Morley BFW LM741 fuzz

[thermionix]

So I was fixing a Morley BFW Fuzz/Wah for a friend of mine.  He said the fuzz was cool, but had quit working.  Easy repair, just a loose wire, actually never soldered to the PCB at the factory back in 1991.  Oops!

Well I agree that the fuzz is indeed pretty cool.  Quite different from a BMP or FF, more midrangey and can do overdrive pretty well.  Sounds really cool going into my Dynacomp.  I had an LM741 handy so I built a clone on the breadboard.

Here's a not-so-great copy of the schematic:



A slightly better .pdf version can be downloaded here:

https://elektrotanya.com/morley_bfw_sch.pdf/download.html

So my clone works fine, and sounds pretty much like the original.  I just copied the top part of the schematic, with a 100k pot for a volume control instead of a trimmer.  For the fuzz control I used a 25k pot because I had it handy with wires on it already.  I put an 18k resistor in parallel with it to make 10k, but I ended up pulling the 18k for more fuzz range.

This thing is pretty loud.  I don't know how the overall volume compares to the original because that is set internally with the trimmer, and I don't want to mess with that setting since it's not my pedal.  I guess I could add a voltage divider before the volume pot if I want to tame the volume some on mine.

I'm wondering if anybody else has messed with this circuit.  Maybe it's based on something else, I don't know.  I'm wondering if it would be worth the trouble to try different opamps.  I have mostly duals (ie 4558) available so obviously different pinout.  From what I gather 741s are kinda trashy sounding by nature, but maybe that's the charm of this circuit.  Or maybe a "better" opamp would be an improvement.

Any suggestions?

[thermionix]

Wow, geez, I just noticed that the schematic has 2 battery snaps in series.  This is supposed to be an 18V pedal!  The one I repaired is modified and only has one snap.  It works on 9V but I suppose it's not sounding like it should.

[Mark Hammer]

The Vref for the 741 is a little odd;  i.e., not "half V+".  But, like the 741-based Distortion+, it also provides negligible bias current to the + input through that 4M7 resistor.  I don'tknow if that is a requirement to get the 741 to "behave" or a requiement to get it to misbehave indesirable ways.

[thermionix]

I noticed that too, the 470k/680k divider.  Getting beyond my level of expertise here, but does it matter as long as you're not hitting either rail?  Maybe even more breathing room with higher operating voltages?  Asking these questions, I have no idea what signal voltages are present at the + input.  I would think very small with no gain stage before the opamp, but my understanding of it all is still very limited.

[Mark Hammer]

My understanding is also limited.  But here is something that sticks way out for me:  a great many op-amp-based overdrives will use a voltage divider for Vref like a 22k/22k or even 10k/10k pair, and a biasing resistor from the op-amp input to Vref of maybe 330k-470k.  In the Dist+, a 1M/1M divider is used with another 1M resistor used for biasing the input.  So, precious little current.  In the Morley, the 470k/680k pair provides the Vref that feeds a 4M7 biasing resistor.  Again, only teensy-weensy biasing current.

I don't know if this is something special to simply getting 741 op-amps to work, or something special to getting them to sound a certain way.  Certainly we had many reports in past years from folks who said "I completely took the clipping diodes out of circuit and I can't get anything resembling clean out of the circuit".

[duck_arse]

I've been working at a different circuit using an MC1458 [a dual 741, near enuff]. on 9V, it has a 2V output offset, so I rebiased it w/ 10k over 12k to centre the available swing. the ratio is about the same as yr 470k over 680k. the 4M7 must be close to the usable limit for the 741. I wonder f they used that value to match the f/back value.

[I ran scared from your .pdf linked, therm, too many INSTALL THIS OR ELSE's for my likings.]

[thermionix]

[I ran scared from your .pdf linked, therm, too many INSTALL THIS OR ELSE's for my likings.]

I don't get any "install" requests from that site, but IIRC you are using some third world OS. 

I don't blame you for being wary though.  I would normally avoid posting a link like that.  I tried to snag the image from the .pdf and make an uploadable .jpeg or .bmp, but couldn't get that to work this time.  Seems it's segmented, I could only select a portion of the schematic at a time.

Gonna try this fuzz on 18V today, see what that sounds like.

[mac]

Dan Armstrong Blue Clipper has asymmetric biasing, iirc.

[I ran scared from your .pdf linked, therm, too many INSTALL THIS OR ELSE's for my likings.]

I don't get any "install" requests from that site, but IIRC you are using some third world OS. 

mac@mac-pc:~$ echo "I run (and live in the) third world OS too "

I downloaded lots of schematics from elektrotanya, no extra installs, only 3 per day free.

mac@mac-pc:~$ exit

[thermionix]

Well I tried 18V on the breadboard project.  Instant mojo killer.  9V gets much better fuzz.

I tried a voltage divider before the volume control.  That works great, doesn't seem to mess with the tone.  Problem is, there's way too much volume difference between the highest and lowest fuzz settings.  So if I get the volume range reasonable at full fuzz, it won't make unity at minimum fuzz.

So my next experiment is to switch to a 5k fuzz pot (I'm using 25k right now) with a 22k resistor in series, and just have the top portion of the fuzz range.  This circuit is pretty useless on the cleaner end anyway.

[PRR]

> I ran scared from your .pdf linked, therm, too many INSTALL THIS OR ELSE's for my likings.

ElecktroTanya is a fine site, but you do have to know what to click.

There's several ads; gotta be, to support the hosting site.

To make you look at the ads (and maybe click some), they give you a delay before you can download.

"This file is downloadable free of charge: ...processing..."

Eventually (30 seconds??) changes to

"This file is downloadable free of charge: Get Manual"

Now click that.

[PRR]

> provides negligible bias current to the + input through that 4M7 resistor.

Note that there is an equal 4.7Meg on the other input.

'741 input bias current may be 0.1uA to >1uA. Taking 50nA to 200nA for most of a crate, this causes 0.25V to 1V of drop in the 5Meg bias resistor, and the same in the other 5Meg, pretty much cancelling error from this source. The current offset error will be less. As the inputs are NPN, the inputs will sit a part-Volt low of the output.

I think it is 5Meg to keep overall impedances high to use smaller (cheaper) capacitors for ample Vref filtering and bass extension (not much).

[thermionix]

Thanks for posting that schematic Paul.

Since you're here (or were)...what do you make of the uneven voltage divider for Vref?

Do you think a different opamp would be worth trying?  Would the circuit need to be modified for 1/2 of a 4558 or similar?

I'm still experimenting on the breadboard, but for a different opamp pinout I'd basically have to re-do it entirely.

[PRR]

> what do you make of the uneven voltage divider for Vref?

I had a thought, but it was wrong, so I edited it out.

> Do you think a different opamp would be worth trying?

'4558 is a marginally improved '741, 2-pack. I'd expect it to drop-in (revised pinout) and work-same, within the huge variations of a '741.

But all this is better answered on breadboard than on keyboard.

[thermionix]

But all this is better answered on breadboard than on keyboard.

True dat.

I'll likely get back into it tomorrow.  I'll try evening out that voltage divider and see if the universe implodes.