Attack Pot Issue With Tone Bender MKII Build?

[debrad]

Hi everyone,

I just completed a NPN silicon based Tone Bender circuit using the vero layout at http://tagboardeffects.blogspot.ca/2013/07/npn-si-tonebender-mkii.html.

After playing a little, I started to audition transistors and wound up going with three 2N3903's (Q1 at 68, Q2 at 80 and Q3 at 84). This combination seemed to give me my favourite raucous tone with full Attack and just the right amount of hair at minimum Attack.

One minor complaint I have is that the linear 5k pot I used for Attack seems relatively subtle in its adjustment right up until the last 1/4 turn at which point I get a very dramatic jump in fuzz. I could understand this a little more if it was a logarithmic pot, but shouldn't it be smoother with a linear pot? Is this because of the pot mod from 1k to 5k or could this be a sign of some issue with the pot or my circuit (note that I also sub'ed a 10k pot for Q3 bias adjustment as recommended in some of the earlier builds)?

[Mich P]

Try a Reverse Log.
Mich P.

[induction]

That circuit calls for a 1k linear pot, not 5k. Some people prefer 2k here, but 5k is going to create exactly the problem you describe. It also will throw off the DC bias of Q3, but your 10k bias pot will probably help with that.

You could try a 5k reverse log pot, but if you're swapping the pot anyway, I recommend using a 1k or 2k linear.

[debrad]

That circuit calls for a 1k linear pot, not 5k. Some people prefer 2k here, but 5k is going to create exactly the problem you describe.

Thanks Induction...makes sense now that I think about it:  I've increased the resistance 5 times over the recommended 1k so the bulk of the adjustment is now in the last 1/5 of the pot's travel, correct?  Does that mean if I "split the difference" and go with a 2k linear pot, I should expect that jump in fuzz to occur closer to the mid-point of travel?

[peterg]

The best solution is to use a 1K reverse audio C pot.

[induction]

I used a 2k lin and I don't notice a jump at all, but YMMV. 1k reverse log is a very common substitute. I'd suggest breadboarding it and seeing what you like best.

[debrad]

Thanks everyone!

I see one of the previous builders of that circuit says that he "used a 5k linear modified to 2.5k"...I'm going to assume he simply soldered a 5k fixed resistor parallel to the pot - is there any concerns about that approach as opposed to replacing the 5k altogether?