Morley Power Wah > Bad Horsie 2 conversion

[Prostheta]

A bit of a different mini-project for me at the moment. As of writing, I have two guitars awaiting paint and out of project sanity I shouldn't take on any new instruments. Too many projects around the house (skirtings, covings, window frames....) plus studies and financial concerns mean that I absolutely need to focus on my endeavours!

Long story:

So back in the UK when Nina was a manager of a music store, I somehow accidentally maybe managed to become the owner of a Morley Power Wah. I only recently started to become comfortable with wah on my Helix, and in fact really like the vocal tone and ability to focus on specific frequencies to induce long sustain/feedback. The pedal on my Helix is great, but the mechanism seems to have developed a weird click....so I figured that I'd dig out that old Morley pedal and see if I could repurpose it into an external controller for the Helix, but then I remembered that it was optical rather than driven by a potentiometer. Oh well.

Cut to the present day:

I don't care what other people think, but Morley are just cool from my perspective and opinion. Optical pedals might be a bit difficult to dial in and hell, I've seen some YouTube hacks brutally messing up the optical curtain as a "hack" in the truest sense of the word. The build is great, and I like the width, low centre of gravity and general form factor. Add into this that Morley designs their PCBs to be common across a number of pedals and you've got yourself a project.

These are the guts of the Power Wah.

 

In spite of my terribly focused phone camera, you can see that this is a very sparsely-populated board produced after the Bad Horsie 2 wah design; this PCB can be converted into the circuit for the Bad Horsie 2 simply by populating it fully, adding in a few pots and drilling out the case. Damn right.

My reference for this project is going to be from a number of sources. The Bad Horsie 2 schematic is easy to find, plus a guy called Keith repaired a Bad Horsie 2 and documented this on his blog. This provides excellent photo reference in addition to the PCB being fully documented as it stands. Awesome!

I haven't yet done a full inventory of the parts needed, however everything looks very standard and easy to acquire other than the Morley LDR. These seem out of stock everywhere I've looked so far, and I even came across a comment about them not being allowed in the EU (RoHS non-compliant?) but have nothing else backing that up. Physically, the enclosure matches other than the switch having been moved from one side of the body to the other. I don't yet know whether I will block this hole up and move it across or not. For the moment I will likely retain the position and figure out in future whether I like the ergonomics of it or not.

Beyond this mini project being not much more than a case of sourcing components and populating the board, I don't see much room for improvement or modification at this stage. I'm sure that I will though. Initially I think that the battery connector could do with basic strain relief, so drilling two holes in the PCB to thread the cables through will be a must. It may be good practice to do this for the switch as well, however the ground plane interferes somewhat. Depending on what I have on hand, I may also swap out all of the carbon comp resistors for metal film. It all looks good though!

[Prostheta]

Conundrum #1

Reading chatter around the Interwebs about Morley repairs, it seems that the LDRs are a sought-after part that only seem to come from Morley. In this instance I'm good for the LDR2/L2 pair which comprise an optocoupler for the wah sweep. What I'm needing is LDR1 and L1 which are for the switchless aspect of the pedal. Since there is no specific sweep for this, I think I can get away with a lot more.

Looking at the schematic, LDR1 is labelled as having a "value" of 150-200. In circuit it comprised half of a simple resistive voltage divider into the input of a CD4069UBE CMOS hex inverter. These are pretty simple logic devices, in that each inverter outputs the opposite of the input....pull the input high (9v) and the output sinks to 0v and vice versa. LDR1 and R1 act exactly the same way as a potentiometer, with the same mathematics behind them;

Vout=Vin.(LDR2/(R1+LDR2))

The values of 150-200 may represent the resistance of the photocell when excited by photons from L2, otherwise the resistance may be in the high kOhms or MOhms. So really, any LDR with an excited resistance of less than the value of R1 should produce the desired response. So maybe not as much of a conundrum as you might expect.

Conundrum #2

Ah, that old school wah user thing. The Power Wah is set up so that the pedal has a "home detent" at the front of the pedal's travel rather than the back, so users can pivot their foot and stomp on the switch with their heel. That's cool too, but not what I want, and interferes with both the switchless operation and the switch is for the contour mode anyway. I haven't disassembled the pedal mechanism yet, however I am guessing that there is some sort of way to reconfigured that detent and bring it to the front so the pedal "detents" at the back of the range, or remove it entirely. That's how I set up the expression pedal on my Helix wah models anyway, and the optocoupler curtain is configured to switch off the wah with the pedal rocked to the back. The "real" Bad Horsie 2 is sprung to bring the pedal back, so I may experiment to see how I like this and whether either mode bites me better.

 

[Prostheta]

After disassembling the pedal, I couldn't find any sort of "detent" at the front end of the travel. It seemed a bit weird to me anyway, so it must have been some sort of stickiness in the mechanism that I accidently fixed through disassembly. I'm not too bothered about adding a spring to the mix at this stage, however the option will always be there to retrofit one.

[Prostheta]

I have most of these components on hand! The complete electronics component BOM for the Bad Horsie 2 is as follows:

R1    4k7
R2    47k
R3    6k8
R4    15k
R5    4k7
R6    15k
R7    6k8
R8    1M
R9    1M
R10    1k
R11    1M
R13    1k5
R14    100k
R15    33k
R16    47R
R17    470k
R18    1k5
R19    22k
R20    220R
R21    470k
R22    1k
R23    10k
R24    100k
R25    10k
R26    10k
R27    1k
R28    1M
R29    470k
R30    470k
R31    470k
R32    470k
R33    1k
R34    10k
R35    100k
R36    82k
R37    1M
R38    1M2
R39    1M2
R40    470k
R41    1k
R42    33k
R43    6k8

 

C1    220u    16v elec.
C2    4u7    50v elec.
C3    47n    mylar
C4    47n    mylar
C5    10n    mylar
C6    30n    mylar
C7    10n    mylar
C8    4u7    50v elec.
C9    10n    mylar
C10    220n    mylar
C11    220n    mylar
(C12 missing?)
C13    1u    50v elec.
C14    10n    mylar
C15    47n    mylar
C16    47n    mylar
C17    100n    mylar
C18    100n    ceramic
C19    1n    mylar
C20    4n7    mylar
C21    1u    50v elec.
C22    5p    ceramic
C23    2n2    mylar
C24    2n2    mylar
C25    1u    50v elec.

 

D1    1N4003    http://www.banzaimusic.com/1N4003.html
D2    1N4148    http://www.banzaimusic.com/1N4148.html
D3    1N4148    
D4    1N4148    
        
        
T1    2N4401    https://www.banzaimusic.com/2N4401.html
T2    2N4401    
T3    2N3904    https://www.banzaimusic.com/2N3904.html
T4    2N3904    
        
F1    2N5484    http://www.banzaimusic.com/2N5484.html
F2    2N5484    
        
TP1    1M    vertical preset resistor    http://www.banzaimusic.com/PT10-vertical-1M.html
        
P1    B50k    http://www.banzaimusic.com/Alpha-16mm-split-shaft-50k-lin.html (need to check threaded section length)
P2    B100k    http://www.banzaimusic.com/Alpha-16mm-split-shaft-100k-lin.html (need to check threaded section length)

U1    CD4069   http://www.banzaimusic.com/CD4069.html (ICs will be socketed)
U2    TL064CN   http://www.banzaimusic.com/TL064CN.html

[henrim]

52 minutes ago, Prostheta said:

U2    TL064CN

You just happen to have them or some other reason for using a TL064 over a TL074?

[curtisa]

Nice! My first and only wah was a Morley Pro Series Wah (the first version with the red label). I seem to recall that the PCB in that was also partially populated, so I assume it could be reconfigured to be other products in the Pro Series range, like the volume and distortion variants.

 

6 hours ago, Prostheta said:

it seems that the LDRs are a sought-after part that only seem to come from Morley

LDRs are notoriously variable in their characteristics, even within the same batch. Morley probably buy some generic component in large volumes and do some kind of screening on them to make sure they're using LDRs with vaguely similar characteristics, so the pedals behave in a predictable way. You're right that any LDR will probably work; it's just that you might find some work better than others. This could be affecting the limits of the 'heel' and 'toe' positions, or the taper as the treadle is moved from one end to the other. The apparent rarity of the specific component is probably down to Morley doing all the leg work to weed out the non-compliant versions for their products.

Edit: LDR1 is only used to switch the effect on and off, so yes - any LDR should work.

 

11 hours ago, Prostheta said:

them not being allowed in the EU (RoHS non-compliant?)

LDRs are usually made using cadmium sulfide - it's the cadmium that makes them unwanted in the EU. Not sure if there's enough Cd in them to make it impossible to buy in your neck of the woods, though?

 

2 hours ago, henrim said:

You just happen to have them or some other reason for using a TL064 over a TL074?

TL064 might be preferred in situations where battery longevity is a concern - they're about a quarter as hungry as their bigger brother, even if only marginally noisier.

[henrim]

19 minutes ago, curtisa said:

TL064 might be preferred in situations where battery longevity is a concern - they're about a quarter as hungry as their bigger brother, even if only marginally noisier.

Yes, I know but is powering pedals from battery a real use case? If it is then I guess it’s a reasonable trade off. 

[Prostheta]

I plan on using a power supply as this will be a secondary wah in ond of my Helix's loops. I'll be using a TLE2074 since I have a few, although I have TL064s, TL074s and OPA2134s kicking around unused.

The spec shows TL064, so for the purposes of quantifying what needs to be done I'm describing in those terms for simplicity.

I've decided that this in its finished form will not be a Bad Horsie 2. It will be an Amazing Horse, including case art.

 

[Prostheta]

Not all the case art.

[curtisa]

1 hour ago, henrim said:

Yes, I know but is powering pedals from battery a real use case? If it is then I guess it’s a reasonable trade off. 

It's a pedal with a 9V battery clip  Battery usage was definitely considered by the designers.

Although as @Prostheta mentions, he plans to use this on a power pack, so he's free to use whatever opamps he feels like and throw caution to the wind with regards to power consumption.

 

33 minutes ago, Prostheta said:

Not all the case art.

Probably a good thing.

For some reason I'm more reminded of this:

 

[henrim]

17 minutes ago, curtisa said:

It's a pedal with a 9V battery clip  Battery usage was definitely considered by the designers.

Yes. Quite a many pedals come with battery clip. And it can possibly be handy in some odd situation. I was just curious because there is TL064 in the BOM and the schematic above mentions TL074. 

[curtisa]

Could be loads of reasons - TL074 might have been cheaper, Morley might have already had a standardised stock item with several hundred thousand on hand, maybe something about the sound..?

[henrim]

29 minutes ago, curtisa said:

maybe something about the sound..?

Now that’s the thing I’m interested. If there was/is a reason to pick 064 for some reason other than the price point or power consumption. If the (marginally) noisier chip would actually have  somehow favorable sound. 

[Prostheta]

In my experience it never made any observable difference other than noise in high gain situations. Slew might be an issue, as that seems to be a common one with micropower op amps, but I'd really need to look at specs to draw any real conclusion beyond what I'm pulling out of my ass. I've always used them interchangeably and gone for TL06x when it's specifically a battery powered device. Not sure why the TL08x exists, however a TL07x is a good known quantity for most things.

[Prostheta]

Alright, I dug around a few basic specs and the general order of things is that the TL08x is the "standard" that nobody really uses, the TL07x is the lower noise version that has more or less established itself as the common general standard, and the TL06x backtracks towards the TL08x by having lower power consumption but is more noisy than the TL07x.

I just had a dig around and I have a few TLE2064 and TLE2074s. I'll probably use the latter since a few mA isn't going to break the bank. Wall warts lose more than this in their general level of inefficiency anyway I bet.

Since I'm going to be buying so many components, I'm going to draw up a total swapout if better quality components present themselves. In general all the resistors will be swapped out for Vishay Beyschlag MBB-SMA 0,6W metal film. This is a single unit, so a few cents on the component for buying on-brand makes sense. In time I may change out the battery box area and fit Li-ion battery with a buck boost charging circuit, simply because I've been doing non-guitar projects with Li-ion stuff recently. This is an afterthought though, and one that rates lower than the Amazing Horse case art.

At some point this weekend I'll review a few modifications on the Bad Horsie 2 and see what may be useful here. It seems that a few mods to the original Bad Horsie came from the Bad Horsie 2 so beyond a few basics I think it'll be fairly plain.

[Prostheta]

So it seems that there's a few simple mods that are just component changes.

Desoldered the battery snap and added in a couple of strain relief holes. I'm not reattaching a snap until the rest of the board has been populated. The two wires go through the first hole, then out through the second. Simple and effective. I use desoldering braid to remove old solder, then clean the board from flux residue with DNA.

 

The same applies for the four footswitch wires. These are in lesser need of strain relief since they don't get waggled around like a battery snap does.

 

Time to stop procrastinating and get the component order in.

[Prostheta]

There's a wealth of info on mods here:
http://www.goethert.com/FrostBottomBoys/AndyHowe/WahPedals.htm

It's best not to go too crazy with these at first, however for the purposes of buying in components for this project I'm selecting a few mods so at least I have the components on hand to experiment. I have limited space for additional switches and pots, so I'm either going to have to bake in some ideas or select those that I want to be adjustable.

• Swapping out R12 (68k) to 56k is a popular one to "give the wah some balls". This can be switchable with a 56k and shortable 12k in series. Easy enough to construct the resistors around the switch bring out flying leads from location R12 on the board.
• R15 can be altered in the same way out to a 50k pot with a 10k resistor in series to provide a 10-60k range over the original value of 33k. This alters the Q value of the filter, however I am going to read into this a little further or directly experiment later in the game to get an idea of feel.
• The DC blocking input capacitor C7 (10nF) can be increased by either adding more capacitance in series or switching between different values. This allows more low frequencies through into the circuit, equating to a "deep" switch of a sort. Again, switches are cheap real estate compared to pots so this one is likely.

[Prostheta]

Minor progress. I depopulated the PCB other than the parts I don't want to replace. Probably ordering parts on Monday since I've got a stock enquiry in....

[Prostheta]

A bit of an update on this one.

Firstly, parts are on order. I'm populating the entire board with 0,6w 1% metal film resistors as mentioned, plus replacing all of the capacitors with Nichicon and poly boxes.

The first thing I'll be doing is bringing in the Bad Horsie 2 "into spec" by populating it with the original factory values. As nice as having a series of modifications is, I don't particularly want to go straight into modified form and have no frame of reference to what changes were made. I'll likely dial these in anyway, but not initially.

I'd also like to take this board slightly further by incorporating the modifications that I settle on into the PCB design. At my place of study, we have access to a PCB router which is perfect for this sort of prototyping. Couple that with an educational licence for EAGLE CAD and we're golden. This will allow me to take the original schematic and layout, then modify it with the component changes (for things like bringing resistors out to pots, adding in switches, strain relief, etc) plus baking in specific lead pitches of say, WIMA and Nichicon caps for a V2 drop-in.

The difficulty is the parts that I cannot source too easily. Morley LDRs are out there, however they're not cheap or standard. The inductor is specific to Morley. In spite of this, it would be a nice idea to try out a few different inductors. I prefer a more vocal-sounding wah with a wider Q to pick and hold notes into sustain, especially the Vox V847. If the Morley sounds too modern for my liking, I could drop in a Fasel yellow or a Halo inductor to try these out to taste. Or make them switchable on the same board.

Whichever direction I take a V2 "Amazing Horsie" board, there's a bunch of things to learn here....

[Prostheta]

Alright, so the board is populated, assembled, set up and appears to be functioning! Not bad for first time. I was dreading fault-finding a board that is not my own, but I'll take it. A rough playthrough last night revealed a few changes that I think I would make in addition to the basic build issues that I found.

Firstly, the "quality of life" adjustments I made. These are general popular modifications often seen with the Bad Horsie 2. Things like the input cap being raised in value so the wah doesn't lose so much low end, etc. Nothing that really affects the basic function or character of the product. The LDR/LED positions were set up so that the wah turns off within the last few mm of heel travel, and the range is better from the in point through to the toe. It's pretty nice, however I agree with a lot of people's comments that the range isn't that linear through the movement, with a lot of the action being bunched up into a very small end of the range. This can be fixed to a degree by moving the appropriate LED/LDR pair, but is mostly a job involving modification of the optical silhouette.

It's also worth mentioning that the the BH2 went through at least two revisions. I have a board that is labelled Rev. B that doesn't fully correlate to the schematic, which was Rev. A. A couple of blogs out there referring to modifications/repairs of the BH2 that I used as a guide for the project were based off Rev. A, however the real differences are simply component values.

The most relevant of which is R42 which is the resistor providing the base resistance for the contour control in series with the 100k potentiometer, giving a value of 33k-100k. In Rev. B this was increased to 68k providing wider range. This control doesn't seem to be as effective as one might imagine, so it may provide some scope for future character modification.

The second difference I noted was R37; a resistor that is part of a voltage divider setting a bias voltage for the contour portion's op-amp. Oddly enough, in Rev. A the pair R37/R38 are 1M and 1M2 respectively; not identical values as you'd expect for a voltage divider in this sort of circuit. In Rev. B this was changed to the expected value of 1M2, however swapping out R38 to 1M would yield the same result. The offset doesn't make any difference in practice, so just a minor note.

-----

So, the changes that I would make if I were designing this pedal.

The two pots added are soldered directly to the board using pins soldered flat to pads which is terrible. The mechanical retention of the pots is based entirely on how strongly the copper will stay on the PCB! Ideally the pots would be of the "solder tag with flying leads" sort of configuration, or pins in through holes at worst. I may have had lifting traces because of this, which may explain the less than effective natural of the controls as they stand....

The same applies to the LEDs. I understand that an integrated PCB is a great idea for component reduction and manufacturing simplicity, but only to a point. I would say that some of these decisions undermine the durability of the end product by placing too many demands on the PCB that they're just not designed for.

-----

What next?

I'm in danger of taking this project way too much to heart when in reality I think my Line 6 Helix' built-in expression pedal is more flexible and provides more than an external pedal can. I do like the switchless option of the BH2, however this is also something that the Helix' pedal can be configured for. My go-to model in the Helix is the "Chrome Custom", which is itself a modified version of a Vox Wah and that thing just screams feedback sustain when you dwell in the sweet spot of specific notes. I would love to outboard this to an analogue pedal that I can leave in a send/return to save DSP, but a wah is rarely that much of a hog that I need to consider this.

So about me taking projects to heart....I'd love to improve and modify this pedal, even to the point of breaking up the PCB into 2-3 boards to improve functionality and durability. What can I say? The product designer in me wants to see how far this can be taken, and watertight perfection being sought-for. I love the optical design, and that appeals to me on a level that probably has no real world use. It's just cool.

I don't think that I want to disassemble and reassemble this pedal continuously. The stresses on the PCB from the board-mounted components is going to hit a point where something will break. I would however like to make my own "ultimate" BH2, even if it's just for the sake of it. Whether that's buying a second-hand Power Wah for the mechanism or building up a wah around a Crybaby-style diecast enclosure....not sure.

Modifications I would introduce would involve making the range of the pedal controllable; once the silhouette is tweaked, making the heel and toe frequencies controllable. These are the Helix models' strong points. Additionally, pushing the inductor into producing even order harmonics would make it scream more. That matters. I'm not a WANK-WANK-WANK style wah player, but more of the vocal expression during leads sort. Catching the vowel.

-----

Let's see whether I get the bid on this cheap PW that I put in.....

[Prostheta]

Oh FFS. I seem to have accidentally bought myself a second Power Wah

Alright, so let's look at what I can do with this one. I am unsure as to which PCB revision this board is, however that is pretty much academic since I won't be using it. I'll recover the inductor from the board, and perhaps the footswitch also. Beyond this, we're looking at the pedal housing and optical mechanism alone.

Looking back at the schematic, we've got distinct areas of the board that can be split up amongst smaller PCBs. Primarily, the optical side which deals with the signal routing aspect of the pedal and the effect section itself. The two 2N5484 N-channel JFETs are controlled by the CMOS inverter logic that interprets the state of the L1/LDR1 heel-position detection pair via U1, charges up C2 when "on" and provides a falling charge when "off" via R2/TP1 to ground. When this falls below the threshold of the next inverter, the JFETs select between the unmolested (well, as unmolested as "buffered" gets) signal and the effected signal. Cool.

So in principle I can isolate the switching into a smaller PCB mounted using the four bolt locations from the pedal. This is pretty much what this guy did when converting a BH2 into an external optical controller and rackmount unit: http://griffineffects.blogspot.com/2014/05/morley-bad-horsie-2-wah-into-rack-wah.html

The processing part of the circuit can also be shrunk and fitted inside the enclosure using small standoffs. The jack sockets, DC socket, pots, LEDs and footswitch can then be brought out via strain-relieved cables to their respective locations.

I'd like to try out a different inductor in this circuit, perhaps even switchable between the Morley inductor and alternatives if space permits. To my understanding, the Morley circuit is based on the Crybaby, which itself is based on the original Vox circuit. Given that I have a lot of love for that Chrome Custom model in the Helix, being able to bring this character out into an external analogue pedal sat one of the four send/return paths of the Helix is just too cool.

https://catalinbread.com/blogs/kulas-cabinet/vox-cry-baby

Morley Bad Horsie 2 Schematic (1).pdf

[Prostheta]

So here's our fully populated board. I ordered in components that were easier to acquire without the usual massive delivery fees, otherwise I'd have gone for top notch components across the entire board. Kemet poly boxes are fine.

[Prostheta]

Splitting the board into two halves is fairly logical, and the way I look at it agrees with the method used by the guy making the rackmount BH2 with a pedal for control only.

The components in this section of the board, plus LDR2 (wah position) can be brought out with five wires. Two for power, two for logic outputs to drive the JFET bypass section and one for the LDR itself. Sensibly, the author of that modification also added in a power smoothing cap to that board which makes good sense rather than not.

So in principle this means that this board could provide the basis for any sort of optically-controlled pedal. All that is required in the "effects" board is the JFET switch section and the processing itself. Very cool. It would be nice to see what is possible for this arrangement beyond just a wah....

[curtisa]

4 hours ago, Prostheta said:

So in principle this means that this board could provide the basis for any sort of optically-controlled pedal. All that is required in the "effects" board is the JFET switch section and the processing itself. Very cool. It would be nice to see what is possible for this arrangement beyond just a wah....

Your biggest challenge in that case is working out how to manage the taper of the LED/LDR combo. I would imagine any effects box with a continuous control input is probably expecting a linear taper 'pot'. Morley would have done their best to make the taper of the BH frequency sweep appear linear to the ear, but is probably something different in reality (some sources quote an 'S' taper?).

Many years ago I used to have a Digitech RP6 floor unit that had a built-in expression pedal. It also used a LED/LDR mechanism for the variable resistor control. It had the annoying habit of drifting slowly over time so that the toe position would gradually lose some of its maximum range. Was less of an issue with the wah and volume functions, but made it totally unusable for the built-in Whammy effect. The only way to get it back to normal was to do a factory reset, thereby wiping out all my stored patches!

[Prostheta]

I'd put some thought into that. The immediate yet flawed solution would be to make a new silhouette curtain with a different aperture than the standard thin triangle. All sorts of people on YouTube (including the most obnoxious one) recommend modifying the aperture with knives, files, etc. Given just how small of an area this is, I don't think it's all that useful due to unpredictability. It seems that for many, any improvement is still an improvement!

My thought is that perhaps it really does not matter too much as long as there is an actual range. I'm in the process of designing a new board for the opto area in EAGLE CAD (bonus of getting AutoDesk educational access) so this should in principle allow me to perform a range test by measuring the resistance of LDR2 at various points in the travel of the pedal. Tweaking the L2/LDR2 positions on the board should reveal as to whether or not the aperture shape is in fact poorly implemented....I can imagine that this is not the easiest thing to set up in the factory and likely isn't done so optimally, however at the design level it may still be valid. Realigning L2/LDR2 may be all that is required.

Similarly, it shows how sensitive this design is to misalignments/tolerances. The aperture card is attached to the underside of the pedal, the pedal attached to the case using two bent metal brackets, the PCB to the case via these bolts spaced with a shake washer and a nut. The optical components are soldered to the PCB. This is a LONG chain of tolerances that could creep in, so misalignments are to be expected even if that is only in the range of half a mm. For an optical aperture measuring 5mm or so in length, that adds up. 

So in that respect, I think that adding in some adjustability to dial that out, which isn't bending wires makes sense. By fitting the new optical board using two of the four pedal bolt locations, I can add in a thumbscrew or set screw at the far end (think, fine adjuster on a Floyd) that bears against the inside of the case, the entire board can be deflected accurately to bring it into location. Using rubber o-rings or washers in the main mounting bolts should provide enough give so as not to stress the PCB.