This Week on the Breadboard: The Pushover Drive

[Chuck D. Bones]

I'm been following some of the Grover / 5Lb Bag build reports and thought it was time to breadboard something. As others have stated, it's basically a Si PNP FF driving a Zen Drive. There seemed to be a few extra parts, which often happens when one sticks two circuits together. After a couple weeks of fiddling, this is where I landed.

Knobs (L-R): LEVEL - BASS - TREBLE - VOICE sw - GAIN


I kept the basic topology, but removed a few redundant parts, tweaked the voicing and replaced the treble-cut control with a Baxandall tone stack. Taking a cue from the Diamond Peak, I used low-HFE transistors (Q1 & Q2) in the FF section and added a couple of caps (C5 & C6) to smooth out the very top end. The main difference between the Grover & the 5Lb Bag is the resistor and cap(s) between the input & Q1-B. Folks have complained about squealing with the Grover and that was mostly down to the 100nF cap at the input. The VOICE switch allows the user to select either the 5Lb Bag's fat voicing or a tamed version of the Grover's bright voicing. As is the case in many Lovepedal designs, the GAIN control also varies the bias. This causes the symmetry to change as the GAIN knob is rotated. When setting the Bias Trim, check the tone across the full range of the GAIN knob.

I figured out that the main characteristic of MOSFET body diodes when used in soft clippers is the relatively large capacitance. The two BS170s I used measured around 50pF each on my CCTT. 2N7000s are a little higher. The capacitance varies with reverse bias, which can also affect the voicing. I tried a few different clipping arrangements (notice the forest of spare diodes hanging off of the lower left side of the breadboard). Replacing the MOSFETs with 1N4148 or BA482 caused the tone to be slightly brighter. I ended up keeping the MOSFETs. I found that a BAT46 and a 1N34A in series, or two BAT46's or a BAT46 + 68Ω all sounded the same. Replacing the BAT46 + 1N34A with a 1N4148 (D2) also sounded about the same, so I went with that.

I thought the tone was a bit muffled, so I added R12 & C9 to create a little "presence" above 1KHz.

The VOICE switch up front and the Baxandall tone stack provide a broad tonal palette. Cork-sniffers may notice that the circuit inverts the signal. BFD.

One interesting facet of this design is that the gain from Q2-C to U1-1 is approx unity. The designer (Jeorge Tripps?) deliberately drives the soft clipping stage very lightly. The soft clipping stage is about voicing & dynamics, not gain. The majority of the sustain comes from the FF stage.

As always, all component values are negotiable.

 

[cooder]

And I thought I had enough drive pedals for awhile... nope...

 

[Chuck D. Bones]

Ha ha ha ha!

I have another rev coming...

 

[Feral Feline]

I knew you would.

 

[temol]

On mybreadboard now... it goes from light drive to distortion.
As a Q2 i tested Si and Ge transistors: 2n1040 (Ge, hfe 20), 2N2905(Si, hfe 115), TG5 (Ge, hfe 57), ASY34 (Ge, hfe 11), Af139 (Ge, hfe 44). All transistors measured with DCA75. Slight differences in sound between transistors but nothing dramatic. I took germaniums because I don't have too many low hfe silicon transistors.

Strat (bridge pickup) -> breadboard -> audio interface -> Mesa 4x12V30 IR. 3 gain settings for each transistor (min, half, max).

Dropbox

www.dropbox.com

 

[Chuck D. Bones]

Thanks for the demo.
Did you check the bias for each Q2 transistor?
Did you use 2N2907A for Q1?
What were the BASS & TREBLE settings?
What do you think of the ranges of BASS & TREBLE? I found I was setting BASS from 5 to 10 and running TREBLE from 0 to 6.

 

[temol]

Did you check the bias for each Q2 transistor?

Yes. Gain in the middle position.
4v for the 2N2905, TG5 and ASY34.
3.7V for the 2N1040 and Af139.

Did you use 2N2907A for Q1?

Yes.

What were the BASS & TREBLE settings?

BASS around 1:30 PM, TREBLE @ 3 PM. Voice switch in Tight position.

What do you think of the ranges of BASS & TREBLE? I found I was setting BASS from 5 to 10 and running TREBLE from 0 to 6.

The range of adjustments is quite wide. An additional advantage is the Voice switch. Take into account that I did not test it with an amplifier that has its own EQ, only into the audio interface. That's probably where the TREBLE settings come from. Another thing is that I like more top end.

How about MID adjustment, even simple one, like a switch? Mid boost? Or mid shift switch?

 

[Chuck D. Bones]

I'm almost ready to release the next rev., just fiddling a couple of capacitor values. In the rev, I replace the FAT switch with a pot, basically the same setup as the Lovepedal Silicon Fuzzmaster. I redid the BASS & TREBLE controls. TREBLE is back to the original circuit with some component value tweaks. The last stage is variable BASS boost. There's a passive Presence boost at the end. I thought about a mid control, but after playing it I wasn't sure it needed one. I'll think on it further, maybe try adding VarioBoost between stages to see it it does anything useful. I did add one more knob, but I'll save that as a surprise.

 

[temol]

A new knob... we'll see Waitng patiently.

 

[Chuck D. Bones]

I could make a variable mid-cut like in the Dr. Robert quite easily. That would get us up to seven knobs.

Haven't tried it yet, so I can't vouch for the tone...

 

[Chuck D. Bones]

Here's v1.3.2. The PUSH knob controls the bass going into the pedal. It replaces the FAT switch. Basically the same as the BLEND knob in the Silicon Fuzzmaster, except I made the pot C100K and increased C1 to 10nF. The tone controls at the back are completely redone. TREBLE is cut only, just like the original Dover Drive. It has a little more range than the Dover Drive's TREBLE knob. BASS is boost only. A little goes a long way. For flat response thru the tone controls, set BASS to zero and TREBLE to 10. I added a 6dB treble boost at the end (C13 & R15) to make up for the fact that TREBLE is cut only. I removed the R & C in parallel with R12, they didn't do enough. The BIG change is the addition of the DRIVE knob. At zero, we have the same signal level going into U1A as the Dover Drive. Crank DRIVE up and we get more clipping and more clean-bleed, which translates to a fatter tone and less compression. Various combinations of GAIN & DRIVE settings yields a broader tonal range than we get with only the GAIN knob. Try it, I think you'll like it.

 

[temol]

So..

Dropbox

www.dropbox.com

Settings for the above
Strat (bridge pickup) -> breadboard -> audio interface -> 4x12 V30 Mesa IR.
First three parts - 2N2905 as Q2, then another three parts with ASY34. Gain settings for each part - min, half, max. Push potentiomerer max CCW.
TRE@ 3PM,BASS@ 9AM.


#1 Drive @ min, 2N2905
#2 Drive @ noon, 2N2905
#3 Drive @ max, 2N2905
#4 Drive @ min, ASY34
#5 Drive @ noon, ASY34
#6 Drive @ max, ASY34

My comments - @ high gain and drive setting sound is fat, fuzz like. Not my type.
To me useable range of the Push pot is between min and noon. Past noon there's too much low end. At least with my recording/listening setup.
BASS pot - useable range between minimum and 11/12 o'clock. But The IR I'm using here has a lot of bottom end.
7 pots would be too many - assuming there was a MID adjustment. If I were to get rid of one.. PUSH or DRIVE. I would replace DRIVE with a trimmer.
Oh.. It seems to me that the voltage on Q2 now changes more depending on the transistor. With one I even lacked the adjustment range.

 

[Chuck D. Bones]

Changing R9 from 220Ω to 1K should cause the trimmer's "sweet spot" to move by <10% (27°). Were you that close to the end of rotation before? My trimmer setting is pretty close to the middle. The other changes should not have affected Q2's bias. You might want to check the connections to C1 & C2.

I usually don't set PUSH past noon either. But some folks might like a more "Doomy" setting. Same goes for BASS.
NB: PUSH interacts with the guitar and with the GAIN setting.

I could see making DRIVE a trimpot or a switch.

Do you get enough TREBLE range with this setup?

I have a passive mid-cut circuit that works well in simulation and on the bench. Goes from flat to a -24dB notch at 650Hz. It has a trimpot that varies the notch from 500Hz to 870Hz. Time for a listening test.

 

[temol]

Changing R9 from 220Ω to 1K should cause the trimmer's "sweet spot" to move by <10% (27°). Were you that close to the end of rotation before? My trimmer setting is pretty close to the middle. The other changes should not have affected Q2's bias. You might want to check the connections to C1 & C2.

Not sure about previous setting but not in the middle of rotation. Rather something like 10-11 o'clock.

Do you get enough TREBLE range with this setup?

Yes. 15-16 o'clock is the max I'd use with bridge pickup on the strat.

 

[Chuck D. Bones]

Latest rev had too much mid scoop. Tweaking some component values and simplifying it.

 

[temol]

Latest rev had too much mid scoop.

I agree.

 

[Chuck D. Bones]

I didn't show anyone my latest rev. Are you saying v1.3.2 has too much mid scoop?

 

[temol]

1.3.2. Maybe a little bit too much.. at least to me.
Btw - I'll try to upload raw recording, without IR applied. Then everyone interested will be able to try their favorite IR.

 

[Chuck D. Bones]

It's mid-forward when I listen to it.

 

[Chuck D. Bones]

Rev 1.4.3. The MID knob is cut only, centered at 700Hz. I tried a twin-T notch filter, like in the Dr. Robert, but it was too extreme for my tastes. The bridged-T version here is just right. For a 6-knob version, I'd make MID a trimmer and set it in the middle. This circuit is getting close to the finish line.