PCB ground planes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I don't want to enter the telescoping versus star grounding strategy argument, but I do take issue with the use of anti-parallel connected diodes to protect against fault currents. It is a fact that, when a ground-fault occurs, the short-circuit current is limited primarily by the resistance in premises wiring. According to studies done by UL (Underwriter's Laboratories) in the US, fault current can range from 150 A to over 1,000 A until the circuit breaker opens - and this can take from 50 ms to over 2.5 seconds. This will quickly turn a 6 A or 8 A rectifier into silicon vapor and an open circuit! There is a product called "Hum-X" that does exactly this. It's been on the market for at least 15 years. Every time I see them at a trade show, I stop and ask where their UL or other agency approvals are. And, for 15 years, the answer has been "Oh, we're working on that." And I tell them not to hold their breath. I would never use this product or use it as a "cure" for grounding issues ... it's a deadly accident waiting to happen. But then audiofools will buy almost anything if the advertising contains enough buzz-words. Many have been told that only one piece of gear in your system has to actually be connected to safety ground because the rest of the system connects to it through signal cables. Can you imagine passing 100 A through the shield of an ordinary RCA cable? Have a fire extinguisher ready!
 
I don't want to enter the telescoping versus star grounding strategy argument, but I do take issue with the use of anti-parallel connected diodes to protect against fault currents. It is a fact that, when a ground-fault occurs, the short-circuit current is limited primarily by the resistance in premises wiring. According to studies done by UL (Underwriter's Laboratories) in the US, fault current can range from 150 A to over 1,000 A until the circuit breaker opens - and this can take from 50 ms to over 2.5 seconds. This will quickly turn a 6 A or 8 A rectifier into silicon vapor and an open circuit! There is a product called "Hum-X" that does exactly this. It's been on the market for at least 15 years. Every time I see them at a trade show, I stop and ask where their UL or other agency approvals are. And, for 15 years, the answer has been "Oh, we're working on that." And I tell them not to hold their breath. I would never use this product or use it as a "cure" for grounding issues ... it's a deadly accident waiting to happen. But then audiofools will buy almost anything if the advertising contains enough buzz-words. Many have been told that only one piece of gear in your system has to actually be connected to safety ground because the rest of the system connects to it through signal cables. Can you imagine passing 100 A through the shield of an ordinary RCA cable? Have a fire extinguisher ready!
I wrote a letter to them years ago asking about UL approval, they did not answer me either. 🤔

It seems we have discussed the diode trick here before. Heres a post from a decade ago... post mentioning diode ground trick :cool:

JR
 
Your point is well taken. How many times have you dealt with a piece of gear with PT short to chassis, hi side?

The diode trick was done by Acoustat in there direct drive tube amps for their estatic speakers in the 70s with 1A diodes so it's not new. Hum-x got a patent on it but no one contested when the patent was applied for. I almost did by sending a copy of the Acoustat schematic to the USPTO.

When I tested mine several times I could only blew out the 10A fast blow fuse in the amp, not even a circuit breaker. That's with a dead short. Bigger diodes could be used to survive longer. Also while the diodes could go open diodes tend to short so you are in no worse a situation than a direct connection to the 3rd prong. Or it's running on a 3 to 2pin adapter from the hardware store. In the meantime - no hum.

The alternative is to spend hours on the phone helping the customer get rid of the ground loop. It's not mandated that a hifi amp needs a 3 wire plug.

God knows there are thousands of ungrounded Amp B15s, Fenders, KithenAid mixers, Old tube and transistor radios. I have a Hallicrafters S38 that is the All American 5 with short wave. One side of the line connect to the chassis. If I were to use it I would put an iso xfmer in it but that was common in the 30-50s. All my original amps used a 2 wire plug. Never had a problem.

In my 40 years of manufacturing I've never had any PT primary short or catch fire. Could it happen? Not sure. Could you guarantee everything you've designed not go haywire? No one can.
 
Your point is well taken. How many times have you dealt with a piece of gear with PT short to chassis, hi side?
That is not common these days thanks to rigorous safety standards. Primary wiring enjoys wide spacing standards, and suitable insulation.

Old school 2 wire line cord guitar amps used to have a polarity switch to alternately connect the chassis to one polarity of the mains or the other. Connecting the chassis to the 0V neutral provided lower hum levels. This was obviously dangerous as in the wrong switch position the chassis was connected to line, through the infamous "stinger caps". :rolleyes:
The diode trick was done by Acoustat in there direct drive tube amps for their estatic speakers in the 70s with 1A diodes so it's not new. Hum-x got a patent on it but no one contested when the patent was applied for. I almost did by sending a copy of the Acoustat schematic to the USPTO.
You still can. I came up with that diode approach independently back in the 70s mainly to deal with conducted ground current between rack mount chassis. I stopped using this approach after learning safer ways to deal with input signal integrity.
When I tested mine several times I could only blew out the 10A fast blow fuse in the amp, not even a circuit breaker. That's with a dead short. Bigger diodes could be used to survive longer. Also while the diodes could go open diodes tend to short so you are in no worse a situation than a direct connection to the 3rd prong. Or it's running on a 3 to 2pin adapter from the hardware store. In the meantime - no hum.
yes the typical failure mode for overheated diodes is to melt, resulting in a low resistance short circuit, as long as the molten silicon puddle is intact. I did my share of bench testing and too small diodes could explode/evaporate. I ended up using diode rectifier bridges for the extra thermal ruggedness. I do not advocate doing this today.
The alternative is to spend hours on the phone helping the customer get rid of the ground loop. It's not mandated that a hifi amp needs a 3 wire plug.
yup, customers can be a PIA... ;)
God knows there are thousands of ungrounded Amp B15s, Fenders, KithenAid mixers, Old tube and transistor radios. I have a Hallicrafters S38 that is the All American 5 with short wave. One side of the line connect to the chassis. If I were to use it I would put an iso xfmer in it but that was common in the 30-50s. All my original amps used a 2 wire plug. Never had a problem.
Modern safety agency rules would not approve making that connection with an only two wire line cord. Consumer gear using two wire line cords depend on double insulating all primary wiring to insure human safety.
In my 40 years of manufacturing I've never had any PT primary short or catch fire. Could it happen? Not sure. Could you guarantee everything you've designed not go haywire? No one can.
It is the mission of UL and other safety agencies to anticipate component failures and prevent fires or shock hazards.

While working at Peavey we experienced one customer killed by one of our UL approved guitar amps plugged into a mis-wired (reverse polarity bootleg ground) mains outlet. The energized outlet ground lead, energized the guitar amp chassis and player, leading to his death when he came in contact with a real ground path.

It was helpful to have UL defend our guitar amp design for safety in court (the judge condemned the house until the wiring was repaired).

JR
 
That's horrible! But not Peavey's fault. 1 in many million chance. 2 wire pwr cord would have saved him.

As a boutique manufacturer my liability is not as exposed as a company the size of Peavey. I don't have that many units in the field.
My transformers were modern from manufacturers that had regulatory approval. Never had a primary short as far as I knew.

And I dare say my customers are more educated and affluent as to afford proper wiring. Many have dedicated oversized wiring to their listening rooms, but properly
fused. It's a different customer base all together.
The diodes are 6amp with 200 Amp peak rating. The 10k r bleeds off leakage current bringing it down to a few millivolts or less. Of course it's not the only piece of gear in the hi fi. But no ground loop hum.
 
True. But, in UK/EU at least, anything with a "two Wire" L & N only mains input without an "Earth" connection has to be designed to be "Double Insulated" and marked accordingly.
How would you double insulate a standard power transformer and the primary wiring inside the chassis Spaghetti on the wiring I can see but inside the PT? Mine had a electrostatic shield in them. These are toroids rated for about a KW power.

Don't most EU countrys run on 230 v?
 
As a boutique manufacturer my liability is not as exposed as a company the size of Peavey. I don't have that many units in the field.
My transformers were modern from manufacturers that had regulatory approval. Never had a primary short as far as I knew.

Are you saying that your manufactured products fail to meet statutory standards ? Or are they properly designed wrt primary insulation to meet applicable standards.
The affluence and attitude of your customers isn't relevant to this.
 
That's horrible! But not Peavey's fault. 1 in many million chance. 2 wire pwr cord would have saved him.
Ironic.... Over the decades I have killed a lot of brain cells thinking about human safety with regard to mains power. I even developed a prototype using a 3 pole relay to not only disconnect power but disconnect the ground when it detects a hazard.

There is a potential scenario with live sound performance where there can be significant ground potential differences between the front of house power drop, and on stage back line. The stories of singers getting shocked by microphones are common, while not many deaths from that. There is a reason pastors use wireless mics in baptismal pools. 🤔

I am a strong advocate for musicians using GFCI protected outlet strips on stage. While that would not protect a singer from an energized microphone (like RPBG at FOH).
As a boutique manufacturer my liability is not as exposed as a company the size of Peavey. I don't have that many units in the field.
My transformers were modern from manufacturers that had regulatory approval. Never had a primary short as far as I knew.

And I dare say my customers are more educated and affluent as to afford proper wiring. Many have dedicated oversized wiring to their listening rooms, but properly
fused. It's a different customer base all together.
The diodes are 6amp with 200 Amp peak rating. The 10k r bleeds off leakage current bringing it down to a few millivolts or less. Of course it's not the only piece of gear in the hi fi. But no ground loop hum.
I do not encourage using the diode trick, but if doing so, a proper diode bridge will be harder to kill.

JR
 
How would you double insulate a standard power transformer and the primary wiring inside the chassis Spaghetti on the wiring I can see but inside the PT? Mine had a electrostatic shield in them. These are toroids rated for about a KW power.
Double insulation is a pretty mature technology.

I had one oddball transformer insulation issue inside an unconventional digital audio amplifier (Peavey DPC). The way the power amp modulated the power supply, a transformer insulation failure could expose the amp outputs to mains voltage. My problem was passing the agency approval testing for the EU market. Their test involved very elevated temperature for prolonged duration. Checking this final agency approval box prevented me from shipping to clear my increasing back order. When the transformer failed it its first test, it would have taken me months more to test again and maybe win approval to get shipping. Instead I changed the transformer design to use double insulted wire inside the transformer primary (I didn't design the actual transformer, Peavey had people for that, and the switching transformers used different people again). As I recall the wire change cost me a few dollars per unit, but I was willing to eat that cost increase to satisfy my backorder. 🤔
Don't most EU countrys run on 230 v?
yes... give or take...

JR
 
Ironic.... Over the decades I have killed a lot of brain cells thinking about human safety with regard to mains power. I even developed a prototype using a 3 pole relay to not only disconnect power but disconnect the ground when it detects a hazard.

There is a potential scenario with live sound performance where there can be significant ground potential differences between the front of house power drop, and on stage back line. The stories of singers getting shocked by microphones are common, while not many deaths from that. There is a reason pastors use wireless mics in baptismal pools. 🤔

I am a strong advocate for musicians using GFCI protected outlet strips on stage. While that would not protect a singer from an energized microphone (like RPBG at FOH).

I do not encourage using the diode trick, but if doing so, a proper diode bridge will be harder to kill.

JR
Thanks. I appreciate you and your approach.

Singers who got shocked were usually e-bass or e-guitar players using amps with the Fender style ground switch. Or working outdoors in a wet area.

I would think GFCIs would trip with a guitar amp with Fender style ground switch but those amps are 2 wire. It would be when the player kisses the mic running in the PA system.
 
Thanks. I appreciate you and your approach.

Singers who got shocked were usually e-bass or e-guitar players using amps with the Fender style ground switch. Or working outdoors in a wet area.

I would think GFCIs would trip with a guitar amp with Fender style ground switch but those amps are 2 wire. It would be when the player kisses the mic running in the PA system.
The mic generally provides the ground path for leakage from a "stinger" cap. The GFCI (works with two wire line cords too) does not open the ground connection (my rube goldberg design did), so a hot mic could still do harm through the still grounded guitars. Wider use of GFCI for all power drops can save lives.

These lip shock events are rarely deadly, but some musos get stuck to the mic.

JR

edit- if a guitar stinger cap trips a GFCI the typical advice is to rewire the guitar amp to use a 3 wire line cord. Hopefully there are less of these left in the wild./edit
 
Don't most EU countrys run on 230 v?

It's 230V with a tolerance of -6% / +10%.

This means that in UK where nominal standard voltage was previously 240V then nothing really had to change.
fwiw whenever I've had cause to measure the mains voltage at the wall socket it has measured 240V min. Usually a couple of volts above (typically measured with a "True RMS" Fluke DMM).
 
It's 230V with a tolerance of -6% / +10%.

This means that in UK where nominal standard voltage was previously 240V then nothing really had to change.
fwiw whenever I've had cause to measure the mains voltage at the wall socket it has measured 240V min. Usually a couple of volts above (typically measured with a "True RMS" Fluke DMM).
I believe that prior to joining the EU and the wish to harmonise mains voltage across the EU the quoted 'tolerance' HAD been + - 10% but by a bit of wrangling it was changed to plus 10 and minus 6 % on a new nominal 230 Volts where the majority of european countries had aimed with others being 220 and the UK 240.
Having wrestled with a fair few 'American' designed (Mexico assembled) 'industrial' power supply units they do allow ripple when fed from 50Hz mains at 208 Volts when on the 230 Volt setting which is fair enough. I suspect using ithem in USA or Japan would allow slightly greater leeway due to increased frequency. Of course transformer manufacturers design and make power transformers with the intention that will meet (or exceed by a suitable margin) the requirements for international safety approvals because getting it 'right first time' is a lot better than the fallout from faulty product.
As an observation, the number of USA sourced mains transformers that 'buzz' mechanically when powered with 50Hz is rather disappointing.
 
I have seen "bootleg ground" wiring advice for some appliance hook up instructions (my house wiring does not include a dedicated safety ground wire.) Tying a chassis ground to neutral "should" be safe as long as the 0V neutral path is solid. If that neutral connection opens up, the chassis could be energized by leakage from the line connection. o_O

JR
 

Latest posts

Back
Top