Dr. Quack and True Bypass

[yano]

I'm attempting to build a Dr. Quack (schematic here), and I'm trying to figure out how to convert it to true bypass.  The dpdt switch is sort of in the middle of the whole circuit. If I'm going to use true bypass, should I just connect the leads to each other as though the switch wasn't there or what?
Thanks for your time!

[Peter Snowberg]

Welcome Yano,

The switch in the middle looks like it actually controls the frequency of the filter and not the bypass. It's fairly common for DIY schematics to concentrate on the core of the effect and leave the switching off. Sometimes this is because the switching logic can be as complex as the actual effect circuit.

If you need more background information on switching, take a look at this article:
http://www.geofex.com/Article_Folders/bypass/bypass.htm

...and if you have any questions, please ask.

Best of luck with your build!

Take care,
-Peter

[amz-fx]

Yes, the switch shown on the schematic is to change the filter mode and you can use a dpdt toggle switch.

The bypass should be wired as shown in one of the diagrams at:

http://www.muzique.com/schem/dpdt.gif

regards, Jack

[yano]

Thanks for all the input. I just wish I lived in a better area. The radioshack here closed about a month ago and the nearest one is a two hour drive both ways. So far I've been ordering everything from the internet, but its not fun to wait.

[yano]

On the Dr. Quack schematic, there are two capacitors of 10 uF, and they have a plus "in front" of them.

Does that mean that they are non-polarized or simply installed "backwards"?

Also, since I already have polarized 10 uF caps, is there any way to substitute them if the schematic does indeed call for non-polarized caps?

Thanks

[Peter Snowberg]

On the Dr. Quack schematic, there are two capacitors of 10 uF, and they have a plus "in front" of them.

Does that mean that they are non-polarized or simply installed "backwards"?

Also, since I already have polarized 10 uF caps, is there any way to substitute them if the schematic does indeed call for non-polarized caps?

Thanks

Those are regular polarized capacitors. Picture each one as having a minus marked on the opposite side from the plus on the schematic. I am guessing your capacitors only have a minus marked on them. If this is the case, treat the unmarked side like they have a plus on them and install as the schematic says.

Since you asked, if you ever do need a non-polarized capacitor but only have polarized ones, you can make a non-polarized capacitor by connecting two caps of twice the value you want in series with one of them reversed. That is a terrible sentance so here is a small schematic:

Code Select Expand


  + | |     | | +
-----| |-----| |-----
    | |     | |

each cap is 20uF, total = 10uF

I hope that helps,
-Peter

[yano]

The op amps I'm going to use are MCP602s from Microchip. I'm assuming that all op amps are the same in that they have power supply + and -, then two inputs per "channel" one inverting and one non-inverting, and then an out  for both channels.
As far as this goes, how do I read the schematic?

Also, the leads on the schematic that start in an empty triangle and have "+9v" next to them, are these simply power inputs?

Thanks for everyone's help!

[Peter Snowberg]

Microchip's opamps are only made for 5 volt MAXIMUM operation, so I would use a different opamp. If you connect 9 volts to them, the blue smoke will escape.

My favorite for clean circuits is the NE5532. You can also use TL072, TL082, 4558, or many others but the noise is higher. Most dual opamps follow the same pinouts. Just wire the pictures up the same as this datasheet and use a socket.

http://www.fairchildsemi.com/ds/NE/NE5532.pdf

And for the 9V question.... Yes, just connect together the four places marked +9V and connect those to the battery + terminal.

Good luck,
-Peter

[idlefaction]

as an addendum, there are tons of cool mods you can do to this circuit    i did a whole bunch of them last night.

i don't like the way the filter is voiced, so i changed the two 0.005 caps in the filter opamp feedback loop to a 0.0068 and a 0.01/0.0022 switchable.

i also changed the 10k/diode string on the + of the envelope detector opamp to a diode-100k log pot - 100k resistor string, for a 'filter cutoff' knob.  i'd also like to try using this point as a place to inject a control voltage, so i can use this pedal as a wah as well.  

the above mod went haywire while you turned the knob, and i still don't know why, but i know i fixed it accidentally by changing the input cap to the envelope detector from 0.05 to 0.005.  this also acts as a 6dB/oct high pass filter, helping the envelope respond to high notes about the same as low notes since the low strings on a guitar tend to put out much more level.

i changed the 2M2 feedback resistor in the envelope detector to a 4M7.  i'd like to figure out how to put a capacitor in this feedback loop to make it a 6dB/oct high pass filter rolling off at 600Hz-ish, but haven't done the maths yet.  the slope from the above HPF mod isn't quite enough.

i changed the 100ohm attack resistor to a 47-ohm+1k pot as an 'attack' knob - this is really useful!

also i found i didn't use the pot on the output of the envelope detector, just the input one cos it affected the LED brightness.  so i changed it for a fixed 27k resistor.

i really like this circuit now!!!  it sounds great.    i did some recording with it and i'm now actually more expressive using the envelope filter than with a wah pedal.

[yano]

the 10 uF caps, I assume they are installed "normally", not reversed? The reason I'm asking is because on the schematic, the + sign is placed where i would expect to put the negative side.

[Mark Hammer]

Dear Idlefaction,

Sounds like you did some useful mods and had a blast doing them.

Note that the gain/Q of the filter section can be altered by varying the value of the 470k feedback resistor.  Higher values get you more gain and increased emphasis at the centre frequency.  This mod tends to shift the centre frequency down a bit, so you may want to rethink your range-setting cap values if you do it.

Probably a better way to get more emphatic resonant sweeps is to cascade two filter sections.  I've been pleased with the sound of two DQ's in series, although since there are two envelope detectors involved in that configuration  the responsiveness of the second depends too much on the action of the first.  Far better to have a single envelope detector driving two transistors (à la Baseballs) and two filters in series.  That way they sweep synchronously over the same range, assuring greater resonance.  

Obviously, doing it this way means that range-switching with caps starts to get more complicated.  One of the limitations is that both filters need to be tuned identically.  One doesn't really have the option of range-switching a single filter section and leaving the other as is.  Of course, there is always the possibility of having range switching for one section and 1-stage/2-stage bypass for the added section.  That simplifies or at least sidesteps the problem of matching stages for multiple ranges by limiting matching requirements for only one range.  You only need a 2-pole switch for the one section, instead of a 4-pole to switch them both at once.  Of course, you can always use a pair of DPDT switches for range switching.

One of the things about goosing the resonance is that one tends to perceive even limited sweeps as being more pronounced.  I.E., if the resonance is low, a wider sweep is needed to sound as "responsive" as a more limited sweep at much higher resonance.

As for setting initial sweep point, Tim Escobedo has some nice suggestions for ways of doing it in some of his assorted transistor-to-ground-filter-tuning designs.  These essentially use the transistor as a CV mixer stage.  In such instances, the 25k trimpot is handy to have as a way of balancing out CV sources (envelope follower vs static voltage).

There are actually several spots you could limit low end follower sensitivity.  One is, as you note, the .05 cap on the op-amp input.  A second is to insert a cap just ahead of the 100k sensitivity pot.  Consider using both.  Also consider the advantages of using a parallel cap and variable resistance along with the 47k/.05uf network to tune the low-end responsiveness on an as-needs basis.  In some of his fuzz designs, Joe Gagan uses this as a way of determining which of two possible paths (a more bass and a less bass path) an input signal favours.  You can adapt that to this context as well.  So, for instance, a .001 cap and a 100k variable resistor in parallel with the 47k/.05 network would let you dial in how much gain is applied to any signal above the cutoff determined by the .001 cap, delivering a sort of added high-end advantage tot he envelope follower.  Naturally, you'd want a fixed resistor in series with the pot to set a minimum resistance.  Try 22k-33k.

Putting a cap of any sort in the feedback loop, in parallel with the 4.7M resistor will work against what you are trying to achieve by producing a lowpass filter.  For instance, a 10pf cap will roll off high-end responsiveness around 3.4khz and a 22pf cap will roll it off above 1.5khz.  Under those conditions you'd have to slam the unwound strings to get any sort of sweep.

On the other hand, trimming low-end responsiveness in the manner you suggested can help a great deal for those whose instrument uses single-coil pickups or who use much lighter gauge strings at the treble side (e.g., E - .009, B - .011) and much heavier ones (E - .052, A - .042) at the bass end.

Playing with the decay time of the envelope follower also yields different feels.  There are two ways you can do it.  One is to simply increase the overall capacitance of the time-constant cap (e.g., from 10uf to 22uf or 33uf).  Another is to switch in one or more caps in parallel to vary decay time.  A third is to use a much bigger cap (e.g., 39-47uf and have a parallel variable bleed resistor to vary "drain time" and achieve different decay rates.  One of the considerations is that the capacitance chosen will also alter the attack time, since attack time is a function of the time it takes to charge that same cap up with that value of attack resistor.  Make the cap bigger to increase decay time, and you need a smaller value resistance to achieve the same charge-up time.

What use is varying decay time?  Well, for starters, you get noticeably less envelope ripple since slower response time smoothes out small fluctuations in envelope signal.  It will actually sound cleaner.  Second, you get a different feel that is more suited to rhythm chord playing rather than single ntoe responsiveness.  Third, the conciseness (or drawn-outness) of a sweep can provide a better match to the timing/rhythm of the tune.