Too Much Bass in 4049 (Kubota TSF Variant)

[jafo]

Basically a mix of the Kubota TSFs in the gallery, except with 100k resistors in the feedback loop, and 0.01 mF input, and 0.1 mF elsewhere. Actually, let me be more precise: Input is grounded via a 100k resistor to reduce noise, followed by a 0.01 mF cap into the first inverter. The first inverter has a 1M pot for feedback. The second inverter has a 100k resistor for feedback. The third inverter has a 0.1 mF cap and a 100k resistor for feedback. Output goes through a 0.01 mF cap. The second inverter also has an output tap via a 0.1 mF cap.

I tend to get mis-biased sounds (dropouts, sudden drops in volume, yadda yadda) and some weird fuzz (normal, then a half second later it's like a ring modulator on heroin), but I'm pretty sure that's due to the tap after the second inverter -- these are symptoms of unused inverters that aren't tied to ground, right?

OK, now to the point of this post. TSFs are supposed to be fizzy and almost painfully bright, but this is exceptionally dark. What can cause this? There are no low pass filters in the path at all, and the high pass filters should be pretty bright. Most builds are markedly bottom-heavy with this guitar anyway (PRS Santana SE, degaussed SD Custom 5), but this takes it to an extreme; there's a pronounced thunk on the bottom even on my Strat.

[stringsthings]

does this circuit have caps in the feedback loop of the inverters?  if so, that will cut high frequencies ...

[jafo]

does this circuit have caps in the feedback loop of the inverters?  if so, that will cut high frequencies ...

Yep, a 0.1 mF in the third inverter, which is smaller than in most 4049 designs. Unfortunately, even the tap off the second inverter is too boomy, so that's not it. I probably just got a hunk of really bassy wood.

[WGTP]

The 100k/.1uf cap form a high pass filter that rolls off at 16Hz, which is extremely low for this type of circuit.  Try a .001uf which will produce a more typical roll off at 1600Hz.  If there is still too much bass, reduce the input cap until you can here less bass.  Not sure of the effect of the 100k resistor at the input, or the tap of the second inverter.  Also, make sure you have the ube/unbuffered variety of 4049. 

[jafo]

The 100k/.1uf cap form a high pass filter that rolls off at 16Hz, which is extremely low for this type of circuit.  Try a .001uf which will produce a more typical roll off at 1600Hz.  If there is still too much bass, reduce the input cap until you can here less bass.  Not sure of the effect of the 100k resistor at the input, or the tap of the second inverter.  Also, make sure you have the ube/unbuffered variety of 4049. 

Hmm, I'm out of tiny caps right now, but I think that 0.001 mF in the input cap might do it (and I'll try joining inverter stages with small caps while I'm at it). The 100k resistor precedes the cap (which is 0.01mF); doesn't a resistor have to follow a cap to create an LPF? Anyway, disconnecting said resistor didn't change anything anyway.

Yah, the buffering was one of my first thoughts, too, but they check out -- which is to be expected, since I bought them from the forum store.

[PRR]

> a 0.01 mF cap into the first inverter.

And no build-out resistor? Fed direct from guitar?

The raw input impedance of an inverter is nearly zero. Working against guitar, the bass/mid will push it, the highs won't.

Add 100K in series with that 0.01u. Is that different?

[DavenPaget]

A 1M to ground (followed by) a 1M in series with the (input) cap .

[WGTP]

Paul is right, it works better with a 100k resistor between the input cap and the inverter.  A really small input cap might work well too.  The .01uf cap (or 100pf replacement) should be in parallel with the 100k resistor, not series. 

Kubota's original had some unusual features that are different than typical practice. 

[jafo]

You guys are awesome. Thanks!