Hum issue

  • Thread starter Thread starter famous beagle
  • Start date Start date
famous beagle

famous beagle

Well-known member
So I was recording some drum tracks today and noticed a hum when listening through headphones. Sounds kind of like 60-cycle.

I was running mics through a phantom power supply through my Tascam M-216 mixer and into the CPU. The CPU's output runs through a small Behri mixer, which is just used as a line mixer so I can use the same power amp to listen to the CPU, the Tascam mixer (when using my R2R), or my cassette 4-track without repatching.

I spent about two hours trying to track down that hum. It was not recorded on the tracks. I sent the computer’s output to my boombox aux in and listened to the tracks, and the hum wasn’t there. I also did a mic check using the Tascam mixer's pres while listening through the phones on the boombox, and the hum was not there.

So I thought, “Ok, it’s the Behringer.” So, I unplugged everything from the Behri except the power cord. Plugged a 57 mic into channel 1 and listened through the headphone jack on the Behri. Clean as a whistle! No hum. So I thought “WTF?” I plugged all the CPU output back into the Behri and then systematically replaced every cord involved: the sub outs from the Tascam mixer to the patchbay, the cords from CPU’s output to the PB, the cords from the PB to the Behri inputs, etc. I even tried swapping patch points to make sure it wasn’t the patchbay! The hum never went away.


So I don’t know what the hell. When I listen to the CPU by itself (through the boombox, for instance), there’s no hum. When I listen to the Behri all by itself (just a mic plugged straight into it), there’s no hum. But when I listen to the CPU through the Behri, there’s hum. Problems like this drive me f-ing nuts!


The good news is that it’s not ending up on the recording, so when we actually record, we can just send the output straight to the power amp/phones and bypass the Behri altogether.


It’s really weird, though, because I just recorded a bunch of ukulele tracks for this book I wrote, and I listened from the CPU through the Behri, and I don’t remember hearing this hum at all. And I was using the same signal chain: mic into phantom supply into Tascam into CPU. It’s possible that the hum was there and was just masked by the AC noise or something. I let the AC run while I was recording the uke and used a noise removal plug-in to get rid of the background noise. (It worked pretty well too.)



I don’t know. Anyway, hopefully something will reveal itself. It’s an annoying problem. Anyone got any ideas that I haven’t thought of?

Thanks
 
Yes, This sounds very much like an earth(ground) loop.
Make up a couple of patchleads wth the shield dissconneted one end but remember to mark them up as such.
Do not lift mains earths.

Dave.
 
I'm afraid I don't entirely understand this. Which cords would I need to replace with these newly made patch cords? Just the two output cords (L and R) from the PC to the Behri?

What exactly is happening by doing this?

Thanks!
ecc83 said:
Yes, This sounds very much like an earth(ground) loop.
Make up a couple of patchleads wth the shield dissconneted one end but remember to mark them up as such.
Do not lift mains earths.

Dave.
Click to expand...
 
famous beagle said:
I'm afraid I don't entirely understand this. Which cords would I need to replace with these newly made patch cords? Just the two output cords (L and R) from the PC to the Behri?

What exactly is happening by doing this?

Thanks!
Click to expand...
Sorry I'm late on parade! It probably does not matter where you break the signal earth. Two devices (at least) in your setup are earthed twice. Once via the chassis safely earth. DO NOT TOUCH! And again by the linking audio shield. Thus a small current can flow around this circuit and, since there is no such thing as a wire with zero resistance (above about -100K!) this give rise to a small AC voltage that gets injected into the audio input. Break the loop and you kill the hum (usually! Ground loops can be a bugger to sort sometimes!)

A more "sophisticated" way to break earth loops is to use a 1:1 audio transformer box (NOT a guitar DI box!) such as the Art Cleanbox 2. A much cheaper device but one that seems quite good is the "Shark" ground loop isolator model # VW43W. Have a Googe.

Dave.
 
Heh! Just Googed, there are a million of them!
I suspect that they all use much the same transformers (the only bit that matters) in a variety of pimped up cases.

Dave.
 
"Would I just insert this between the CPU's output and the Behri?"

Probably but it pays to experiment. No smoke I promise you!

Dave.
 
Ok, someone es'plain this to me:

I purchased the ground loop isolator I linked above. I inserted it between the CPU output and the patchbay (the points of which are connected to the Behri's input). I put on headphones and listened for the hum. It's gone! Awesome. Buttttt, for some odd reason, the volume on the right side is much quieter than the left.

So, here's the strange part. I switched the cables on one side of the isolator (plugged R into L), and the right side of the phones was still quieter. Then I switched the cables on the other side of the isolator, and the right side was still quieter.

Just to be sure, I took the isolator out of the loop completely and ran only my cables. The stereo image in the phones was perfect again, but the hum was back.

What in the world?

Anyone got a clue?

Thanks
 
Try each "leg" separately, i.e. keep the hum on one channel, isolate the other.

Just come to me that you might have a strange source impedance on one channel. Try the isolator in a completely different line.

Dave.
 
ecc83 said:
Two devices (at least) in your setup are earthed twice.
Click to expand...

Not necessarily. A so-called ground "loop" need not be a loop. In fact, nearly every time I've seen this, it has involved only one piece of gear with an earth ground and lots of gear with two-prong cables or wall warts. There are only two ways to absolutely avoid ground hum problems: ground the chassis of every piece of gear (e.g. with star grounding) or have no grounds on anything (potentially dangerous).

What's probably happening is that some of your gear in the output path (the computer, the Behringer, and/or the power amplifier) is properly grounded and other pieces aren't. The device with the good ground is acting as a noise sink for the 60 Hz chassis hum of one of the other devices, and all that hum is flowing through the audio cables' shields. As it does so, it bleeds into the audio that is running through those cables.

You can hack around the problem by lifting the signal ground on one end of certain cables (whichever cables happen to be carrying the noise) or by adding an isolation transformer or optocoupler inline, but a better fix is to properly ground the piece of gear that is bleeding so much mains hum into its chassis in the first place. This is better not only from a noise perspective, but also from a safety perspective. (Ever get a shock when you touch a piece of gear while walking barefoot on a concrete floor? That's caused by the same sort of badly ground gear that is producing your hum.)

To fix this, if you're comfortable working around electrical wiring, first make yourself a custom power cord that has only a large gauge (e.g. 12 AWG) stranded ground wire (connected to the third prong ground, not the neutral) and with no hot or neutral wire.

Important: Use lots of electric tape over the hot and neutral contacts to ensure that even if the ground wire comes loose, it cannot physically short against the hot or neutral contacts inside the plug. :)

Next, plug the cable into an outlet and short the signal ground (at any of the jacks on the back) of each piece of gear to your grounding wire, one device at a time. When the hum goes away, make it permanent by wiring up a custom audio cable with that extra ground wire attached to its shield, and use that in place of the normal cable. For maximum noise reduction, ground each piece of gear in this way.

If you aren't comfortable working with household AC wiring, find someone who is. I don't want to get anybody electrocuted here. :)
 
Dgatwood.
The "earthed twice" model is THE classic way that hum gets into systems. To suggest otherwise is I believe unhelpful and misleading.

The most common scenario that the projjy studioist finds is that of hum in active monitors when connected to an earthed AI (usually grounded by connection to a PC or laptop*). The hum can even still be present when source and sink are balanced, but I will come back to that point. "Classical" again is the fix. Either a 1:1 transformer in the audio lines to break the second earth path or special lead which has the shield disconneted one end (and RF "keeper" cap is advised here).

There must be thousands of instances of this most common hum problem reported in forums in any given year and one or other of the above fixes almost always works.

That hum and grounding problems are far from simple can be seen in..

http://www.jensen-transformers.com/an/generic seminar.pdf
and I note on page 21 that Mr Whitlock states "DO NOT ADD UNNECESSARY GROUNDS (his emphasis) Additional grounding of equipment tends to increase system ground noise rather than reducing it."

That said, the network industry does often employ parallel earth cables (PECs) to bypass the very high shield currents that can flow in systems but that is for a different problem to do with "RF".

Most "domestic" audio/video equipment is class 2 insulation and thus does not have a mains earth but even so multiple interconnected systems of CD players, cassette decks, minidisc recorders, video, DVD/HDD recorders even my Teac A3440 tape machine, give no hum problems whatsoever. Such systems all untimately get earthed via a power amp or a computer.

Balanced connection does not always eliminate earth LOOP noise. Depening upon how good the balance is from source to sink the system can reject hum induced on the lines but much equipment has a very poor input common mode rejection ratio (the classic 10k+10k+10k op amp) and this does not remove the residual earth "situp" voltage. A very good quality INPUT transformer has a CMRR some 2-3000x better than most electronic inputs and are far better at such rejection but transformers are bulky, costly and degrade audio perfomance to some degree.

*This is proof of the "loop"! Laptops are notorious for generating noise, hum and often mouse movement related clicks. But run them on battery power and all is silent. The very best solution here is to buy a power supply with class 2 insulation, i.e. not earthed.

It is a fact of course that I am in the UK and circumstances might be different. For instance a very well meaning chap published a piece in a noted audio magazine here a couple of years ago about the benefits of a balanced mains supply. Such a supply could leave equipment in a very dangerous condition should an internal mains fuse blow.

Be SAFE ALL!
Dave.
 
ecc83 said:
Dgatwood.
The "earthed twice" model is THE classic way that hum gets into systems. To suggest otherwise is I believe unhelpful and misleading.
Click to expand...

The term "ground loop" is a very, very poor way to describe the problem. A ground "loop" is caused by having two different potentials between two things that are ostensibly supposed to both be grounds (and by ground, in this case, we mean signal grounds, which are not necessarily earth grounds).

The potential on either device can be anything from zero (shorted to ground) up to infinity (not connected to ground at all). What matters is that the signal grounds are not at the same potential. In no way is it required that either "ground" actually be grounded. The problem occurs in exactly the same way even if every ground but one is completely floating. To suggest otherwise is seriously misleading and can lead people to make very wrong assumptions about where ground lifts can be needed. In the absence of a proper star grounding scheme, ground lifts can be required between any device that has an actual earth ground and any other device, regardless of whether that other device has an earth ground.


ecc83 said:
That hum and grounding problems are far from simple can be seen in..

http://www.jensen-transformers.com/an/generic seminar.pdf
and I note on page 21 that Mr Whitlock states "DO NOT ADD UNNECESSARY GROUNDS (his emphasis) Additional grounding of equipment tends to increase system ground noise rather than reducing it."
Click to expand...

That advice is wrong. Period. Star grounding is a very well established way to fix hum in audio systems, both at the macro level (studio-wide) and internally within electronic devices themselves. Anyone who would suggest otherwise either doesn't know what he is talking about or is intentionally twisting reality to sell more transformers. Either way, bad advice is bad advice.

Sure, you can make hum problems worse by adding grounds, but only if you ground some of your devices instead of grounding all of them, and only if the devices you choose to ground are the wrong ones. :)


ecc83 said:
*This is proof of the "loop"! Laptops are notorious for generating noise, hum and often mouse movement related clicks. But run them on battery power and all is silent.
Click to expand...

The majority of that noise is caused by poorly designed power supplies. When running on battery power, all is silent because the noise is actually caused by the external power bricks not responding quickly enough to the rapid changes in power utilization by CPUs as they turn cores on and off, GPUs as they power cycle, etc. I would talk a bit more about this, but to be honest, the details of how inductors in properly designed supplies damp the VDC jitter are a bit beyond me. Either way, I've seen nearly identical machines go from electrically noisy as heck to electrically dead silent just by moving to a better-designed PSU. The presence of the ground is usually not the culprit.

That said, in some cases, the hum does go away if you lift the mains ground on a laptop. The reason for this is that in an entirely floating system with no grounds, no noise flows through the cables because there's no sink for the noise to flow towards. That said, floating the computer is a bad idea, because if you are grounded (e.g. barefoot on a concrete basement floor), you will get a small electric shock every time you touch the machine. It tingles, mostly, but it still isn't a good thing to do.... :)
 
Last edited:
Well this chap calls them earth (ground) loops and I will take his word, with respect, over yours.
Q. What is a Ground (Earth) Loop?

Bill Whitlock of Jensen, far from promoting his companies transformers has been instrumental in the development of the "THAT" chip which uses some rather clever bootstrapping techniques to give an electronic, balanced input the same CM Z as a good input transformer, i.e. around 50MOhms. He recognises that however good an audio transformer is made it will degrade the signal to some degree.

I agree about "star" earthing. The point is always made by Hugh and others on the forum that equipment should always be fed from a single outlet (or a double in our case, 50mm of copper strap ain't going to hurt anything!). This helps but 5mtrs say of mains cable is still going to have some impedance and hum loops can still be a problem.

Laptops: To suggest that EVERY laptop that had its hum problems cured by using a 1:1 audio transformer or a diss'ed audio shield had a rubbish PSU is plain daft. The "tingle" btw can be fixed with a 100k or so resistor for floating chassis.

But again, we are several 1000 clicks apart. Our supply is done different. Almost always comes in underground in steel armour and most domestic supplies are "Protectice Multiple Earthed" which means the incoming neutral is tied at the meter to the armour. We are also "cursed" with a ring main system which is very economical on copper but not as "quiet" as multiple spurs.

As for touching electrical gear, barefoot on concrete? I don't! I live in 240volt land and don't take those sort of chances!

But I think we can agree that hum problems can be a bugger to sort and can really only be done hands on, on site?

Dave.
 
ecc83 said:
Laptops: To suggest that EVERY laptop that had its hum problems cured by using a 1:1 audio transformer or a diss'ed audio shield had a rubbish PSU is plain daft. The "tingle" btw can be fixed with a 100k or so resistor for floating chassis.
Click to expand...

I never said that. I said that most of the noise from laptops is caused by bad power supplies. 99% of the noise that people complain about hearing when using laptops is clicking and chirping and other bizarre sounds, not hum. Hum from laptops (the remaining 1%) is almost invariably caused by the laptop being the sole three-prong device in the system.
 
dgatwood said:
I never said that. I said that most of the noise from laptops is caused by bad power supplies. 99% of the noise that people complain about hearing when using laptops is clicking and chirping and other bizarre sounds, not hum. Hum from laptops (the remaining 1%) is almost invariably caused by the laptop being the sole three-prong device in the system.
Click to expand...

Ok, well if I misunderstood you I apologise. The problem, hum,clicks and whistles is usually cured over here 99% of the time (if we are bandying percentages!) by a 1:1 traff in the audio lines or a class 2 insulation power supply.

We obviously have a different take and experiences about what can be a very intractractable problem. I think others reading this now have enough information to diagnose their individual situation so in the interest of not confusing them further shall we agree to differ?

BTW. Folks might like to Google for "The Pin One Problem"?

Dave.
 
Sure.

BTW, I'm not sure why that page called pin 1 "Signal" in the old standard. It hasn't ever been used for the signal. Maybe they misinterpreted somebody's discussion in which S stood for shield, but if so, then that's the same as the current standard. :)

The question of whether to ground the barrel or not is a long-standing one, and my experience is that yes, you should, because if you don't, you'll get all sorts of lovely hum if you ever plug the cable into an impedance matching transformer. And anyone who needs to lift a ground will probably end up lifting the signal ground, too, because half the time it is the same as the chassis ground, so those folks have to open up the connector anyway. :)
 
Well we didn't tie the shells to pin one, certainly not on 100volt speaker lines since when laid in a muddy field with mic lines the amps would go instable!

Like I say....Complicated!

Dave.
 
Wow, y'all. To say I'm confused at this point would be an understatement. :)

However, am I to understand that "star grounding" a system simply means to make sure that everything in the entire system is plugged into the same outlet on the wall (or a dual outlet, as we always have in the US)? If so, obviously many power strips will have to be used.

I think I've done that, but I'm not entirely sure. But I can check for sure. Does the quality of power strip matter in this regard? I know that I have a few cheap ones in there, but I've never had this problem before, so I never thought twice about it.

Thanks for all the info.
 
Can I post a question about my ground loop problem in this thread? If not, where could I and how could I create a new thread for it?
 
You must log in or register to reply here.
Back
Top