Metering between shield and ground?
June 29, 2010 6:52 PM Subscribe
What kind of circuit can I use to measure between shield and ground in an electric guitar?
What I'm thinking, is to keep shield and ground separate and bring them together in one place, and then tie them together. By shield, I mean a copper vault Faraday cage that encloses all the other electronic workings. I'd like to have a meter that shows the potential between the two before hooking them together. Can I use diodes to keep them from backwashing into each other, and still get the benefit of tying them together? I'm imaging a display the would change as you position the instrument in space, and the lowest reading would indicate the lowest-noise orientation.
What I'm thinking, is to keep shield and ground separate and bring them together in one place, and then tie them together. By shield, I mean a copper vault Faraday cage that encloses all the other electronic workings. I'd like to have a meter that shows the potential between the two before hooking them together. Can I use diodes to keep them from backwashing into each other, and still get the benefit of tying them together? I'm imaging a display the would change as you position the instrument in space, and the lowest reading would indicate the lowest-noise orientation.
To see how dirty your ground is, again a crude way can be to put a multimeter on AC volts, one probe to ground, another to something big - like you (lick your fingers and clasp it in between).
I just had a vision of someone trying to do this and sticking a probe in the live socket of a mains outlet, or the ground socket of a crosswired room. Don't do that. I was assuming that you already have a grounded exposed wire that you already know is safe. If instead you have to poke stuff into mains sockets, don't, unless you genuinely know what you're doing.
posted by -harlequin- at 9:23 PM on June 29, 2010
I just had a vision of someone trying to do this and sticking a probe in the live socket of a mains outlet, or the ground socket of a crosswired room. Don't do that. I was assuming that you already have a grounded exposed wire that you already know is safe. If instead you have to poke stuff into mains sockets, don't, unless you genuinely know what you're doing.
posted by -harlequin- at 9:23 PM on June 29, 2010
If you ground and shield it properly, you shouldn't get noise, so there should be no need for such a meter.
But the best "meter" is just plugging the guitar in and listening to any noise you get. If it is silent, awesome. If it is noisy, but when you turn the volume to zero it goes silent, you have a wiring problem or just noisy pickups.
posted by gjc at 3:43 AM on June 30, 2010
But the best "meter" is just plugging the guitar in and listening to any noise you get. If it is silent, awesome. If it is noisy, but when you turn the volume to zero it goes silent, you have a wiring problem or just noisy pickups.
posted by gjc at 3:43 AM on June 30, 2010
Response by poster: bringing in meters would kind of seem like one of those audiophile wild goose chases
That's exactly what I'm trying to do, get a signal that is clean enough, that it may sometimes be audible as improved noise, and other times contribute to phase and staging qualities that can't really be "eyeballed." I can't see bacteria on my food, but still have legitimate reasons to not want it there. All of the degradations that occur anywhere are summed up -- one at a time, they might be very small. However electric guitar noise is not all that small, these are antennas that pick up noise, that's what they do.
In particular, I'm building this specialty guitar to be touched very softly with a lot of amplification: sort of an electro-acoustic microscope.
I'm not talking about electrical mains. Even if the thing is not plugged in at all, it can pick up radio interference etc. Or just pretend it can, that's not my question.
Let me clarify:
........................M....................
...................../.....\..................
ground ===G........S=== shield.
I can AC meter between G and S, that's simple.
ground ===G===S=== shield.
Metering with G & S connected is the same as putting both probes together on G or S, no?
I'm looking for any strategy that would let me effectively connect G & S, and still meter something about the potential that would occur at G & S, if they were not connected. Maybe I need to have a breaker between G & S for calibration and monitoring. Is there a way to do this without that breaker?
posted by StickyCarpet at 6:38 AM on June 30, 2010
That's exactly what I'm trying to do, get a signal that is clean enough, that it may sometimes be audible as improved noise, and other times contribute to phase and staging qualities that can't really be "eyeballed." I can't see bacteria on my food, but still have legitimate reasons to not want it there. All of the degradations that occur anywhere are summed up -- one at a time, they might be very small. However electric guitar noise is not all that small, these are antennas that pick up noise, that's what they do.
In particular, I'm building this specialty guitar to be touched very softly with a lot of amplification: sort of an electro-acoustic microscope.
I'm not talking about electrical mains. Even if the thing is not plugged in at all, it can pick up radio interference etc. Or just pretend it can, that's not my question.
Let me clarify:
........................M....................
...................../.....\..................
ground ===G........S=== shield.
I can AC meter between G and S, that's simple.
ground ===G===S=== shield.
Metering with G & S connected is the same as putting both probes together on G or S, no?
I'm looking for any strategy that would let me effectively connect G & S, and still meter something about the potential that would occur at G & S, if they were not connected. Maybe I need to have a breaker between G & S for calibration and monitoring. Is there a way to do this without that breaker?
posted by StickyCarpet at 6:38 AM on June 30, 2010
Best answer: Metering with G & S connected is the same as putting both probes together on G or S, no?
For a voltmeter, yes
I'm looking for any strategy that would let me effectively connect G & S, and still meter something about the potential that would occur at G & S, if they were not connected.
Same setup, but set the meter to amps instead of volts. The meter will directly connect G & S for you through the shunt, and on top of that connection it will watch for current. Ammeters are designed to be invisibly wired in series within a circuit like this, acting as if they're a direct connection so as to not influence the circuit any more than what is necessary to "see" anything. You could think of it this way - you directly connect G and S, insulate that connecting cable, then bend an "observer" wire around the outside of the insulation, which never touches the connecting cable, but listens for the magnetic field generated by any current in the connecting cable. Some high-amp meters work this way (there is an air gap between the probe and the cable you are monitoring), but that's mainly for safety and convenience purposes - using a shunt (ie a regular multimeter) is fine. (There will be some slight resistance on the shunt, but it will be less than the resistance of even some direct cable connections - nothing has zero resistance.)
So it's kind of like this:
......_M_......
...../ \....
==G===S===
That said, I don't know how useful the readings will be, because the only similar stuff I've done was with a voltmeter, and the interference you're trying to watch for is probably stronger in volts than in amps.
posted by -harlequin- at 2:06 PM on July 1, 2010 [1 favorite]
For a voltmeter, yes
I'm looking for any strategy that would let me effectively connect G & S, and still meter something about the potential that would occur at G & S, if they were not connected.
Same setup, but set the meter to amps instead of volts. The meter will directly connect G & S for you through the shunt, and on top of that connection it will watch for current. Ammeters are designed to be invisibly wired in series within a circuit like this, acting as if they're a direct connection so as to not influence the circuit any more than what is necessary to "see" anything. You could think of it this way - you directly connect G and S, insulate that connecting cable, then bend an "observer" wire around the outside of the insulation, which never touches the connecting cable, but listens for the magnetic field generated by any current in the connecting cable. Some high-amp meters work this way (there is an air gap between the probe and the cable you are monitoring), but that's mainly for safety and convenience purposes - using a shunt (ie a regular multimeter) is fine. (There will be some slight resistance on the shunt, but it will be less than the resistance of even some direct cable connections - nothing has zero resistance.)
So it's kind of like this:
......_M_......
...../ \....
==G===S===
That said, I don't know how useful the readings will be, because the only similar stuff I've done was with a voltmeter, and the interference you're trying to watch for is probably stronger in volts than in amps.
posted by -harlequin- at 2:06 PM on July 1, 2010 [1 favorite]
Best answer: (It sounds like you're kind of up against the laws of physics - you can't measure a potential difference (ie volts, not amps) across a connection without breaking the connection to introduce some resistance, because if they're connected then there isn't a potential difference, so there's nothing to measure.
Putting a voltmeter in series as discussed earlier, measures the potential difference, but introduces resistance. Using an ammeter preserves the connection, but measures current rather than potential.)
That said, I'm guessing that while an ammeter might not work very well, it may be enough to help you out.)
posted by -harlequin- at 2:18 PM on July 1, 2010 [1 favorite]
Putting a voltmeter in series as discussed earlier, measures the potential difference, but introduces resistance. Using an ammeter preserves the connection, but measures current rather than potential.)
That said, I'm guessing that while an ammeter might not work very well, it may be enough to help you out.)
posted by -harlequin- at 2:18 PM on July 1, 2010 [1 favorite]
Best answer: I still don't have an complete answer to my question/quest but I did just find this document, which points in the direction that I hope to go:
Grounding and Shielding Audio Devices
And yes, it comes from the camp of the dreaded high-end audiophiles.
posted by StickyCarpet at 8:43 AM on September 5, 2010
Grounding and Shielding Audio Devices
And yes, it comes from the camp of the dreaded high-end audiophiles.
posted by StickyCarpet at 8:43 AM on September 5, 2010
This thread is closed to new comments.
For better results, an oscilloscope. For best results, a proper EMF meter (but you're unlikely to have one of these lying around or easily able to borrow one).
For smartest results, use no meter, ground the shield and use your ear while playing a sound on the guitar to find the area with the least interference. If you can't hear any positional change in interference, bringing in meters would kind of seem like one of those audiophile wild goose chases, and if you can hear the difference, are meters that much additional help?
Regarding backwash and diodes, I don't think I follow - do you suspect you have a dirty ground that may be introducing an oscillation in the shield? If so (and this is not solid advice but might be fun) you could consider using an alternative ground (connect a wire to metal pipes, or hammer 6 feet of rebar into the ground), and see if the isolation from other electrical devices gets you a cleaner ground. (Don't use your hobby ground for mains-powered appliances that use a ground for safety)
To see how dirty your ground is, again a crude way can be to put a multimeter on AC volts, one probe to ground, another to something big - like you (lick your fingers and clasp it in between).
As before, disregard the actual numbers on the meter - they're not meaningful under these conditions, but the general magnitude and responsiveness can often give an idea if there is much going on.
posted by -harlequin- at 9:16 PM on June 29, 2010