Something I've been thinking about recently I've just built 2 KOT clones for myself and a friend one on pcb and one on vero both builds are using the same brand and part values though admittedly I didn't meter every part before installing them. The pcb pedal sounds really good and just like the last KOT I made myself the vero, however, sounds absolutely fantastic. I've spent ages over the last few nights going back to back and cannot tell the difference between the two pcb pedals when they've been dialed in whereas the vero build both me and my friend could pick it out everytime regardless of which of us was playing. I get that components have tolerances I've used metal film resistors 1% tolerance, 5% film caps and NPO ceramics in all of them from the same order could the individual tolerances add up to one pedal being spectacular compared with the others?
If that is the case then how often have you had a 'magic' pedal?
Whoever can answer this will be a millionaire. Seriously.
9mm/16mm pots are +/- 20%
Having a very good dual gang pot is supposed to be a big factor I that pedal.
The vero may have been sitting in a bin in the back of some warehouse for decades. In that time, the copper has been allowed to age. The electrons have "sweetened", meaning that they slide past each other in a very fluid way. The electrons in your PCB traces are still "green." They need to be broken in. It's like they have rough edges that still need to be knocked off. Play some frequency sweeps through it for a few days and you will be amazed!
OK, OK. It would be interesting to measure the resistance on both sides of the pot wipers to verify that they are truly set the same. 20% pot tolerance factored with the knob placement on the shaft can easily account for significant differences. A couple dB of volume difference will seriously fool you every time. Gain matching is critical. To a human, louder = better.
Quote from: Leevibe on June 21, 2016, 02:10:09 PM
The vero may have been sitting in a bin in the back of some warehouse for decades. In that time, the copper has been allowed to age. The electrons have "sweetened", meaning that they slide past each other in a very fluid way. The electrons in your PCB traces are still "green." They need to be broken in. It's like they have rough edges that still need to be knocked off. Play some frequency sweeps through it for a few days and you will be amazed!
It's all about those electrons.
Quote from: madbean on June 21, 2016, 01:04:12 PM
Whoever can answer this will be a millionaire. Seriously.
I assumed that would be the correct answer.
Quote from: Leevibe on June 21, 2016, 02:10:09 PM
The vero may have been sitting in a bin in the back of some warehouse for decades. In that time, the copper has been allowed to age. The electrons have "sweetened", meaning that they slide past each other in a very fluid way. The electrons in your PCB traces are still "green." They need to be broken in. It's like they have rough edges that still need to be knocked off. Play some frequency sweeps through it for a few days and you will be amazed!
OK, OK. It would be interesting to measure the resistance on both sides of the pot wipers to verify that they are truly set the same. 20% pot tolerance factored with the knob placement on the shaft can easily account for significant differences. A couple dB of volume difference will seriously fool you every time. Gain matching is critical. To a human, louder = better.
We were very careful to not just set the controls in the same places but to try and match the sound from one pedal regardless of where the pots were set. That's why I said the 2 pcb built pedals were indistinguishable from each other when set up this way. The Vero pedal just sounds better no matter where the knobs were set or even if it was quieter than the others. I might try measuring the pot resistance to set them to the same values for a laugh.
I was wearing my lucky purple shirt when I soldered the Vero board so maybe that added some extra mojo!
Quote from: Leevibe on June 21, 2016, 02:10:09 PM
To a human, louder = better.
(http://www.interfaces.com/blog/wp-content/uploads/2013/05/spinal-tap-poster.jpg)
Potentially the vero has been infused with unicorn tears
Seriously though, Bean nailed it :)
This has happened to me many times. Built several big muffs and one stands out as magical. Same parts, same everything. One sounds sweet, the others harsh. I've gone over it many times and can't find a difference.
You throw in 1-5%, 20% tolerances over 30+ parts and viola, you get the same variation you find in biological species. There are always those outliers that sound amazing (or awful, on the other end of the spectrum).
And the lucky shirt makes a difference.
Because bean has some "STANK RIFFS"!!!
-Mike-
you think it happens with pedals a lot, imagine how it is with analog synths.
that's the beauty of analog
Quote from: DPTX on June 21, 2016, 07:33:20 PM
This has happened to me many times. Built several big muffs and one stands out as magical. Same parts, same everything. One sounds sweet, the others harsh. I've gone over it many times and can't find a difference.
I've had this happen with a Zen Drive on vero. It's maddening to re-create it.
Happened to me a couple of times, and always vero sounds better. It's a bitch to build on vero in comparison to pcb but there is something about it, seriously. I've heard some stories that vero being what it is can produce parasitic capacitance or something like that which influences the sound. A dude that's designing pcbs for living for a huge company/factory told me that. Dunno.
Best muffs (pharaoh) I've made have been on vero.
the tolerance is more believable than parasitic capacitance or induction. I say this because the parasitic capacitance between tracks is going to filter at best ultrasonic frequencies in the MHz range. if we were building a radio transmitter, it would be a problem.
Quote from: Cortexturizer on June 21, 2016, 09:49:20 PM
It's a bitch to build on vero in comparison to pcb but there is something about it, seriously.
Vero isn't to bad, so long as the parts count isn't to high. I hadn't built on vero for over a year until my last two builds. Going back I was actually surprised at how easy it was to work with. It is certainly more convent than pcbs, because I don't have to wait for a board in the mail. The off board wiring can get a little tiring though.
Moral of the story, get some parts and build another one on veroboard and see if it has that magic mojo.
Layout can make a difference in high gain, or circuits with clock signals.
I'd like to know whether the vero layout worked from the same schematic and values as the pcb, if it sounds better :)
Interesting, i wonder if its down to the trace widths against the track width on vero, not sure about the copper weight and quality difference between the two but i had always assumed (most likely wrongly) that the sort of currents we're dealing with in pedals, this shouldn't be an issue :-\
I think layout does play a role, especially in high gain. My vero board stuff looks much different than a PCB layout. I know when I build amps layout plays a huge part.
I usually use vero for everything (except my phase 90 clone, that's a PCB), and I always change the layout between builds for various reasons.
I'm guessing it's parts tolerance.
Quote from: alanp on June 22, 2016, 03:36:31 AM
Layout can make a difference in high gain, or circuits with clock signals.
I'd like to know whether the vero layout worked from the same schematic and values as the pcb, if it sounds better :)
The pcbs are a Queen of Bone (in 9V), a ballbreaker from fuzz dog and Ivlark's vero layout from Tagboard effects.
ok that sheds some light ... if you followed the individual BOM for each of those, you could end up with pedals considerably different(I'm not going to compare, you can). if each of them are identical, then tolerance would come into play
beside the obvious one that is part tolerance, there is also the silicon lottery. the third thing that I think people are mostly missing or skeptic is the induction. all kinds of induction. everything is an antenna. we are playing with high gain audio amplifiers. ceramic caps create and are influenced by magnetic fields. but lets also be realistic, the induction depends on the presence of AC magnetic fields from both inside and outside the case. a shielded enclosure will mostly only get RF on the cable so there is a length of cable there that is radiating RF before it reaches the first RC low pass. RF + high gain can steal headroom if it is not filtered properly. even if it is it is not since it radiates from the jack to the switch. the more you RC filter this the more you increase the RF load through the wire (also an antenna) this also increases the current through the wire. I built a pedal with a LM386 into a TL072 and it was notorious for induction. this is a function of sensitivity to picking up noise and currents being used by the LM386. it makes no difference if the ripples are generated and picked up in the signal or in the power. for single ended signals the feedback loops are the same result with high gain amps. it does not matter if the noise is coupled to the power rails or to the signal and one of the power rails. it will be amplified at each gain stage. there is ground at 0 Ohms theoretically ideal and then there is ground at G > 0 Ohms. there is wire at 0 capacitance and then there is real wire including the leads of the part have C > 0. don't forget L > 0. all your transistors and opamps have different ground this is exactly what is coupling the induction and power ripple between builds. usually this is very very low less than 1 Ohm but a solder joint can vary depending on oxidation of the PCB and component leads. no one ever notices or records measurements of oxidation that can be as high as 10 Ohms or more even worse on the power or a transistor with beta 100+. you can sort all these things out but you really can't sort the silicon lottery past a point of Hfe and I know that manufacturer, part number and Hfe do NOT tell the story when you are trying to make consistent guitar pedals. it is much deeper than that. you need 4 dimensional data sets, not worth it to check a batch or to check a part. thats where I give up.