as with measuring loudspeakers with REW/TrueRTA/etc. and a mic by:using an exciter coil as described at Github as 50 turns #26 wire on a single coil
(Strat/ Jazz bass bobbin) - there's a series resistor (value?) with the exciter coil and the coil driven by an amplifier
The exciter coil is placed against the pickup under test. The output of the pickup is terminated with a load like it would see from a guitar's pots and external amplifier input impedance. (some cable capacitance might be added to evaluate effect))
The pickup under load's output is fed to the audio interface to be evaluated by swept sine.
OK? - any mods to the exciter coil? - how much R in series with the exciter coil?
I need to find a cheap domestic source for the bobbins.
(Strat/ Jazz bass bobbin) - there's a series resistor (value?) with the exciter coil and the coil driven by an amplifier
The exciter coil is placed against the pickup under test. The output of the pickup is terminated with a load like it would see from a guitar's pots and external amplifier input impedance. (some cable capacitance might be added to evaluate effect))
The pickup under load's output is fed to the audio interface to be evaluated by swept sine.
OK? - any mods to the exciter coil? - how much R in series with the exciter coil?
I need to find a cheap domestic source for the bobbins.
The magnetic field produced by the exciter coil is proportional to current.
So you want constant current drive.
To get close, the series resistor (or source resistance) should be large.
More exactly it should be much (say at least 10 times) larger than the coil's impedance at the highest frequency of interest.
E.g., if your coil inductance is 25mH and the upper frequency limit is 10kHz, the series resistor should be larger than 10*2*pi*10k*25m = 16kOhm.
Too low series resistance will cause a HF drop.
I think you could use less turns, resulting in lower inductance and smaller series resistor but maybe lower PU output.
So you want constant current drive.
To get close, the series resistor (or source resistance) should be large.
More exactly it should be much (say at least 10 times) larger than the coil's impedance at the highest frequency of interest.
E.g., if your coil inductance is 25mH and the upper frequency limit is 10kHz, the series resistor should be larger than 10*2*pi*10k*25m = 16kOhm.
Too low series resistance will cause a HF drop.
I think you could use less turns, resulting in lower inductance and smaller series resistor but maybe lower PU output.
Can't edit my post above, but a very rough estimation of your 50 turns coil shows that inductance should be much lower than the 25mH I assumed in my example above. More like 0.25mH.
So the minimum series resistance might be as low as 160 Ohm.
To be on the safe side I'd use something like 1k.
Of course it would be best to measure inductance.
So the minimum series resistance might be as low as 160 Ohm.
To be on the safe side I'd use something like 1k.
Of course it would be best to measure inductance.
Electrical frequency response is measured straight away as a series inductor, loaded with the guitar circuitry and cable, no need for exciter coil. Such a coil just adds uncontrolled variables (for example, what does it do with humbuckers?). Alas, even a with an exactly known behavior like from a current-driven Helmholtz air coil pair, the measured frequency response (of a single coil) is pretty useless in that it does in no way reflect what happens with a real string vibrating over a pickup.
This geometric/magnetic aspect of the response is a really complex 3D magnetic problem, nonlinear as well. And of course it is not a single response, rather its literally hundreds, a whole set. There simply is no way other than using a real (part of a) string vibrating, resembling typical conditions.
Both responses multiply. You can have two pickups with the same electrical response and even the same free-field exciter-induced response which still may sound quite different because what counts is how the AC field lines are deflected by the string (which often gets magnetically saturated right over the magnet) and how those bent AC field lines exactly cut through the actual windings.
This geometric/magnetic aspect of the response is a really complex 3D magnetic problem, nonlinear as well. And of course it is not a single response, rather its literally hundreds, a whole set. There simply is no way other than using a real (part of a) string vibrating, resembling typical conditions.
Both responses multiply. You can have two pickups with the same electrical response and even the same free-field exciter-induced response which still may sound quite different because what counts is how the AC field lines are deflected by the string (which often gets magnetically saturated right over the magnet) and how those bent AC field lines exactly cut through the actual windings.
Last edited:
What is the purpose of trying to measure FR of a guitar or bass pickup in the first place? There is no reason why it should be flat, and probably plenty of reasons why it shouldn't be (especially if playing in a band and or with other instruments).
I agree this measurement only reveals a part of the truth.
Nonetheless measuring the impedance peak frequency is not useless.
If this is lets say 1.5kHz, you can not expect much above this frequency.
Thus this sets an absolute upper limit no matter of any magneto-mechanical structure.
I have single coils with resonant peaks close to 10kHz, that do sound more "transparent" than those with 4.5kHz.
This is a simple test for a first discrimination.
A real frequency response measurement with some kind of a ferromagnetic string would be nice, but I have no idea how to achieve this.
That is not what I've said. I said using an exciter coil isn't very helpful.
Measuring impedance, similar to how we do if for speakers, is IMHO the best way we can do for the electrical part. From the impedance we can derive the R, L and C and thus can model the electrical response with any loading and actually get useful information of how it will shape the sound. And just like you said, it will show the limits of high frequency response.
For bass pickups, I've experimented with sections of bass strings glued to the edge of a piece of thick cardboard, and then gluing this to the cone of small wideband driver, with some additional cardboard ribs to stabilize things laterally. But I ran into two obvious issues, dealing with unknowns: The actually frequency response of the obtained excursion and influence of the DC and AC magnetic field of the driver on the pickup (the cardboard helps to increase distance but there is still enough leakage that must be factored out).
@freddi , don't let this discussion spoil the fun you will have and the insights you will gain by still trying the exciter method. But I would suggest that you also measure electrical response as described, for a base line you can compare to. With the exciter coil, you should be able to see the exact same general shape of the response wit regard to the peaking, just more or less skewed and deformed by the combined geometric/magnetic response of exciter and receiver.
I watched the responses and it's time to enter the discussion.
A long time ago, 40 + years, I had a similar idea. I wanted to measure the response of guitar pickups but again, as mentioned, using another pickup or whatever inductor didn't give a good result.
There was no reference whatsoever and rendered the results useless.
So I made a setup with strings, switchable pickups and a rotating set of wheels to create the gentle strumming of the strings.
I do not remember the name of that wheel assembly but it's been used in the late seventies by 10CC.
The strings and strumming were always the same and considered the reference.
The only variable is the the pickup.
So, I used a modified cheap electric guitar, routed a grouve so that I could easely change the pickup. The pickup was inserted in the space under the strings. The strings of the guitar were tuned as for a regular guitar.
Then the wheels started to turn and the response was sent to a graphic spectrum analyser. The rectified RMS output set out on graph paper.
Very crude but at least I was able to compare different pickup. And the setup was identical for any type of pickup.
I have no doubt that with the modern technology it can be refined.
Now it's up to you folks...
A long time ago, 40 + years, I had a similar idea. I wanted to measure the response of guitar pickups but again, as mentioned, using another pickup or whatever inductor didn't give a good result.
There was no reference whatsoever and rendered the results useless.
So I made a setup with strings, switchable pickups and a rotating set of wheels to create the gentle strumming of the strings.
I do not remember the name of that wheel assembly but it's been used in the late seventies by 10CC.
The strings and strumming were always the same and considered the reference.
The only variable is the the pickup.
So, I used a modified cheap electric guitar, routed a grouve so that I could easely change the pickup. The pickup was inserted in the space under the strings. The strings of the guitar were tuned as for a regular guitar.
Then the wheels started to turn and the response was sent to a graphic spectrum analyser. The rectified RMS output set out on graph paper.
Very crude but at least I was able to compare different pickup. And the setup was identical for any type of pickup.
I have no doubt that with the modern technology it can be refined.
Now it's up to you folks...
Was the device this (or something like it):
https://www.pmtonline.co.uk/e-bow-p...VaY9QBh0nKRThEAQYAiABEgJaC_D_BwE&gclsrc=aw.ds
Other (cheaper) alternatives are available.
https://www.pmtonline.co.uk/e-bow-p...VaY9QBh0nKRThEAQYAiABEgJaC_D_BwE&gclsrc=aw.ds
Other (cheaper) alternatives are available.
Hello Keithj01, the idea is about the same but no, it was mounted just before the bridge as far as my memory is telling me... It's been a long time...
Lots of info on pickup theory, measurements and data here:
https://guitarnuts2.proboards.com/thread/7723/measuring-electrical-properties-guitar-pickups
https://guitarnuts2.proboards.com/thread/7723/measuring-electrical-properties-guitar-pickups
Many thanks for all those references ! ~50 years ago someone gave me a bobbin with matching alnico bar magnet. I wound a pickup with around 600 ohms DCR - can't remember if #28 or something else. I hooked to a miniature low Z to high Z transformer and fed it to my Traynor amp. It sounded good imo and had more highs- more hi-fi range on top than a P-bass pup and it seemed to have enough voltage output with that transformer. I didn't mount it as the bobbin had deep burn marks on it - maybe from a previous owner's soldering iron. I never see pickups done that way - (why?)
There were and still are low to medium impedance PUs using a step-up transformer or active electronics.
This thread might interest you:
https://music-electronics-forum.com...kup-makers/5622-low-impedance-pickup-research
This thread might interest you:
https://music-electronics-forum.com...kup-makers/5622-low-impedance-pickup-research