Bourns MN blender: Help "Flipping" the pot

[stevewf]

Nutz, help me out here because I'm stuck needing too many poles in my switch.

It has to do with a dual-gang blend pot, specifically Bourns PDB182-GTRB (manufacturer data sheet). That pot can blend a pair of signals with either of two curve sets: what I'll call "AND" and "NOR". Here's what those curve look like:


The difference between the curves is what happens at the center detente: there, "AND" will include input from both sources, while "NOR" will include input from neither source. But notice that in both graphs, I've depicted the blue signal as 100% on the left and 0% on the right (and mirrored for the green signal). That's to keep the control physically intuitive (e.g. "Always, CW=Blue and CCW=Green"). This has added added some complication. To me, this is the heart of the challenge: sticking to the law of "Always, CW=Blue and CCW=Green".

I'd like to toggle between those two curve sets - maintaining the consistent Left/Right control - by using a two-position switch that requires 4 poles or fewer. So far, I can only design for 6 poles, unless I ignore the above law. Any Nutz got a good answer?

[stevewf]

Here's what I've been working on as a base:



The "Left=Blue, Right=Green"" is respected, and it's done by swapping the positions of the pot's ganged wafers. I'd like a two- position switch that can change between "And" and "Nor", but all my draft designs end up needing 6 poles.

I thought: if there are brilliant ideas to be found they'll be found here at the Nutz House. Anybody?

Bonus points if the solution can be used in conjunction with a separate 2-position switch that can swap serial/parallel. Here, ChrisK posted a 2P2T Serial/Parallel switch scheme for the blend pot.

In case it might help, here's a proposed diagram of "AND"-vs-"NOR" in parallel:

(Or is that an abomination?)

Anyway, I'm mostly after a switch that can go between my "AND" and "NOR" outputs.

[Yogi B]

Dec 5, 2022 3:21:20 GMT -5 stevewf said:

It has to do with a dual-gang blend pot, specifically Bourns PDB182-GTRB (manufacturer data sheet). That pot can blend a pair of signals with either of two curve sets: what I'll call "AND" and "NOR".

Whilst that pot does provide the shown change in resistances between an outer terminal & the wiper (and complementary change between the wiper & other outer terminal, shown in the opposite colour in the second diagram) — this however does not directly correlate to the blending of signals. You have to consider that: a pickup itself has internal impedance, thus its contribution to the resultant signal is affected by the load placed upon it (by the pot and/or the other pickup); and since this impedance is dependent upon frequency, the variation of the response will also (in most circumstances) be frequency dependent unless the variable load mirrors the pickup's impedance (i.e. is not a purely resistive potentiometer subsection).

(This particularly applies to the parallel blending version, but is therefore additionally relevant to ChrisK's series/parallel blending toggle. It's a neat trick to be able to swap between the two circuit topologies on paper, but in reality is undermined by the necessity of maintaining the same pot value/taper between the modes.)

---

Aside from the prior, I also question the desire to switch between the specific "AND" & "NOR" modes of operation. "AND" is a blend: giving any intermediate shade between blue and green — fine, desirable even. On the other hand "NOR" is just a volume control, albeit one which alternates between levels of only either fully blue or fully green.

Splitting the functions of blend & volume control between separate controls seems more logical to me, and affords greater flexibility. (In your scheme, with its single toggled control, what if you want turquoise but quieter?) Separate controls also means you can specialise each to their purpose — rather than, for example: mandating (inferior) independent volume wiring for the halves of the dual-gang pot when in the parallel NOR configuration, when it is literally impossible to vary both 'volumes' simultaneously.

[sumgai]

steve,


Yogi beat me to several points, but I want to address one thing in particular: what's the useful purpose of the NOR circuit?

One of the things we do here is try to make the final results useful, and I for one fail to see any use-case for having a dead spot in the middle of a pot's rotation. Rule #3 here in The NutzHouse - avoid dead spots in any circuit application that controls output levels. (The corollary being, avoid duplication in switching combinations.)  Of course such a thing might be useful when it comes to tone controls, but that's not the thrust of your request.

If you're just aiming to blend the two coils, you can certainly do that equally well with either option. So why have a switch to go either way.... it will just be one more thing to get in your way when actually playing on stage. That's the usability part that I'm harping about, the fact that you're violating the K.I.S.S. principle with no practical benefit for either your tonal palette or your stage presence.

And that's before we get to Yogi's point about loading on the pickup coils.... He's covered that, I don't need to rehash it. Still, if you're going to insist on beating your head against the wall, I'll help you out with a clue: You can safely "violate" that alleged law about which color should go where, because it applies only when the abstract representation is seen on paper. When all the hooking up of things is done, and sealed up inside of the enclosure, no one will be able to tell green from blue, not even you. It'll be the physical reality that the circuit operates as intended that will "make your day".

And that reminds me....  I have an old saying that is probably the reason I'm alive today. It goes:

"The day you stop learning is the day you start dying."

Well, hot damn if I ain't alive today, 'cause I just learned something I've never heard of before. This "law" you speak of.... I've been in Electrical Engineering, in one specialty field or another, for over 50 years now, and I'd never heard of this 'til you brought it up. So please tell us, where does one find a valid source for it? And by valid, I mean more than just "I read it on a blog somewhere", I'd like to see a white paper from a reputable source, if you please.  TIA.

HTH




sumgai
 

[stevewf]

I'll quickly admit that the loading of a coil is something that I need to internalize, and that the resulting sounds are something that I need to be able to predict better.

However, let's not presume that the whole guitar will consist of two coils only. This time, I'm aiming this module at a guitar that has three coils. Perhaps I could have avoided a detour by explaining the context in the first place... but I didn't want to change the focus away from the module in question. Anyway, here are some details about the coils and controls that I want in the guitar.

Aside from the blend function, I also plan on using a multi-pole, three-position switch that will, in essence, select one of the three coils as the "main coil". Probably a rotary. Each of the 3 positions will choose a different one of the coils as "main". Then, it's the other two coils that will be treated by the module I'm asking about. So, for example, I'll choose the neck coil as the "main" coil. Then, to the Main coil's signal, the blend knob is meant to blend the middle, the bridge, both or neither, at controlled levels. 

The reason I'm interested in limiting the number of poles in a switch is that I already have a switch in hand. I modified a Bourns MN pot by adding a push-pull switch to it; further, this push-pull switch has also been modded to include 4 poles.

(by the way, it was really painstaking to make that, hence my reluctance to mod it further. Plus, stacking another 2 poles on it would make so tall that it would fit only in the hollow body guitar who's meant to get this particular blend setup; as a 4-pole, it presently fits in my LP copy, but I'd lose that versatility... the switch module can be swapped onto any Bourns P/P).

Here's where my "law" comes into play. It's no law I ever heard from anywhere; rather it's constraint that I want to apply to a wiring solution. Specifically, that when the blend knob is twisted fully to a given direction, then signal is accepted from the coil positioned accordingly... and that law, by my caprice as king of this guitar, shall apply to AND as it does to NOR, and so any wiring solution shall abide!

If I clarify "Blue and Green", the above might become less murky. While the Blue and Green names do relate to the physical layout of the coils, I didn't pin a position (like "neck" or "middle") to them in my drawing. Maybe now it's visible why -- I want to rotate which coil plays what role (via the aforementioned multi-pole, 3-position switch). So Blue could mean "the neckward of the two coils [that are being treated by the AND/NOR blend control]" and Green "the bridgeward of the two coils". (The middle coil is neckward in relation to the bridge coil, in my realm).

Building on the example where the Neck coil is the Main coil, here's a walkthrough of the progression of the blend knob:
To the neck coil's signal, the blend knob fully CW (knob=10) should add signal from the middle coil. From there, the effect of twisting it depends on AND/NOR. In AND, from knob=10-5 we should hear the bridge coil's signal fade in, and at 5, all three coils are active; in NOR, we should hear the middle coil's signal get faded out, and at 5, only the neck coil is active. Onward, from 5 down to 0: in AND the middle coil's signal fades out; in NOR, the bridge coil's signal fades in. Now at knob=0, the neck coil and bridge coil signals are blended.

All my draft solutions need 6 poles, minimum.

I tried to avoid long-windedness in my original post. It seems that I end up going too sparse or too wordy. Sorry.

Onward.

It seemed to me that it'd be great if the AND/NOR switch were on the same physical control as the blend knob. That's why I dug that modded pot out of the box.

I haven't bothered drawing a schematic of my 4-pole solution because it breaks the Law. In AND, Clockwise=neckward, but in NOR, Clockwise=bridgeward. Look at the original drawings, and imagine a 4P2T switch that simply swaps the Blue and the Green coils (instead of swapping the two wafers).

Solutions:
- A design more clever than mine
- Mod my switch to 6 poles
- Use my current scofflaw solution and risk a hanging
- Go all Pete Townshend and smash the thing

[stevewf]

Oh, and....
thanks for the super-speedy replies, you two! Faster'n I can bend a string out of tune!

[stevewf]

Dec 6, 2022 1:46:04 GMT -5 stevewf said:

- Use my current scofflaw solution and risk a hanging

Hmmmm. What if I kept the 4P2T solution and just renamed things, achieving consistency that way? Like instead of AND/NOR, how about Pass/Kill? With a switch that controls whose signal gets blended in (pass) and whose gets shorted out (kill).

 That way, an example walkthrough of the blend knob, again with Neck=Main, could be thought of as:
Pass mode:
knob=10: neck coil plus middle coil (middle coil, where the knob is pointed, is passed into the signal)
knob=10 down to 5: neck coil plus middle coil, and bridge coil gets faded in
knob=5: all three coils
knob=5 down to 0: neck coil plus bridge coil, and middle coil gets faded out
knob=0: neck coil plus bridge coil

Kill mode:
knob=10: neck coil plus bridge coil (middle coil, where the knob is pointed, is killed)
knob=10 down to 5: neck coil plus bridge coil, which gets faded out (killed)
knob=5: neck coil only (middle and bridge coils are killed)
knob=5 down to 0: neck plus middle gets faded in (bridge remains killed)
knob=0: neck plus bridge middle [Edit thanks to @angellahash for spotting my error] 

It's only a change in name (i.e. a change in my "law"), so that my brain can cope with the controls. I can live with this. But still, it'd be delightful to see a solution with 4 poles.

[Deleted]

ok wanna find out more whats going on here.
I will take me a few days to catch up but keep on running with it.

is it possible to show what each state looks like in circuit format
maybe a lot of drawings but i think it will show and see if there is any way to match things up to work a way around this

------------------
Trying to ge my head around all this .. so im miles behind but i did come up with the same icon look

Zig Zag US Resister with a line and the Arrow in the Middle look
ahh i see Chris has the design of that Icon

---------------
Reading more im not sure what good it will do
I am intrested in the Switch on the PUSH PULL as the Push pull is normally just a block with the Pull bit just attached at the top so i am wondering how you got them to join and



based on

[stevewf]

The original post would have been too long for my own taste, so I keep adding info along the way instead. Hope that's not frustrating people.

So here's a close-ip of the guitar's control area:

I'm ready to redo all the controls. I don't want to add any new holes, though.

The current vision for what control would go where:
A: a 3-position selector (probably rotary) that chooses which coil is the "main" one - and of course, which two are to be treated by Control B.
B: will blend the two "non-main" coils into the signal. This control is meant to be the focus of this thread.
C: Master Vol
D: Master Treble (cut)
E: Master Bass (cut)

So the player would choose a general scheme by using A, then blend to taste using B, then adjust Vol and tone with C, D & E.

I really feel like A & B should be right next to each other, and that spot up on the upper bout seems like a good one for that. Note that the hole at location A is 1/2-inch diameter, and all the others are 3/8-inch. I don't want to enlarge any of the holes.

Those latter three controls C, D & E -- I'm flexible as to the functions and locations, but I don't want that subject to take over this thread. Before ironing out these controls, I first wanted to explore the feasibility (and usefulness) of A & B.

About  the Control A that I'm proposing, I'll admit that I haven't yet even cracked open a feasibility study for it. Don't even know how many poles I need, but I feel like is likely that I could achieve that function with a rotary switch, as they can offer lots of poles. I think it's also more likely that I'll be able to find a rotary that fits a 1/2-inch hole -- another factor that points to putting A&B in that upper bout. While I don't want Control A to hijack the thread, I see that it might be necessary to solidify this control in order to finalize Control B.

[Deleted]

Neck always on
Pass mode:
10 to 5 = middle
9.9 to 5.1 = bridge gets faded in
5 = all three
4.9 to 0 = middle gets faded out
5 to 0 = bridge

Kill mode:
10 = bridge (middle=killed)
9.9 to 5.1 = bridge, which gets faded out (killed)
5 = neck only (middle and bridge coils are killed)
4.9 to 0.1 = middle gets faded in (bridge remains killed)
0 = Middle

(Guess there was a error there on your post as it says Bridge again)






123 .. 1
456 .. 6
789 .. 7
ABC .. C
4P2T on/on/on where the number to the right is the one that is active in middle state.

Neck hot to 5 with 4 to Output, 6 to Left Pot
Middle hot to 8 with 9 to Right Pot, 7 to B, A to Left Pot and C to Output.
Bridge hot to 2 with 1 to Right Pot, 3 to Output

RIGHT / BRIDGE / OUTPUT
OUTPUT / NECK / LEFT
LINK / MIDDLE / RIGHT
LEFT / LINK / OUTPUT

U) Neck>Output, Middle>Left & Bridge>Right
C) Neck>Left, Middle>Output & Bridge>Right
D) Neck>Left, Middle>Right & Bridge>Output

That will flip your N,M&B around with a toggle look switch than a rotary


i was half a sleep when i dreamed of this 4P2T (On/on/on) for your blend system

[stevewf]

Thanks, @angellahash!

It seems so simple. Why couldn't my brain do that? Sigh.

As to usefulness, I can't guarantee that part. While an understanding of the wiring can help predict, the ultimate test is, as always, done by the ears.

And that ear testing requires a guitar as subject. I wasn't about to tear apart the guitar's wiring until/unless the wiring could be proven feasible. Now ang has done that, at least for that module.

Two more feasibility steps before I dismantle the guitar's present wiring in order to do sound tests:
A) Converting the above solution to series, in case the parallel sounds bad. As mentioned above, my understanding of the loading of coils needs improvement, hence the "in case" in the previous sentence. At a glance, conversion to series seems simple, but my brain is apparently busy being a tire chock .

B) Designing another control (and possibly planning the creation of that control) that places the three coils properly, i.e. which coil becomes "main" and which coils get assigned CW/CCW on the blender.

PS - thanks, ang, for also spotting a typo above, which I've now corrected in the post.

PPS - I'll try to add a post somewhere about that Push/Pull switch, i.e. how I made 2P2T+2P2T = 4P2T. I think I have som photos of that. Maybe not in this thread, though.

[Yogi B]

Dec 6, 2022 1:29:01 GMT -5 stevewf said:

However, let's not presume that the whole guitar will consist of two coils only. This time, I'm aiming this module at a guitar that has three coils. Perhaps I could have avoided a detour by explaining the context in the first place... but I didn't want to change the focus away from the module in question.

Context, or the external effect of a module, is as critically important to the overall intended operation as the module's own internal effects. (e.g. an on/off switch which works by shunting a pickup is fine on a single pickup guitar or as a global kill-switch but entirely inappropriate for selecting from multiple coils wired in parallel.)

Dec 6, 2022 3:20:33 GMT -5 stevewf said:

That way, an example walkthrough of the blend knob, again with Neck=Main, could be thought of as:
Pass mode:
knob=10: neck coil plus middle coil (middle coil, where the knob is pointed, is passed into the signal)
knob=10 down to 5: neck coil plus middle coil, and bridge coil gets faded in
knob=5: all three coils
knob=5 down to 0: neck coil plus bridge coil, and middle coil gets faded out
knob=0: neck coil plus bridge coil

Kill mode:
knob=10: neck coil plus bridge coil (middle coil, where the knob is pointed, is killed)
knob=10 down to 5: neck coil plus bridge coil, which gets faded out (killed)
knob=5: neck coil only (middle and bridge coils are killed)
knob=5 down to 0: neck plus middle gets faded in (bridge remains killed)
knob=0: neck plus bridge middle [Edit thanks to @angellahash for spotting my error]

Wait, I thought you specifically didn't want the above behaviour (wherein the outer selections are differing) and instead wanted the outer pickup selections to be equal? That is, for example, neck & middle when at 10, and neck & bridge when at 0 — regardless of the setting of the blending mode toggle.

[stevewf]

Dec 7, 2022 8:58:09 GMT -5 @angellahash said:

Wow, all in a DPDT. For me, I would want to switch the two wafers so that Sweeper-to-Left applies to the Middle coil. Impressive nonetheless.

I prefer the 4P2T diagram, @angellahash, because it obeys the so-called Law that I tried to decree in the OP. That is, "Regardless AND or NOR, turning the knob CW increases the neckward coil or decreases the bridgeward coil." The 4P2T solution adheres to this, and so may be named "AND/NOR"; the 2P2T solution falls under the "PASS/KILL" type.

Both solutions are elegant, so thank you! Now I'm trying to steal a block of time to see if I can convert them to a serial blended pair, in case that's needed... and indeed, perhaps the capability to switch between serial and parallel could be useful, depending on some future listening tests.

[stevewf]

Dec 7, 2022 16:33:19 GMT -5 Yogi B said:

Context, or the external effect of a module, is as critically important to the overall intended operation as the module's own internal effects. (e.g. an on/off switch which works by shunting a pickup is fine on a single pickup guitar or as a global kill-switch but entirely inappropriate for selecting from multiple coils wired in parallel.)

I wouldn't contradict that example, for sure. I had figured (perhaps erroneously) that in this case, context didn't matter as long as there was a + and a - at the interface. Module-like, whatever happens within that +- can be put in parallel or series, in phase, HOoP or however, with whatever else lies outside of it, with all the benefits and pitfalls that come with it.

Wait, I thought you specifically didn't want the above behaviour (wherein the outer selections are differing) and instead wanted the outer pickup selections to be equal? That is, for example, neck & middle when at 10, and neck & bridge when at 0 — regardless of the setting of the blending mode toggle.

I've been waffling. Yes, that was the original "law" that I'd royally decreed, thank you for understanding. That's related to my "AND/NOR" mindset and the hardware at hand. What you see in the "PASS/KILL" mindset is me conceding that I couldn't design a solution that behaves "correctly" with fewer than 6 poles, and so instead renaming the function for mnemonics at playtime. @angellahash has now done better, so I can now affirm my intent to stick to the behavior that you described. Sorry for my back-and-forth.

[stevewf]

Dec 7, 2022 8:58:09 GMT -5 @angellahash said:

123 .. 1
456 .. 6
789 .. 7
ABC .. C
4P2T on/on/on where the number to the right is the one that is active in middle state.

Neck hot to 5 with 4 to Output, 6 to Left Pot
Middle hot to 8 with 9 to Right Pot, 7 to B, A to Left Pot and C to Output.
Bridge hot to 2 with 1 to Right Pot, 3 to Output

RIGHT / BRIDGE / OUTPUT
OUTPUT / NECK / LEFT
LINK / MIDDLE / RIGHT
LEFT / LINK / OUTPUT

U) Neck>Output, Middle>Left & Bridge>Right
C) Neck>Left, Middle>Output & Bridge>Right
D) Neck>Left, Middle>Right & Bridge>Output

That will flip your N,M&B around with a toggle look switch than a rotary

Wow.
@angellahash, you're prescient. Or, you edit your earlier posts 
In either case, it's "wow" here. I'm gonna have to diagram that in order to get a feel for how it works. Thanks!

It just so happens that I own one of these (ebay):

which is a 4P2T 3-position switch that will fit exactly in the guitar's existing "A" position (1/2-inch hole).

A hitch is that it's On-Off-On:

but I may be able to tweak it into behaving like the one you've described (which I think is typical 4P2T On-On-On toggle switch behavior).

On the other hand, I was thinking about using HalfOutOfPhase for some combos. And also toying with a system Serial/Parallel (meaning switching between Main + Blended vs Main x Blended). Those features would necessitate more poles, I think. I'm not 100% fixed on those features, at least not until a test hearing.

About phasing, the magnets are:
Neck P90=North
Neck Rail=South
Bridge P90=South

I haven't done any listening yet of how the coils sound together. But with the designs so far, I'm getting closer to undoing the current guitar, putting gin the coils in question, and then I can Ash Lab it for ear tests before deciding what OoP/HOoP and Ser/Par could/should be wired in.

[Deleted]

4P2T (On/On/On)





looking at that switch im not sure how to change it to suit
i would guess to when its in MIDDLE to connects to one of the sides
depends if can stop it moving from Middle Postion to another side, if you could stop it then i guess it would work

the connection Arms.. i see im going to have to buy one to take it apart to see what i can do with it

[stevewf]

That's nice (and quickly produced, again). Thanks!

I'm thinking of starting a different thread for the selector control, because the original blender request has been fulfilled (see angellahash's post, above, with a 4P2T solution)! I'll see if I can get to that new thread tonight.