Next Tech: Ground Loop/Lift

To many ads? Support ODJT and see no ads!

Proformance

DJ Extraordinaire
Nov 6, 2006
8,444
5,969
I've moved onto my next tech project. I'm inserting a small circuit on the ground lead of all my quad boxes and power strips. It essentially has the effect of lifting the AC ground (to a limit of 0.5V) without actually breaking the connection. I'm hoping it will reduce the instance of my having to chase down ground loops that originate in the AC distribution of certain venues. It does what the HUM-X does for about $0.30 instead of $79.

Has anyone every tried this...
 
  • Like
Reactions: Ausumm and steve149
No .. have a link to a schematic?
 
I've installed them, but haven't been able to field test them yet. I'm going to do some testing here in the shop with respect to how they interact with GFI circuits, dead shorts to ground, etc.

If anyone has any test ideas to add, let me know. I won't get to evaluate their effectiveness at reducing AC ground loops until I can locate AC sources with differential grounds. I'm curios about the voltage drop in loops that exceed 0.5v as to whether it produces a meaningful reduction in noise. I also want to find out how they react to environments with a lot of static electricity.

The NEC provides for up to 3% ground differential within a facility - which at 120V could be as high as 3.5V. I think it might be common to find up to 2V but I've never actually measured any of the ground differentials I've encountered.
 
Similar, with an explanation how it works -
Novacek301-4.jpg

"This shows how to break ground for appliances, such as a PC, with three-prong plugs. You can build this circuit into a computer or another appliance, but I find it better to build it as an independent break-out box. The diodes provide open loop for signals up to about 1.3 VPP. A hum is usually of a substantially lower amplitude. C1, 0.01 µF, provides bypass for high-frequency EMI to ground. The loop would be closed for voltages higher than 1.3 VPP, such as the ones due to isolation fault of the hot wire to the chassis. For 120 VAC distribution, D1, D2, and C1 should be rated for 250 V at a minimum. In a circuit branch with a 15-A breaker or fuse, the diodes need to be rated for a minimum of 20 A so that the breaker opens up before the diodes blow. If the appliance takes only a fraction of the rated fuse current, say 2 A, you could use 5-A diodes and include an optional fuse rated for 2 A. For countries with 230-VAC power, the components must be rated accordingly."
 
Similar, with an explanation how it works -

As I understand it (and not that well) there's some things about that appliance model that I wouldn't want if the target is a ground loop. The capacitor for example, isn't desirable so a resistor sits there instead. (You could still put a capacitor in parallel for dual purpose but, it isn't needed for the ground loop and is less desirable if audio is the thing we want to protect.)

The diodes only need to be rated for 10A because of the twin path. It's highly unlikely they could both fail open before the line breaker reacted to the condition. (You also can verify this in actual testing.) The goal is that it fit within the existing enclosure and meet the requirements that the enclosure is typically deployed for, which is 120V 10A nominal, 15A max in dry locations only.

The HumX for example, likely uses 10A diodes and the 6A rating provides more than enough safety margin to make it commercially insurable. (We can certainly anticipate average DJs everywhere plugging their whole rack into this thing including their amplifier. The rating has to account for liability and the design for reality.) I'd love to hear from DJs who fried a HumX by overloading it - and how they were using it. It's also possible to fry a HumX and not know it, save for the fact that it no longer prevents ground loops.

For computers, the dominant hum issue is EMI and there's bound to be a capacitor involved Ground issues are easily isolated on the audio line by a transformer. In pro audio you'll never see a PC connected to a console without first putting it through a PCDI or USBDI. DJ controllers which route audio via USB are doing this internally.

Keep in mind - this filter is at a breakout position (power strip) and not on a feeder line. Amplifiers are not something I would connect to a power strip or quad box with limit protection. Retail strips are all circuit protected for 15A (the max sustainable long term draw being as low as 10A.) My rule of thumb is always 10A per leg for any setup that is not fed directly from a qualified distro. We don't always have complete access to a system so there has to be headroom to account for lower voltage, 15A rather than 20A breakers, or other unexpected draws on a circuit we can't identify.
 
I used to get hum when I'd plug powered speakers into the closest outlets (and probably on separate circuits - especially in bars). Since I don't have things spread out, nor do I need more than 15A these days, I've gone to pulling AC from a single circuit (and most of the time from the same outlet), and I feed the speakers with combo AC/audio cables (mine are CBI).

I have been building some Powercon equipped AC extensions cords, though I bought the boxes (1 from Blizzard and the other an ADJ Pow-R Bar65, though the ADJ doesn't pass-thru).

IMG_2546.JPG

I have thought of getting some project boxes and building some more .. might be useful to drop in a circuit like you did. These boxes should have enough space for the circuit as well as a small breaker.

1589126915510.png
 
I looked for testing options .. didn't find anything of note. This video (may be where you got the info from because the circuit is identical) at least shows electrical testing ..

I think I used that guys video for how he built it, but I eliminated the nails because they're not necessary.

The schematic is very common. My source for making tech determinations was this article:

Secrets of Hum Eliminator

It's a pretty small part in the end:
GLI_block.JPG

I'm only adding this to strips or quads with metal boxes. If the resistor were to completely burn out I would want any resulting fire condition on the insulating jacket to be contained within a metal enclosure. They're also marked so I can easily identify which boxes have it and which ones don't.
 
Last edited:
  • Like
Reactions: steve149
I think I used that guys video for how he built it, but I eliminated the nails because their not necessary.

The schematic is very common. My source for making tech determinations was this article:

Secrets of Hum Eliminator

It's a pretty small part in the end:
View attachment 51030

I'm only adding this to strips or quads with metal boxes. If the resistor were to completely burn out I would want any resulting fire condition on the insulating jacket to be contained within a metal enclosure. They're also marked so I can easily identify which boxes have it and which ones don't.
The ProCraft project box above is stamped steel, as is the ADJ box. The Blizzard is a rubber encased quad. Also might use metal single or 2-gang handy boxes to make a stringer.
 
I have thought of getting some project boxes and building some more .. might be useful to drop in a circuit like you did. These boxes should have enough space for the circuit as well as a small breaker.

View attachment 51029

Radio Shack used to sell 4 and 6 outlet power strips that came in lightweight aluminum enclosures just like the one in your picture. I have a few of them and they've been through a few evolutionary cycles.

I formerly refitted the 6 outlet versions with 19 pin Socapex tails to use as 6 circuit breakouts for lighting. After DMX eliminated the need for multi-channel electrics I returned them back to duty as power strips. Yes, they are the ideal candidate for this kind of project.

I wouldn't spend a lot of money using power cons on these boxes unless you have a constant need to adapt the cable lengths, or fly stuff. The issue that comes up with power cons is that we now have a proprietary plug on our AC supply and are no longer compatible with ready available alternatives. It just adds to the list of things you have to carry for backup. Using a shorter Edison tail allows various cable runs while also preserving easy access to alternatives in a pinch.

IMHO PowerCons are better suited where you need dead front connectors or the absolute surety of a twist lock connection (like speaker arrays and anything that gets flown.) In DJ applications I prefer a small piece of gaffer tape instead. The cord connection stays tight , dry, and discreet. I've sometimes rescued DJs who as a result of trying to make their rigs more "professional" actually painted themselves into a corner with respect to solving otherwise simple issues.
 
Last edited:
The ProCraft project box above is stamped steel, as is the ADJ box. The Blizzard is a rubber encased quad. Also might use metal single or 2-gang handy boxes to make a stringer.

I like the rubber encased quads, and they're required outdoors and in most touring applications but, at that standard buying them pre-made is a better route. You can include a lamp indicator and other cool features including power-cons, and pass thru, or dual phases. Given the cost I rent quads in that class when needed rather than try to own them. My personal clients are typically hotel ballroom bound and not in arenas.
 
I like the rubber encased quads, and they're required outdoors and in most touring applications but, at that standard buying them pre-made is a better route. You can include a lamp indicator and other cool features including power-cons, and pass thru, or dual phases. Given the cost I rent quads in that class when needed rather than try to own them. My personal clients are typically hotel ballroom bound and not in arenas.
The Blizzard Quad has a LED light and Powercon in/out .. the downside is they went way up in price for a simple box. I bought mine maybe 18 months ago and it was $68 .. now they are $100+ .. just not worth it. These guys make the boxes and at around $39, might be more feasible to make a bunch ..
 
Outdoor Tents is where the I'd like to have those. But, this is where your Power cons come in, because the GFI exclusion for outdoor theatrical stages presumes your hot connections are dead front twist locks, and any 125V Edison outlets are inaccessible to the public and/or in lock boxes. Ideally - your end box would be Power-Con input/feed-thru and the outlets dead front twist locks to Edison tails at the actual equipment. When you add up the cost - rental looks really good. :)
 
I bought mine maybe 18 months ago and it was $68 .. now they are $100+ .. just not worth it.

Sounds like they've discovered the Government issued Mastercard. LOL. It's easy to make those price increases once you start supplying government agencies - everything from FEMA to the military.

The pandemic has really hiked prices on all the gadgets that can be used for video-conference/chats, etc. Even little things like the Elgato Cam Link went from $120 to over $400. Good luck finding the Logitech 920 Webcam in stock inside of a month or more. It was easy to take some of these nicely performing jackknifes for granted before the whole world suddenly wanted them.
 
  • Like
Reactions: steve149
I'm still waiting to be able to test with real life ground loops that turn up on the job. What concerns me is that if the NEC allows up to 3V of Ground differential in a facility - then what is the actual voltage I'm experiencing when an AC loop turns up in these large ballrooms? by comparison a GFCI requires only like 5.0mA (0.005V) to activate, and my diodes are probably blocking only up to 0.5V or 0.7V at best.

Has anyone ever pinned down an offending loop and measured the source value?
 
Keep in mind that such circuits are a violation of electrical code. You simply cannot place circuitry in the path of the safety ground. This is why HumX does not and will never have a UL rating despite EbTech saying for years that such a rating is coming.

Another thing to consider is fault currents. Momentary fault currents can be huge (100+ amps) and diodes will blow faster than circuit breakers. In other words, should you have a catastrophic fault, you will be left without any protection whatsoever.

The presentation below was given by one of the world's foremost experts in audio grounding, Bill Whitlock. Bill is an AES and IEEE fellow and a good friend of mine. We have even co-presented on a few occasions. The pertinent information appears around slides 25, 125 and the HumX is specifically mentioned on slide 209. This presentation is probably more than you want to know (it was a half day workshop if I recall) but this covers every possible hum issue.
 
Keep in mind that such circuits are a violation of electrical code. You simply cannot place circuitry in the path of the safety ground. This is why HumX does not and will never have a UL rating despite EbTech saying for years that such a rating is coming.

Another thing to consider is fault currents. Momentary fault currents can be huge (100+ amps) and diodes will blow faster than circuit breakers. In other words, should you have a catastrophic fault, you will be left without any protection whatsoever.

The presentation below was given by one of the world's foremost experts in audio grounding, Bill Whitlock. Bill is an AES and IEEE fellow and a good friend of mine. We have even co-presented on a few occasions. The pertinent information appears around slides 25, 125 and the HumX is specifically mentioned on slide 209. This presentation is probably more than you want to know (it was a half day workshop if I recall) but this covers every possible hum issue.

Thanks, good information in that article, though a lot of it is better applied to premises NEC (509). Care has to be taken to know when NEC (520) or NEC (525) are the applicable standards. This can also be a lesson in HOW to apply standards so that they do in fact, promote safe practice rather than an ideal for the ideal's own sake.

While there's nothing actually wrong in that presentation - it would be wrong to apply it as you suggest. Let's start with the HumX and consider what it is and what it isn't. Whether or not the HumX is UL listed is of no real impart. It's 6 amp capacity (that's 6 amps on the ground leg, the plug itself is a 15A device) makes it a novel appliance not electrical magic. It's rating is consistent with the typical (retail) DJ gear or casual musician.

The HumX addresses an issue that absolute standards don't - which is the intended USER of this device. In the absence of the HumX what is that user most likely to do? What have they traditionally done? In any meaningful ground fault condition the HumX rectifiers fail to short providing a level of safety to this type of user in conditions where habitually safety was discarded..

UL Listing is an underwriting endorsement, not an electrical code. The device is legal and likely well built within it's 15A design. When you consider what it does for the kind of users I just described - I think it's a win. For all we know it may have already saved someone's life. At the same time - there's likely frustration from user's who have placed it in persistent overload conditions that resulted in some embarrassing moments. HumX doesn't have a corner on that market - excess is often applied to all manner of setups.

Issues with age of the information are relevant when presenting "worst case" scenarios or hand picking numbers without any context. How do you suggest a 100A fault to ground persisting for 2.5s on a modern branch circuit to occur? That's neither momentary nor likely and If that's what surges down my 15A Edison cable a smoking HumX will be least of my concerns. We're not dealing with mains, fuses, feeders, or even a toaster. At 5 to 7 times the nominal current a modern branch breaker is going to win the race with a HumX every time.

Here's the thing - large modern diodes rarely fail open because they are made with solid thru conductors. If they are rated consistent with the stranded wire devices into which they are installed then they will pass exceptional faults to the breaker with time to spare.

So, when DOES the HumX fail and what does that look and sound like? It will fail in a severe fault condition that loads the ground with more than 6A. We're talking however, about the permanent failure of it's noise cancelling "lift" not a failure to pass fault to a GFCI or impede current overload protection at the breaker.

The plug assembly itself is consistent with any other 15A Edison so, if the fault was non destructive - the DJ resets the breaker or GFCI and continues on perhaps unaware that the buzz from any previously blocked ground loop has now returned. A week later he plugs in his HumX to resolve a ground loop but the buzzing doesn't go away. Yanking it from the wall he exclaims "Piece of crap doesn't even work!" ...and promptly cuts the ground lug off his extension cord. :)

My goal, is to build a bridge that is on par with the current carrying capacity of the existing internal wiring but with a threshold capable of blocking minor ground differentials from looping through the system. By building the bridge into a specific end point box I'm eliminating the ambiguities around application and rating that make people uncomfortable with the HumX. This ends up positioned appropriately at the appliance end rather than the head end of any AC source lines.

As I said, I'll be doing some testing, and if anyone has done such tests, or experienced failures of such a device, please chime in. One test I will NOT be doing however, is the application of 100A spike to a 15A cable and breakout box. It's just not realistic.
 
Last edited:
I would only consider a HumX type device on a GFCI circuit. I do carry short GFCI equipped extension cords with me (mostly for outdoor gigs but also for bad premise wiring).

Let's assume a failure mode where the line wire came loose inside a DJ mixer and touched the case. If the case is properly grounded, the breaker should trip after so many milliseconds (unless it is an FPE or Zinsco :monooh:). However, even on the better brands there is a fairly high degree of variability in trip time. This is why we have to worry about the diode failure. Diodes fail to open circuits in over-current situations and this would be an over-current situation. They fail to short circuits in over-voltage situations (aka punch through) which is common in motor and dimmer circuits where there are inductive spikes. If the breaker doesn't trip fast enough you could open the diode and be left with the mixer case at line potential.
 
BTW, if the HumX diodes fail due to over-current, the resulting circuit would be no different than cutting off a ground lug. In other words, there would be no hum (as well as no safety). Your only indication of failure may the release of smoke and odor. GIven that it encased in plastic, it might melt as do many plastic surge suppressors when their MOVs pop.
 
  • Like
Reactions: ittigger