Royal RFC-1 Double Effect Machine wah issue

[pure_analog]

Hello,

This unit has been tricky as it's hard to know what it's supposed to do. I build effects, but following vero layouts and am not good at translating schematic to the board I'm looking at. I'm learning.

I've been over the whole thing. The fuzz(super) seems to work fine as does the fuzz tone effect though I wish it didn't have SUCH a scooped "dropping into the 7th layer of hell" sound. A little mid clarity would be nice. I'd like to keep it stock-ish so maybe there is a resistor swap I could do just to tame that?

The inductorless wah doesn't seem to work, and I can't find a schematic. When I engage it, it sounds REALLY scooped with no high end. Like it has a big wet blanket on it, along with a solid volume drop. You can hear the fuzz engaged, and hear the fuzz tone engaged, and if I roll the wah pot you can JUST hear a change in sound. Sort of like a tone knob rolled all the way back and then maybe up a couple clicks. I measured the pot and its creating variable resistance 5.9ohm-20 though the pedal sweep is much less. Both of the C828 transistors on the wah board seem to measure OK (.46-1.76-1.06) and (1.15-5.10-1.76). I've checked all connections, solder joints, etc. Replaced the 5-10uf electros with good ones. Doesn't look like anyone had been in there before. I'm not sure where I should be looking.

Any ideas for the wah or the fuzz scoop taming?
Thanks!
Jeff

[pure_analog]

Anyone? I'll dig back into it tomorrow with renewed fervor. I'm going to start by bypassing the switch just to double check that, but as mentioned it seem to turn the wah on and off . . . there is just a FAINT audible of the wah working.

I'm looking for help with a clue of what I should look at for the wah, and any mod that could reduce the amount of scoop when the fuzz tone is engaged.

[bluebunny]

Welcome, Jeff.

Without any kind of schematic, this is gonna be very difficult to debug (or mod!) - unless you strike it lucky and one of the learned fellows around here is familiar with this beastie.

[Rob Strand]

Is the fuzz part anything like this?

http://www.experimentalistsanonymous.com/diy/Schematics/Fuzz%20and%20Fuzzy%20Noisemakers/Shin-ei%20Fuzz%20Wah.gif

I remember it was a bit too scooped.

The part of the circuit that does the scoop is just after the 2xdiodes and the 10uF.  It comprises of the 22k + 10k + 1nF + 100nF.   Some variants might have slightly different values.

The way to reduce the scoop is to increase the 1nF by a factor K   and decrease the 100nF buy a factor K.   
Try K = 3 to 4.   
1n     replaced by 2.7n, 3.3n, 3.9n
100n replaced by  39n,  33n, 27n

Once you do that you can play with the 1n position to trim the highs and the 100n position to trim the lows.

You could trace the wha circuit.   If you have pics of the wha PCB (top and bottom) it is possible for us to trace  the circuit.

Here's a few examples:
http://4.bp.blogspot.com/-p5zwwQmEPcI/VUVnU52N2RI/AAAAAAAAOi0/X8EEncbtE9g/s1600/Colorsound_Wah_Schematic_Stock.gif
http://parasitstudio.weebly.com/uploads/2/4/4/9/2449159/9700209_orig.png?485
http://www.geofex.com/Article_Folders/wahpedl/twinteewah.gif

The inductorless wha's with the twin-T circuits were the most common.

[duck_arse]

interesting. I havva Teisco Wau Wau Fuzz here, it is only 7 transistors with the fuzz Q1 and 2 replaced by a K30A jfet. the wah uses an audio transformer style inductor. both sections are on a single board, tho.

[pure_analog]

Rob - I saw that schematic and it looks very similar. The super fuzz section might even be the same. My wah board is separate and has 5 - 10uf caps on it, which I've not seen anywhere. The fuzz board has an additional 12 - 10uf caps on it.

Hang on . . . I JUST found an old Ebay auction for this exact effect box that has a picture of the inside and the schematic! Boom.
Let me figure out how to host just the pic after I download and clean it up in PS. I'll get some close up board shots both sides.

I've taken the image of the schematic I found and redrew it. The stuff in red are values that were illegible, I will see if I can cross reference with my board but I'm not good at translating these directly.

Another link with an inside shot, looks exactly like this one:
https://www.guitarscanada.com/index.php?threads/nopd-wacky-fuzz-content.78541/

[Rob Strand]

Yeah some of those parts are hard to read.

Look at this one,

http://www.lynx.net/~jc/pedalsRoyal.html

For the most part, when I read the unclear schematic I see what you see.   However, I think the actual values are like that on the link I gave.  (To some degree because I've seen these values on other variants time an time again.)
The one I'm having the most problems deciding on resolving is the 100p or 1n on the scoop circuit after the diodes.

The input to the wha is probably 47k.

[Mark Hammer]

If I'm not mistaken, I fixed one of those for a buddy.  It uses a cam, rather than a rack and pinion, right?

It's yet another Superfuzz variant that is a couple of components different than the Shin-Ei original.  Most of them are identical, but for the front end.

One of the things that has come up in discussion here before is that these things generate plenty of distortion without the clipping diode pair you see just ahead of the scoop filter.  I've come to the conclusion that this oft-found diode pair is not used to distort, but rather as a crude compressor to hold the output at a steady level a little longer, such that the octave appears to "bloom".  Unfortunately, the side effect of that sub-circuit is to introduce more harmonic content that obscures the octave.  Dino/digi2t has recommend inserting a small-value resistor between the diode pair and ground to "soften" the clipping a bit.  I've tried it and it works.  If you are using germaniums for that diode pair, the resistor is even more recommended.

If you want a gentler midscoop, there are two things you can do:
1) Insert a 1k resistor or pot between ground and the 0.1uf cap to ground in the midscoop filter.
2) The midscoop component values seem wrong.  Make the 47k resistor 10k or even 6k8.  It will be louder, but you'll have more lower mids and a little more definition.  You can always turn the volume down to compensate.

[pure_analog]

That other schematic is certainly helpful. And I see this part of the circuit to attend to. It looks like that 100P is really the main difference.

I'm curious why some caps are designated 10uF 10v, and others 10uF 15v. That seems odd to me.
Also, wondering if this helps with finding where to look regarding the wah not being heard?

I really appreciate your help!

Mark - Do you have a recommended resistor value for the diode to ground? Would I put one on each diode in series to ground?
Also, on the 47k scoop resistors - should I lower both of them? And both the same value?

[Rob Strand]

I'm curious why some caps are designated 10uF 10v, and others 10uF 15v. That seems odd to me.

It was common in the 70's to use different voltage caps for size and cost.   It also indicates where you can substitute a lower voltage cap or not.  Often you will find units are built with all the same voltage caps, despite what the schematic says.

Also, wondering if this helps with finding where to look regarding the wah not being heard?

Yes, getting back to the problem.

After simulating the circuit I am fairly confident the input resistor is 47k.   The 200nF cap  looks OK too.

Both of the C828 transistors on the wah board seem to measure OK (.46-1.76-1.06) and (1.15-5.10-1.76). I've checked all connections, solder joints, etc.

Those bias voltages agree with my simulation.

I measured the pot and its creating variable resistance 5.9ohm-20 though the pedal sweep is much less.

The pot is the main suspect that would clearly make it "stuck".   
You should see close to 0ohm to 5k change.
As Mark indicated, there might be a mechanical issue.

You could pull the pot out, or you could just wire a new one temporarily to see if it sweeps.  However, I'd be looking at the mechanical parts first.


[BTW; when you get it working, the fact the sweep can go from low to high means it is probably working.
The inductorless wha's always sound weaker than the ones with inductors.   You can try to tweak it but you
  can only go so far.]

[Rob Strand]

The midscoop component values seem wrong.

IIRC,  there's two versions around a 1kHz notch and a 500Hz notch.

[pure_analog]

I'll work with the pot some more and try a sub to see if anything changes. FWIW - while inductorless wahs are lower in volume, this one is for sure not working correctly. This is nearly inaudible, though you can just tell that the fuzz circuit engages, and you can almost hear a tone sweep when you manually turn the pot from one end to the other. The mechanical operation does indeed use a kind of cam mechanism and has a far shorter sweep, but I have heard videos of this working and mine is not even close. It almost seems like it's an amplification issue of the wah circuit.

[Rob Strand]

I'll work with the pot some more and try a sub to see if anything changes. FWIW - while inductorless wahs are lower in volume, this one is for sure not working correctly. This is nearly inaudible,
.. t almost seems like it's an amplification issue of the wah circuit.

Something is wrong. 

The circuit should be more than unity gain.  The level not much different to a normal wha in the higher parts of the sweep.

You bias points look good so the circuit is probably working.

Other issues are:
- dogdy switch contacts:  can try forcefully wiring it on.
- dodgy output pot. can forcefully bypass it with wires.
- dodgy electrolytic caps:  you have already replaced this

[pure_analog]

Gentlemen - we have success!

First thing I did was take out the pot and sub in the closest thing I had on hand which was a mini Alpha 100k log. Interestingly, there was a Cosmos 50k in there. So, either someone tried fixing it before, or the stock installed pot value is different from the schematic. But, I fired it up and it works! The 'usable' sweep is in a very small range (10-20% of the shaft rotation) with the rest being sounding like low pass ~80Hz and below. But considering the narrow range of the cam mechanism it seems like you could adjust them together about right (ie. pedal range = usable pot range). I'll try to get a pic in here of the mechanism - basically the pot shaft rides inside of a housing with a setscrew. You can pick the spot on your pot shaft and tighten, so it's sort of adjustable that way.

Tomorrow, I'll also do some experimenting with the fuzz scoop and see if I do something with that.

One question - What effect do we get using a 100k log vs a 50k log vs a 5k log pot in this circuit?

[Rob Strand]

So, either someone tried fixing it before, or the stock installed pot value is different from the schematic. But, I fired it up and it works! The 'usable' sweep is in a very small range (10-20% of the shaft rotation) with the rest being sounding like low pass ~80Hz and below. But considering the narrow range of the cam mechanism it seems like you could adjust them together about right (ie. pedal range = usable pot range).

Cool.
Yeah, hard to know.  Interesting dilemma what is "right".

There's probably a correct pot taper so there is a natural relationship between the foot position and the sound.

[Rob Strand]

One question - What effect do we get using a 100k log vs a 50k log vs a 5k log pot in this circuit?

The range of frequencies depends on the span of resistance.   From what I can make out that circuit produces roughly the correct frequency range of sweep when the pot varies from a small (non-zero value) to 5k.

Wha's with inductors are less susceptible to the pot value.   The inductorless what *need* the correct resistance.

If you arranged for the mechanics to set the pot to the small non zero value when the pedal was fully in, then in order to get 5k resistance amount of angular rotation when the pedal is fully up needs to be  *much* less for the 100k and needs to be the full range of the pot on the 5k.

If you look at the mechanics you might be able to estimate the amount of rotation you are getting. 
Another way would be to make the any pot 0 ohms when fully then set it to fully out and measure again.   If you have a 100k pot and you measure say 10k then you know 10k is too much.   The span is 10% of the pot value, so if you only want 5k 10% of 50k pot would produce the correct span.



Also a *log* pot works best on the inductored wha's because when the pedal is fully in  the mechanics are such that the pot rotates *counter clockwise*.   The mechanics kinds of makes the log pot into an anti-log pot.   If you use a log pot make sure the pot rotates *counter clockwise* when it is fully in.   If it rotates clockwise then a linear or antilog(/reverse audio) pot would be better.

[Mark Hammer]

Mark - Do you have a recommended resistor value for the diode to ground? Would I put one on each diode in series to ground?
Also, on the 47k scoop resistors - should I lower both of them? And both the same value?

Since the diodes are tied together at both ends, all you need is one resistor.  Unsolder the ground side of the two diodes, leaving those "free" ends connected to each other, and install a 470R-1k resistor from the two connected free ends to ground.

If you wanted to get fancy, you could install a 2k-5k trimmer in there and fine-tune it, although that seems like overkill to me.

As for the scoop filter, the Tone switch select between a 47k/1k pair, and the network formed by the 100pf/47k/22k/0.1uf combination.  The second combo forms the midscoop filter.  The first pair provide no filtering, but simply drop the level down to match what happens as a result of the midscoop filter (remember that the scoop is produced by bleeding off selected portions of the signal).

In the midscoop filter, the combination of 47k and the 0.1uf cap to ground form a lowpass filter that rolls off treble very low in the spectrum.  When that bass-only part gets combined with the mosquito sounds that pass through via the 100pf cap, the end result is a sound that has some deep bass, some high highs, and very little mids.

But looking at the values in that Royal schematic, they just seem wrong.  47k and 0.1uf rolls off treble starting around 34hz (), and the vast majority of Superfuzz clones use a treble bypass cap of 1000pf, rather than 100pf.  Of course, the values shown in the drawing would remove so much signal, that the use of a 47k/1k no-scoop pair sort of makes sense.  Normally, in other SF variants, that same o-scoop pair is 47k/10k, which drops the signal less.  If the scoop produces an output that is a LOT lower, then any no-scoopalternative also needs to be dropped in level considerably to match it, which is what a 47k/1k pair would do.

So, while the drawing is "helpful" in a way, I am not convinced the component values in the drawing are accurate.  Ideally, when taking all other SF types into account, the scoop-bypass pair should be 47k/10k (i.e., replace the 1k resistor with 10k), the 100pf cap in the scoop filter should be 1000pf, and the 47k resistor just ahead of the 0.1uf cap to ground should probably be something between 10k and 22k at maximum, but NOT as high as 47k.  It's possible that whoever traced out the circuit and drew it mistook 15k for 47k (the little "hat" on the 1 may have made it look like a 4),

With all relevant components changed to their proper values, there will be more mids, but you may still find that it lacks enough mids for you in the scoop tone position.  IN which case, the suggested 1k between the 0.1ufcap and ground still applies.

[pure_analog]

THANK YOU so much to the both of you. I'm going into the lab now, and will report back with results.

[pure_analog]

Mark - you were dead on right. Regardless of what the schematics said, there was a 10k there instead of the one 47k, and a 1000p instead of the 100P.

- I ended up using a 50k LIN pot for the pedal because thats what I had on hand, and then did some component swapping for experimenting to learn.
- I switched the one 220k resistor to a 100k which seemed to make the Q spread wider. I wish the "break" from "oow" to "waa" was still a bit wider.
- I swapped the 250P for a 330P which seemed to take some of the unwanted fizz off the top end.
- The schematics .2uF was actually a .1, so I changed that to a .22uF and it made the wah much more vocal.
- I did the 1k resistor to ground by the .01 cap mod and it works excellent. JUST enough but keeps the scoop sound.
- I also did the resistor on the diode pair which is subtle but it does seem like it smooths it some.

Thanks again everyone. Nice to have this thing actually working, it's a beast! I'll make sure to get gut shots up soon.