|
|
Post by geo on Jul 19, 2022 6:50:32 GMT -5
 Updated the Rhythm control to switch between 500k and 250k tone+volume controls. Just a quick update, hopefully I didn't make any errors. Looked into automatically fixing the phase so the top pickup and bottom pickup always noise cancel. I know I can't do it by automatically flipping the bottom coil of the top pickup and the top coil of the bottom pickup (when either is selected alone), but I might be able to do it by automatically flipping the top coil of the top pickup and the bottom coil of the bottom pickup (when either is selected alone). |
|
|
|
Post by geo on Jul 19, 2022 7:38:09 GMT -5
Alright, I set up the automatic hum-cancellation in all positions. Updates since 4.1: - Rhythm tone automatically switches to 500k pots when humbucker is connected, otherwise 250k - Lead has hum-cancellation in all pickup selections  Yogi B, please let me know if I screwed something up again! |
|
|
|
Post by Yogi B on Jul 19, 2022 20:19:03 GMT -5
Yogi B, please let me know if I screwed something up again! I think it's functionally fine, but there is at least one thing I'd change. With S2 (bridge mode select) now being a four pole switch you can completely disconnect the unused coil in the split positions, something like this: | C | lower − | lower + | upper − | upper + |
|---|
| 1 | | | purple | grey |
|---|
| 2 | purple | grey | purple | grey |
|---|
| 3 | purple | link together | grey |
|---|
| 4 | grey | purple | | |
|---|
Also, I'd consider swapping the 500k/250k switching around such that ground was always connected and the output (red) wires are switched instead. |
|
|
|
Post by geo on Jul 19, 2022 23:00:42 GMT -5
When using the neck and bridge pickups with both coils split, swapping the phase works, however since my switching lets you grab either pickup, you just want the outside coils (topmost and bottommost) or the inside coils (S of neck and N of bridge, by your nomenclature). You could do two north coils or two south coils, but as you've pointed out, you'd need to invert the phase or you'll get noise. My point is that it should be possible to have that happen automatically: i.e. you select a north coil & and south coil while in phase, then just by flipping the phase switch one of those coils is swapped out to get either two norths or two souths. It turns out you were right about this! (Unless I've made a mistake.) And it was easier than expected; since each pickup only exposes unpaired coils in two modes, it's as simple as inverting that selection in half of the modes on each; the neck pickup always looks like a south coil, the bridge pickup always looks like a north coil. (Unless you swap the phase, in which case the bridge always looks like a south coil as well.) This is done by swapping + and - of the top coil split of the neck humbucker and swapping + and - of the bottom coil split of the bottom humbucker. Doesn't affect the rhythm mode at all. Double, triple checked my work and I think it's actually done. I'll try to write up a clean explanation and a parts list, then put this in the finished designs forum. In this post you talked about disadvantage to just adding a resistor to switch from 500k to 250k, but I wasn't able to follow the explanation. I'd been kicking this idea around for the Lead circuit, but it sounds like you're saying it won't work as I would hope. |
|
|
|
Post by geo on Jul 19, 2022 23:05:19 GMT -5
Yogi B, please let me know if I screwed something up again! I think it's functionally fine, but there is at least one thing I'd change. With S2 (bridge mode select) now being a four pole switch you can completely disconnect the unused coil in the split positions, something like this: | C | lower − | lower + | upper − | upper + |
|---|
| 1 | | | purple | grey |
|---|
| 2 | purple | grey | purple | grey |
|---|
| 3 | purple | link together | grey |
|---|
| 4 | grey | purple | | |
|---|
Also, I'd consider swapping the 500k/250k switching around such that ground was always connected and the output (red) wires are switched instead. Good suggestion! I'll try to do that. One less antenna on this thing... Keeping ground connected instead of the hot out makes sense, since that way you can ground things to the body of those pots, it just didn't look as pretty on my diagram. I'll do it over and do it right. |
|
|
|
Post by geo on Jul 21, 2022 10:04:29 GMT -5
Wait, I think I have it wired so if the neck has both coils in parallel and the bridge has both coils in parallel, the hum will add instead of cancelling...
|
|
|
|
Post by newey on Jul 21, 2022 11:04:56 GMT -5
Wait, I think I have it wired so if the neck has both coils in parallel and the bridge has both coils in parallel, the hum will add instead of cancelling... A humbucker "bucks hum" whether wired in series or in parallel. Two parallel humbuckers, wired together in parallel, won't be any noisier than 2 series humbuckers wired together in parallel. While hum-cancellation is never perfect, in a design such as this we are really only concerned about noise with splitting the coils of the humbucker. The "full HB" settings should be just as quiet as, well . . . 2 humbuckers.  |
|
|
|
Post by geo on Jul 21, 2022 13:56:42 GMT -5
Oh! You're absolutely right. I'm sorry, my brain is melting. The coils are still wound opposite, so it doesn't matter if you're adding the current from the inductors in parallel or in series; the noise will cancel.
|
|
|
|
Post by Yogi B on Jul 22, 2022 5:45:46 GMT -5
My point is that it should be possible to have that happen automatically: i.e. you select a north coil & and south coil while in phase, then just by flipping the phase switch one of those coils is swapped out to get either two norths or two souths. It turns out you were right about this! (Unless I've made a mistake.) And it was easier than expected; since each pickup only exposes unpaired coils in two modes, it's as simple as inverting that selection in half of the modes on each; the neck pickup always looks like a south coil, the bridge pickup always looks like a north coil. (Unless you swap the phase, in which case the bridge always looks like a south coil as well.) I was talking about making the phase switch also swap which coil is selected, rather than what you have where switching to the other coil swaps the phase — similar, but different. Basically along the lines of this, but somehow combining the binary switching into your single 4-position mode switch. The resistor trick only gives exactly equivalent results when the controls are at 10 (and technically, dependent on the variation used, also zero). Since the standard wiring only uses the pot as a variable resistor, it's easiest to illustrate this with a tone control. Assuming a regular 10% log taper, a 250k pot at 50% rotation is at 25k. With the 500k pot plus 500k resistor combination you'd need to turn the pot further, down to approximately 30% (basically the same as without the resistor), in order to get the same resistance. When it comes to volume controls, the difference is lesser because the volume reduction is primarily based on the ratio of resistances which is essentially unchanged. Therefore, other than the change in loading — which also affects the severity to which cable capacitance acts as the volume is reduced, and where 250k actually performs better — the only real difference amounts to the fact you have a reduced variable resistance in parallel with a treble bleed, if present. I don't know what you do hope for, but in my opinion the 500k/250k thing is often a hypercorrection (borrowing that term from linguistics) — a split or paralleled humbucker is not an AlNiCo single coil. (The reason why my opinion may seem to differ here versus the thread you linked is that there the swap between the two values wasn't tied to pickup selection — the difference, in and of itself, being seen as a worthwhile 'effect'.) |
|
|
|
Post by geo on Jul 22, 2022 7:16:29 GMT -5
The resistor trick only gives exactly equivalent results when the controls are at 10 (and technically, dependent on the variation used, also zero). Since the standard wiring only uses the pot as a variable resistor, it's easiest to illustrate this with a tone control. Assuming a regular 10% log taper, a 250k pot at 50% rotation is at 25k. With the 500k pot plus 500k resistor combination you'd need to turn the pot further, down to approximately 30% (basically the same as without the resistor), in order to get the same resistance. Yeah, these parts I was able to figure out pretty well on my own. When it comes to volume controls, the difference is lesser because the volume reduction is primarily based on the ratio of resistances which is essentially unchanged. Therefore, other than the change in loading — which also affects the severity to which cable capacitance acts as the volume is reduced, and where 250k actually performs better — the only real difference amounts to the fact you have a reduced variable resistance in parallel with a treble bleed, if present. Can you elaborate on this? I assume this is where the darker/brighter bit is coming from, but this is the part that I have trouble understanding. I get how the potentiometer and series/parallel resistance bit works; I'm not so clear on cable capacitance and the effect of loading on it. I don't know what you do hope for, but in my opinion the 500k/250k thing is often a hypercorrection (borrowing that term from linguistics) — a split or paralleled humbucker is not an AlNiCo single coil. What would you advise? |
|
|
|
Post by Yogi B on Jul 23, 2022 5:30:00 GMT -5
When it comes to volume controls, the difference is lesser because the volume reduction is primarily based on the ratio of resistances which is essentially unchanged. Therefore, other than the change in loading — which also affects the severity to which cable capacitance acts as the volume is reduced, and where 250k actually performs better — the only real difference amounts to the fact you have a reduced variable resistance in parallel with a treble bleed, if present. Can you elaborate on this? I assume this is where the darker/brighter bit is coming from, but this is the part that I have trouble understanding. I get how the potentiometer and series/parallel resistance bit works; I'm not so clear on cable capacitance and the effect of loading on it. I probably should have said output impedance, rather than loading, to make sure I was including the series component (which is the relevant part to this discussion). Ignoring the 'lower' half of the volume pot, the 'upper' half is in series with the signal and is followed by the cable capacitance between the signal and ground: this is a capacitive low pass filter, i.e. what is shown below. Note that the cutoff frequency is inversely proportional to the series resistance, R. And also that, except at maximum, the series resistance is always larger with a 500k pot versus a 250k pot. Obviously this is a small part of a more complex circuit, but is the principle that underlies the overall effect. For a sense of the difference this makes to the complete circuit, here's a plot of the output a PAF-esque pickup through the two volume setups into the standard 1Meg/470pF load: the coloured trace is 250k, the grey is 500k plus additional 500k resistance. (You can interpret this as purely volume controls without tones, or where the tones are present and maximised but the 250k volume is paired with a no-load tone.) The line colour changes every 10% (e.g. bright red lasts from 100%—91%). The maximum difference is around 3 or 4dB, not huge, but definitely not negligible. Well, I'd just keep the nominal 250k (combined 500k volume & 500k tone) across all positions. I don't really understand the logic behind attempting to get a brighter sound via splitting only to simultaneously partially undo that same change via placing more load on the pickup — it just leaves you with less options. Side note: personally I much prefer the taper of a 250k tone pot paired with any pickup, so I'd actually go with the aforementioned 250k volume & 250k no-load tone. Additionally, if I were using an AlNiCo single with a HB on the same guitar I'd do the same, but specifically use a homemade no-load pot (where lug 3 is connected but only when rotated to maximum, i.e. by cutting a single slice through the track rather than removing/insulating the entire end) and use that to add the 250k-ish resistor when the single was selected. |
|
|
|
Post by geo on Jul 24, 2022 13:26:26 GMT -5
Alright, please tell me I got it this time! Per Yogi B's advice: - Rhythm pots can be soldered to ground - Bridge pickup connects no extra antennas  (The bit about the bridge pickup carrying no extra antennas is slightly dishonest, because whatever part of the bridge pickup is selected on S2 will still touch the ground. We could theoretically use the spare 2 poles on S3 to fully disconnect whichever pickup is inactive, and the only extra antenna would be the bottom coil of the neck pickup, but that feels like a lot of unnecessary soldering.) It looks like you could also use MSSA4450 if you can't get SS44D04G4's at some point in the future, but MSSA4450's don't have common poles, so you'd have to tweak the circuit. S44D04G4 is discontinued as of 2020, but DigiKey has a couple thousand left. I'll see if I can get a PCB schematic for this, since S44D4G4's are through-hole mounted; you can put standoffs on the PCB and mount that to the control plate (or pickguard), rather than soldering these switches to the control plate and trying to solder a dozen connections to the tiny little pins. Just in the unlikely scenario that anybody else ever wants to build one of these things. |
|
|
|
Post by geo on Jul 28, 2022 17:42:35 GMT -5
Side note: personally I much prefer the taper of a 250k tone pot paired with any pickup, so I'd actually go with the aforementioned 250k volume & 250k no-load tone. Additionally, if I were using an AlNiCo single with a HB on the same guitar I'd do the same, but specifically use a homemade no-load pot (where lug 3 is connected but only when rotated to maximum, i.e. by cutting a single slice through the track rather than removing/insulating the entire end) and use that to add the 250k-ish resistor when the single was selected. I often play with tone all the way open or most of the way open (rarely lower than a 7), but I'm a little reluctant to cut a no-load pot because that sound is raw, and just a tiny bump in either direction on the tone knob and you're either cutting the sound or straight to the no-load output. I'm a big fan of switches and toggles; safe things that an oaf like me won't mess up. Hair triggers are for people with better discipline and coordination. |
|
|
|
Post by geo on Jul 31, 2022 20:16:47 GMT -5
Wait, am I going to have phase issues from my hum cancelling scheme? I'm assuming top of each humbucker is an N magnet with positive winding.
S1:
N- (phase problem!)
N+S-
N+S-
S-
S2:
N+
N+S-
N+S-
S+ (phase problem!)
So in those hum-canceling positions, the signal is going to be out-of-phase... I should just deal with the hum and use SZ to switch phase, yeah?
|
|
|
|
Post by geo on Aug 3, 2022 10:52:43 GMT -5
OK, did the proper cuts on the rhythm tone section (leaving ground connected at all times) and reverted the hum-canceling that swapped phases.  It's done. I'll wait until PCBs are printed to upload a full spec, because even with the simplified circuit, you don't want to wire this without a breakout board at the very least. |
|
|
|
Post by geo on Jan 12, 2023 15:06:18 GMT -5
Not printing a PCB for this, so I guess it's all done. Just have to write it up neatly! A lot of switch combinations, and all in the original Jaguar profile.
|
|
|
|
Post by gckelloch on Mar 13, 2023 20:30:06 GMT -5
You sure you need 250k pots for the SC and parallel modes? Fe core pickups like PAF's are down a few dB by 4kHz as well as having a lower Q than a typical Fender SC. You might try just a 500k and 250k pot for everything. I bet that would be fine. I have an HB35+ guitar. The pickups are very close to the SD PAF's you want, including the Ni covers and weakened (maybe rough cut or unoriented) AlNiCo V magnets. The 24.75" scale guitar has a Mahog center block and a Maple neck with a Pau Ferro FB (sounds a bit brighter than Indian RW). It has 500k pots and coil splitting. The coil split sounds are nice with 500k pots-- sweet and Chimey. I could see changing one pot to 250k, but I think it would be too dark with two 250k.
|
|
|
|
Post by geo on Apr 18, 2023 20:03:16 GMT -5
You can get 2x 500k by splitting the coil on the lead circuit instead. There's really a lot of flexibility here.
|
|
|
|
Post by kitwn on Apr 23, 2023 23:37:33 GMT -5
You can get 2x 500k by splitting the coil on the lead circuit instead. There's really a lot of flexibility here. geo, You did say 'mad scientist' in your other recently updated thread and I have to agree. I'm going to need some time and a very large mug of tea to work my way through all the above diagrams.
Kit
|
|
|
|
Post by geo on Aug 15, 2024 8:35:13 GMT -5
OK, did the proper cuts on the rhythm tone section (leaving ground connected at all times) and reverted the hum-canceling that swapped phases. It's done. I'll wait until PCBs are printed to upload a full spec, because even with the simplified circuit, you don't want to wire this without a breakout board at the very least. If anyone with software and experience would be willing, could they please turn this into a more traditional wiring diagram for use with SS44D04G4 switches? (Pin out below.) Having trouble explaining the wiring to a luthier.  |
|
