"Elevated ground" Good idea?

[Diabolical Artificer]

What are the benefits if any of "elevating" the ground from chassis with a low value resistor on an amp? I've read it reduces hum/noise but also that it's a daft idea.

Andy.

[mark2collection]

I've seen this in the past.

100 ohm 5 watt resistor, in series with 2x 1N4007's (wired top-to-tail) from signal 0V to chassis (mains) earth.

I also read somewhere this was indeed to reduce hum/earth loops when using indifferent Classes of equipment, true I suppose with those nasty wallwart SMPSU's where the negative lead is usually at half-mains with respect to real mains earth.

This also being a good reason to make/break any equipment connections with everything switched off at the mains socket.

As to being a daft idea, I believe (though could be wrong) so long as any metalwork which the operator can make contact with under normal operating conditions, is earthed via the mains lead, it's ok, but, the signal ground lift is more of a design flaw 'workaround'.

A DAB tuner once had a tantrum when I removed my 'scope crocodile clip (true ground) from the DAB chassis, which suddenly went from 0V to half mains!

Mark

[thermionic]

Hi Andy.

Primarily used to help prevent earth loops when different items of equipment are connected together signal wise, and also have mains supplies taken from different sockets and circuits.

‘Lifting’ the equipment ground from the chassis through a low value resistor, helps prevent earth loop circulating currents which can cause hum, noise & distortion.

Usually the mains earth is connected straight to the equipment chassis, then the ‘ground’ of the amp circuitry is connected to chassis earth through a low value resistor of say,10 to 47 Ohms. Sometimes this resistor will be bypassed by a capacitor as well.

Cheers. SimonT

[Radio Wrangler]

The full solution is to let safety earths handle the safety, and then handle signals properly. This is what balanced connections are for. Do not have earth connections as part of your signal flows. They have different priorities.

But everyone has single-ended signals, phono plugs etc.

Doesn't make them right!

David

[mark2collection]

The supervisor where I used to work was playing with a fairly big (2kW's RMS per channel) power amplifier & for what ever reason, decided to move the rear-mounted slide switch from 'Grounded' to 'Lift' whilst the system was running, unbalanced.

The resultant (very loud) 'screech' from the system suggested he may have managed to receive all the available local radio stations, at the same time!

We learnt an awful lot through his mistakes.

But yes, ideally balanced is the way to go, luckily in my escape room, the serious system is all balanced. The house 'HiFi'? Modern phono-leads as interconnect, as that's all it has.

Mark

[Diabolical Artificer]

Thanks chaps. I have come across earth lift in pro gear to stop earth loops on stage or in a studio. I'm more thinking of it's application in a straight hifi amp.

BTW, how does this work? Is the resistor there to reduce current through the chassis or simply to lift ground above chassis earth? Need do do some reading up on this subject.

Andy.

[GrimJosef]

The problem with an earth loop is that it will be threaded by all the magnetic flux noise that exists in the modern environment. A lot of that will be at 50Hz, coming from the nearby mains wiring. When magnetic flux threads through a loop it induces a voltage around that loop and if the loop impedance is low that voltage will drive a current around the loop. Some bits of the loop will have higher impedance than others - there might be excess resistance at a plug-socket contact for example - and the flowing current will generate a voltage across that resistance. Basically the total voltage around the loop will be shared between the different impedances with the highest impedance points getting the largest share of the voltage. If the flux is at 50Hz then the current and voltages will be too. A 50Hz voltage at a plug-socket contact will add in to the music signal which is also flowing through the plug and socket and will appear as 50Hz hum on the audio. That's how it works.

The ways to reduce it include i) making the loop area as small as possible so it is threaded by as little flux as possible, ii) excluding flux from places we don't want it to get to by using shielding, iii) adding a relatively large impedance to the loop in a place where the audio signal isn't flowing - this large impedance will drop almost all the loop voltage leaving very little to get into the audio signal. An earth lifter is an example of iii). If you just use a resistor then the larger the resistance is, the more effectively it works. The problem is that if you use too large a resistance then the fault current which flows if the signal ground accidentally rises to dangerous voltage isn't sufficient to blow the safety fuse. It might well, however, blow the resistor instead and now the signal ground is disconnected completely from the safety ground. As stage musicians know only too well disconnecting the safety ground can kill people. It has.

A pair of diodes wired in anti-parallel work better because they have an extremely high impedance at low voltages - up to about 0.7V per junction - but can have very low impedances at voltages above that. So they drop all the (low voltage) earth loop noise but still allow the safety ground to protect against dangerous (high voltage) faults. It's important to use diodes which are beefy enough to survive longer, when the fault current is flowing through them, than the fuse.

Cheers,

GJ

[lesmw0sec]

Originally Posted by Radio Wrangler

The full solution is to let safety earths handle the safety, and then handle signals properly. This is what balanced connections are for. Do not have earth connections as part of your signal flows. They have different priorities.

But everyone has single-ended signals, phono plugs etc.

Doesn't make them right!

David

Entirely agree. I'd just add that the common-mode rejection of an unbalanced line can also be improved with a differential amp. or transformer coupling.
In my audio days we used to construct mixers with two 'earths' technical(signal) earth and safety earth for the metalwork. The two were joined at one point only. There is nothing worse than a virtual-earth summing node for sniffing out a bit of stray earth signal!
Les.

[David Roberts]

To illustrate GrimJosef's excellent post, I am attaching a diagram I recently found on the web whilst researching this topic. It incorporates a bridge rectifier package and recommends a minimum of 35A capability.

[llama]

I've been toying with this proposition for a few months. Video is the problem.
My ceiling-mounted projector is earthed in the loft, the signals feeding it are earthed in the lounge. I used to use a video sender for connection but avoiding wifi frequencies became too much of a problem so I made a direct connection. Result - hum bars.
I removed the mains earth from the projector thus curing the hum bars but creating a safety hazard now.
I'm now thinking I should use the diode idea (probably a diode bridge with a diode across the middle) to reconnect the mains earth. 3 diode drops in either direction using 5 diodes.
Graham

[PJL]

An earth lift resistor was used in the Beam Echo Step 11 phono amplifier but was for a slightly different purpose. The power for the amplifier was taken from the pre-amp and the resistor prevented the PSU current running through the coax shielding.

[Diabolical Artificer]

Thanks all - GJ, grand explanation. David, a 35A BR seems over the top and not cheap.

Andy.

[Radio Wrangler]

You need that large rectifier to be sure that in the case of a fault, it can handle sufficient surge current to pop fuses fast enough.

You are going to trust your life to this part.

I'd also test whatever I bought at full current just to cover the risk that a big case hides a tiny fake. It ought to be unlikely that this part will need to do its job, but if it does, you become critically dependent on it working.

David

[GrimJosef]

I'm with David when it comes to the safety aspects. A 13A fuse will take anything from less than second to more than 100s to blow at 35A. 350A should see it off inside 10ms https://www.pat-testing-training.net...cteristics.php.

I use these https://uk.farnell.com/vishay/pb3510...-pb/dp/1815637 which are rated for 35A continuous and 350A for an 8ms pulse. Since you're paralleling the diodes you should do even better than this. They're £4 each including the VAT or £3 each if you can think of a reason for buying 10.

Cheers,

GJ

[Synchrodyne]

I can see the utility of using a diode array to provide an “elevated” functional earth connection between internal circuitry and the chassis, case or other part which is directly connected to the supply protective earth, thereby minimizing hum loop problems. But would it be appropriate to use a diode array to connect (to the supply earth) a part (chassis, case, etc.) that requires a protective earth for safe operation? It seems highly desirable that protective earths should have “hard” connections to the equipment that they protect.

In the video case, rather than interfering with the protective earthing at either end, a video isolating transformer might be an alternative solution. Such are available, e.g. see: http://www.nhsignal.com/pdf/appnote-wb-151.pdf. I have a vague notion that back in the days when multiroom systems were fairly new, there were available line amplifiers (for both audio and video) that provided ground isolation, I think by using opamp buffers with differential inputs and single-ended outputs, but right now I can’t put my hands on any information about them that I might have been retained.


Cheers,

[GrimJosef]

Originally Posted by Synchrodyne

... would it be appropriate to use a diode array to connect (to the supply earth) a part (chassis, case, etc.) that requires a protective earth for safe operation? It seems highly desirable that protective earths should have “hard” connections to the equipment that they protect ...

This is a very good question.

In one sense it could be argued that either the diode link is capable of providing the protection or it isn't. If it isn't then it's not safe to use it anywhere. If it is then it doesn't matter whether it's protecting some of the metalwork (in an audio system typically the input and output grounds) or all of it.

But in the real world of course there are no absolutes. The diode link might be 99+% safe, in that the fuse almost always blows first. But there will be the odd duff component. And there is always the chance that the installer will have made some error in assembling and fitting it - it's more complex than just a tag bolted and star-washered to the chassis. It might even have got damaged somehow. In any of these cases we'd have been better off if as little as possible of the metalwork earthing depended on the diodes.

Significant problems can occur if a ground-lifter is retro-fitted to provide at least some protection to a previously unprotected device. Yes, I'm aware that equipment shouldn't be manufactured without being safety-earthed as appropriate. But the hard fact is that it is. I've got a customer's item in the workshop right now which has an all metal chassis and control knobs and arrived with no earth at all. It dates from the early 2000's and was manufactured by a small company in Japan. There is no realistic prospect of getting it brought up to standard by the manufacturer. So if the owner can't find someone who will make it acceptably safe then he will either have to scrap it or carry on running it un-earthed. I fear I know what he would do. So I'm fitting a ground lifter having explained the issues as clearly and thoroughly as I can. The problems are a) that the equipment has more than one connection between the signal grounds and the chassis and b) that even if it didn't have there are components, e.g. metal cased volume control pots, where there is capacitative coupling between the chassis and the audio signal wiring. These facts mean that the best chance of keeping the hum low is to leave the chassis and signal wiring coupled as they always have been and to lift all the grounds, including the chassis. I'd prefer not to. But to paraphrase the Rolling Stones, we can't always have what we want.

Cheers,

GJ

[Argus25]

In the past I've built opto isolators with a 4MHz+ bandwidth for PAL video and its dead easy for audio with the low bandwidth (but for low level audio there can be S/N issues), to eliminate earth loop issues every time. Also I have used a differential amp at the inputs as an option, but these are not as good or as safe as opto-isolation especially for high voltages differences, say with cables between two buildings. I have never had to, or suggest that a chassis earth should be lifted. The safest method is opto isolation.

The ground lifter diode idea is ok, but I would not trust the diode junctions 100% to beat the fuse or breaker, even if they had a 100A rating and looked like they escaped from the land of the Giants.

[Radio Wrangler]

For the above reasons, I always provide solid ground connections which will easily take a 25A PAT test or pull a 13A fuse.

If I have noise problems, I remember that ALL signals are differential whether unbalanced or not. I design amplifiers to sense signal ground from the preamp etc. Let hum loops circulate and buzz, they don't get to my speakers.

I spent years designing lab equipment where everything had to have solid metallic ground paths. It was not my decision, it was required. THEN I had to handle and move signals, sometimes right down to the thermal noise floor. You know what? It worked. Lifting grounds gives you an advantage in keeping signals clean, but it doesn't go all the way. Start by handling signals carefully and you don't need to touch the grounds. Also you can now handle complex installations with potential for multiple and unavoidable ground loops.

David

[GrimJosef]

Yes, good design and implementation of that with due regard to the regulations (they're there for a reason !) is absolutely the correct thing to do. If we start from scratch then there shouldn't be any need for an earth lifter. The problems that land on my bench are caused by someone else having started out with a poor design and fixed the consequences by playing fast and loose with the regs. If I want to make the best of this bad job then an earth lifter can be the least of all the available evils.

Cheers,

GJ

[Argus25]

I was quite surprised to find this (see attached circuit fragment) on the input of a video monitor. This sort of thing is seldom seen even in professional studio video monitors, but it is obvious they wanted to avoid ground loop issues and its a good method. The monitor typically runs from 400Hz 3 phase power in a plane and would be connected to other equipment, possibly on an alternate power system.