Bassballs sensitivity control, the saga keeps on going

[Gabriel Simoes]

Hello all

I've built a bassballs clone some weeks ago, and it sounds great! The only mods I did were ci 1 = tlc2262, ci 2 (filters) rc4559 ... and pots instead of trimpots so I can adjust the filters ....

But, the sensitivity control, at least to me, is not working properly ... well maybe working but I'd like to mod the circuit so I can play not that hard (less amplitude in the original signal) and get the same effect ..
I think maybe the problem is my bass output volume, but I'd like to be able to use the pedal with guitar too so any adjustable control for fine tune this amplitude problem (since the effect work with the amplitude) would be perfect!
Any tip of what should I change so make it work well with instruments with less output volume ?
Thanks
Gabriel

ps: schematic

http://topopiccione.atspace.com/pjimages/EHBassBalls.old.jpg

[Gabriel Simoes]

Where's the guy from topopiccione ?

[nelson]

I think the IC's in this circuit are critical, same with the nurse quacky. replace the IC's with those recommended and it should be more responsive to lower input signals.

[Zero the hero]

Where's the guy from topopiccione ?

Here I am, man.
I'll check it at home, I believe you have to increase the gain of the second op-amp, but I'm not sure.
I have no internet connection at home, so I rarely appear on the forum during week-ends.

[Zero the hero]

Try increasing R5 from 220k to 470k, or put a 220k pot in series with it.
Let me know if it works!

[Gabriel Simoes]

I tried using a 330k and the sound started to get distorted ... but I will try it again ..... and will hope my pcb can handle this .... since the heat in this area (trying to make this work by myself) have damaged it ....

I will take the hole stuff out of the box this week since I will refinish it, the box has a small space and is hard to work, so Ive spent a hole afternoon to make everything fit in it .... and it made me give up the possibility of improving it ... but now I will get the chance ....

Zero, please, if you have time, take a look at the schematics (I will do the same) and lets think about it ... I will try changing the opamp, but since the rc4559 doesnt work in this place, and so I had to use a tlc2262 (dont ask me why, I have no idea, but changing from one ic to the other made the hole think work) and I do not want to use a dual 741 :/

Thanks for your help guys
Gabriel

[Mark Hammer]

... the sensitivity control, at least to me, is not working properly ... well maybe working but I'd like to mod the circuit so I can play not that hard (less amplitude in the original signal) and get the same effect ...Any tip of what should I change so make it work well with instruments with less output volume ?

The suggestion to increase the 220k feedback resistor is a reasonable one, though it may be too much.  One thing you might try is to replace the 1N4148-type diode in the envelope follower with a germanium type.  That will subtract less voltage from the envelope signal during rectification, which has the same effect as increasing the gain of the envelope follower a bit.

Because the 1458 distorts easily in this circuit, the Baseballs uses the output of the envelope follower gain stage for its "fuzz".  If you boost the gain of the envelope follower, this will have the side-effect of increasing the volume difference between fuzz and non-fuzz settings.  Personally, I do not like the sound of that fuzz, so I replaced the 100k and 47k resistors at point E with a 10k resistor and a pair of 1N914 diodes to ground.

Although changing the 10k trimpots for chassis-mounted pots is a good idea (I did it too and am happy with the results), please note that changing the pot settings has an effect on the amount of sweep.

Since the transistors in the filter sections are being used a bit like voltage-controlled resistors, consider the following possibility.  The resistance to ground is the combination of the transistor and the 47k resistor in parallel with it.  As the envelope signal is fed to the transistors, they produce a lower-impedance path to ground and the filter frequency changes.  How *wide* a sweep is produced depends on how big a change in resistance is produced.  If the 47k resistor is a larger value, the transistor will play a greater role.  With the power on (but no signal), measure the resistance to ground from the collector of one of those transistors, and that may give you some ideas about what value to use in place of 47k.  Perhap something like 68k or 82k may be enough.

Finally, the op-amps do matter in this circuit.  The Baseballs is really two parallel Dr. Q filters.  I have found with the Dr Q that some op-amps work well, some provide very little sweep, and others provide so much sweep that the pedal becomes unusable.  You may find that a simple change of op-amp will improve the pedal without having to change anything else at all.

[Gabriel Simoes]

I see we share the same opinion about the fuzz mode, I really do not like it and If i rebuild the hoel circuit for me I wont spend my time with it, unless I try your mods, ...

I just do not know why lowering the resistors at E .... but I will use the germanium diodes to ground for sure ... or even try to stick a germanium diode in place of the 1n4148, since I have some 1n60 here ... for this mod should I just place them 2 to ground in place of the 1n4148 and do not mess with the rest of the circuit ?

As I told you, the rc4559 works perfect in the filter section, but not in the gain stage ... and do not ask me why, I just know it doesnt (I tried five different ones, and they all work in the filter but not in the gain stage) .... I will try to use a tl072 ....

I do not play strong and some friends want to use it with guitar so Improving this will be real good .... I just want to know if what increase the sensitivity of this effect is the act of clipping the signal, and how much clipping (in this case the germanium diode will be the solution) or the amplitude of the signal that passes thru it (maybe this is a noob question)

Thanks
Gabriel

[Gabriel Simoes]

Have somebody done it this way ?? using ge diodes

So lowering the clipping voltage will make the effect more "sensitive" ?
And signals with smaller amplitudes will work ?

How about using 2 diodes to ground instead of one in the signal path maybe a dpdt switch for using one or the other ...

PLease, give me a hand here!

Thanks
Gabriel

[Mark Hammer]

Here's a hand.

The envelope follower gain stage puts out a higher amplitude version of the instrument signal.  The diode in series cuts out one half cycle of that signal so that only voltage changes in one direction are fed to the transistor.  This is why it is referred to as "half-wave rectification" - it cuts out half the wave.  The cap to ground smooths out or blurs that unidirectional signal so that small rapid changes are eliminated or made less noticeable.  The bigger the voltage, the greater the change in the transistor.  

The series diode not only eliminates one half of the audio signal (above or below ground), but also subtracts a fixed voltage from the amplitude.  In the case of silicon diodes, it will subtract between about 450 and 700mv, (though 550-600mv more typically) so that a peak of 3.3vac that leaves the output pin of the op-amp is going to be around 2.7vac on the other side of the diode.  A germanium diode is going to subtract around 200-300mv so that the amplitude of the signal being fed to the transistor will be greater.

That's all.  No great magic.  I've done it before and it works and does not ruin anything...although some might consider it a waste of a germanium diode! :wink:  The reason I mention it is because if you're going to have wires running everywhere, you don't want to have too much gain sitting in any one section.  You also want to have a reasonably well-matched level for Fuzz and nonfuzz settings.  I figured that you can either increase the gain of that op-amp (not really optimal) OR you can conserve more of what the op-amp already does by use of the GE diode.

[Gabriel Simoes]

Thanks Mark!

i will try it this weekend ...

[Zero the hero]

... the sensitivity control, at least to me, is not working properly ... well maybe working but I'd like to mod the circuit so I can play not that hard (less amplitude in the original signal) and get the same effect ...Any tip of what should I change so make it work well with instruments with less output volume ?

The suggestion to increase the 220k feedback resistor is a reasonable one, though it may be too much.  One thing you might try is to replace the 1N4148-type diode in the envelope follower with a germanium type.  That will subtract less voltage from the envelope signal during rectification, which has the same effect as increasing the gain of the envelope follower a bit.

Because the 1458 distorts easily in this circuit, the Baseballs uses the output of the envelope follower gain stage for its "fuzz".  If you boost the gain of the envelope follower, this will have the side-effect of increasing the volume difference between fuzz and non-fuzz settings.  Personally, I do not like the sound of that fuzz, so I replaced the 100k and 47k resistors at point E with a 10k resistor and a pair of 1N914 diodes to ground.

Although changing the 10k trimpots for chassis-mounted pots is a good idea (I did it too and am happy with the results), please note that changing the pot settings has an effect on the amount of sweep.

Since the transistors in the filter sections are being used a bit like voltage-controlled resistors, consider the following possibility.  The resistance to ground is the combination of the transistor and the 47k resistor in parallel with it.  As the envelope signal is fed to the transistors, they produce a lower-impedance path to ground and the filter frequency changes.  How *wide* a sweep is produced depends on how big a change in resistance is produced.  If the 47k resistor is a larger value, the transistor will play a greater role.  With the power on (but no signal), measure the resistance to ground from the collector of one of those transistors, and that may give you some ideas about what value to use in place of 47k.  Perhap something like 68k or 82k may be enough.

Finally, the op-amps do matter in this circuit.  The Baseballs is really two parallel Dr. Q filters.  I have found with the Dr Q that some op-amps work well, some provide very little sweep, and others provide so much sweep that the pedal becomes unusable.  You may find that a simple change of op-amp will improve the pedal without having to change anything else at all.

Mark, may I paste your suggestions into the "parts substitutions" of the bassballs article on my site (with credit to you, pf course)?

[Mark Hammer]

[quote="Zero the hero]Mark, may I paste your suggestions into the "parts substitutions" of the bassballs article on my site (with credit to you, pf course)?[/quote]

Be my guest.

[Gabriel Simoes]

Thank you all for your suggestions ...

As I told you, changing the 220k for a 330k resistor gave me a lot of distortion, and almost no improvement in the sensitivity control, so I will try germanium instead of si ...

The sweep is not a problem at all, to tell you the truth the effect works and sounds sweet ... but unfortunately I have to play hard, and it's kind of impossible to use this with guitar ....

I'd like to know how to dispose the diodes to ground, just the same way they are now but with oposite directions to ground

The bassballs is a damn good envelope, but the fuzz sounds like crap, my concern is that maybe changing the diode to 2 ge to ground will change the vocal effect too ...

I will try tl072 instead of tlc2262, ... and maybe this will help ... as I don't think the rc4559 for the filters is the problem at all .... at least for this sensitivity problem =)

[theguy316]

Anyone know how to bypass the filters in the circuit so i can only run the distortion from the pedal without the filter effect.

[Zero the hero]

Referred to the layout and schematic of my site:
The fuzz effect is directly taken from the envelope follower and it drives the filter section. If you want to catch the fuzz effect only, connect SW1c (from the envelope follower section) to the positive pin of C8.
Maybe the effect will be too high, use a pot to control the output volume.

[Mark Hammer]

As I told you, changing the 220k for a 330k resistor gave me a lot of distortion, and almost no improvement in the sensitivity control, so I will try germanium instead of si ...

I'd like to know how to dispose the diodes to ground, just the same way they are now but with oposite directions to ground

The bassballs is a damn good envelope, but the fuzz sounds like crap, my concern is that maybe changing the diode to 2 ge to ground will change the vocal effect too ...

The "fuzz" from the overdriven 1458 is not particularly good.  My guess is that the only thing that makes it sound even remotely tolerable is the presence of the filters.  My suggestion to replace the resistive divider network with a diode pair will have no effect ion the "vocal" effect.  This is a function of the filters.  The diode pair only influences what gets fed to the filters, not the filters themselves.  It is essentially the same as feeding your guitar to a Distortion+ then feeding that to the Bassballs.  The main difference is that the envelope signal is extracted before the clipping removes the dynamics.  You will note that this is why the E-H Q-Tron includes a loop.  This allows you to feed a more interesting signal to the filter while keeping as much dynamic variation in the control signal as possible to permit more user control.

I found the 2nd Bassballs I made for myself a few weeks ago had too big a difference in volume between fuzz and normal when I used 1N914's.  Consequently, I would recommend a pair of germaniums like 1N34, 1N60, 1N270.  Additionally, since it is impossible to vary envelope sensitivity and amount of clipping separately, consider sticking a 1k trimpot between the diodes and their junction with the switch.  This will let you dial in the hardness of clipping, as shown in the warp control article at AMZ.

[Gabriel Simoes]

In the end I´ve built my bassballs with the original buffer section, a switch for choosing 1n4048 or 1n60 (just one) and replaced the envelope ci for the original 1458.
In the end: Now I have no gated sound .... and the pedal works great but! the quack from the 1458 is good, kind of weak.
So I found a post here where someone built a dr.quack, wich has almost the same structure as the bassballs, and he did a great job trying a lot of differenct cis. His project showed almost the same results as mine: with most of the cis you get no quack or gated quack, 1458 has no problems but not a great sounding quack, and ne 5532 shows the best results.
The filter ci doesnt seem to be problematic though.
I hope zero the hero can add all this information in his site so people know how sensive the change or this ci is in this circuit.
OK? please , at least TRY the effect first with a 1458 (mc or lm!) or a ne5532!
Thanks,
Gabriel