Box of rock clone, volume problem

[ricci]

Hi everyone,

I built the box of rock pedal in the version without boost and therefore only the distortion, following the scheme
I compared it with another clone and with the same components, the clone sounds a little louder and with a little more of gain and also slightly fatter. Could this difference depend on the quality of the components?

[Helmholtz]

I don't think "quality" of passive components can make much difference.
Caps typically have a tolerance of 10% which might add up to a slight sound difference.

Was the comparison done using the exact same equipment (guitar, cable, amp, speaker)?

[ricci]

Was the comparison done using the exact same equipment (guitar, cable, amp, speaker)?

yes.

Here is an audio sample, you hear first the clone and then the best performing one. The controls ( level, tone and drive) are at maximum.

https://vocaroo.com/15aAJC6v8TuJ

The second one has more volume, brighter and more distortion.

[Helmholtz]

O.k., second one def. has more gain.

Start with comparing Mosfet voltages.

[ricci]

O.k., second one def. has more gain.

Start with comparing Mosfet voltages.

First the values with 2N7000 and in brackets those with BS170


-----------------Vd-----------------Vg------------------Vs
stage I 2.99 ( 4.27)-------1.48(2.13)--------0.012 (0.002)
stage II 3.15 (4.61)-------1.57(2.30)--------0.079 (0.080)
stage III 3.47 (4.87)-------1.73 (2.43)-------0.250 (0.258)

Readings with BS170 are very similar to the values of the best performing model, in which the BS170 mosfets are present. I put the 2N7000 because they seemed less noisy to me.
The volume and gain gap is the same with the Fairchild BS170s too.

[Helmholtz]

Do you have a scope and sig. gen.?

[ricci]

Do you have a scope and sig. gen.?

Yes I have a 20Mhz and I can generate 1khz

[Mark Hammer]

NEVER underestimate the impact that component tolerances can have on the resulting sound. When a critical cap in the signal path could be +10% or -10% from its stated/nominal value, that can have a big impact. Gains of transistors can vary widely, and forward voltage of diodes can as well. Thankfully, 1% resistors are inexpensive and widely available these days, but resistors aren't the whole story, when it comes to achieving consistency in a product.

There was an interview with Mike Matthews in a business magazine in the early 2000s in which he stated that, during the 1970s, you could take any four consecutively built Big Muff Pi pedals off the assembly line and they would all sound different from each other. The design hadn't changed, but Mike would get his capacitors off available sources in the NYC area, some of which I imagine may have been surplus.

This is why testimonials, about this issue having such and such a property that this other issue of the same pedal lacks, rankle me. If a person had 10 randomly selected pedals of each issue, and there was a consistently audible difference between the two issues, irrespective of which individual pedals one tried out, THEN we're talking about issue-to-issue differences that are real and likely related to either design or component choices. But if it's one pedal against one other pedal, with no other point of reference, than we could be comparing two versions whose design or componentry is different in some substantive way, OR we could be comparing two identical circuits whose component values are different but are within tolerances for those particular parts and their nominal value.

[ricci]

NEVER underestimate the impact that component tolerances can have on the resulting sound.

Yes, I had thought about the tolerances of the capacitors which are around 8%, the resistors are more precise at 2% with metal film.

[g1]

You checked tolerance on resistors and caps, but how about the pots? If they are not precision type they can have quite wide variations.

[ricci]

NEVER underestimate the impact that component tolerances...

You made me curious and I spent this time looking for capacitors with values as close as possible to those in the diagram, but also stable in repeated measurements, I found them and I put the two 100nf and the three 22nf but the gap remains with both 2n7000 and bs170. In the high-performance PCB the traces are very thin, perhaps the geometry in general could also influence it ?

Anyway, the gap is only volume because the grain of the distortion is identical. Keeping all the controls at maximum, to have the same volume as the performance one, I have to set the volume of the latter at 3 o'clock.

[ricci]

You checked tolerance on resistors and caps, but how about the pots? If they are not precision type they can have quite wide variations.

I only know that the pots are identical to those of the original Zvex pedal, all Alpha brands and identical also in values, of course.

[Helmholtz]

Mosfet voltages look ok.
2N7000. and BS170 seem to have somewhat different transconductance.
But as the circuit uses lots of NFB, the AC gain will be close.

With the scope you can actually see gain.
Use the same settings as with the sound comparison.
Feed a 400Hz/10mV sine signal to the input and scope the output and check if it's undistorted.
If it's distorted lower the input signal until the output is clean.
It's essential that the output is clean as gain is only defined without distortion.

Now,
1) read the peak to peak voltage (Vpp) of the input signal.,
2) read Vpp at the drain of the first stage,
3) do the same at the drains of the second and third gain stage.

Dividing drain Vpp by input Vpp gives gain.

[ricci]

For my scope 10mv is too low a sensitivity, I start from 100mV.

If I adjust the scope well, I get
Vpp= 2V as input
6v i stage,
7v ii stage and
6.7v iii stage

[Helmholtz]

Is the signal at the third drain still an un-squashed natural sine?