As I said earlier ... connect a 1K0 resistor between a good ground and the outside metallic parts. Measure that with your meter on ACV. This is the standard way to check leakage current, check service manuals and this is what you will see.
Although I would cetainly agree with you that anything conductive should be earth bonded, it doesn't necessarily breach safety legislation if it isn't. It's perfectly possible to make a double insulated piece of equipment having exposed metalwork. The top cover of a VHS machine (remember them?) springs to mind. Such machines used a two-pin mains connector and bonded their internal 0V to the junction of a pair of 4n7 capacitors between L and N, so the metalwork floats at half mains voltage. In the UK, that puts it at 120V from a Thevenin source capacitance of 2 x 4n7, or 339k at 50Hz, limiting short-circuit current to 0.35mA, which is below the generally accepted perceptible limit of 1mA. But that 1mA is an average. Years ago, I investigated an incident at work exactly like this where three people were known to have touched earth and the floating conductor. Only one noticed the shock. I first noticed it when I plugged a phono lead from a VHS machine to an earthed amplifier.
There's a wonderful Midsomer Murders episode where Suzi Quatro is deliberately electrocuted via her microphone.
Measure capacitance between the metalwork and L.
Ok I did that and measured 83v ac on the microphone and metallic parts of the PA... seems high, right ?
I tested at home with my lips on the microphone, I could feel a very light electric sensation (no shock) when touching metallic parts of my guitar. The sensation stopped as soon as I removed my hand from the guitar ( makes sense).
I tried the same thing with a wire connecting the chassis from the front panel to the middle metal plate (itself connected to the back panel chassis and to ground - see before) - electric sensation gone.
Also, test with 1k resistor with that wire installed: no voltage anymore.
I think I will just keep my modification: I don't get shocks anymore and I don't hear any sound problem.
To, me the fact that
- There are 2 screws on the middle plate that look like they are supposed to have ground wires connected to them (paint removed around them),
- The only ground wire (coming from the back chassis) is not even connected to one of these screws (see picture above),
strongly suggests that there was a mistake when this was built.
It seems that both the front and the back panel are supposed to be connected to these 2 screws.
Anyway, I really see no way how my modification coud bring a safety issue, and it seems to solve everything.
Now, even if my solution fixes the problem, Is this normal to measure 83V AC like I did ?
Was your 1k measurement with the 1k resistor connected between earth and the metallic parts and the meter across the 1k? I suspect your 1k resistor was actually in series with your meter because 83V across a 1k resistor would dissipate almost 7W and get very hot. In theory, you should have seen half mains voltage, but the 4n7 capacitors in a mains filter are likely to be 10% tolerance (or more), so 83V is entirely possible. And if it's caused by a mains filter as I've suggested, entirely legitimate.
I do not advise testing for shock risk on a piece of equipment that is suspected of being unsafe with your lips. 🙁
Moving on, it is possible that there was an assembly mistake. In 1990, I did electrical safety tests on two desktop PCs we had bought and they failed the earth bond test. On investigation, I found that the failure was because the serrated washer was above the earth tag on the earth terminal, not between the tag and the chassis so that it would cut into both surfaces and form a gas-tight connection. My findings caused a lot of fuss because our company had ordered a lot of these computers and demanded that the manufacturer put them right. Even today, earth bonding is not always done properly, especially where anodised aluminium surfaces are involved.
I do not advise testing for shock risk on a piece of equipment that is suspected of being unsafe with your lips. 🙁
Moving on, it is possible that there was an assembly mistake. In 1990, I did electrical safety tests on two desktop PCs we had bought and they failed the earth bond test. On investigation, I found that the failure was because the serrated washer was above the earth tag on the earth terminal, not between the tag and the chassis so that it would cut into both surfaces and form a gas-tight connection. My findings caused a lot of fuss because our company had ordered a lot of these computers and demanded that the manufacturer put them right. Even today, earth bonding is not always done properly, especially where anodised aluminium surfaces are involved.
Yes indeed ! How should I do it ? I have to admit i don’t know why a resistor is needed… can you please explain ? Sorry if it’s a very basic question 😬.
For lip testing the thing, you’re right, I wouldn’t normally do it but we’ve been having this issue for a while so I figured the risk was limited…
At least the issue seems to be solved.
83 VAC across a 1K0 resistor would have smoked it for sure, so you must have been in series with your meter.
83 VAC across a 1K0 resistor would have smoked it for sure, so you must have been in series with your meter.
No, it's not a basic question. A standard DMM has a very high input resistance (10M) in order not to load circuits when measuring. The purpose of the 1k resistor across the meter was to make any voltage caused by stray components (such as nearby wiring, or mains filter) collapse, allowing you to distinguish between a genuine hazard and none.
I suspect a full explanation would require you to have more knowledge than you presently have. I suggest you find a basic electronics book in your local library (if it hasn't been shut down) and read about current, voltage, Ohm's law, reactance, impedance, and potential dividers. If it's the first time you've met this sort of stuff, that will take you some time to fully understand. When you understand that little lot, you could then have a look at a book about electrical house wiring and safety. And if you've a strong stomach, you could read about the effects of electric shock and where the specifications for earth leakage circuit breakers, earth bonding, and fuses came from. It's all grisly stuff and proper engineers are frightened of electricity. Oh yes.
83 volts is very, very dangerous and would likely feel like much more than a small shock. I've seen sparks from 120 volts........
It's good that it's working and seems to have fixed the problem. Is there no longer any voltage between the microphone and guitar pickups?
As I said, I would be surprised if they manufactured it wrong, butter certainly does seem like they might have, which is pretty wild.
I've been working in live sound across the world for 20 years and one of my main jobs is to make sure that the power is safe every day and I've never seen anything as bad as 83 volts. I have seen 60 volts before (from a bad bucket transformer in India). I have seen many different problems in power distribution and it's usually from a a tube amp where someone has lifted the ground on the power cable or the generator doesn't have a bond between ground and neutral. I have also seen a lot of improperly made power cables. I've never encountered a problem where it's the mixer being manufactured wrong but there's always a first for everything. Make sure that your mod has given you no voltage between the microphone and guitar though. These things are dangerous and we take so much for granted on a daily basis with power.
Ok I definitely did the measuring wrong... I did it in series as you said... Sorry, I'm learning ! (and will continue with books as you suggested!). I measured again. If I understood correctly: water pipe -> 1k resistor -> metallic part (chassis) on the PA. Each probe of the multimeter on each side of the resistance. Here I measured 0.03 V AC. Did I do something wrong?
Something else I tried: measuring DC voltage between guitar and PA chassis: I see around 50mV right when I connect the two, and then it drops to almost 0 quite quickly. With the wire I installed that connects to front chassis to ground, it is immediately around 0.
Thank you so much to you all for your patience and your explanations, you're really helping me a great deal !
Something else I tried: measuring DC voltage between guitar and PA chassis: I see around 50mV right when I connect the two, and then it drops to almost 0 quite quickly. With the wire I installed that connects to front chassis to ground, it is immediately around 0.
Thank you so much to you all for your patience and your explanations, you're really helping me a great deal !
That sounds reasonable, and would correspond to 3nA if we were to believe the meter. I say that because only a quite expensive meter will measure AC accurately below 1V. But what it does tell you is that there isn't a hazard.
The direct voltage you saw was a small capacitance discharging into your meter. Again, nothing to worry about.
The direct voltage you saw was a small capacitance discharging into your meter. Again, nothing to worry about.
So I tested everything yesterday at my rehearsal space, and everything is fine. I tested all the plugs with a small tester and they all have a ground, all the other devices (amps) are also connected to ground. No more shocks as well. As a bonus, my guitar amp was previously picking weird sounds (radio ?), and it doesn't anymore ! I don't know if it's related, but it's all good 🙂 Thanks again to all 🙂
My friend had a tube amp on loan for his guitar, was about to buy it, and got shock from vocal mic. Turned out the power cord was tight, under the amp, tight enough safety ground didn't make proper contact. It was fixed right there just by lifting the amp up temporarily and rerouting the wires. And more permanent fix by improving on the tension protection inside with more slack. Luckily no harm, but this is serious stuff, keep safe! Always have safety grounds working.
Last edited:
FWIW I regularly measure 90V ac or higher from either + or GND to PE (in floating situations) because of leakage current translating to voltage. The high voltage won’t smoke resistors as the current is very limited. A relatively high ohmic connection of GND to PE and it will be 0V both at + and GND. Floating GND is OK but less so when stuff can be plugged in or out. Of course PE to metal casings is a must.
Things do get problematic when plugging stuff in and out (energized or not) that does have PE connected and GND referenced to PE. Many ICs are allergic to such voltages. This is a reason why HDMI interfaces sometimes fail.
Here one of the most occurring errors is to use power strips that have PE to wall sockets that don’t have PE. Then many devices will have half the mains voltage on the casing (I invite to try to understand why this is!). For some reason this is apparently hard to understand even for electricians.
Things do get problematic when plugging stuff in and out (energized or not) that does have PE connected and GND referenced to PE. Many ICs are allergic to such voltages. This is a reason why HDMI interfaces sometimes fail.
Here one of the most occurring errors is to use power strips that have PE to wall sockets that don’t have PE. Then many devices will have half the mains voltage on the casing (I invite to try to understand why this is!). For some reason this is apparently hard to understand even for electricians.
Last edited:
We are talking about decades old Schuko sockets without PE and the average consumer plugging today's stuff in. Wall sockets with PE were introduced somewhere in the seventies and stuff was made that good! Kitchen sinks, water piping, shower and bath tubs were connected to PE way before that. So in very old houses one has PE but not a yellow/green wire in the vicinity of wall sockets. Odd. Worse are wall sockets without PE but a PE wire rolled in the wall box.
There are many things illegal in the UK except some things that should be illegal 🙂
There are many things illegal in the UK except some things that should be illegal 🙂
Last edited:
I have seen yellow/green wires in France going from wall sockets towards consumer unit, then coiled neatly and not connected to anything. As you say, that's worse than no yellow/green wire at all!
There are doubtless some things here that should be illegal, but aren't. But that's going to be true anywhere, and if we're not careful we'll stray into politics.
There are doubtless some things here that should be illegal, but aren't. But that's going to be true anywhere, and if we're not careful we'll stray into politics.