Box of Rock, only dirty: headroom problem

[menstrel]

Hi everyone,

i made the Box of Rock only dirty for a small pedal board, following the pcb described here

http://effectslayouts.blogspot.com/2015/01/zvex-box-of-rock.html

I already own the original one purchased a few years ago, including the boost. In this one I used 2N7000 transistors instead of BS170s because the ones I had didn't fit, and all the other components are the same as in the diagram above but I find it strange that it has little headroom, in fact the pedal is already a bit distorted with gain at zero. Could it be due to the 2N7000s?

How can I intervene to have more headroom while always remaining with 9V ?

thx !

[GibsonGM]

Can you post pics of the front & back of the board.

Yes, it could be the 2N7000s. I use them in place of 170's all the time. You have to increase the drain resistor with 7000s for them to bias 'the same' as the 170.  On an AMZ mosfet boost, I have to increase the 2.7k drain R to 22k!   A trimmer could be used to do it by ear - I typically get it in the ballpark and trim is for best sound or waveform on the o-scope.  If you're lucky the bias won't have to be messed with much, but we can cross that bridge after (tweaking).  Drain should sit at about 4.5 to 5.5v.

As you've noticed, it sounds crap if you don't do this modification for having used a different component.   

Since we apparently don't get schematics from layout sites (radically hindering the ability of others to help), here it is so you can see what I mean.  I am talking about the R5/R19 combo, needing to add up to about 22k.  A pic of your board would let us identify which they are in your build so you can do this.

PS - before digging in, be SURE the 2N7000 pinouts are not different than the BS170's! 

[duck_arse]

and ...... welcome to the forum.

Can you post pics of the front & back of the board.

Yes, it could be the 2N7000s. I use them in place of 170's all the time. You have to increase the drain resistor with 7000s for them to bias 'the same' as the 170.  On an AMZ mosfet boost, I have to increase the 2.7k drain R to 22k!

.....

PS - before digging in, be SURE the 2N7000 pinouts are not different than the BS170's! 

the 2N7000 will almost certainly be reverse pinout of BS170. I didn't know about the resistor changing business. if I were doing this, I'd just rebias the gate voltage on the breadboard for best swing, and then go nearest fixed values.



ooohhh, post count!

[GibsonGM]

Don't recall why I needed to increase the drain R, Stephen...perhaps it wouldn't bias any other way or something - it was years ago and in my 'notes' on the LT Spice ckt I mocked up at that time.   I'm sure I read it somewhere.

[marcelomd]

As others have said, 2N7000 and BS170 have different pinouts. Check this.

Re drain resistor. I've built a few preamps based on the Box of Rock. I've used BS170, 2N7000 and 2N7002. One of them had sockets to test the difference between transistors. I don't have golden ears, but I've heard no difference.

[GibsonGM]

I've always had to adjust the bias and/or drain resistor when using other transistors - maybe it's just me...(?) 

Jack Orman notes this here:  [url]http://www.muzique.com/schem/mosfet.htm[url]

Maybe my specific 7000s are off or something, wouldn't bias without increasing Rd.  Or I didn't mess with "R2", but rather the drain resistor.  Can't recall.

Only point being - maybe the thing needs to be adjusted for the 2N7000s to sound good. 

[marcelomd]

I've always had to adjust the bias and/or drain resistor when using other transistors - maybe it's just me...(?) 

Jack Orman notes this here:  [url]http://www.muzique.com/schem/mosfet.htm[url]

Maybe my specific 7000s are off or something, wouldn't bias without increasing Rd.  Or I didn't mess with "R2", but rather the drain resistor.  Can't recall.

Only point being - maybe the thing needs to be adjusted for the 2N7000s to sound good.

Note that mr Orman's MOSFET booster linked is biased with a voltage divider (R1, R2, R3). SHO and Box of Rock use a self bias method: One resistor from drain to gate and one from gate to ground. See mr Arse's schematic above.

[GibsonGM]

One might suspect then, that OP has the transistor in wrong (pinout).

It may be naive of me, but honestly - how does applying a bias voltage from a resistor network like AMZ differ from the divider formed by the 2 1 Meg resistors of the BOR?

If anything, I would 'instinctively' (and perhaps naively) assume that the former would hold the bias voltage more constant, since it has the bypass cap and is composed of lower value resistors (thus presents a more 'stiff' voltage to the bias resistor).   There is negligible current draw on a mosfet gate, after all.   

Why would changing to a '7000 alter the bias in the AMZ version and not the BOR, if not for some intrinsic property of the mosfet model?   (not an engineer)

Really, if the quiescent current of the device changes due to swapping the part, shouldn't both methods of bias 'adapt' to this?

[marcelomd]

I have no idea why 2N7000 and BS170 work differently in the MOSFET booster.

I know they do work the same on the BOR-based boards I have with me.

The first stage of the BOR changes gain by changing the source resistor directly, which rebiases the transistor. Measuring one of my units here at min and max gain I get:
Gate: 1.74V - 3.08V
Source: 2.44V - 0V
Drain: 8.3V - 3.66V
I don't think a fixed bias would work here.

The MOSFET booster has a big decoupling capacitor which "separates" the bias-setting effect of the source resistor from the AC gain set by the potentiometer.

I think the other fixed gain stages would work the same with either biasing schemes.

I don't remember all the theory, so I'll stick with what I can measure here.

[GibsonGM]

If I didn't have 5 things on breadboard needing to be built I'd just make one, using both methods, and examine why one alters bias and one appears not to.

Probably has to do with the source resistor as you pointed out, a feedback effect (triodes exhibit this...).  The bypass cap is probably why it's different with AMZ boost.  But that's only conjecture.

EDIT: Ya, I need new glasses....I took the top 1Meg of the divider in BOR as coming from 9V rather than the drain, LMAO...sorry.   In that configuration it is "self-adjusting", where if fed only from a divider like AMZ it's not, so it is less finicky of device differences.  Sorry 'bout that.   5 minutes in LT Spice confirms this.

[menstrel]

Wow, I didn't expect to find all these answers. Thank you !

Well, for those who told me to check the pinout of the 2N7000, don't worry, I rotated it 180°, I already knew this. In the meantime I checked the welds and did another A/B comparison with the original BOR and now they sound the same except for the tone, mine is better, it ranges up to being brighter than the BOR. I can't explain this though since the components are identical.

So the headroom problem solved itself, but even the original BOR with zero gain has a very slight crunchy.

A little while ago after checking it for the last time, I tried it on the pedalboard and after almost half an hour of fun, it suddenly started croaking by itself, producing almost continuous electric discharge noises. This is the last problem I need to solve.

I wanted to point out that the scheme posted at the beginning by GibsonGM is the correct one but in all the potentiometers you have to invert lug 1 with lug 3 otherwise it won't work.

Attached are the photos of the pcb

https://ibb.co/y8CGcVv
https://ibb.co/HzQ4Z3R
https://ibb.co/NNhNSjN

[GibsonGM]

The 'must invert pots' is a little troubling to me, menstrel...it should work either way - perhaps in an odd, crappy way because of the pot taper but still...do you find there is only a small 'window' the gain works at, and the rest is silence or something?

Might be brighter just due to component tolerances, shorter wire runs inside the box, things like that.  Some copies are more susceptible to those things than others

[antonis]

Might be brighter just due to component tolerances, shorter wire runs inside the box, things like that. 

"that" including interelectrode capacitances as well as dynamic characteristics..

[marcelomd]

In the posted schematic the potentiometers are indeed backards.

For the gain potentiometer, less source resistance == more gain. So turning clockwise should move pin 2, ground, to the source pin.

For the tone potentiometer, we usually want more treble when turning clockwise. This should move the pin 2 to the high pass side.

[GibsonGM]

otherwise it won't work.

The inverse-log taper would be bunched. But the circuit should 'work', just not how you want the controls to act.   The schematic (not mine) should indicate the need to reverse the lugs.

[marcelomd]

otherwise it won't work.

The inverse-log taper would be bunched. But the circuit should 'work', just not how you want the controls to act.   The schematic (not mine) should indicate the need to reverse the lugs.

Fair enough.

[GibsonGM]

No argument here   I just wasn't clear from what the OP said is all.  And I messed up by not noting how the thing is actually biased lol

[menstrel]

The 'must invert pots' is a little troubling to me, menstrel...it should work either way - perhaps in an odd, crappy way because of the pot taper but still...do you find there is only a small 'window' the gain works at, and the rest is silence or something?

Might be brighter just due to component tolerances, shorter wire runs inside the box, things like that.  Some copies are more susceptible to those things than others

I know that by reversing the lugs of a volume pot, in the clockwise direction the volume decreases instead of increasing and vice versa.

But by my "it won't work"...I meant that it sounded awful, i.e. very low volume sound, weak distortion, RF noise with gain rotation and continuous crackling, similar to Crackle Ok.

I don't remember how and where, but I had found that the tone lugs of the pot were reversed. As soon as I swap lug 1 and lug 3 of the tone, I remember it worked fine, then suddenly died. The soldering seemed ok and after a lot of stress I redid the pcb and checked the BOR wiring, noting that in fact the tone lugs 1 and 3 were reversed with respect to the diagram but also those of the gain and volume. In short, now it works but last night it started making crackles, equal to the "Crackle Ok"

[GibsonGM]

Seems that you've achieved success, then!  I think we can leave it there, ha ha.   

As you see, there is no cap between the gain pot and source of the transistor (and there shouldn't be), and it goes to ground. There is a DC voltage present there. Usually it doesn't matter, as it would be a fixed resistor, no changes, so no noise.  As you adjust the resistance, you're changing the DC level and the wiper of the pot, which isn't perfect, 'jumps around' and you hear the discharge of the DC as static.  Like early Fender amp presence controls.
It's a 'feature', ha ha, which is why they label it "Crackle OK". 

[menstrel]

Seems that you've achieved success, then!  I think we can leave it there, ha ha.

  Yes, I know how "Crackle Ok" works, the bias is changed hot. Surely I didn't know how to explain, excuse my english, but I wanted to say that the noise is "similar" to the "Crackle Ok" that you get with this pedal on the gain and boost pots, but the crackle that I hear occurs without turning the gain pot and it suddenly comes on without me touching anything, then it goes away on its own and after a while it comes back up again, and so on.