Home |
Search |
Today's Posts |
#1
|
|||
|
|||
Ground Loop Hum Problem Not Solved by Balanced Audio Interface
The situation is that I have an input on the main sound system that comes
from a mixing board that has several synths and computer audio devices as inputs. They, unfortunately, are on a separate AC circuit. Even with the mixer powered off, the hum still exists, due to the various ground connections through all the input sources. I built the equivalent of a Henry Matchbox, to solve an annoying ground loop hum problem, but it does not solve the problem, other than giving me another 6dB of signal. The input of the sound system is an unbalanced input on a Carver C4000 preamplifier. I added the differential input stage as an externally-powered dual op-amp, with differential inputs for the balanced output of the mixer on the other side of the studio. (Prior to this change, I used to take the + and G output leads off the mixer and feed the unbalanced input.) Now the full +/- G connection is made to this level converter with the differential input stage, outputting an unbalanced signal to the preamp. Since the converter runs off the rack power that the preamp does, I figure that should eliminate the ground source problem. I've tried lifting the ground connection, making the system totally floating, but the hum remained and even picked up some harmonic frequencies of 60hz. Years ago, when there was only one computer, I had a not so elegant solution of using a line isolation transformer for the PC's power. However, with four synthesizers, a mixer and three PCs connected to a UPS, the isolation transformer no longer makes practical sense and would not handle the load anyway. Before I compromise my audio by adding audio 1:1 isolation transformers, I wonder it a reality check wouldn't be a good idea. Shouldn't balanced connections between systems on different AC circuits be able to cancel out the hum? (Especially if only the + / - leads are used and ground is not connected?) Just wondering what I should do, before I run out and buy a couple of transformers. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - |
#2
|
|||
|
|||
Before I compromise my audio by adding audio 1:1 isolation transformers, I
wonder it a reality check wouldn't be a good idea. Shouldn't balanced connections between systems on different AC circuits be able to cancel out the hum? (Especially if only the + / - leads are used and ground is not connected?) Just wondering what I should do, before I run out and buy a couple of transformers. Yes, balanced connections should dramatically reduce *common mode* noise. The shield lead should be connected at one end, usually at the downstream end, by convention. If RF is present, you may also want to ground the other end through a 0.01uf ceramic capacitor. However, your noise may not be common mode. If the cable runs past a strong alternating magnetic field (a power transformer, large motor, etc), it might pick up enough differential noise to be a problem. Common mode noise, if it is strong enough, can still get through the op amp, especially if the + and - inputs are not precisely matched. Also, op amps can only reject common mode voltages that are significantly less than the op amp supply rails. Audio line transformers have no such limitations (within reason). You might want to use them instead of op amps in extreme situations. Here is a good "how to" reference for balanced program feeds: http://www.dself.dsl.pipex.com/ampin...d/balanced.htm It goes on for pages, covering practically all anyone would ever want to know on the subject. |
#3
|
|||
|
|||
The situation is that I have an input on the main sound system that comes
from a mixing board that has several synths and computer audio devices as inputs. They, unfortunately, are on a separate AC circuit. Even with the mixer powered off, the hum still exists, due to the various ground connections through all the input sources. On a different tack, the problem might not be between the console output and the main amplifier at all (or the amp is just the victim). Are the console inputs also balanced? If so, you should be able to isolate the grounds here also, even if the sources are unbalanced. When ground loops are a problem, the key is to use balanced lines, and *not* connect the shield at both ends. Again, this site (http://www.dself.dsl.pipex.com/ampin...d/balanced.htm) shows how to connect unbalanced outputs to balanced inputs without inducing ground loops. I have had good luck with wide-area connections where ground loops were a problem, using the following hookup over shielded twisted pair (STP) cable: Balanced to balanced cable connections: red: src+, dest+ black: src-, dest- shield: src n/c, dest gnd Unbalanced to balanced cable connections: red: src hot, dest+ black: src gnd, dest- shield: src n/c, dest gnd You might also try a process of elimination. Disconnect all sources. Start with the simplest configuration, and work up from there. See if the noise increases incrementally with each new source, or if it jumps up suddenly with a particular source. I'd be interested to know if your console sounds clean in the program monitor (headphones) but not in the main amp. It might take some sleuthing to figure this out. |
#4
|
|||
|
|||
Mark & Mary Ann Weiss wrote:
The situation is that I have an input on the main sound system that comes from a mixing board that has several synths and computer audio devices as inputs. They, unfortunately, are on a separate AC circuit. Even with the mixer powered off, the hum still exists, due to the various ground connections through all the input sources. SNIP Years ago, when there was only one computer, I had a not so elegant solution of using a line isolation transformer for the PC's power. However, with four synthesizers, a mixer and three PCs connected to a UPS, the isolation transformer no longer makes practical sense and would not handle the load anyway. SNIP Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION This may be a clue to your problems. UPSs tend not to be too clean on the output. OK if your going to feed a PC PSU, but not so good for audio kit, and radiated interference can be a big hassle. I do use one, but had to work quite hard to get rid of hum problems. |
#5
|
|||
|
|||
"Mark & Mary Ann Weiss" wrote in message
nk.net The situation is that I have an input on the main sound system that comes from a mixing board that has several synths and computer audio devices as inputs. They, unfortunately, are on a separate AC circuit. Even with the mixer powered off, the hum still exists, due to the various ground connections through all the input sources. I built the equivalent of a Henry Matchbox, to solve an annoying ground loop hum problem, but it does not solve the problem, other than giving me another 6dB of signal. The input of the sound system is an unbalanced input on a Carver C4000 preamplifier. I added the differential input stage as an externally-powered dual op-amp, with differential inputs for the balanced output of the mixer on the other side of the studio. (Prior to this change, I used to take the + and G output leads off the mixer and feed the unbalanced input.) Now the full +/- G connection is made to this level converter with the differential input stage, outputting an unbalanced signal to the preamp. Since the converter runs off the rack power that the preamp does, I figure that should eliminate the ground source problem. You might be running into the common mode dynamic range limit of your buffers. Transformers don't have this problem but of course they have other problems. |
#6
|
|||
|
|||
Thanks for the reference material, Karl.
The problem here is due to the mixer and its sources being connected to a different AC outlet. There is one Carver M-500 amplifier connected directly to the outputs of the mixer, which drives a separate set of studio monitors. There is no hum problem there, as the M-500 is plugged into the same branch circuit as the UPS driving all the synths and computers. However, the big sound system runs off a separate 150-amp service, and bridging audio from the mixer to this system is where the problem lies. If I disconnect all inputs from the mixer, the hum remains. If I connect the mixer to a 110VAC line isolation transformer, with no inputs, the hum is gone. As soon as I plug ANYTHING into the mixer that is powered by the outlet that the mixer WAS plugged into, the hum returns at full amplitude. It is clearly a ground loop between outlets or things powered by those outlets. As for RF, we should be clean--the whole studio is Faraday shielded. No electric motors near the wiring. The only possibility is that my balance to unbalanced converter is not a true instrumentation amp. I used the - / + inputs with appropriate resistor values to give each input equal (unity) gain. Perhaps I should add a non inverting buffer in front of each differential amp. Thanks for your help! -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION • FILM SCANNING • DVD MASTERING • AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - On a different tack, the problem might not be between the console output and the main amplifier at all (or the amp is just the victim). Are the console inputs also balanced? If so, you should be able to isolate the grounds here also, even if the sources are unbalanced. When ground loops are a problem, the key is to use balanced lines, and *not* connect the shield at both ends. Again, this site (http://www.dself.dsl.pipex.com/ampin...d/balanced.htm) shows how to connect unbalanced outputs to balanced inputs without inducing ground loops. I have had good luck with wide-area connections where ground loops were a problem, using the following hookup over shielded twisted pair (STP) cable: Balanced to balanced cable connections: red: src+, dest+ black: src-, dest- shield: src n/c, dest gnd Unbalanced to balanced cable connections: red: src hot, dest+ black: src gnd, dest- shield: src n/c, dest gnd You might also try a process of elimination. Disconnect all sources. Start with the simplest configuration, and work up from there. See if the noise increases incrementally with each new source, or if it jumps up suddenly with a particular source. I'd be interested to know if your console sounds clean in the program monitor (headphones) but not in the main amp. It might take some sleuthing to figure this out. |
#7
|
|||
|
|||
Yes, but I had this problem years ago, before the UPS came.
-- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - This may be a clue to your problems. UPSs tend not to be too clean on the output. OK if your going to feed a PC PSU, but not so good for audio kit, and radiated interference can be a big hassle. I do use one, but had to work quite hard to get rid of hum problems. |
#8
|
|||
|
|||
Conceivable, since the voltage difference between grounds might be higher
than 12vdc, but I'd be surprised if it was. I'll measure it. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION • FILM SCANNING • DVD MASTERING • AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - You might be running into the common mode dynamic range limit of your buffers. Transformers don't have this problem but of course they have other problems. |
#9
|
|||
|
|||
Mark & Mary Ann Weiss wrote:
Conceivable, since the voltage difference between grounds might be higher than 12vdc, but I'd be surprised if it was. I'll measure it. I would be surprised if you find any DC potential difference between grounds. The hum that results from ground loops is caused by AC voltages. -- ================================================== ====================== Michael Kesti | "And like, one and one don't make | two, one and one make one." | - The Who, Bargain |
#10
|
|||
|
|||
"Michael R. Kesti" wrote in message ... Mark & Mary Ann Weiss wrote: Conceivable, since the voltage difference between grounds might be higher than 12vdc, but I'd be surprised if it was. I'll measure it. I would be surprised if you find any DC potential difference between grounds. The hum that results from ground loops is caused by AC voltages. If the op amp is running on +/- 12vdc, you'd be looking for AC noise voltage approaching 12vac peak-to-peak to exceed common mode range. I think this is unlikely. I would use a dual-trace oscilloscope on the + and - lines. That would show whether the noise is common mode or differential mode. Your approach to solving the problem must be completely different, depending on which it is. |
#11
|
|||
|
|||
If the op amp is running on +/- 12vdc, you'd be looking for AC noise
voltage approaching 12vac peak-to-peak to exceed common mode range. I think this is unlikely. I would use a dual-trace oscilloscope on the + and - lines. That would show whether the noise is common mode or differential mode. Your approach to solving the problem must be completely different, depending on which it is. DOH! Make that peak, not peak-to-peak. |
#12
|
|||
|
|||
Part of the problem was related to the fact that I used the + / - inputs
directly. Shifting source impedance may have unbalanced the differential gain. I just added unity gain non-inverting buffers ahead of all inputs (good thing I had plenty of unused room on the PCB). The gain of the inputs is perfectly matched, such that I cannot measure a difference on my H/P 400E. I used all 1% metal film resistors throughout. This has improved the situation, however, it's not as quiet as I expected (which was to remove ALL hum). I used a DVM to measure potentials between grounds and floating pins of both systems and found none at all. Nothing I could measure on AC *or* DC ranges. Nothing between grounds. Nothing between - to - pins, nothing between + to + pins. The hum is reduced about another 10dB with the addition of the buffers. It appears that switching from unbalanced to balanced bought me 16dB improvement, all-told. Not bad, but could be better. Lots of hours into this, but a small win in the battle against hum. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION • FILM SCANNING • DVD MASTERING • AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - If the op amp is running on +/- 12vdc, you'd be looking for AC noise voltage approaching 12vac peak-to-peak to exceed common mode range. I think this is unlikely. I would use a dual-trace oscilloscope on the + and - lines. That would show whether the noise is common mode or differential mode. Your approach to solving the problem must be completely different, depending on which it is. |
#13
|
|||
|
|||
On Sun, 09 Jan 2005 22:40:15 GMT, "Mark & Mary Ann Weiss"
wrote: Part of the problem was related to the fact that I used the + / - inputs directly. Shifting source impedance may have unbalanced the differential gain. I just added unity gain non-inverting buffers ahead of all inputs (good thing I had plenty of unused room on the PCB). The gain of the inputs is perfectly matched, such that I cannot measure a difference on my H/P 400E. I used all 1% metal film resistors throughout. This has improved the situation, however, it's not as quiet as I expected (which was to remove ALL hum). I used a DVM to measure potentials between grounds and floating pins of both systems and found none at all. Nothing I could measure on AC *or* DC ranges. Nothing between grounds. Nothing between - to - pins, nothing between + to + pins. The hum is reduced about another 10dB with the addition of the buffers. It appears that switching from unbalanced to balanced bought me 16dB improvement, all-told. Not bad, but could be better. Lots of hours into this, but a small win in the battle against hum. You should also try your luck at sci.electronics.design There are some very helpful people there. When you say that you measured no potential difference between grounds, just how close to zero did you get? Regards, Boris Mohar Got Knock? - see: Viatrack Printed Circuit Designs (among other things) http://www.viatrack.ca |
#14
|
|||
|
|||
You should also try your luck at sci.electronics.design There are some very helpful people there. When you say that you measured no potential difference between grounds, just how close to zero did you get? Regards, Boris Mohar Good suggestion. The measurements I took would have detected differences as low as 1mV. I was surprised that the difference was unmeasurable. I could get a reading of 1.000 volt by touching the probe, so the meter is definately sensitive enough to pick up small currents. The potential difference between the board and the converter box (powered by separate circuits) was far less than the potential difference between the meter probe and my finger. The hum probably wouldn't be considered severe by most standards, but it is still somewhat noisier than my phono input stage, and due to the buzzing nature of it, more annoying than the hiss of a phono preamp. It's probably down a good 55dB, but in this digital world, that's just not enough. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - |
#15
|
|||
|
|||
On Mon, 10 Jan 2005 03:19:09 GMT, "Mark & Mary Ann Weiss"
wrote: You should also try your luck at sci.electronics.design There are some very helpful people there. When you say that you measured no potential difference between grounds, just how close to zero did you get? Regards, Boris Mohar Good suggestion. The measurements I took would have detected differences as low as 1mV. I was surprised that the difference was unmeasurable. I could get a reading of 1.000 volt by touching the probe, so the meter is definately sensitive enough to pick up small currents. The potential difference between the board and the converter box (powered by separate circuits) was far less than the potential difference between the meter probe and my finger. The hum probably wouldn't be considered severe by most standards, but it is still somewhat noisier than my phono input stage, and due to the buzzing nature of it, more annoying than the hiss of a phono preamp. It's probably down a good 55dB, but in this digital world, that's just not enough. "due to buzzing nature" would indicate a frequency dependence of the common mode rejection in some part of your system. In other words you have good DC common mode but not so good at higher AC frequencies. The buzzing indicates that harmonics of 60hZ are more successful in getting through due to poor CMMR. Regards, Boris Mohar Got Knock? - see: Viatrack Printed Circuit Designs (among other things) http://www.viatrack.ca |
#16
|
|||
|
|||
"due to buzzing nature" would indicate a frequency dependence of the common mode rejection in some part of your system. In other words you have good DC common mode but not so good at higher AC frequencies. The buzzing indicates that harmonics of 60hZ are more successful in getting through due to poor CMMR. I decided to experiment with making a common ground, taking advantage of the hot water heating system's copper piping that flows throughout the studio. So I tied the UPS chassis to a nearby radiator. That reduced a LOT of the hum. Then I tied another ground strap from the outlet ground of the sound system to the radiator. That reduced additional hum. The hum is now way down. Such that I hear more hiss than hum when the volume control on the preamp is turned all the way up. At normal listening levels, it is inaudible now. This is good, and the level of noise is acceptable now. It has been a long month of modifications to amplifiers, rewiring, rearranging equipment by separating power amplifiers from preamps and sources in the racks and chasing down ground loops and even sources of hiss. A lot of optimization has been achieved. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - |
#17
|
|||
|
|||
Mark & Mary Ann Weiss wrote:
"due to buzzing nature" would indicate a frequency dependence of the common mode rejection in some part of your system. In other words you have good DC common mode but not so good at higher AC frequencies. The buzzing indicates that harmonics of 60hZ are more successful in getting through due to poor CMMR. I decided to experiment with making a common ground, taking advantage of the hot water heating system's copper piping that flows throughout the studio. So I tied the UPS chassis to a nearby radiator. That reduced a LOT of the hum. Then I tied another ground strap from the outlet ground of the sound system to the radiator. That reduced additional hum. The hum is now way down. Such that I hear more hiss than hum when the volume control on the preamp is turned all the way up. At normal listening levels, it is inaudible now. This is good, and the level of noise is acceptable now. It has been a long month of modifications to amplifiers, rewiring, rearranging equipment by separating power amplifiers from preamps and sources in the racks and chasing down ground loops and even sources of hiss. A lot of optimization has been achieved. -- Take care, Mark & Mary Ann Weiss VIDEO PRODUCTION . FILM SCANNING . DVD MASTERING . AUDIO RESTORATION Business sites at: www.dv-clips.com www.mwcomms.com www.adventuresinanimemusic.com - I came upon this thread a bit late, but that's exactly what I was going to suggest. Have everything connect through one common ground. I had a similar problem on a much less sophisticated setup. My computer DAW was connected to my stereo. Upon attaching my VCR to the stereo, hum galore. Apparently my system then had 2 gorunds, one via the computer and monitor's ground plug, and the other via cable TV connected to my VCR. I solved it by removing the ground plugs from my computer and monitor. Been totally humless for about a year now CD |
#19
|
|||
|
|||
"Dale Farmer" wrote in message ... Of course by removing the safety ground pins from your power cord you are defeating a major safety feature of the computer and monitor. It is your own life that you are putting at risk, I hope you know. But he still has one ground (instead of two) Problems can arise of course if he disconnects the audio cables and ends up with NO grounds. MrT. |
#20
|
|||
|
|||
I discovered something the other day that made me feel like an idiot. No wonder this problem was so hard to solve: I made an assumption that the 1/4" TRS jacks on the output of the mixer were balanced. I noticed some XLR connectors next to these jacks and they were labeled 'balanced outs'. I had assumed that the TRS jacks were also balanced, just like on much of my other equipment. Wrong. I checked the schematic of the mixer and discovered that only the XLRs are. I ended up chopping off the TRS plus from my cable and installed XLR receptacles so I could use the balanced outs. Problem solved! On another note, we just got a Behringer Ultra Drive Pro DCX2496 Loudspeaker Management System http://www.behringer.com/DCX2496/index.cfm?lang=ENG What an amazing piece of gear this is! It replaced four discreet boxes we used to use: 2 EQs, 1 crossover, 1 optical limiter. And it added features that I wished existed years ago, such as delay lines for aligning sound arrival times for better transient response and less energy cancellation at the intended listening area. The first thing I did was put this unit on the bench in our engineering lab and run some signals through it, with a scope and RMS voltmeter on the outputs. The box is flat from practically DC to 39KHz and cuts off like a stone falling through vacuum above 40KHz. The unit is VERY quiet and does not add any noise to our system. The filters are ACCURATE. Raising 0.5dB @1KHz on the display, causes our RMS voltmeter to rise by exactly 0.5dB only at the bandpass frequency set. The crossover points are accurate and the adjustable slopes are flexible. There are three filter topologies to choose from. The crossover points can be linked or independantly adjusted. There are limiters, dynamic EQs, and multiple EQ filter types that can be enabled and tuned to perfection. All settings can be saved to internal memory and backed up to PCMCIA card adapter with flash memory or smart media card. It all works like the fairy-tale-perfect engineer's dream of the ideal crossover network. When I connected it to our 10kW sound system, I was frankly worried that I'd hear a lot of hiss. I was much relieved to find that it was quiet enough that the speaker arrays remain silent when no program content is being fed through. It doesn't color the sound, and it doesn't harm the stereo imaging. All concerns I had during the purchase phase. In fact, due to the steeper Linkwitz-Riley filter slope that I chose, imaging improved, because directional cues were removed from the vast wall of woofers and effectively confined to the midrange drivers which are positioned for a coherent stereo image with Carvers' Sonic Hologram Generator. The steeper slopes also remove the possibility of IM distortion of the midrange drivers--their excursions were limited greatly by removing all traces of bass from their input. The whole system sounds more neutral and clean than ever before. The time alignment has improved detail on things like kick drums, as well as the impact and lower bass. Moving the timings around, one can fix some nasty phase cancellation problems. And there is a separate phase adjustment menu where the phase can be continuously rotated through 360º. In short, this box is like magic! I now have speaker protection, better sound, more bass with less power consumed, and better definition. And the opportunity to tweak, save settings to 64 memory locations and experiment. It's amazing what you get in a box under $500 these days, because of DSPs and digital electronics. Analog is finally finished. My homebrew Linkwitz-Riley crossover now sits in storage. RIP. :-) -- Best Regards, Mark A. Weiss, P.E. www.mwcomms.com - |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
OT Political | Pro Audio | |||
Artists cut out the record biz | Pro Audio |