Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Stephen Swaffer]

14 bucks, not a bad deal at all.

Tomorrow I'm doing a show at the venue in question, so I'm bringing both the original adapter and the new wall-wart based adapter.

The other measurement I've been mulling over is a neutral to ground measurement. Might see some really interesting stuff there.

Sounds like a plan.  I know your focus is on a measurement tool to diagnose the problem.

I would curious to see if the noise is eliminated by using either a separate power cord (to the same "dirty" power) or signal cable.  That would definitively localize the path the noise is taking to the cable.  I would have thought the Meyer engineers would have spec'd the shields connected at only one end-but maybe the manufacturing engineer didn't think it mattered and was easier to manufacture-certainly wouldn't hurt to ohm out the shields just to see.

You mentioned a better line driver.  The problem is, if the noise is in the audio spectrum, how does the circuit discriminate between noise and desired sound?  Some "music" I hear intentionally amplified is "noise" to me .

While noisy mains may be the source, it is unlikely you will be able to fix that.  Even power factor correction won't entirely eliminate harmonics-and it can be a pricey fix.  It certainly won't happen before the show starts.   I think it more likely you will be able to break the path the noise takes-if you do that, then it works in every venue.

[John Roberts {JR}]

Sounds like a plan.  I know your focus is on a measurement tool to diagnose the problem.

I would curious to see if the noise is eliminated by using either a separate power cord (to the same "dirty" power) or signal cable.  That would definitively localize the path the noise is taking to the cable.  I would have thought the Meyer engineers would have spec'd the shields connected at only one end-but maybe the manufacturing engineer didn't think it mattered and was easier to manufacture-certainly wouldn't hurt to ohm out the shields just to see.

To reject RF and very HF interference shield grounds should be connected at both ends. Professional gear should transmit the audio as separate + and - stems (pins 2 and 3), ignoring ground noise (pin 1).

You mentioned a better line driver.  The problem is, if the noise is in the audio spectrum, how does the circuit discriminate between noise and desired sound?  Some "music" I hear intentionally amplified is "noise" to me .

Source impedance of the signal driver can make a difference in noise pickup. Consumer gear can be 1k ohm or more, pro gear much lower.

While noisy mains may be the source, it is unlikely you will be able to fix that.  Even power factor correction won't entirely eliminate harmonics-and it can be a pricey fix.  It certainly won't happen before the show starts.   I think it more likely you will be able to break the path the noise takes-if you do that, then it works in every venue.

Proper gear design should ignore all but the nastiest mains noise. I can imagine troubleshooting mains power quality as being more appropriate for fixed installs or perhaps recording studios, not sure what you can do, on the spot at live gigs. While it's never bad to have more information.

JR

 

[Mike Sokol]

To reject RF and very HF interference shield grounds should be connected at both ends.

Many, many years ago I was in a band playing in the front window of a bar with my back to the street. Yes, a really low-class establishment. Less than 20 away from this window was the road with an intersection and a traffic light. This was back in the day of CB radios and such, so every time the traffic light was red some yahoo in his car would get on their CB and start yapping. My PA system was normally pretty quiet, but somehow the RF from the AM modulated CB radios would show up in my speakers. Of course, this was during the slow ballads, which drove us crazy every time we played there. I did some studying on the problem (1975 style with actual books) and found that sometimes speaker cables could feed RF back into an audio amplifier, where it could be demodulated inside the amplifier's negative feedback loop, then amplified by the output stage as significant wattage. One of my ham radio buddies suggested putting tiny ceramic capacitors right across the speaker wires to shunt out the RF before it got inside the audio amplifiers, and it worked like a charm. I can't remember the exact value cap, but it was in the pico-farad range. Now, I'm not saying that's the same problem here... just that RF frequencies have a way of getting inside of audio gear and being detected by any stray corrosion on a shield bond or a P-N junction in its path. 

[Jonathan Johnson]

My PA system was normally pretty quiet, but somehow the RF from the AM modulated CB radios would show up in my speakers. Of course, this was during the slow ballads, which drove us crazy every time we played there. I did some studying on the problem (1975 style with actual books) and found that sometimes speaker cables could feed RF back into an audio amplifier, where it could be demodulated inside the amplifier's negative feedback loop, then amplified by the output stage as significant wattage.

Would twisted-pair vs. parallel zip cord (for the speaker cables) make a difference in a situation like this?

[Kevin Graf]

Yes twisting will help.
Jim Brown Power Point and paper on the subject.
Just skip the Ham radio sections.
Jim Brown was past AES committee co-chair on EMT/RFI and is a Ham radio operator.

"RF Interference in Audio Systems"
http://www.audiosystemsgroup.com/AES-RFI-SF08.pdf

"RFI, Ferrites, and Common Mode Chokes For Hams"
http://www.audiosystemsgroup.com/RFI-Ham.pdf

About 50 more Jim Brown papers:
http://www.audiosystemsgroup.com/publish.htm

[Geoff Doane]

I'm not really sure at all about solutions to dirty AC, but at least learning to measure it would be a start on the problem.

I chased a problem like this a few years ago, and posted some pictures in an old version of this forum, but it seems to have disappeared.  Here's the 2016 version.

At my day job, we discovered that we couldn't get a clean signal out of some db Technology monitors (model 12-4M).  We have 8 of them, and they all exhibited the same problem, a slight "rizz" when plugged in to the building's AC power.  They hadn't always done this, but the problem showed up shortly after an 80 KVA building UPS was installed in the basement electrical room.  None of the other studio gear seemed to have a problem (the building housed radio studios and a music/drama recording studio) and various portable consoles and other powered speakers all seemed to be unaffected.  The problem was not a ground loop, because it happened as soon as the speaker was powered up, even before an XLR feed was attached.  It was also constant regardless of the gain setting on the speaker's amplifier.

I tried all the usual (and unusual) tricks to eliminate noise: lifting the ground, putting a 2KVA isolation transformer in the AC feed, even bypassing the diodes in the bridge rectifiers with small bypass caps.  Nothing worked.  Since the only thing they were connected to was the building power, it was tempting to blame that, but difficult to eliminate it in an A/B test.  Finally, I came up with the idea of running the speakers off a UPS, and the noise went away.  At the time I only had a fairly ordinary APC UPS (not double conversion), so I had to unplug its input to get it to generate a sine wave, but the speaker was happy when I did.

I didn't have Smaart to look at harmonics, but did look at the outputs with a scope.  I used an ordinary bell transformer to drop the voltage down to about 6V.  I suspect that most 60 Hz transformers will pass frequencies a lot higher than that.  The other interesting point I discovered was that the AC was noisy whether the speaker was plugged in on the line or load side of the UPs, although the character of the noise was a bit different.

Here are the waveforms.  The first is the line side (of the 80KVA UPS), second is load, and the third is the rackmount UPS output, with no load except for the scope and powered speaker.  The building UPS was running at about 60% of capacity.

GTD

[John Roberts {JR}]

Many, many years ago I was in a band playing in the front window of a bar with my back to the street. Yes, a really low-class establishment. Less than 20 away from this window was the road with an intersection and a traffic light. This was back in the day of CB radios and such, so every time the traffic light was red some yahoo in his car would get on their CB and start yapping. My PA system was normally pretty quiet, but somehow the RF from the AM modulated CB radios would show up in my speakers. Of course, this was during the slow ballads, which drove us crazy every time we played there. I did some studying on the problem (1975 style with actual books) and found that sometimes speaker cables could feed RF back into an audio amplifier, where it could be demodulated inside the amplifier's negative feedback loop, then amplified by the output stage as significant wattage. One of my ham radio buddies suggested putting tiny ceramic capacitors right across the speaker wires to shunt out the RF before it got inside the audio amplifiers, and it worked like a charm. I can't remember the exact value cap, but it was in the pico-farad range. Now, I'm not saying that's the same problem here... just that RF frequencies have a way of getting inside of audio gear and being detected by any stray corrosion on a shield bond or a P-N junction in its path.

Not extremely common but yes, speaker wires can act like antennas at RF, and since many solid state power amps have an inductor in series with the output (to decouple from capacitive loads), that output can look like high Z at RF.

Any significant RF on the output line can come back into the amp via the feedback resistor. If the amp isn't fast enough (and few are fast enough to handle RF) that amp can slew rate limit and rectify the radio signal. Good amp design can minimize this with several design strategies that are TMI for this forum.

I've heard of this happening more typically in churches that are in some unusually strong radio station broadcast pattern.   

Be careful about arbitrarily throwing capacitors across power amp outputs, it can literally make some amps break into oscillation and release their magic smoke. I recall Jack Sondermeyer (Mr. CS800) would hang tens of uF on the output of every new amp he designed to protect them from Peavey Customers (don't try that at home). 

JR

PS: Zip cord should make a lousy antenna but enough RF can cause grief, twisting won't hurt but generally this should not be problem. Note: It isn't just power amps that can have RF snake back into their outputs, but again prudent design anticipates and manages this.

PPS: I really discourage people from lifting grounds, shields or otherwise but expect many here to ignore my advice. If you must lift, a "hybrid" grounding scheme, well known in broadcast circles, uses small caps to provide a good RF bond to ground at multiple points, while being modest to high Z at hum frequency. Using a cap in series with a safety ground is only acceptable to me in combination with a GFCI (0.15uF so it can still trip the GFCI), and so far never acceptable to UL or electrical code. 

[John Roberts {JR}]

I chased a problem like this a few years ago, and posted some pictures in an old version of this forum, but it seems to have disappeared.  Here's the 2016 version.

At my day job, we discovered that we couldn't get a clean signal out of some db Technology monitors (model 12-4M).  We have 8 of them, and they all exhibited the same problem, a slight "rizz" when plugged in to the building's AC power.  They hadn't always done this, but the problem showed up shortly after an 80 KVA building UPS was installed in the basement electrical room.  None of the other studio gear seemed to have a problem (the building housed radio studios and a music/drama recording studio) and various portable consoles and other powered speakers all seemed to be unaffected.  The problem was not a ground loop, because it happened as soon as the speaker was powered up, even before an XLR feed was attached.  It was also constant regardless of the gain setting on the speaker's amplifier.

I tried all the usual (and unusual) tricks to eliminate noise: lifting the ground, putting a 2KVA isolation transformer in the AC feed, even bypassing the diodes in the bridge rectifiers with small bypass caps.  Nothing worked.  Since the only thing they were connected to was the building power, it was tempting to blame that, but difficult to eliminate it in an A/B test.  Finally, I came up with the idea of running the speakers off a UPS, and the noise went away.  At the time I only had a fairly ordinary APC UPS (not double conversion), so I had to unplug its input to get it to generate a sine wave, but the speaker was happy when I did.

I didn't have Smaart to look at harmonics, but did look at the outputs with a scope.  I used an ordinary bell transformer to drop the voltage down to about 6V.  I suspect that most 60 Hz transformers will pass frequencies a lot higher than that.  The other interesting point I discovered was that the AC was noisy whether the speaker was plugged in on the line or load side of the UPs, although the character of the noise was a bit different.

Here are the waveforms.  The first is the line side (of the 80KVA UPS), second is load, and the third is the rackmount UPS output, with no load except for the scope and powered speaker.  The building UPS was running at about 60% of capacity.

GTD

I am inclined to blame the gear design, power is routinely dirty and probably getting worse with home automation and utilities trying to piggyback control data on the mains.

Years ago (way last century) I investigated transmitting audio over mains power wiring. Just like those cheap hifi speakers that did that. This would have been a huge cost saving for the fixed install business if they didn't have to pull wire for every install they did. Most buildings had mains wiring already in place.

The ugly secret is they don't work all the time. It would be a disaster for a sound contractor to bid on a job without the cost of pulling wire, only to find out after the customer turns on their copy machine, or whatever, that the system is noisy and they have to go back and hard-wire the job.

I shut down that product development after about a year with a junior engineer working on it... It worked acceptably well under ideal (clean mains voltage) conditions, but the real world is not ideal, and loves to bite design engineers in their ASSumptions. 

JR

[Scott Helmke]

Good news - I did head-to-head comparisons between the original adapter (voltage divider down from 120v) and an AC 120v to 15v wall wart into a voltage divider. The transformer in the wall wart showed all the same harmonics all the way up - not an exact match, so frequency response isn't that great, but definitely enough to show the problem. 

I might try to figure out how to take a video (maybe just with my phone camera), because the noise in the right half of the display tend to move around in interesting ways. Almost like there's a big elevator nearby, though I don't think this building has one.

[Mike Sokol]

Good news - I did head-to-head comparisons between the original adapter (voltage divider down from 120v) and an AC 120v to 15v wall wart into a voltage divider. The transformer in the wall wart showed all the same harmonics all the way up - not an exact match, so frequency response isn't that great, but definitely enough to show the problem.

Aren't these all odd-order harmonics at 180, 300, and 420 Hz? Since this is 3-phase power, methinks you might be looking at Triplen currents generating voltage spikes....

[ProSoundWeb Community]