help me understand some bits on the BB schem?

[jplebre]

Hey guys

Trying to understand what some bits do.
I think I'm alright until after o/p1 of the opamp.

1 - The series of 0.22uF and 4.7Kohms resistors into i/p 2, what exactly is this doing?
2 - Also, on Marshall schematics the order is swapped around. Any change?

I remember hearing a old teacher of mine saying that on a resistor+capacitor series, wherever you tapped  from (cap or resistor) you could have a LPF or a HPF. Is this true? if so, wouldn't change the order affect the signal? (effectively turning it into a lp/hp?)

3 - The R17 (Volume pot) doesn't have any active component. Will the 1M resistor to ground effectively work as a feedback circuit to the opamp, but the opamp is inverted so this shouldn't make much sense (the feedback is the path through the diodes right?)
So how does it get a "boost"? or doesn't it? (effectively working as an attenuator only?)
Also, if it was affecting the op-amp o/p it would change the (sonic) character of the filter circuit ahead as well!

Schems from:
http://www.generalguitargadgets.com/pdf/ggg_bluesbreaker.pdf   - The true bypass one!
http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=3818&g2_serialNumber=2

Thanks for your help in advance

[Astronaurt]

An LP or HP filter is when you take the output from the junction where the Resistor and capacitor meet, like so:

http://www.muzique.com/schem/filter.htm

as it is in this BB schematic, it doesn't matter much what order those 2 4K7 resistors and .22uF caps are in. Just remember that resistors in series add together, but Caps in series end up doing this:

(C1 x C2) / (C1 + C2)

so puttin' in the values...
(0.22 x 0.22) / (0.22 + 0.22)

you get 0.11 uF for both the capacitors in series. If there's any Filter action going on, it's between that 0.11 uF combination and the input impedance of the Op-amp. I'd Look up the input impedance on the TL072 and put that in the "resistor" box of Filter calculator and put 0.11uF in the "capacitor" box and then you'll get the frequency it starts to roll off at. Op-amps by definition have ridiculously high input impedance, So there's only gonna be a slight roll-off of VERY high frequencies going into the second op-amp stage - which is usually good if used sparingly because it'll keep the distortion generated from being the "nails on a chalk board," unpleasantly screechy type, and also make sure radio frequencies don't get in there and mess everything up.

As for the volume pot, all that's doing is attenuating the signal coming out of the op-amp. It works like a boost because the two op-amp stages have already boosted your guitar signal well above what it was before coming in. If you set the volume knob so that the guitar is the same volume as when the pedal's in bypass then of course there'd be no boost!

[jplebre]

Hey Astronaut!

Cheers for this!
This explain why there are some versions out there that only have one resistor and one cap in that section.

So, you recon that it's only attenuating and doing a slight roll off at 1mHz (using the mentioned calculator?)


As for the volume - confirms my darkest theories about the pedal
It's just I thought I'd encounter a "make up" circuit after a EQ section not an attenuation

[anchovie]

Op-amps by definition have ridiculously high input impedance, So there's only gonna be a slight roll-off of VERY high frequencies going into the second op-amp stage

In this case the op-amp is set up as an inverting amplifier, so on the GGG schem input impedance is set by the 4.7K resistor.

[jplebre]

Right there's still a couple of things that don't really make sense in my head:

So the opamp is operating on an inverting amplifier set-up. This means the i/p is a "virtual earth" and everything else happens on the feedback loop (distortion, etc.).
No issues there.

I'm still not quite there with the bit between the pot and the i/p2.

Why boost something, to bring it down? (it looks like a sort of voltage divider as it meets the i/p and feedback loop). This is against everything you do in audio. By boosting you raise the ground floor, by bringing it then purposely down, you are just making noise more noticeable - ie. closer to the signal. This is a big no-no even in macro signal flow procedures (talking about line level (+4dBu nominal) between several processors for example), in the instrument level range... it seems like pushing your luck. Maybe I'm just not getting it right?

Then, those caps. I can understand how they sum up now (thanks Astronaut!!!!!!) but still don't understand how that component series works, or it's propose.

(Hey, understanding is half the fun!

[ashcat_lt]

Those caps, being in series with the signal flow, act as highpass filters, and attenuate the very low frequencies.  Most importantly they block DC (0Hz) completely.  I'm not completely sure why this scheme has two of these, but that's why they're there.

[Thomeeque]

Then, those caps .. but still don't understand .. it's propose.

Check the schematic with original wiring, there is an explanation. They wanted to mute the circuit in bypass mode (good practise in case of high-gain circuits to minimise noise inducted into passing signal and into supply rails). For some reason they decided to mute it by grounding signal in the "middle" of the circuit, on it's way between those two op-amp sections. So they divided simple R+C path (which forms "R_in" of IC1b based inverting amp and partly sets gain and low frequencies roll-off*) into two halves and they are grounding the middle (by right half of Sw1). There was no other way then to divide both, resistor and capacitor, try to findout why (it's fun too ).

In schematic with true-bypass wiring it does not have sense (there could be one resistor and one capacitor). GGG left it this way probably to stay as close to the original as possible (and to allow original wiring). Mashall would just put one 100n and one 10k there

T.

* Btw. right "half" of GAIN pot is part of R_in too (so it affects gain and roll-off frequency of IC1b stage too)

[Thomeeque]

Why boost something, to bring it down?

IC1b section is about clipping, not boosting..

3 - The R17 (Volume pot)

There's nothing magical about Volume pot, only difference to "usual" setup is that it's not grounded (bottom lug) to the ground but to V_REF (due to DC level equal to V_REF at top lug). From AC point of view V_REF and real ground are equal.

R13 does not form any feedback to op-amp, it just pulls DC level at very output to 0V.

T.

[ashcat_lt]

Check the schematic with original wiring, there is an explanation. They wanted to mute the circuit in bypass mode (good practise in case of high-gain circuits to minimise noise inducted into passing signal and into supply rails). For some reason they decided to mute it by grounding signal in the "middle" of the circuit, on it's way between those two op-amp sections. So they divided simple R+C path (which forms "R_in" of IC1b based inverting amp and partly sets gain and low frequencies roll-off*) into two halves and they are grounding the middle (by right half of Sw1). There was no other way then to divide both, resistor and capacitor, try to findout why (it's fun too ).

 In schematic with true-bypass wiring it does not have sense (there could be one resistor and one capacitor). GGG left it this way probably to stay as close to the original as possible (and to allow original wiring). Mashall would just put one 100n and one 10k there

 T.

You know, I think I saw that before I started typing.  Then I got distracted and when I looked back at it, I had lost it. 

The caps keep the bias voltage on the opamps from shorting to ground to avoid that whole thing with the smoke and the smell and the pedal never working again.  C6 blocks the bias on the first opamp, and C7 blocks from the second.  Cool.

Seems to me like Rin on the second opamp is smaller than the feedback path at least until the diodes open up.  So there is gain there for small signals and that gain is reduced/becomes attenuation for large signals which is what distortion is all about, right?

There's some other weirdness here.  Don't we usually try to avoid DC across a pot in the audio path to avoid that scratchy sound?

Also it seems C3 could just as easily connect to chassis ground.  Shouldn't make any difference either way though, huh?

[jplebre]

Hey guys!

Thanks for the feedback on the post! this is great stuff!
Thomeeque:

Quote from: jplebre on Yesterday at 04:44:47 PM
Why boost something, to bring it down?

IC1b section is about clipping, not boosting..

I was referring from IC1a section. As in, it's boosting it up, so it's then attenuated (in some way, the distortion acts as compression) on the IC1b feedback and on the volume pot.
just didn't make much sense to me in an audio sig flow (overall).

Check the schematic with original wiring, there is an explanation. They wanted to mute the circuit in bypass mode (good practise in case of high-gain circuits to minimise noise inducted into passing signal and into supply rails). For some reason they decided to mute it by grounding signal in the "middle" of the circuit, on it's way between those two op-amp sections. So they divided simple R+C path (which forms "Rin" of IC1b based inverting amp and partly sets gain and low frequencies roll-off*) into two halves and they are grounding the middle (by right half of Sw1).

Ka-ching!!!! that sound like you hit the big money! that makes perfect sense.

There was no other way then to divide both, resistor and capacitor, try to findout why (it's fun too ).

didn't really understand if you are saying or asking :S

Ashcat_it also raises some interesting questions.... more headscratching :S

[slacker]

There's some other weirdness here.  Don't we usually try to avoid DC across a pot in the audio path to avoid that scratchy sound?

There is no DC across the pots. For the volume and tone pots, the output of the opamp (pin 7) is at 4.5 volts DC and the pots then connect to VREF, which is also 4.5 volts DC. Therefore the DC voltage is the same on both sides of the pots, zero volts across them.

[Thomeeque]

I was referring from IC1a section. As in, it's boosting it up, so it's then attenuated (in some way, the distortion acts as compression) on the IC1b feedback and on the volume pot.
just didn't make much sense to me in an audio sig flow (overall).

Well, this is overdrive pedal - it clips (distorts) input signal. As diodes are used for the clipping, you need to amplify input signal enough to reach diodes clipping level (cca 1V here, as two 1N4148 are in series) and then you need to attenuate it back to guitar signal level (as loudness feel depends on distortion level /and other factors/ and sometimes you want different volume then unity with bypass, the VOLUME pot is the best solution).

T.

[jplebre]

cca 1V here, as two 1N4148 are in series) and then you need to attenuate it back to guitar signal level

Very true. 1.23v is 0VU for +4dBu devices. That is VERY hot specially for a preamp.

[Thomeeque]

Also it seems C3 could just as easily connect to chassis ground.  Shouldn't make any difference either way though, huh?

Yep, funny that you ask - I think the same, but I've seen they use V_REF in this situation in many circuits. I wonder if it's only because it was easier in given PCB layout (V_REF track did go by, ground was too far) or if there's something deeper..

T.

UPDATE: Schematics linked here use ground there

[teemuk]

Also it seems C3 could just as easily connect to chassis ground.  Shouldn't make any difference either way though, huh?

Yep, funny that you ask - I think the same, but I've seen they use V_REF in this situation in many circuits. I wonder if it's only because it was easier in given PCB layout (V_REF track did go by, ground was too far) or if there's something deeper..

As long as that shunt branch is AC coupled and the Vref is effectively like a short circuit for audio frequencies and low noise (like it should be) it doesn't really make a noteworthy difference. I think it really just falls down to layout convenience in almost every circuit.

BTW, if you check out the layout it actually shows the C3 connecting ground instead of Vref, likely because connecting C3 to Vref would have been much more difficult layout wise.  

[jplebre]

because since caps block DC there's no difference if that was ground or Vref?