The SOS team purge the unwanted buzzes and hums from reader Jim Fish's new studio.
This month's Studio SOS visit came about because of a post in the SOS Forum in which reader Jim Fish complained of serious hum problems in his studio. He had recently moved house and was in the process of rebuilding his studio equipment into a large room above a timber-framed garage, but he had started to get such high levels of noise that it could clearly be heard in the next room! He was suspicious that the problem might have been caused by two new LCD computer screens, with which he had recently replaced the previous pair of 'legacy' CRT monitors.
He described his equipment as being based on a Dell PC coupled via Firewire to an original MOTU 828. Mixing was performed on a large Soundtracs Topaz eight-buss analogue mixer and monitored on a pair of Fostex PM2 active nearfields, with a range of outboard effects units and synths hooked into the system via various patchbays. His pair of 18-inch LCD screens were mounted on the wall behind the Topaz desk and hooked directly to a Radeon dual-head graphics card in the PC.
The symptoms of the problem were intriguing, as the hum apparently remained if the mixer or MOTU 828 were turned off, and it became louder if the Firewire link between computer and monitors was removed. The only thing that silenced the hum was turning off the LCD monitors or removing their video inputs (in which case they shut down anyway, of course). One monitor seemed to create a louder hum than the other, and they produced slightly different pitches too!
Jim had found that jiggling the monitor video connectors or power leads affected the hum (suggesting poor earthing perhaps), and after relocating the monitor VGA cables along the back wall (to move them away from mains and audio cables) the hum level was reduced — although that could just have been a coincidence of remaking the connections, and thus improving the earthing contacts in the process.
A number of other Forum posters had offered help, but to no real avail. The usual suggestions of moving the LCD screens' power supplies around, using balanced connections to the Fostex speakers, and changing the computer graphics card had already been tried and found to make no difference. Questions over noisy mains supplies appeared to have already been addressed as well, since Jim was feeding all his equipment via ETA power conditioners. Clearly, this was a complicated problem and it sounded like a suitable challenge for Studio SOS.
However, when Reviews Editor Mike Senior suggested a visit, Jim confessed to being a 'Hobnob scrooge', and we all know that Paul White doesn't get out of bed nowadays unless there's a pack of chocolate Hobnobs in it! Fortunately for Jim, though, it turned out that Paul was going to be away visiting a certain Antipodean microphone manufacturer, and since Mike doesn't suffer the same debilitating condition as our Editor In Chief, it fell upon him to join me in attempting to find a solution to this thorny problem.
SOS To The Rescue!
On our arrival in Bungay, Jim demonstrated the loud buzzy hum he was getting on the system, and how it disappeared completely if he turned the LCD monitors off. That suggested that the buzz was definitely being generated by his two LCD screens, but exactly where the noise was getting into the system was far from obvious.
To confirm the source, we substituted the two GNR-brand screens for a different LCD screen of a different make, borrowed from his office, and that produced a substantially lower level of buzz and noise. However, the replacement monitor wasn't capable of the very high screen resolutions that Jim was using on his two GNR monitors, and it is certainly possible that the additional demands placed on the monitor and its power supplies to provide such high screen resolutions were a significant factor in the level and quality of noise generated. Jim was very keen to maintain the current high screen resolutions, though, and, while switching to less demanding resolutions may have helped, it clearly wasn't a proper solution to the problem.
Although it had already been tried, our first step was to experiment with moving the LCD screens' power supplies around. The specific angle, orientation, and distance of these separate switched-mode power lumps from other equipment and cabling can make a huge difference, as they usually emit electromagnetic radiation in specific patterns. However, after carefully turning them through 360 degrees in all directions we found no significant difference to the level of buzz. So direct radiation didn't appear to be the cause, and it looked much more like a complicated ground-loop problem.
Tracking Down Earthing Problems
The key to solving any kind of ground-loop problem like this is to work through the system logically and patiently. So the first step was to unplug the audio input to the Fostex speakers to make sure that they weren't picking the noise up directly through their mains power supplies, or via radio-frequency interference. Sure enough, the hum disappeared as soon as both speaker inputs were unplugged, proving that they were not faulty and that the hum was reaching them as a genuine audio signal.
At this stage, given the nature of the buzz and the confusing symptoms, it seemed fairly certain that the problem was the result of one or more earth loops. Jim had a lot of equipment hooked up to the desk, some balanced and some unbalanced, and the rat's nest of mains, audio, and computer cabling under the large custom desk probably wasn't helping either. So we decided it would be worth proving whether or not the suspect LCD monitors could be used in a minimalist system without creating the excessive hum apparent in the full system.
We disconnected the MOTU 828, one of the Fostex speakers and one of the LCD monitors from the main system, and re-connected them to a laptop computer, all powered from a separate mains extension fed from a different wall socket. An audio connection between one of the MOTU outputs and the Fostex input created no buzz at all — so we could now be confident that, while the LCD monitors were undoubtedly the source of the buzzing noises in the main system, sensible mains earthing and lack of ground loops could produce a quiet system.
Next we re-plugged the speaker, interface, and screen back into their original mains sockets under the desk... and the buzz returned! We hadn't changed any signal connections, so we hadn't created a ground loop that way, and the implication was therefore of a problem in the mains power distribution for the main system. A quick glance under the desk revealed an evil tangle of mains and audio cables, and Jim also admitted that the equipment was fed from two different wall sockets on opposite sides of the room. At this point we politely suggested a bit of rewiring...
Jim had no objections, so we were soon scrabbling around under the desk — a regular feature of Studio SOS visits! After a bit of pruning back of audio cables to reveal the mains distribution boards we spotted the first problem. A European mains plug had been forced into the first four-way plugboard coming off the left-hand wall socket. It turned out that this was feeding a (metal cased) switched mains distribution unit bolted into the left-hand desk rack — a unit purchased from a European supplier. This distribution panel was feeding both the MOTU interface and the Topaz mixer, amongst other things, and although the European plug was equipped with an earth terminal, the action of plugging it into a standard UK mains distribution board meant that it was left dangerously unearthed. So, neither the MOTU nor the console were actually earthed properly, which not only made them potentially lethal, but also made them a significant contributing factor in the buzz problem.
Mike volunteered to replace the European mains plug with a standard UK-style plug, and five minutes later when we re-powered the system the level of buzz was immediately improved, although it was still unacceptably audible. We were making some progress, and the system was definitely safer now, but clearly there were still some further problems to resolve.
Re-configuring The Mains Wiring
Since I was already under the desk, I set about unplugging all the equipment mains plugs and rewiring everything in a more sensible 'star' configuration, stemming from the wall socket on the left of the studio. All the ancillary equipment — lights, phone chargers, and so on, were fed from the right-hand wall socket to keep them from degrading the mains quality of the audio setup.
While doing this, we came upon a second bit of comedy mains wiring — the mains plug of a Roland U110 sound module had its earth wire protruding from the plug casing. Someone had obviously isolated the earth wire, probably in an uneducated attempt to cure a hum problem. At least they'd had the sense to leave the earth wire showing so that we could all see that it was in a dangerous condition! I re-connected it and made a mental note to check on the hum level when we got the rest of the system sorted out.
At the same time as rewiring the mains, we also re-patched the audio cabling to the rackmounted equipment and suspended it all well away from the mains wiring using a cup hook and some velcro strips. Several eight-way snakes had been used to hook up the console inputs and outputs with the MOTU and patch bays, and in places these were coiled on top of mains cabling and power-supply transformers, which wouldn't have helped the system noise floor.
So, with the mains distribution re-organised, audio cables re-routed, and the MOTU, LCD screens, and Fostex speakers reinstated to their original configurations, we fired everything up again and had a listen. The bad news was that the buzz was still there, but the good news was that it was even quieter than it had been previously. The restructuring had obviously improved things further, but there was still a major problem somewhere.
Hunting Down The Final Earth Loop
After further head-scratching, we returned to the idea that there must still be earth-loop problems, and our suspicion initially fell on the right-hand side of the studio, where a number of unbalanced rack synths were being fed via a balanced patchbay to the mixer. However, removing the eight-way multicore link between the patchbay and desk made no appreciable difference to the level of buzz.
However, coincidentally at this point we noticed that the level and quality of the buzz varied a little if the Firewire or mains connections at the back of the MOTU interface were wiggled. Furthermore, we found that the buzz increased dramatically in volume if either the Firewire or mains cables were removed — something Jim had mentioned in his original forum post. Mike felt that this pointed the finger at possible earthing problems in the Dell PC, as it suggested to him that the computer was relying on the MOTU for its earth connection. We tried grounding the PC via a jack cable to the mixer (with the MOTU Firewire cable disconnected), but that yielded no improvement. We were also able to exonerate the PC by replacing it with the laptop — which also made no difference to the buzz.
With the computer re-connected, our attention finally turned to the potential for earth loops created via the MOTU's audio connections. The first step was to unplug all the audio cabling between the MOTU and the Topaz desk (we had already removed the synth rack connections), and that resulted in blissful silence. So the desk coupled to the Fostex speakers formed an inherently quiet system, which was encouraging.
Plugging the audio cabling from the synth racks on the right of the studio reinstated a very low level of buzz, and this was traced to one particular cable fed from one channel of an Emu sound module. The other channel was quiet, and the noise varied in level as the plug in the back of the unit was wiggled, so we suspected that the output socket was making a poor earthing connection — perhaps one of the internal contacts was bent or slightly corroded or dirty. This was a relatively minor problem, though, and the buzz was barely audible once we'd fiddled with the connections a bit. In the longer term the Emu unit would benefit from having the socket cleaned and/or replaced, but it obviously wasn't the cause of the annoying buzz problem.
We continued re-plugging various cables back into the mixer until we reached the loom carrying the eight outputs from the MOTU interface, which plugged into the first eight mixer channels. Everything remained delightfully free from hum until the moment we inserted the first of these plugs, when the buzz was immediately reinstated in all its glory. It looked, then, like the problem was one of a ground loop with the MOTU connections, and I suggested the simple test of lifting the earth in one of the audio cables between the MOTU and the desk, to see if that would solve the problem, given that these connections were balanced.
Mike had brought a DI box with him just in case we wanted to try something like this, so we connected one of the MOTU's outputs to the DI box input and its XLR output back to the desk. Flipping the earth-lift switch on the DI box instantly cured the buzz, without affecting the signal quality at all, so we had finally found the ground-loop culprit.
A permanent solution would involve making or modifying a cable loom with the screens disconnected at the console end. All of Jim's ready-made looms had moulded connectors, which were impossible to modify in the way we needed, but he was able to find six individual balanced cables with Neutrik jack plugs on the ends, which we could dismantle and modify fairly easily. I showed Jim how to snip the earth connection at the desk-input end of each one, and we re-plugged the MOTU using these modified cables. With all six connected the buzz was completely gone, and on playing back one of Jim's mixes from the computer everything emerged in a nice clean state — much to everyone's relief! Jim resolved to locate another couple of cables and modify them to isolate the screen at the desk end after we had gone, completing the setup.
As Jim was still in the process of expanding his studio we had a few further suggestions for him to consider. Firstly, any other connections made to the mixer (he was planning to rig up the MOTU's inputs as well) which reinstated the hum should be tackled in a similar way. Secondly, if connecting any unbalanced gear to the mixer caused hum, then isolating transformers or DI boxes would be needed. In neither case should any mains earth be disconnected!
The synths on the right of the studio desk were being connected via a balanced patchbay to the balanced inputs of the desk, but many of the synths had unbalanced outputs. Sometimes this arrangement can create a 'one-legged' unbalanced signal for the desk, with low signal levels and obtrusive noise as a result. Proper unbalanced-to-balanced (pseudo-balanced) cables between the unbalanced synth and balanced patchbay would avoid this potential problem.
A Job Well Done
All in all, this had been a very challenging problem, with a number of interacting issues combining to create a seemingly illogical situation. The lack of a proper mains earth on one of the mains distribution units — because of its European plug — was worrying and proved to be a significant facet of the buzz problem. Other aspects included the unstructured mains distribution, the close proximity of audio and mains cabling, and the ground loops between the MOTU 828 audio interface and the mixing console. Clearly, the LCD screens were the source of the buzzing noise, and different monitors might have alleviated the problem completely, despite the other wiring issues, but with some logical experimentation and attention to detail we were able to cure the buzz completely and leave Jim with a clean and quiet system.
Jim's Comments
"After a long day of head-scratching the guys did a fantastic job and got the buzz so it was nestled right down with the hiss from some of the effects units and such (which is only audible when the system is turned up twice as loud as it would ever need to be). Well, as you can imagine, I was over the moon with this result, after two months of crippling my ear drums and literally giving myself a migraine whenever I tried to work on something. For the next week I was in a state of bliss, and I was finally able to get back to work. Unfortunately my happiness was short-lived, because the buzz crept back, although luckily still not even half as bad as it was before — I could just about keep on working.
"A few weeks after this, though, I got so peeved with my Dell PC for its constant crashing that I replaced it with a new custom-built model. Guess what? No buzz! Hooray! So the problem is finally solved, but really the whole thing still seems totally illogical. Encouraged only by my girlfriend, who likes to treat an illogical problem with an illogical solution, I have come to the conclusion that it was definitely the pixies playing tricks on me... yes, the pixies... that's what it was..."