|
|
Post by ms on Dec 22, 2018 12:06:46 GMT -5
The SD SH4 is a 16K pickup (overwound, no highs when loaded with a typical cable) with a lot of eddy current losses. The first attachment shows that the real part of the impedance (with the effect of the C removed, green line) has more than doubled by 3000 Hz.  The purpose of this thread is to once again make a comparison between the gain as measured with an exciter coil and the gain computed from the measured impedance. The impedance is used to make a voltage divider. The series leg is the measured impedance with the impedance of the C removed, and thus includes the eddy current effects from current induced in the pickup circuit. The shunt leg is the measured pickup C in parallel with the input impedance of the Apogee Element 24 used in the exciter coil gain measurement, about 5.25ZM, 210pf. That is, no buffer is used in the measurement with the exciter coil, but instead the gain from the impedance is compensated. The two results are close, indicating that there is not a lot of loss from the "shielding" effect of eddy currents, at least out to 7KHZ. This is not surprising since there is no cover and the base plate low conductivity.  |
|
|
|
Post by stratotarts on Dec 22, 2018 12:17:57 GMT -5
That's a really valuable confirmation. To make it a complete statement, though, I think you would have to also perform the same process on a pickup with expected cover/baseplate losses. Forgive me if you've already done that elsewhere, and I've missed it. This might reveal the extent of coil to cover eddy current coupling (vs. exciter to cover coupling).
This is kind of cool because it would allow you to reliably separate cover/baseplate losses from what I assume are pole material losses without a complete pickup disassembly.
|
|
|
|
Post by ms on Dec 23, 2018 14:01:36 GMT -5
That's a really valuable confirmation. To make it a complete statement, though, I think you would have to also perform the same process on a pickup with expected cover/baseplate losses. Forgive me if you've already done that elsewhere, and I've missed it. This might reveal the extent of coil to cover eddy current coupling (vs. exciter to cover coupling).
This is kind of cool because it would allow you to reliably separate cover/baseplate losses from what I assume are pole material losses without a complete pickup disassembly.
Good idea  The cover used is nickel silver. At the 3.2KHz peak, the gain from the impedance is a little over 1 db down from the gain without the cover. The gain from the exciter is just over another 1.5 db down from that. So both the coupling of current in the coil to the cover and the direct shielding effect of the cover matter, but the latter is larger. |
|
|
|
Post by ms on Dec 31, 2018 17:04:47 GMT -5
Now let's look at the effect of the steel around and behind a telecaster bridge pickup. First the impedance:  The difference with and without steel is very much like the difference between an Alnico single coil and a humbucker. The other thing to look at is the gain from exciter coil and computed from the impedance:  With no steel there is no significant difference between the exciter coil and impedances predictions of gain, until very near the resonance. With the steel, the gain from the exciter coil falls off with increasing frequency near the bottom, like a humbucker with a lot of eddy current effect. It would appear that eddy currents in steel go a long way to explaining how two guitars, tele and strat, both with maple necks and alnico single coils, can sound so different when using the bridge pickup. |
|
|
|
Post by antigua on Jan 1, 2019 23:44:30 GMT -5
I saw a much smaller effect upon the resonance when I tested a Fender NoCaster with and without the included steel base plate. It only decreased by about 2dBV, from about 13dBV down to 11dBV. Usually PAF type humbuckers show a resonance no greater that 5dBV, without a cover, when Ive tested them.  |
|
|
|
Post by ms on Jan 2, 2019 6:59:29 GMT -5
I saw a much smaller effect upon the resonance when I tested a Fender NoCaster with and without the included steel base plate. It only decreased by about 2dBV, from about 13dBV down to 11dBV. Usually PAF type humbuckers show a resonance no greater that 5dBV, without a cover, when Ive tested them. My "with steel" measurement includes both the steel base plate and the pickup is mounted in a steel bridge-tailpiece assembly. Did you use both? I found that the bridge has the larger effect, and so I just lumped the two effects together. Also, this is not a true Fender bridge, but some brand X I had lying around; the steel might be a bit different. The pickup is one I wound from a kit years ago, not a true tele bridge. I do not know about the size of the peak. Maybe there is something off in my calculations. The fact that two different kinds of measurements agree suggests that they both are OK, but that does not prove it, of course. Also, I am not using a buffer with the exciter coil measurements, and so I have a bit more than 200pf of input capacitance. I include this in calculation of the gain from the impedance measurement for the comparison. |
|
|
|
Post by antigua on Jan 2, 2019 12:40:11 GMT -5
I missed the part about it also featuring the surrounding mounting plate. It would be good to have a picture of the plates, or a measurement of their thickness, because I've noticed that both vary widely, and 'm certain the heavier damping coincides with thicker plates. Most of the newer Fender Teles come with a rather thick, flat link this:  Where as the vintage "ashtray" with the wrap around sides looks to be somewhat thinner:  And then just for a laugh, I saw this comically thick base plate from armadilloguitar.com  |
|
|
|
Post by ms on Jan 2, 2019 14:52:10 GMT -5
The skin depth of steel is about 1 mm at 100 Hz, as explained and illustrated here: en.wikipedia.org/wiki/Skin_effect. Of course it gets smaller as frequency increases. Bridge plates thicker than 1 mm will all be similar in effect at all guitar frequencies. . |
|
|
|
Post by ms on Jan 2, 2019 18:12:31 GMT -5
The bridge plate is about 1.06 mm thick; so you would not expect much difference from a thicker plate. Less effect at lower frequencies from a thinner plate.
|
|
|
|
Post by antigua on Jan 2, 2019 22:30:00 GMT -5
|
|
|
|
Post by stratotarts on Jan 3, 2019 8:38:31 GMT -5
I proposed an untested mod to eliminate that eddy loss. I'm sure it will work. But it doesn't make practical sense, because the effect is so small:
I suppose you could at least say that a non-wraparound bridge like a Strat hardtail doesn't have this effect.
|
|
|
|
Post by ms on Jan 3, 2019 9:14:37 GMT -5
There is more to it than the thickness. In the test you reference, one plate was magnetic, the other not. The permeability decreases the skin depth. Also the conductivity was different, I believe, and this affects the skin depth. So the overall effect is quite complicated. It is simple only if you compare plates of two thicknesses made from the same material. |
|
|
|
Post by ms on Jan 3, 2019 9:18:32 GMT -5
I proposed an untested mod to eliminate that eddy loss. I'm sure it will work. But it doesn't make practical sense, because the effect is so small:
I suppose you could at least say that a non-wraparound bridge like a Strat hardtail doesn't have this effect.
I do not think the effect is small. It makes the impedance of an Alnico single coil look more like a ;pickup using steel cores, etc., such as a humbucker,. |
|
|
|
Post by antigua on Jan 3, 2019 11:18:51 GMT -5
That's a good point about the steels having differing permeability, but what about the coupling coefficient between the pickup coil and the theoretical eddy current coil? Wouldn't that ratio increase, having more of the flux path of one enveloped by the other?
|
|
|
|
Post by ms on Jan 3, 2019 12:08:20 GMT -5
That's a good point about the steels having differing permeability, but what about the coupling coefficient between the pickup coil and the theoretical eddy current coil? Wouldn't that ratio increase, having more of the flux path of one enveloped by the other? I would expect a thicker permeable plate to have a greater effect on the inductance, but not much beyond that. But I am not sure that is the whole story. |
|
