FAQ: High Frequency Noise on Auxes

[nottooloud]

Yes, that's what the hummer test confirmed.

[RoadRanger]

I don't think the issue is a traditional "pin 1" problem so much as a digital circuit RF infiltration issue. It seems mackie did a poor job keeping the digital RF hash out of the analog domain.

Actually I seem to remember that the noise goes away if you lift the ground to the power supply. That implies that the problem is that they tied the AC ground to the power supply negative output in the power supply - they really should have brought out three wires from the power supply to the DL1608 .

[nottooloud]

the noise goes away if you lift the ground to the power supply.

Unfortunately, so does the safety path to ground for the musician with his lips on the microphone.

[TimC]

My experience running auxs to powered monitor with straight TS to TS 1/4 was unbearable, but fully resolved using TRS to TRS 1/4 cables (expect TRS to XLR would work fine as well).  Still working on eliminating high freq noise on synth (Yamaha XS8) input (output from synth is unbalanced). Was hoping to avoid investing in DI box, but doesn't look hopeful. Anyone have recommendation on what has worked?

[Greg C.]

My experience running auxs to powered monitor with straight TS to TS 1/4 was unbearable, but fully resolved using TRS to TRS 1/4 cables (expect TRS to XLR would work fine as well).  Still working on eliminating high freq noise on synth (Yamaha XS8) input (output from synth is unbalanced). Was hoping to avoid investing in DI box, but doesn't look hopeful. Anyone have recommendation on what has worked?

Is it the same noise as patching the monitor with an unbalanced cable? IOW, it the synth generating the noise or is it the mixer once again dumping ground path noise into it's own input?

[WK154]

My experience running auxs to powered monitor with straight TS to TS 1/4 was unbearable, but fully resolved using TRS to TRS 1/4 cables (expect TRS to XLR would work fine as well).  Still working on eliminating high freq noise on synth (Yamaha XS8) input (output from synth is unbalanced). Was hoping to avoid investing in DI box, but doesn't look hopeful. Anyone have recommendation on what has worked?

Is it the same noise as patching the monitor with an unbalanced cable? IOW, it the synth generating the noise or is it the mixer once again dumping ground path noise into it's own input?

The two problems are totally different as well as the solutions. One is a unbalanced output ( consumer level XS8) into a balanced pro level input of the DL. The Monitor hookup ( unknown monitor pro or consumer?) from the balanced aux output of the DL into the monitor. There is no magic fix-it solution that would properly handle all cases. For the XS8 I would use a ISOMAX DB2PX from Jensen Transformers or something equivalent. Not inexpensive but worth while addition to your toolbox. It's a two channel unit. For the monitor it could be a ISOMAX PC2XR if the monitor is consumer level. There are a lot cheaper DI boxes out there but in the end you get what you pay for. Read AN003 from Bill Whitlock at http://www.jensen-transformers.com/apps_wp.html

[sam.spoons]

The safety earth is indeed tied to the low voltage -ve. Removing the safety earth does not remove the noise (at least not on my system, see earlier post). Have you tried jack inputs with a TS-TS lead? If they are noisy then you almost certainly need a DI box, if you're in the UK buy one of these http://www.soundonsound.com/sos/may12/articles/orchid-di-boxes.htm Micro DI is only £25 posted, brilliant!

[Greg C.]

The two problems are totally different as well as the solutions.

Not necessarily. It depends on how the noise is coupled to the audio circuity. My suspicion is that the noise is riding on DL's internal audio ground bus. If you use an unbalanced connector that shorts ground to cold, which is exactly what happens in a 1/4" TRS connector when you plug in a TS cable, noise riding on the ground would then be introduced to the cold input on a differentially balanced input. Since the noise would not be common mode on the hot connection, it would not cancel. In the case of plugging in the unbalanced cable on on the input, the noise would be introduced into the the DL's channel input as opposed to the self powered monitor scenario where the noise would not be cancelled at the monitor's input using an unbalanced cable. In both cases though, the source of the noise would be the DL's signal ground bus. When using a balanced cable in both instances, the cold connection is never tied to ground and the non-common mode noise would not be injected into the balanced input though the noise would still be coupled to the upstream or downstream device's audio ground. So long as said device has proper isolation to prevent that noise from coupling within it's circuitry, you wouldn't know it's there. Of course, this is just a theory, but it seems logical based on reports. At my day job, we design a variety of FPGA based digital A/V equipment and go through great pains to ensure that switching PSU noise and various digital clock circuits don't get into the analog audio circuity. It's not always easy & I suspect Mackie didn't do their due diligence in that regard. FWIW, this isn't the first time I've noticed a whine in Mackie digital mixers. Their d8b board also had a whine on the control room outputs of a similar nature, though it may not have been as bad.

As for line level transformer isolation, I agree that the Jensens are the way to go if you can afford it. In lieu of those, I'd recommend the Whirlwind ISO1 or ISO2 transformers for line levels as they have very good specs for a lot less money. They can take fairly hot signals (+20dBu @ 20Hz) without saturating and are pretty flat to that point. The devices come with both 1/4" and XLR connectors on inputs and outputs so you can balance and unbalanced lines as needed along with the isolation.

And to the OP, get a decent quality passive DI for the keys and be done with it. As long as the transformers are good, it will sound great and you won't have to worry about noise problems. I generally use Radial JDIs for keys and other line level devices. They have Jensen transformers in them and sound great. If you can't afford the JDI, the Whirlwind Directors work ok.

[WK154]

Read Bill Whitlock's explanation AN003 it's too long to repeat here but has valuable info on the subject. I also considered the signal differences (pro vs. consumer level) which you never touched on which clearly makes for different solutions. Some of the solutions you proposed (low $) use an output transformer, a clearly inferior solution compared to a input transformer as used in the ISOMAX . As to noise on the signal return in the DL, that would cause balanced to balanced (XLR) to react as well and it doesn't. A quick check of the DL indicates no direct connection of chassis ground and signal return. The power supply on the other-hand has safety ground tied to the power supply's ground output as opposed to a more direct path from the chassis. Clearly a poor design. Since it takes two kits to make noise we can't exclude the interaction of that design either. The DL chassis leakage current should be drained by a different more direct path. We are also getting into an area where schematics are a must.
To the OP buy the best solution that you can afford not the cheapest. Ask what kind of transformer is used and stay away from output transformers.

[Greg C.]

Read Bill Whitlock's explanation AN003 it's too long to repeat here but has valuable info on the subject. I also considered the signal differences (pro vs. consumer level) which you never touched on which clearly makes for different solutions. Some of the solutions you proposed (low $) use an output transformer, a clearly inferior solution compared to a input transformer as used in the ISOMAX.

I'm very familiar with Bill's writings and often refer folks to them myself. I've had email conversations with him on various grounding subjects. He's a smart cookie. And I agree that input transformers generally offer better CMMR. However, there are cases where using an input transformer isn't always possible to accomplish certain tasks. For instance, if you need to balance an unbalanced line level output for a relatively long run and maintain line levels rather than step down to mic levels. You cannot use an input transformer for that task due to the inherent issue of input transformers requiring close physical connection of their output to the input they're feeding due to capacitive loading. And while Bill's method of isolation does offer more noise rejection on paper, output transformers have worked in pretty much every application I've ever used them in to get rid of ground noise usually caused by loops from running gear over distance on different branch circuits in given venue. By all means, use the ISO-MAX stuff. It's great (except the dip switches suck to deal with on dark stages in tense situations) but not readily available or affordable for everyone.

As far as consumer vs. pro levels not being addressed, I'm not sure what you're talking about. The DIs referred to are designed to handle a variety of levels from weak consumer signals to hot "pro" levels. Yes, they must drive a mic input rather than line input which is par for the course. The ISO1 can handle most line levels as well.

As to noise on the signal return in the DL, that would cause balanced to balanced (XLR) to react as well and it doesn't. A quick check of the DL indicates no direct connection of chassis ground and signal return.

Why would a balanced to balanced connection cause noise? And I'm not sure what you mean by "no direct connection between chassis ground and signal return." The chassis is connected to the sleeve connection on the input, no?

The power supply on the other-hand has safety ground tied to the power supply's ground output as opposed to a more direct path from the chassis. Clearly a poor design. Since it takes two kits to make noise we can't exclude the interaction of that design either. The DL chassis leakage current should be drained by a different more direct path.

Agreed.

To the OP buy the best solution that you can afford not the cheapest. Ask what kind of transformer is used and stay away from output transformers.

While I will always agree that you should always purchase as much as you can afford, output transformers are not as bad as you make them out to be and have applications that can't be addressed with input transformers. A decent quality output transformer is far superior to having no transformer at all and will provide one key element regardless of it's lesser CMRR as compared to an input transformer - you can still completely isolate the ground between devices which is the root of most noise issues.

[WK154]

I was addressing the specific OP issue about the Yamaha XS8 not a bunch of other hypothetical situations. I try not to comment on situations that have little to no useful information ie. powered monitor (no model # or other helpful info.). I can give lots of scenarios were I would use much different solutions such as optical or wireless. I'm guessing his distance is quite short (Keyboard to mixer ~10 ft.). Input transformers are named that way because their location is near the input so why would I locate them near the output?  In fact Jensen had a mod for the Cr1604 that located the transformers in the unit where they really belong (unfortunately it doubled the cost of the Cr1604). Bill's solutions work and not just on paper that's why he is one of the Worlds foremost authorities on the subject. He is one Valley away and I have attended his seminars and spoken to him on several occasions at AES meetings and at his shop. I try not to find myself in situations that require last minute panic setup changes of this order. The pro to consumer level of 12dBu is what I was referring to which is not always handled by the mixer. As to the noise issue on unbalanced external kit I will try to reproduce the noise condition and see if there is a simple solution such as a more direct chasis ground. I agree that output transformers give you the isolation and in most cases solve the noise problem but give you little further protection against other noise sources especially with long runs of unbalanced 2 wire cable. I did measure the DL leakage and remember somewhere around 400uv or -64 dBu for the unbalanced cable use. That would be the noise floor which is not good. I will revisit that measurement.

[TimC]

Thank you all for the great insights and recommendations.  Sincerely appreciated.  The powered stage monitor I am using is a Behringer Eurolive 205D and with balanced TRS cable from DL1608 aux to main input the noise level is greatly reduced and satisfactory. With regards to the my main synth, a Yamaha Motif XS8, I do think it is a contributor to noise (FYI, I am not using the FireWire connection for connection to a PC). I do get some lower frequency noise/hum when using an analog mixer as well - Yamaha MG16/6FX, just not the same magnitude.  I have two other keys that plug into the DL1608 as well (Yamaha MO6 and a DSI Prophet 08) and they are clean.  I am using 15 and 20 foot cables into the mixing board.

[Greg C.]

I am using 15 and 20 foot cables into the mixing board.

15-20 feet can be problematic for unbalanced cables in a RF and ground loop hostile environment, especially since that model of keyboard uses a 3 pin AC cord. I would suggest purchasing a quality passive DI and using a shorter run of unbalanced cable to the DI, then use regular XLR cables between the DI and mixer. When using a DI, you will need to use mic inputs rather than line inputs. If you plan on providing stereo feeds for various venues, you might consider a stereo DI or 2 single DIs. If for some reason you must use the line inputs on the DL, then get line level isolation transformers instead. Get the ISO-MAX units for the best possible quality. Get the Whirlwind units if you're on a tighter budget.

[TimC]

I decided to go with the Radial Pro D2 to address the synth noise since my primary application is live sound as opposed to studio.  I will say the Radial JDI looks to be the top of the line solution, but hoping my choice will be satisfactory. Will provide an update on how things work out when I receive the unit. Thanks again for the excellent advice.

[Greg C.]

That DI will work great. They rate maximum input at 20Hz at +18dBu which is pretty good. That means it will handle even more input at higher frequencies and it's doubtful the keyboard's output is anywhere near that hot anyway.