Treble booster questions

[zbt]

I want to get a treble booster, one with selectable range. I see this gets implemented either by a switch (I assume switching between capacitors, e.g. Fulltone Ranger) or with a potentiomeneter (RC filter? e.g. Naga Viper). Any recommendation on what is the best way to select the boosted range? How does the pot thing work?

Or both 

[zbt]

Master Bibl Rocks say it useful

replace Boost Pot 10K with 10K resistor, use guitar volume

Mr Brett Kingman personal taste

Elliot Sound info

Have fun 

[zbt]

for consistency the value for

C1 = C x 1
C2 = C x 2
C3 = C x 4

Code Select Expand

hfe rpi zin Hz 5nF C1 nF C2 nF C3 nF C1 Hz C2 Hz C3 Hz
50 6250 5655 5629 9.13 18.3 36.5 3082 1541 771
60 7500 6659 4780 7.75 15.5 31.0 3082 1541 770
70 8750 7627 4174 6.77 13.5 27.1 3082 1541 770
80 10000 8559 3719 6.03 12.1 24.1 3082 1541 770
90 11250 9459 3365 5.46 10.9 21.8 3082 1541 770
100 12500 10327 3082 5.00 10.0 20.0 3082 1541 770
110 13750 11165 2851 4.63 9.3 18.5 3082 1541 771
120 15000 11976 2658 4.31 8.6 17.2 3082 1541 770
130 16250 12759 2495 4.05 8.1 16.2 3082 1541 770
140 17500 13518 2355 3.82 7.6 15.3 3082 1541 770
150 18750 14251 2234 3.62 7.2 14.5 3082 1541 770
160 20000 14962 2127 3.45 6.9 13.8 3082 1541 771
170 21250 15651 2034 3.30 6.6 13.2 3082 1541 770
180 22500 16319 1951 3.16 6.3 12.7 3082 1541 770
190 23750 16966 1876 3.04 6.1 12.2 3082 1541 770
200 25000 17595 1809 2.93 5.9 11.7 3082 1541 770
210 26250 18205 1748 2.84 5.7 11.3 3082 1541 770
220 27500 18797 1693 2.75 5.5 11.0 3082 1541 771
230 28750 19373 1643 2.67 5.3 10.7 3082 1541 770
240 30000 19933 1597 2.59 5.2 10.4 3082 1541 770
250 31250 20477 1554 2.52 5.0 10.1 3082 1541 771

according to my notes

for hfe 100 without leakage rpi wouldbe 12500 ohm and zin = (rpi//470K//68K) = 10327 ohm
for 5nF C the frequency would be 3082 Hz

< zin > Hz or > zin < Hz
< hfe > Hz or > hfe < Hz
< C   > Hz or > C   < Hz

C depend on hfe and for germanium also leakage

< leakage < Hz or > leakage > Hz (directly proportional)
this a bit confuse me, I have ge tranny that leak about 500uA I test is good no hiss sound bright 

Now the question is how tolerant our hearing is of frequency?

[ElectricDruid]

Now the question is how tolerant our hearing is of frequency?

That depends *a lot* on the frequency. We have pretty hopeless frequency detection at the top and bottom of the range, but it's pretty sharp in the "vocal" range in the centre. The just-noticeable-difference for pitch is usually reckoned as 6 cents, 6/100th of a semitone. If my sums are right, that's a frequency that is only x1.003472 higher than another.

The wiki page has some interesting stuff:
https://en.wikipedia.org/wiki/Just-noticeable_difference

In general, we're way more sensitive to pitch than volume. The JND for volume is reckoned as 1dB, which is roughly equivalent to two semitones! (x1.12)

HTH

[Rob Strand]

Now the question is how tolerant our hearing is of frequency?

In the case of a treble booster it's not about hearing a frequency.  It's more about the perceived level and tone.  A filter acts over a  band of frequencies and for a wide-band signal like a guitar you hear the overall effect.

For a treble booster, decreasing the input cap pushes up the cut-off frequency but a bigger effect is it lowers the gain at low frequencies.  If you change the input cap by about 10% it's around the point where you hear a difference.  For some people even 20%.  Also it depends if you A/B two circuits side by side, which lets you pickup smaller changes.   For the most part you will hear the level drop with a smaller cap.

If you look at a tube screamer gain circuit,

https://www.electrosmash.com/images/tech/tube-screamer/tube-screamer-block-diagram.png

Decreasing R4 and decreasing C4 both raise the filter cut-off.  However, decreasing C4 will reduce the low frequency gain in a similar way to the input cap of the a treble booster.  Whereas decreasing R4 will keep the low frequency gain the same and increase the high frequency gain.   A listening test will show for a given filter cut-off the experiment which decreases C4 will be more audible than decreasing R4.

[zbt]

Thank You Sir Tom and Sir Rob, I would digest it first

correction first for C3 = C x 4, it would be C3 = C x 3 cause it parallel with C

Code Select Expand

hfe C1 C2 C3 C1 Hz C2 Hz C3 Hz
50 10 20 33 2814 1407 853
60 8.2 16.4 27 2915 1457 885
70 6.8 13.6 22 3069 1534 949
80 5.6 11.2 15 3321 1660 1240
90 5.6 11.2 15 3005 1502 1122
100 5 10 15 3082 1541 1027
110 4.7 9.4 15 3033 1516 950
120 3.9 7.8 12 3408 1704 1107
130 3.9 7.8 12 3198 1599 1039
140 3.9 7.8 12 3019 1509 981
150 3.3 6.6 10 3384 1692 1117
160 3.3 6.6 10 3223 1612 1064
170 3.3 6.6 10 3082 1541 1017
180 3.3 6.6 10 2955 1478 975
190 3.3 6.6 10 2843 1421 938
200 2.7 5.4 8.2 3350 1675 1103
210 2.7 5.4 8.2 3238 1619 1066
220 2.7 5.4 8.2 3136 1568 1033
230 2.7 5.4 8.2 3043 1521 1002
240 2.7 5.4 8.2 2957 1479 974
250 2.7 5.4 8.2 2879 1439 948

here is my data for using available cap value 5% tolerance
seem higher hfe there is not much difference
with Si with hfe 100 BD139 still good candidate available, so I limit to germanium
because most of the ones I have are under 60

Even if we use decade capacitors, it seems like too much



so I set by data for lowest cap 4.7nF and max 33nF, with 1nF precission

Code Select Expand

zin 10327 ohm
C Hz Diff Hz
4.7 3279
5.7 2704 575
6.7 2300 404
7.7 2001 299
8.7 1771 230
9.7 1589 183
10.7 1440 148
11.7 1317 123
12.7 1214 104
13.7 1125 89
14.7 1048 77
15.7 982 67
16.7 923 59
17.7 871 52
18.7 824 47
19.7 782 42
20.7 745 38
21.7 710 34
22.7 679 31
23.7 650 29
24.7 624 26
25.7 600 24
26.7 577 22
27.7 556 21
28.7 537 19
29.7 519 18
30.7 502 17
31.7 486 16
32.7 471 15
33.7 457 14
34.7 444 13




My perception of tone is still a work in progress
as this picture