|
|
Post by pablogilberto on Jun 12, 2020 5:48:47 GMT -5
|
|
|
|
Post by newey on Jun 12, 2020 7:02:49 GMT -5
Sure sounds that way from the description. He discusses Fender's noiseless SCs, which emply a second coil below the string-sensing coil (there are many other types that do this as well). He says they didn't want to do that because it affects the sound somewhat (it does).
Years ago here, someone did a lot of research on using a dummy coil (the thread is around here somewhere, if you search way back for "dummy
coil"). IIRC, the upshot of those experiments was that positioning of the dummy coil made a big difference in the amount of noise reduction he achieved.
Some years ago, someone also posted a commercially available dummy coil that was just a few strands of wire that ran around the whole underside of a Strat pickguard, so as to give a good deal of length to the dummy coil without taking up space. I forget who made that.
|
|
|
|
Post by newey on Jun 12, 2020 7:05:42 GMT -5
|
|
|
|
Post by JohnH on Jun 12, 2020 15:24:38 GMT -5
the dummy coil system that I remember discussing is by Ilitch: www.ilitchelectronics.com/product/bpncs-fender-stratocaster/They do it with a large area of coil with a fairly small number of turns, built into a special backplate on the guitar. The concept has a nice basis in theory. The trouble with most dummy coil systems and stacked systems for hum cancelling is that to get enough reverse hum induced in the extra coil to cancel out the hum in the main coil, it usually needs a second coil of a comparable number of turns. Hence, in addition to the hum cancelling, we get extra resistance and inductance and this changes the tone. The wide Ilitch coil collects the required amount of reverse hum with relatively few turns (signal induced is dependent on flux x area x turns). This also promotes low inductance and low resistance too. There is actually not that much wire length involved, so there is not much tonal effect when it is added to the main coil. |
|
|
|
Post by thetragichero on Jun 12, 2020 15:46:55 GMT -5
is it same number of turns or same surface area/mass/some other physical measurement of the coil?
reason I'm asking is that i have some 26 gauge wire I'm looking to use for my dummy coil project.... that or some crazy thick wire from a shorted ups transformer (as i flipped the switch to measure the secondary voltage it tripped a breaker)
|
|
|
|
Post by JohnH on Jun 12, 2020 19:08:33 GMT -5
is it same number of turns or same surface area/mass/some other physical measurement of the coil? reason I'm asking is that i have some 26 gauge wire I'm looking to use for my dummy coil project.... that or some crazy thick wire from a shorted ups transformer (as i flipped the switch to measure the secondary voltage it tripped a breaker) Sure have a go! I think there's scope for a good DIY version. So I reckon the basic formula is to keep turns x area constant. So let say you are trying to balance against a normal coil of 4000 turns, and say its equivalent to 60mm x 10mm, but you are making a dummy coil that is 200mm x 150mm. Then the turns should be something like: 4000 x (60x10) / (200x150) = 80 But, lots of tweaking needed around that. Who knows what the real effective area of a coil with a core really is in terms of how much hum flux it sucks in? This coil would go in series with the main coil. It could have a pot to control it. |
|
|
|
Post by stevewf on May 31, 2022 14:19:04 GMT -5
So I reckon the basic formula is to keep turns x area constant. Thanks for the ballpark method! Lemme see if I can apply it to my Strat with P-90's in it. I want see how many turns I'd need in a dummy coil that lines the pickups' "bathtub" cavity. That bathtub is about 85mm x 145mm (excludes the control cavity). Some givens: The bridge P-90 probably has about 10,000 turns, if my my wiki-research is accurate. My eye measured its coil (under the pickup cover) at about 75mm x 25mm. Let me try to apply the formula: (10000 x 75 x 25) / (85 x 145) = 1521 turns. Lacking a winder (and a cavity-shaped bobbin, for that matter), it looks impractical. Rats. Might have to skip this idea, unless Nutz can help. Is my formula applied correctly? |
|
|
|
Post by MattB on May 31, 2022 17:40:39 GMT -5
P90 coils are really wide. You should probably be using an average value for the coil area.
 |
|
|
|
Post by stevewf on May 31, 2022 21:11:34 GMT -5
Thank you of posting the nice photo. It shows more than the glance I got of my p90's coil. Here's an attempt at refining my measurements on that basis plus measuring my own P90's cover for scale.
The bundle of wire appears to be about 17mm thick where it's pulled taught on the ends, and a touch wider along the long sides. I figure the coil width to span from about 10 to 29mm, and the length to span from 58 to 75mm . It'd be tempting to take the halfway points of both... and I don't see a problem with doing that, as the number of turns inside the half-thickness mark will be the same as outside of it. So the new dimensions: 19.5mm x 66.5mm
So the formula with refined parameters (this time I'll include the units for clarity): (10000 turns x 19.5 mm x 66.5 mm) / (85 mm x 145 mm) = 1052 turns.
Better, but not enough to motivate me to try to wind it around a bobbin that has yet to be created. It it got down to 100 turns or so I might make a stab...
...but not until I also calculate what gauge coil wire to buy, given space constraints (thinner wire=better) versus reduced resistance (thicker wire=better)...
...and also not until I convince myself that the dummy coil really does have to be in serial with the real coil(s). I need convincing because of humbucker coils in parallel, which, despite not being in series, do cancel hum; what hum-canceling principles are applying in that case, versus the principles applying to real + dummy coils?
|
|
|
|
Post by JohnH on May 31, 2022 21:33:24 GMT -5
Id say the rough maths look ok, but it definitely needs testing and fine-tuning in order to get a null in the hum. Only a precise match will really zero the hum and that can only be done by tweaking in some way.
Parallel wiring of the dummy coil is not good. Two equal pickup coils work in parallel by providing an average of the two signals, and cancelling out hum. But if the coils are unbalanced, then the signal contributions are weighted in favour if the low impedance coil. If that was dummy coil, producing no signal but with very low impedance, then the signal from the main coil is greatly reduced.
But in series, its just an addition, a sum of the two signals. So the signal from the main coil is added to no signal from the dummy, resulting in not much change, but with hum cancelled.
|
|
|
|
Post by stevewf on Jun 4, 2022 10:05:43 GMT -5
Parallel wiring of the dummy coil is not good. Two equal pickup coils work in parallel by providing an average of the two signals, and cancelling out hum. But if the coils are unbalanced, then the signal contributions are weighted in favour if the low impedance coil. If that was dummy coil, producing no signal but with very low impedance, then the signal from the main coil is greatly reduced. Ah, so. If I absorb correctly, JohnH points out that imbalanced impedance will result in one coil's signal -- and hum -- dominating the pair. That addresses the 2nd part of my question, underlined here: " what hum-canceling principles are applying in that case [of the parallel HB coils], versus the principles applying to real + dummy coils?" Here's a try at comparing it to the first part of my own question: In contrast to a real & dummy coil pair, a humbucker usually features two coils that are identical (or at least very close to identical), so they also have identical # of turns and impedance; with no coil whose impedance is [relatively] low, no coil dominates the pair. Since the # of turns are similar, the hum and anti-hum levels are also similar, which results in a balance between hum/anti-hum. |
|
|
|
Post by ashcatlt on Jun 4, 2022 10:33:58 GMT -5
But in series, its just an addition, a sum of the two signals. Impedance doesn't really play into the series situation except to the extent that the impedance is tied to the actual signal level. In a string-sensing coil, the output voltage usually rises with the impedance, but if you can get the same actual voltage out of a smaller Z, then the signals will balance out despite the Z difference. You can look at the parallel connection as a voltage divider. Or, I guess, two voltage dividers. From the perspective of one pickup, it is the "top resistor" and the other coil is the "bottom resistor". Each pickup's contribution is divided down by the appropriate ratio and the results are added together. The smaller impedance is divided down less than the bigger one because that's how voltage dividers work. |
|
|
|
Post by stevewf on Jun 6, 2022 9:02:54 GMT -5
Impedance doesn't really play into the series situation except to the extent that the impedance is tied to the actual signal level. In a string-sensing coil, the output voltage usually rises with the impedance, but if you can get the same actual voltage out of a smaller Z, then the signals will balance out despite the Z difference. Thanks! Curious, and still learning: Given the same string motion, can you two coils that would produce equal/similar voltages but with different Z? What physical differences would there be? |
|
|
|
Post by ashcatlt on Jun 8, 2022 10:15:22 GMT -5
Well…sort of??? We often talk about Impedance as though it was just one simple number, but it’s actually complex. And I mean complex as in there are imaginary numbers involved, but also as in it’s usually frequency dependent. The things that one might do to change the impedance of a coil could also change the frequency response. So that kind of brings up the question of what you mean by “equal/similar voltages”. But maybe I’m thinking too hard on that. Rickenbacker single coils use fewer turns of smaller wire in order to get hotter overall output without quite so much inductance.
Edit - So but honestly I’m not sure how far you can practically take it. You can change impedance by changing number of turns or wire gauge, but that also changes voltage output. You can try to compensate for that by changing the strength of the magnet, but I’m not sure how far you can really go with that to create a significant Z difference while keeping V about the same.
|
|
