 and J201 have different Vgs off but I though this could work out as the J201 is used in the RAT circuit instead 2N5458 too.

Could be this the problem? I have no any other Jfets types just some 2N5458 (and no way to get others where I live).


Thank you.

[PRR]

Are there any DC Voltage readings to review?

[percyhornickel]

Thanks for the answer, this are my readings:

Vcc 9.01 (filtered / tonepad´s power supply versión)

Q1g 0
Q1s 1.05
Q1d 4.6

Q2g 0
Q2s 1.15
Q2d 6.21

Q3g 0
Q3s 1.11
Q3d 6.15

[antonis]

Are all Source resistors of equal value..??
(1k or so..)

If yes, there is a spread on Drain current in relation with V_GS (assuming equal Drain resistors) so trim Q2 & Q3 Source resistors for respective Drains to come as close as possible to Q1 Drain voltage..
(although your particular measurement readings are quite OK for a high gain circuit..)

[percyhornickel]

Yes all source resistors are 1K

http://effectslayouts.blogspot.com/2016/08/wampler-plexi-drive.html (This is the layout I made).

All trimmers are used to set the bias from the drain, not from the source (100k instead 50K) but even with 4.5V in all drains I just get clean sound, no gain present.

I have checked the solder and pots..  ..can´t find a solution.

[bushidov]

I'm using anotherjim's schematic as a reference


The 15K resistors are the ones I've fiddle with trim pots.

[anotherjim]

Not my schematic Bushidov!
But, if you can take the JFET's out, measure the drain trimpot resistance, especially Q1 & Q3, if they are close to 1k, they have only unity gain.
It may help to get more drive at the front end, fit a cap across Q1 source resistor as Q2 has.

[antonis]

JIm is by far faster on typing.. 

[bushidov]

Not my schematic Bushidov!

okay, I got it from a post of yours, 
https://www.diystompboxes.com/smfforum/index.php?topic=115814.0

[percyhornickel]

Ok  thank you. I will check the traces with the shematic first, I can take out the jfets cause I used 3 sockets I had in here. All three source resistors are 1K.

Bushidov I even used your video yesterday to match the jfets.

Maybe I could use a pair of 10k trimmer with 4.7k to set the bias easily. Let me see what I can solve with what you guys wrote in here.


Thank you

[antonis]

BTW, with an estimated 2N5458 tranconductance of about 1000μmhos, grounded CS amp shouldn't exhibit gain greater than 15k/1k..

[PRR]

Thanks for the answer, this are my readings:

OK, nothing really wrong there.

So I build a sports-car, but have to use truck tires. It isn't as sporty as I had hoped.


Those 2N5458 are "trucks", good for brute switching but not good for fine amplification. J20x are high-strung little darlings and can make some gain.

Yes, bypass the first source resistor for a little more gain. Then add a boost in front of that if still needed.

[percyhornickel]

Ok Schrödinger's cat is alive!!!...       

After biasing the the jfets, following traces, resoldering weak points...     ...the chinesse A500K gain pot was the blame of no getting signal enough, defective shaft. It started working after pressing a little.

I still need to change the trimmers for others more accurate, with 100K is too hard to set bias properly. This ones produce some weird noises too. I hope to find other ones soon.

For tone control I´m using B25K instead A25K, I think I will solder a 4.7k resistor between lugs 1 and 2 for a logaritmic responce, imposible to get A25k pot here.

Antonis right now I am studing about Jfets (I am petroleum/drilling engineer), I have an idea about what transconductance is but not that familiar.

Well, my english is not perfect. Thank you all for the help,  I will make the part changes as soon as I could get them and then try some other bias points to test, I have read Q2 and Q3 can be biased 2/3 Vcc..   ...let´s see...


Thank you

[percyhornickel]

Thanks for the answer, this are my readings:

OK, nothing really wrong there.

So I build a sports-car, but have to use truck tires. It isn't as sporty as I had hoped.


Those 2N5458 are "trucks", good for brute switching but not good for fine amplification. J20x are high-strung little darlings and can make some gain.

Yes, bypass the first source resistor for a little more gain. Then add a boost in front of that if still needed.

Great information PRR...          ...so I just have a bunch of brute jfets, I will try with the resistor bypass trick for sure.

Thank you

[PRR]

100k trimmer is probably far too high for these truck-size FETs. In fact with test voltages we can calculate a happy resistor and use a fixed resistor within 10% of these values.

Q1s 1.05
Q1d 4.6
use Rd = 4.2k

Q2s 1.15
Q2d 6.21
use Rd = 2.5k to 3.9k

Q3s 1.11
Q3d 6.15
use Rd = 2.5k to 3.9k

[percyhornickel]

Ok the last value I was setting the drain for Q1, Q2, Q3 is 4.5V

I will mount the spice model and check what values for Rd I can use instead. I have in here some 3.3K and 4.7K, I hope to get good bias points with these.

[antonis]

I will mount the spice model and check what values for Rd I can use instead.

I'm not aware of any person playing music via spice..

[Vivek]

I will mount the spice model and check what values for Rd I can use instead.

FET's have too much range of their important parameters.

A Spice model normally just roughly models some average value

The actual FET in your hand could act very differently than the FET model.

I measured and made 23 different SPICE models of the same FET part number, based on the actual physical FETs in my hand.


https://www.diystompboxes.com/smfforum/index.php?topic=126702.0

[percyhornickel]

I will mount the spice model and check what values for Rd I can use instead.

I'm not aware of any person playing music via spice..

Antonis just two words......            ...Spice Girls   

Vivek you are right, Vds current/voltages change a lot from jfets spice models to the real one. 2N5458 Vgs in spice is 2.8V and this parameter can vary  a lot in a bunch of same type of jfets..   ..even if they are matched / regular way.

I use Lt Spice to simulate and listen (.wave / .tran statements) the way it´s affected any circuit if I made some of changes..  ..changing caps/resistors values, changing voltages, etc..     It is not exactly what you will get in the end but gives you an idea you are in the right path to achieve what you are looking for. 

Right now I am simulating the behavior of a 50 log pot with 47k resistor between lugs 1 and 3, I don´t have 25K log for plexi tone control so I am trying alternatives with what I have in here..     ...I live in Venezuela so I have to use a coconut to lit a fire if there´s a posibility to.   

 

[percyhornickel]

This thing sounds really good, I changed the Q1 trim for 4.2K (2K+2.2K resistors) = 4.5V drain!!!....      ...Q2/Q3 = 4.5K (still trimmers)..   ...Soon I will change them for the right resistors.

For tone A25k pot I used A50K with 47K resistor lug 1/lug3, works pretty well. I just have on off on switch instead on on so..    ..maybe later I will try to add an aditional cap to the selector.

No weird noises at all