Testing for microphonics with an un-potted PAF clone

[antigua]

An issue that comes up often regarding guitar pickups is the potential for "microphonics", which is the idea that a pickup behaves partially like a microphone, in addition to it's primary function as a magnetic transducer. In particular, it's often stated that a pickup is more microphonic if it's not wax potted. That's not so debatable, after all the reason pickups are wax potted at all is to prevent squealing noises when the guitar and pickup are subject to high gain. But another claim that is much less self evident is the idea that an unpotted pickup might partially act as a microphone at all times, imparting some amount of acoustical component to the output signal, sort of like a pickup / piezo hybrid.

So I've set up a few different tests to measure how much voltage output emanates from a non wax potted pickup when it's physically interacted with using non ferrous materials, since ferrous objects would induce a voltage in the pickup the same way a steel guitar string would. The pickup in question is a Seymour Duncan Antiquity humbucker bridge model. Seymour Duncan Co. says "Staying true to the original Gibson P.A.F., these humbuckers are not wax potted which takes you right to the edge of harmonic breakup." I've opened these pickups before and have verified that there is no wax to be had.

For the test, I've hooked the Antiquity humbucker up to a Velleman PCSU200 USB Oscilloscope, which has a "persist" feature, so the waveform input over a long period of time will overlap in a single view, showing anomalous voltage spikes as well as a broad average over time.


First, the base noise level, with the pickup just sitting on the table, about 15mV.






Next, the realistic in-situ output of the pickup, I held it above steel guitar strings and strummed them, the voltage output gets up to 580mV, but for the most part stayed below 300mV. Therefor the typical signal output is about twenty to forty times higher than the noise level in my environment, which is free of stage lights and the like. Note the larger scale of 0.3 volts per vertical division. 





The first acoustic test was to tap on the cover of the pickup with a wooden spoon, which will not magnetically induce a voltage. I ran it across the tops of the screws and hit it directly a bunch of times. For the most part, the voltage output stayed below 25mV, which is only 10mV above the noise floor, with some very occasional spikes reaching close to 100mV. I hit the pickup rather hard too, since it's supposed to look "distressed", each whack just makes it that much better.







This next test is the closest to "in situ", where the pickup is affixed to a cheap Chinese Tele I use for testing things like this. I have it taped to the body, far enough away from the strings that any magnetic interaction between the strings and the pickup is also below the noise floor, but still in contact with the body along the center axis, as pickups tend to be. Aside from some little anomalous spikes, about thirty seconds of vigorous strumming didn't result in any sustained voltage that exceeded the 15mV noise floor. I held the guitar in different positions while strumming in order to get the body of the guitar to vibrate as much as possible, but it didn't make a difference on way or another.





For the final test, I taped the pickup to the bridge of a nylon string guitar. Even though the wound strings are metal, they are non-ferrous and ungrounded, so they don't interact with the pickup magnetically or electrically. Once again, aside from a few little spikes, vigorous flamenco-style rapping on the strings didn't cause the Antiquity humbucker to produce a consistent output above the noise floor.





Based on this testing, it seems very very unlikely that there is an audible acoustic component to an un wax potted humbucker. On the other hand, it's still true that they experience acoustic feedback with high gain, but in that case, we know it's most often a very high pitched sound, one that it requires a high gain and volume to achieve, and one that it gets louder as the oscillation compounds in a feedback loop, so in the one instance where it's obvious that a pickup is responding to air pressure, the circumstances are also fairly specific. I have personally witnessed some pickups that were so microphonic that I could talk into them and hear myself through the amp, but these were usually especially cheap pickups from especially cheap guitars, not well made pickups that simply lack wax potting. 

As for the specifics of the test, such as how the pickup was physically interacted with, or how the pickup was taped to the top of a guitar, since the result didn't exceed the noise floor, and because the noise floor is about twenty times less the typical voltage output, I can't hypothesize that any variation in the testing method (such as screwing the pickup into the guitar, or rattling the pickup through other means) to produce a significantly different outcome.

Another point to consider is that even if you install a piezo pickup in a guitar, a device which intended specifically for the purpose converting vibration into voltage, still only produces about a tenth of the voltage of a regular magnetic pickup, hence the need for a battery powered pre-amp.

There are also other types of pickups to consider, ones that might show a microphonically induced voltage about the 15mV noise floor, and I will try to test other types of pickups, but since un wax potted PAF replicas are frequently the subject at hand, it seemed like a good place to start, especially a model that is well regarded by the market. 

[ms]

To complete the test you could wax pot the antiquity and see if anything changes.  (Well, that would destroy the historical accuracy, and so it would make more sense to test a wax free cheap pickup and then pot it.)

[antigua]

Mar 5, 2019 6:20:30 GMT -5 ms said:

To complete the test you could wax pot the antiquity and see if anything changes.  (Well, that would destroy the historical accuracy, and so it would make more sense to test a wax free cheap pickup and then pot it.)

Well since the tests were all below the noise floor, I think the wax would just make a difference that is too small to observe, even smaller still. I have some loose wrapped single coils I made myself which I can test in order to see what the potential is for microphony from extreme loose coil windings. 

[newey]

It seems to me that vigorously tapping the (metal) cover with a wooden spoon enough to further "age" it likely slightly distorted the metal momentarily, the moving metal cover then inducing a small current in the windings, hence the slight spike above the noise floor with that testing. So, not sure even the slight difference is truly due to microphonics.

[blademaster2]

So, per my recent post on this topic, would you agree that it seems unlikely that Brian May's Burns TriSonic Pickups on his Red Special could actually work well enough as a microphone for Brian to talk back through them to the studio control room?

This was depicted in the movie Bohemian Rhapsody when they were in the studio recording Brian for that song.

I feel that if those pickups were truly that responsive to his voice it would make them almost useless on his guitar.

[ms]

Mar 5, 2019 11:09:16 GMT -5 antigua said:

Mar 5, 2019 6:20:30 GMT -5 ms said:

To complete the test you could wax pot the antiquity and see if anything changes.  (Well, that would destroy the historical accuracy, and so it would make more sense to test a wax free cheap pickup and then pot it.)

Well since the tests were all below the noise floor, I think the wax would just make a difference that is too small to observe, even smaller still. I have some loose wrapped single coils I made myself which I can test in order to see what the potential is for microphony from extreme loose coil windings. 

I was thinking that you might see a difference with the spikes in the wooden spoon test.

[antigua]

Mar 5, 2019 11:47:50 GMT -5 ms said:

Mar 5, 2019 11:09:16 GMT -5 antigua said:

Well since the tests were all below the noise floor, I think the wax would just make a difference that is too small to observe, even smaller still. I have some loose wrapped single coils I made myself which I can test in order to see what the potential is for microphony from extreme loose coil windings. 

I was thinking that you might see a difference with the spikes in the wooden spoon test.

I can try tapping on an already potted humbucker, that should work as well. It would be interesting to see if wax potting make a difference with direct impacts.

Maybe this is a question you can help me with; clearly in the clean context the microphonic aspect is very tiny, in amongst the noise, but what if high gain is applied, and the acoustic/microphonic voltage is increased to the point of being audible, and occupies a frequency band that its not already occupied by the primary signal, or it's harmonics? Suppose for example the acoustic content is made up of bass frequencies, but the bridge pickup is selected, causing the primary signal to be mostly treble frequencies. Could the highly amplified acoustical bass portion of the combined signal then become audible?

[antigua]

Someone on another forum took issue with the fact that the humbucker was taped to the tail end of the body, and not mounted where a pickup would be mounted, so I tested the voltage output of a piezo pickup in those two locations to put that person's fears to rest. Based on the voltage output, it appears that there is virtually no difference, about 50mV output from the piezo in both locations, indicating the tail end of the guitar body vibrates about as much as the center of the guitar.








I also taped the piezo to the nylon guitar bridge, that averaged about 150mV with some spikes exceeding 400mV (note that the type of piezo that fits underneath the saddle might produce a higher voltage that the round piezo being taped to the top). Therefore the bridge of the nylon is vibrating with at least three times as much intensity as the solid guitar body, and in either case the microphonic component, to whatever extent it existed, was below the 15mV noise floor.







Just for fun I stuck the piezo in the neck pocket, as some people apparently do in practice, and found that the voltage output in that location was remarkably high, averaging somewhere around 250mV, with occasional spikes exceeding 400mV.





The neck pocket isn't back, but the head stock generates even more vibration, nearly double:



It makes me wonder if a head stock mounted piezo could produce enough juice to avoid the need for a pre-amp, but running the wire up there, maybe through the truss rod shaft, would be tricky business. 

[antigua]

Mar 5, 2019 11:47:50 GMT -5 ms said:

Here are a couple more wooden spoon tests.

First, hitting a homemade coil with windings that are extremely loose, to the extent that you can't fit a cover over the pickup. The interesting thing is, like with the PAF replica, not every single strike resulted in output spikes, but in this case, where there was a spike, they tended to be much higher. It might be that the spikes correspond to wire movement, and that the wire more a lot more in some instances than it does in others. Also sort of interesting is that the tall positive spikes correspond to sharp strikes with the spoon, while the broader negative cycle spikes came about from the dull thud of hitting the pickup with the face of the spoon, all while the pickup is sitting on a table cloth, as is seen in the pics.  







Also, just because why not, I tried hitting the piezo with the wooden spoon, and the direct shocks caused it to put out some very high voltages, well over 40 volts it seems. Considering piezos produce such a low voltage in situ, this came as a big surprise. One again, the voltage spikes being on the positive side of the cycle seems to relate to the direction from which the piezo is struck. If I turned the piezo over, the spikes were on the negative side of the cycle. The longer voltage spikes on the bottom came from dull impacts, which the piezo on a table cloth. 

[ms]

Mar 5, 2019 18:08:10 GMT -5 antigua said:

Mar 5, 2019 11:47:50 GMT -5 ms said:

I was thinking that you might see a difference with the spikes in the wooden spoon test.

I can try tapping on an already potted humbucker, that should work as well. It would be interesting to see if wax potting make a difference with direct impacts.

Maybe this is a question you can help me with; clearly in the clean context the microphonic aspect is very tiny, in amongst the noise, but what if high gain is applied, and the acoustic/microphonic voltage is increased to the point of being audible, and occupies a frequency band that its not already occupied by the primary signal, or it's harmonics? Suppose for example the acoustic content is made up of bass frequencies, but the bridge pickup is selected, causing the primary signal to be mostly treble frequencies. Could the highly amplified acoustical bass portion of the combined signal then become audible?

The output of an amp played very high gain is non-linear.  As a first approximation, you have a clipped waveform which changes from positive to negative and negative to positive where the original waveform crosses zero.  With a weak low frequency and a strong higher frequency, the zero crossings are set by the high frequency, but modulated a bit by the low frequency.  Thus, I think you do not hear the low frequency, but might it might produce some audible intermod on the higher frequency.

[antigua]

Having learned from the piezo test that the Tele headstock provides that most vibration anywhere on either the Tele or nylon acoustic, I taped the unpotted Antiquity to the headstock of the Tele, and did manage to get a definite microphonic voltage, about 34mV. I'm certain it's a microphonically induced voltage, because I also taped a Gibson 57 Classic and a Super Distortion to the headstock, and despite the vigorous vibration of the headstock, they both measured much less output, 25mV for the 57 Classic and the 18mV for the Super Distortion, and both are wax potted. 

Of course, having the pickup attached to the headstock is an extreme case, and even in this extreme case, the voltage is very low compared to the principle output, so this doesn't confirm that microphonic content combines with the primary signal source to influence the overall tone, under normal circumstance or even in this extreme case, but it verifies that the un-wax potted humbucker has a some small amount of microphony over a range of audible frequencies, and not just some particular resonant frequencies that might come into play when feedback occurs. It also shows that a potted pickup shows some measurable microphnics, but less so.


Noise level is 15mV


Antiquity unpotted HB taped to the head stock, produces about 34mV, about double the noise level.




57 Classic, showing 25mV, still microphonic, but less so:




For some reason, the Super Distortion showed some occasional spikes, but the overall signal level was only about 4mV above the noise level. Unlike the 57 Classic, the Super Distortion has no metal cover, and the magnet is non conductive ceramic, so it's not surprising that it's a little quieter.

[ms]

So that shows that wax potting has benefits, and is not just a comforting ritual.

[antigua]

Mar 7, 2019 6:16:18 GMT -5 ms said:

So that shows that wax potting has benefits, and is not just a comforting ritual.

I agree with that, but I've also noticed how unpotted pickups tends to squeal more readily when the gain and volume are high.  I think the bigger question is whether not wax potting pickups is a comforting ritual. 

[ms]

Mar 7, 2019 10:03:01 GMT -5 antigua said:

Mar 7, 2019 6:16:18 GMT -5 ms said:

So that shows that wax potting has benefits, and is not just a comforting ritual.

I agree with that, but I've also noticed how unpotted pickups tends to squeal more readily when the gain and volume are high.  I think the bigger question is whether not wax potting pickups is a comforting ritual. 

Hmmm..  Creating a ritual by not doing something is too Zen for me.  It evokes images of the sound of one hand clapping.

[antigua]

The real issue now, I think, is trying to determine how, if at all, a unpotted microphonic pickup might sound different than a less microphonic potted pickup, while you play the guitar. The testing shows that even the potted pickups produce some microphonics, but not as much, perhaps only 70% as much (25mV versus 35mV attached to the head stock). Personally I'm just not seeing how it's possible for such a weak input vector to compete with the much stronger primary input. 

It's true that an unpotted PAF style pickup squeals more readily than a potted one, making that 70% figure seem like to small of a reduction given how much less prone a potted pickup is to feedback, but I suspect that feedback deals with pickup resonances, where as overall microphonics deals with the whole audio range. It could be for example that the loose windings cause the broad range microphonics, but the cover of the pickup resonates, and is specifically responsible for the problematic feedback, and when you wax pot a pickup, the cover become wholly  immobile, while the windings might still be susceptible to movement, especially if the wax penetration of the wings is only surface deep. So to take this investigation further, it's probably necessary to dismantle the pickup and figure out if the microphonic output increases of decreases as parts are removed from around the coils. 

The other factor is that there are some pickups that have historically been far more microphonic than even an unpotted PAF, such as a Teisco pickup, so maybe if we had one of those to test it would show a higher voltage output from microphonics, and lend credibility to the idea that the microphonic component has a real influence, even while playing the guitar. 

[antigua]

Here is a test to measure the amount of vibration in the area around the pickups of three Epiphone HH guitars, when strummed as vigorously as the strings will allow without snapping, with varying degrees of hollowness; a Les Pail, a semi hollow Sheraton II (335 style), and a full hollow Swingster (Emperor-II body).

The purpose of this is experiment is to determine how much vibration exists around the pickups of these different types of guitars, in that seemingly unlikely event that you could hear that input microphonically beneath the primary output signal of the pickup, which appears to be on the other of twenty times greater for a solid body guitar, and about seven times greater with the hollow body. 

Surprisingly, the Epiphone Les Paul shows more vibration, 62mV, between then pickups than the semi-hollow Sheraton II, only 44mV, which has a hollow wings, but is solid down the middle, where the pickups are bridge are mounted. I think the reason is, in part, that even though the Sheraton is partially hollow, it's a longer, winder and more bulky guitar all around. I understand a real Gibson 335 has a solid top instead of a laminate, so the Sheraton II might not be representative of how the 335 vibrates.

Not surprisingly though, the Swingster, which is a legitimate arch top acoustic, showed strong vibration between the pickups, reaching 250mV and higher.

For reference, I get about the same amount of voltage, around 45mV, with the piezo at the center of the Sheraton II's body as I did with the unpotted PAF taped to the headstock of a Tele, so essentially it took the lowest observed voltage with the piezo to reach a highest observed voltage of the unpotted PAF, being applied as if it were a piezo pickup.


Epiphone Les Paul, solid except for control cavities, ~62.5mV





Epiphone Sheraton II, hollow sides ~43.7mV





Epiphone Swingster, fully hollow ~250mV

[antigua]

Today I tried to get a sense of what a neck pocket piezo might sound like with a Strat by plugging it in to an amp why tapping it to the back of the Strat neck near the hear. The output is very low, but since so many guitar rigs have built in gain, that's not necessarily a problem, but the sound wasn't too great, it wasn't a deep, rich acoustic tone like you get from a bridge piezo, it sounded more like a cigar box guitar, very hollow with a lot of mid range, not much treble or bass. I think a low inductance magnetic pickup, like the Bill Lawrence AlNiCo Microcoil, or the Burns TriSonic, have a much more convincing acoustic tone than the piezo, and they have a strong output to boot.

The other problem with a neck piezo is that it picks up lots of hand movement sounds, as well as the string's vibration on the other side of the fret note, so like if you play the 14th fret and up, you will hear the portion of string between the net and the 13th fret pretty clearly, and it's not very musical to say the least.

While I was at it I also tried attaching the piezo in between humbuckers to see how sensitive it is there "in situ", and it was louder than I would have guessed (still far from usable) but what was immediately obvious was that the sound of my nicking the pick guard (a transient stimulus) was a lot louder than the sound of the strings. So if you have a pickup that is really microphonic, one tell tale will be that tapping on the pick guard or the mounting ring will produce a pretty loud click.

This Sheraton II picture has Seymour Duncan Seth Lovers, unpotted pickups. I plugged it in properly, and turned up my amp rather loud, a 5e3 clone, and stacked two tube screamers to get a dense gain signal, and it started feeding back immediately at about 900 Hz, so I turned down the volume just enough to stop that from happening, and clicking on the pick guard made an audible sound, but it was still a lot less audible than the clicking had been with the piezo.

The conclusion I'm coming to by way of triangulation is that the even unpotted PAF clones pickups are not especially microphonic, and at the same time, the string vibrations in a solid body guitar is not especially strong, which taken together doesn't make it real likely / plausible that an unpotted pickup sounds different from a potted pickup. There are surely some pickups that are a lot more microphonic than an unpotted PAFs, but since I don't have any of those on hand, I can't speak to that.