A failed test

[ashcatlt]

And now for today's puzzler.

I had this great idea. I was going to conduct a semi-scientific test to compare the output level and frequency response of the l'il killers in my HHH strat. The idea was to measure the differences on one pickup between internal series, parallel, and single. I was also going to try with different length cables and even through a whole string of buffered pedals to see what we could see about that.

Of course, I didn't figure that I'd be able to hit the strings consistently enough to give any real meaningful results. If I could find my e-bow, that might work fine. Ever have one of those deals where you lose something, you know exactly where it was when you saw it last, but it ain't there, so you keep hoping beyond hope that you'll find it? Even years later after several moves? That's my e-bow. I'm pretty sure my GK pickup was right next to it. Where are they now? Well, whaddaya call a blind buck with no legs and no "boy parts"? (PM for the answer)

I came up with what i thought was the perfect solution. I'd take a cordless drill and strap it down to the guitar over top of one of the pickups in such a way that it wouldn't move as a pulled the trigger. Then I could step through the switch settings and the different cables and record it to the hard drive. From there I'd run various tests in SoundForge and come up with some "definitive" comparisons.

Well, it didn't require any fancy computer analysis, the difference was immediately audible, and about the opposite of what I expected. Both the series and parallel settings were significantly quieter than the single coil settings. Thinking there might be something wrong with the mid-position pickup, I tried again with the neck. Same issue. Then I slapped myself in the forehead.

So, this might be too easy, but I thought I'd throw it out there. Why do you think it came out this way? Where did I go wrong?

Sumgai and ChrisK, I'd ask that you abstain for a couple minutes.

[JohnH]

cool! I love tests, especially 'unscientific' ones that actually help you understand something.

Some thoughts:

1. If both coils of a pickup are fairly similar, then series should be louder than parallel, at least in terms of the max amplitude. If you have it now on disk, you can probably get graphical traces of amplitude v time, out of your editing program.

2. For the fundamental tone of a note, ie without all the harmonics, parallel should be about the same as single coil. I think of an analogy of two dry cells, where each is 1.5V, and putting them both in parallel is also 1.5V

3. But, when you consider all the harmonics and the total sound, the single coil is the sharpest and brightest, with the most cut due to higher harmonics - so it can sound overall more dominant than the parallel setting.

Any good?

John

[ChrisK]

Ok, it's been slightly more than a few minutes.

Hmmm, if I read your description correctly, you placed the drill above a pickup and turned it on. You do not indicate if the drill actually did anything except rotate. You did indicate that of all the things you've lost, you miss your Ebow somewhat.

It appears to me that your drill just sat there running and generating an electro magnetic field (if it actually DID do something, you've cheated by leaving something out).

For some strange reason, humbucking configured pickups were quieter..... This is a good result it you actually didn't have the drill GeWacking the string(s).

If you actually did have the drill GeWacking the string(s), meybe the coils were wired out of phase.


Ok, I'll bite, were you using strings?

Oh, and

Then I slapped myself in the forehead.

was this intentional?


BTW, two coils in series will generate twice the voltage, but have twice the output impedance of a single coil (signals add).

Two coils in parallel will have the same relative output voltage, but half the impedance of a single coil (signals average).

[ashcatlt]

Yeah, no. I guess I should have mentioned that this was not the Mr Big thing where you stick a pick on the end of the drill (maybe that would work?). Just the motor.

So ChrisK is right (of course, you actually waited hours).

I could swear that when I'm actually playing the thing it sounds louder in the parallel position than the single. May be psychosomatic, but I think I hear more overdrive, and it looks louder on meters. Still wish I had that darned e-bow!

[ChrisK]

I could swear that when I'm actually playing the thing it sounds louder in the parallel position than the single.

Yes!

You are driving the amp with two generators in parallel. At a minimum, assuming that the missing coil in the single coil combination isn't the stronger coil, you have effectively cut the output impedance of the generator in half. For a given voltage, it can supply twice the current. It is less affected by the amp input load, and it will have a combined inductance of half that of just a single equal coil. The frequency response curve will go up perhaps half an octave as well.


Now, did you mean louder in parallel than series?

[ashcatlt]

ChrisK, you were talking in that other thread about the right way to do a test. Control all the variables aside from the one in order to produce meaningful data. That's what I was trying to do with this.

There is, though, another way to conduct a scientific test. That is to conduct a series of experiments, changing all variables but one, and watch for patterns in the outcome. This is the basis for the random sample. We can talk all day about the tactics of the pharmaceutical industry, but it doesn't change the fact that this sort of study can be valid.

I've had these pickups installed for a couple years now. Played and recorded them through various cables and pedal chains, amp models (usually one of two modelers), and in several different venues. My experience agrees in full with what you've said.

The series position is the loudest, and is very midrangey and kind of compressed sounding. The parallel is not quite so loud but much brighter. The single coil sound is quieter yet. It seems, though, that the parallel has a bit more natural or full sounding low end. Hadn't really noticed it being any brighter than the single, or at least it sounds natural to me with the louder output. Of course, the cutoff point there is right around where most guitar speakers (and models thereof) start to fail.

Help me to understand this part though. Seems to me like the increased current should be proportional to the decreased impedance. Doesn't that end up multiplying out to the same voltage? And isn't the voltage what we end up perceiving as volume? And doesn't sumgai keep saying that these things are voltage sources, rather than current sources?

In general, how do I know which (if any) of Ohm's variables remain constant in situations such as these?

[sumgai]

ash,

Very technically speaking, all generators produce both current and voltage, that's why you're correct to invoke Ohm's Law. However, there two more forces at work here, in terms of how to calculate the transfer of power (voltage times current) from a generator to a load.

First, Ohm's Law is usually trotted out when thinking of DC, but here in the GuitarNutz world, we're actually speaking to AC, not DC. Fortunately, all the formulas still obtain, but we have to take into consideration how each part (component) reacts at a given frequency.

Second, when we "do the math", we usually try to make the power transfer at a ratio of 1:1 - that's what we call the optimum power transfer ratio, for what I hope are obvious reasons. But not quite so obvious, at first glance, are the two cases away from that optimum, those where the ratio favors either one side or the other..... what happens then?

Well, as it turns out (and I've said this before, but it was too long ago to be relevent now), if you purposely mismatch the impedance ratio in the direction of low source/high load, you actually promote a better voltage transfer ratio. Of course, the current transfer ratio suffers in direct proportion. But when a high impedance load (the amp's input) sees that incoming voltage, all is good in amplifier-land. (For reasons not germane to this discussion, let's leave it at that, eh? )

If you want to promote a higher current transfer ratio, you'd set the output impedance higher than the load impedance. Does this sound familiar to anyone? The main trouble here is, the more current you transfer, the larger your components have to be (physically) in order to accomodate the attendant heat build-up. An analogy would be your house current's feed from the power company - that comes in on a 12,700 volt line, and is lowered by the transformer on the utility pole just outside your house. By sending power at a higher voltage, the current is kept way down, thus the size of the wire can be kept smaller - there's less heat build-up. (Remember, voltage is merely the pressure behind the current, the flow of electons. It's the number of electrons bouncing around that cause the heat, not the pressure of them being bounced.)

I hope this has helped, but if not, ask for clarification.



sumgai

[sumgai]

ash,

Further in answer to your multiple questions..........

In general, how do I know which (if any) of Ohm's variables remain constant in situations such as these?

My final is "none of them" Regis! Since we're talking about AC, we have to remember that impedance varies with the frequency, thus the other two factors will also vary. For any given single frequency, you can rely, somewhat, on the impedance being steady. But that doesn't take into account such things as non-linear string decay, harmonics, resonance, the time of day, when Mr. Murphy last ate, etc., etc. You're never gonna get better than an approximation, sorry.

An eBow is about the best way of getting a steady signal, for testing purposes.

HTH





sumgai

[ChrisK]

My point was, is, and always shall be:

To actually and accurately determine the actual, quantitative difference between "A" and "B" in an A/B comparison, with an electric guitar, one needs to keep as much as possible constant.

For a shorted coil vs an unshorted coil, one should pluck a note with all instrumentation setup (do oneself the favor of documenting said setup), and rapidly switch between state "A" and state "B". Then and only then, since all other things can remain both short-term as well as long-term stable (which includes the SAME recording and scope trace et al), can one precisely determine the difference between.

"Consistency is the virtual of fools."

It's also the most significant virtual of meaningful testing.