I know and have nothing against some more protection, I was just wondering if the usual GFCI used in bathroom were designed to hold a tube amp without tripping needlessly, the industrial one PRR posted seems more suitable, for sure it's a great concept I'm with you on this, please share the result as soon as you did that on your amp.
Yes that's the big fun in DIY, I'm not challenging you about the concept at all, but I wonder if at least one equiped GFCI amplifier has been released on the market. I agree that it should be mandatory, not in amplifiers and gears but in all houses just like they do in Danish and probably other places.
Sometimes the outlets in bathroom are a little isolation transformer.. That's how it is where I live, if I remember they are rated at 20VA.
The OP asked about how to make hot chassis safe, I think it's a good idea to have such tester then make sure to check the outlet wiring is correct, those are warning you when the ground is missing or when the hot neutral is reversed, I always have one of these on my bench, another one inside my guitar case. These are for north america but I'm sure every places have something similar.
Oh, feel free to challenge me, it's something I know very little about. Electronics I'm comfy with. Electrical wiring, especially high voltage, high power stuff - that I know very little about.
And it is an entirely different world, with unimaginably different hazards. For example, a little household dust stirred up by air movement when you open the electrical panel access cover can cause an arc-flash that can burn your face off in a fraction of a second.
I have no electrical certification (though my wife thinks I'm certifiable
), but as far as I know, electrical code in California (where I used to live), and British Columbia (where I live now), both require GFCI outlets if they are within six feet of a sink or bathtub. Here's some info for BC: https://ringelectric.ca/where-are-gfis-required/When I was looking for information on GFCI use recently, I found out they are also required for hot-tub installations: https://www.poolandhottubdepot.com/new-hot-tub-set-up/
I've never seen the bathroom isolation transformer approach you described. Perhaps it is an older system?
I've also never heard of the whole-house system that Rasped mentioned is used in Denmark. I wonder if it is one device for the whole house, or perhaps every circuit-breaker on the main panel is of the GFCI type?
-Gnobuddy
I'm not an electrician, I'm a lineman (power poles), so my $0.02
The gfci monitors the difference between phase and neutral currents in your mains outlet. At a roughly 30mA difference, the gfci opens the phase, which disconnects power from whatever its protecting.
This will protect you during a fault in primary side of your amp. (120v in US, 240v in other parts of the world)
A gfci may not protect you on the secondary side of the transformer (which your guitar is connected to) as a phase to chassis fault on the secondary can produce equal currents on the primary side, and not trip the gfci.
It all depends on if your mains neutral is connected to the chassis or not.
The OP question about making one safe, is easy. Add an isolation transformer, and properly ground the amp. Feeding the whole thing through a gfci will add that last line of defence if something goes wrong with the mains.
At a gig, take a voltmeter and check for potential differences between your guitars and mics. Especially if using multiple power sources and pieces of equipment.
We always assume everything in the building is wired correctly, but know its usually done by the lowest bidder.
Cory
The gfci monitors the difference between phase and neutral currents in your mains outlet. At a roughly 30mA difference, the gfci opens the phase, which disconnects power from whatever its protecting.
This will protect you during a fault in primary side of your amp. (120v in US, 240v in other parts of the world)
A gfci may not protect you on the secondary side of the transformer (which your guitar is connected to) as a phase to chassis fault on the secondary can produce equal currents on the primary side, and not trip the gfci.
It all depends on if your mains neutral is connected to the chassis or not.
The OP question about making one safe, is easy. Add an isolation transformer, and properly ground the amp. Feeding the whole thing through a gfci will add that last line of defence if something goes wrong with the mains.
At a gig, take a voltmeter and check for potential differences between your guitars and mics. Especially if using multiple power sources and pieces of equipment.
We always assume everything in the building is wired correctly, but know its usually done by the lowest bidder.
Cory
Now that is very relevant indeed. Thank you very much!
Not. Otherwise there's the risk of the entire chassis becoming live if the outlet is miswired and live and neutral are cross-connected, no? That was exactly the problem with those horrifying old "hot chassis" guitar amps.
So AFAIK the mains live and neutral are kept completely isolated from the guitar amp chassis, and only the mains earth wire is connected to it. That's how I wire my DIY guitar amps.
Which, if I understood you correctly, means a GFCI might not add any additional safety if a fault occurs on the secondary side of a guitar amps power transformer.
-Gnobuddy
30 mA would be pretty bad. The attached data is from Ohio State University ( Electrical Safety: The Fatal Current ). I highlighted the relevant bit.
-Gnobuddy
-Gnobuddy
Attachments
> I'm stunned that not a single guitar amp manufacturer so far seems to offer a built-in 9V DC power jack for your FX pedals.
One did (does?). Did not catch on. TrueTone?
> called HPFI in Denmark. My entire house is protected by one of those.
And RCB in the UK; and they too sometimes do it whole-house. In more casual US wiring, GFI has grown from per-outlet to per-circuit, but whole-house seems dubious to me. How do you find the problem and reset the breaker if ALL the lights are out? But whatever works.
> I'm not sure how common it is for GFCIs to trip by mistake.
It's a common complaint. Especially when a specific technology (there have been several) is new. Also cascaded GFIs seem glitchy.
> Sometimes the outlets in bathroom are a little isolation transformer.. That's how it is where I live
That is different. Common in the UK, unknown in the US. Dedicated low-voltage supplies for lighting and shaver.
> I think it's a good idea to have such tester then make sure to check the outlet wiring is correct, those are warning you when the ground is missing or when the hot neutral is reversed
Yes; but there are at least two faults these will NOT reveal.
> a little household dust stirred up by air movement when you open the electrical panel access cover can cause an arc-flash that can burn your face off
Never heard that one. Seems unlikely on 120V or 240V lines. Even in quite dusty fuseboxes.
However opening a strange fusebox IS dangerous. The hazards I have seen! Sometimes ready to fall out in your face. If you look on Inspection Horrors sites you can find breakers and even buses and feeders over-stuffed or rusted-through and ready to bite.
> We always assume everything in the building is wired correctly, but know its usually done by the lowest bidder.
Or by slipping a case of beer to some guy who once spent a day as Electrician's go-fer. I'm quite sure much of the framing and wiring in this house was done by beer-buddies.
> electrical code in California (where I used to live), and British Columbia (where I live now), both require GFCI outlets if they are within six feet of a sink or bathtub.
> When I was looking for information on GFCI use recently, I found out they are also required for hot-tub installations
GFI around water has been in the NEC for a long long time, and is pretty universally enforced (in new or major refurb) by local code officers.
Hot tub GFI and a LOT of grounding rules was a hot topic in NEC at the end of the 20th century. There were several revisions and clarifications. If you are wiring a hot tub you want a quite recent copy of NEC. (I have six, from 1919 to 2009.)
> Not sure where the 30mA is from. For a 15A GFCI, it's 5mA
Look at the spec. It is a *Time Curve*. 30mA will trip quickly. 5mA will trip after some time. I strongly suspect that UK-spec "30mA" and US-spec "5mA" are the SAME calibration, but the nominal names taken from different points on the same curve.
One did (does?). Did not catch on. TrueTone?
> called HPFI in Denmark. My entire house is protected by one of those.
And RCB in the UK; and they too sometimes do it whole-house. In more casual US wiring, GFI has grown from per-outlet to per-circuit, but whole-house seems dubious to me. How do you find the problem and reset the breaker if ALL the lights are out? But whatever works.
> I'm not sure how common it is for GFCIs to trip by mistake.
It's a common complaint. Especially when a specific technology (there have been several) is new. Also cascaded GFIs seem glitchy.
> Sometimes the outlets in bathroom are a little isolation transformer.. That's how it is where I live
That is different. Common in the UK, unknown in the US. Dedicated low-voltage supplies for lighting and shaver.
> I think it's a good idea to have such tester then make sure to check the outlet wiring is correct, those are warning you when the ground is missing or when the hot neutral is reversed
Yes; but there are at least two faults these will NOT reveal.
> a little household dust stirred up by air movement when you open the electrical panel access cover can cause an arc-flash that can burn your face off
Never heard that one. Seems unlikely on 120V or 240V lines. Even in quite dusty fuseboxes.
However opening a strange fusebox IS dangerous. The hazards I have seen! Sometimes ready to fall out in your face. If you look on Inspection Horrors sites you can find breakers and even buses and feeders over-stuffed or rusted-through and ready to bite.
> We always assume everything in the building is wired correctly, but know its usually done by the lowest bidder.
Or by slipping a case of beer to some guy who once spent a day as Electrician's go-fer. I'm quite sure much of the framing and wiring in this house was done by beer-buddies.
> electrical code in California (where I used to live), and British Columbia (where I live now), both require GFCI outlets if they are within six feet of a sink or bathtub.
> When I was looking for information on GFCI use recently, I found out they are also required for hot-tub installations
GFI around water has been in the NEC for a long long time, and is pretty universally enforced (in new or major refurb) by local code officers.
Hot tub GFI and a LOT of grounding rules was a hot topic in NEC at the end of the 20th century. There were several revisions and clarifications. If you are wiring a hot tub you want a quite recent copy of NEC. (I have six, from 1919 to 2009.)
> Not sure where the 30mA is from. For a 15A GFCI, it's 5mA
Look at the spec. It is a *Time Curve*. 30mA will trip quickly. 5mA will trip after some time. I strongly suspect that UK-spec "30mA" and US-spec "5mA" are the SAME calibration, but the nominal names taken from different points on the same curve.
Attachments
My lab is in a concrete basement which could be damp by definition and therefore requires protection by GFCI. Every outlet in the basement is wired to a single 20 amp GFCI breaker, and my workbench has its own 30 amp GFCI. I have been experimenting on all sorts of electronics and nothing I have done has tripped the breaker. These were breakers installed by the builder, and likely the cheapest stuff available. The treadmill however randomly trips the GFCI whenever it pleases. I haven't figured out why either.
This is true, however the biggest risk in a guitar amp, a musical instrument, or any DIY device, is a fault connecting live (primary side) mains voltage into any circuit accessible by the user. It is usually the case that something, the PA, mixing console, or other large part of a sound reinforcement system is grounded through the mains ground connection. This should put mics, synths and most of the on stage equipment at ground level electrically. A guitar amp with a shorted death cap could put the guitar at line voltage potential, such that touching both could be disastrous. Stage lighting "cans" especially if they have ever been wet, are notorious "shockers." A properly installed and wired GFCI will trip in this case possibly saving someone's life.
A short somewhere inside the amp connecting B+ or some other high voltage on the secondary side of the power transformer to the input can not be seen or detected by the GFCI, however such a short SHOULD BLOW the amps fuse since all user inputs SHOULD be connected to circuit ground.
It has been debated in the HiFi world as to whether it is necessary to ground one side of the speaker winding on the OPT. I have seen an OPT with a primary to secondary SHORT, and it was in a guitar amp. The amp operated normally with no clue that there was 300 volts on the speaker leads until someone tried to attach a DI box and stuff blew up. We SHOULD GROUND one side of the speaker for this reason.
A GFCI will not protect you from any acts of stupidity on the secondary side of a power transformer. This includes touching the case of your TO-3 output transistors with one hand to see if they are hot while holding the guitar in the other hand.......especially when doing an outdoor gig in the Florida sun with a live mic nearby so that the audience all gets to hear the "F" word!.....No, it wasn't me.
The manner in which we grip the neck of the guitar will make it nearly impossible to let go of it if a AC current fault through the player occurs.
The 30mA is a rough number. The typical resistance for a man is different than the typical resistance for a woman. (Typical man and woman???) Things change as we age, and rough hands from manual labor increase this resistance. The resistance of a person is roughly 500 to 1000 ohms.
Most protective devices are labeled in amps, but that's rarely the technical reality. A gfci provides no current limiting devices, so when my body ends up in series between the phase and neutral, I am exposed to the entire fault current available in the circuit. In reality it's a time curve that we're working with. How fast the fault can be cleared cleared, or in the case of the gfci, how fast the circuit can be deenergized. A 10A breaker just trips faster than a 20A breaker during a fault, but I receive the same amount of current till it trips.
Fuses are the same way. Grab a 1A fuse, a rheostat, and a 12v battery. Slowly increase load through the fuse and you will get it to hold more than 1A. In some cases, 150% more. It eventually melts out instead of blowing. I see 40A fuses hold at 50A every summer as AC loads grow. A 10A difference is well within the margin of error of my ammeter.
If you look at the chart, the big box before DEATH is from what 10 to 100 milliamps? That's a very wide range of currents. It's there to take into account the differences in our bodily makeup.
I am in no way an expert, and I'm not arguing with the experts, but I have spent considerable time with industry experts being trained on this exact subject.
Most protective devices are labeled in amps, but that's rarely the technical reality. A gfci provides no current limiting devices, so when my body ends up in series between the phase and neutral, I am exposed to the entire fault current available in the circuit. In reality it's a time curve that we're working with. How fast the fault can be cleared cleared, or in the case of the gfci, how fast the circuit can be deenergized. A 10A breaker just trips faster than a 20A breaker during a fault, but I receive the same amount of current till it trips.
Fuses are the same way. Grab a 1A fuse, a rheostat, and a 12v battery. Slowly increase load through the fuse and you will get it to hold more than 1A. In some cases, 150% more. It eventually melts out instead of blowing. I see 40A fuses hold at 50A every summer as AC loads grow. A 10A difference is well within the margin of error of my ammeter.
If you look at the chart, the big box before DEATH is from what 10 to 100 milliamps? That's a very wide range of currents. It's there to take into account the differences in our bodily makeup.
I am in no way an expert, and I'm not arguing with the experts, but I have spent considerable time with industry experts being trained on this exact subject.
I'm grateful for your contributions, and I'm sure most people on the thread feel the same way.
Lots of smart and knowledgeable people here, so there is always some discussion and back-and-forth. I don't think any of it is hostile, just a healthy part of the scientific process. (But unsettling to people used to polite non-technical company in which nobody disagrees with anyone else, because that would be rude. Scientists and engineers and mathematicians don't follow that rule, the only way to keep things honest is to have debate and disagreement and scribbled equations and experimental data until the facts emerge.)
So: we should be thinking " a GFCI limits the total energy (joules) delivered by the shock" rather than "a GFCI limits the maximum current delivered by the shock", yes?
-Gnobuddy
Gnobuddy, I appreciate you looking out for me!! Upon rereading, I realize I didn't sound the same way on paper as I did in my head. [emoji3]
I wasn't feeling like things were hostile, this forum is amazingly helpful! I realize that I know just enough about electricity to understand that I have tons more to learn.
Most of us linemen are cautious about being too confident and forward outside of work. We have to work and understand things in such a way that we stay alive. Its common for us to argue and fight when someone has different ideas because I'm the one that gets killed for following or misunderstanding someone else's advice.
I also realize I'm a tradesman in the midst of engineers. Our engineer is the smartest man on earth! So I have serious respect for what many people on this forum know and understand!!!
I teach a transformer theory class, and a grounding class at work in addition to working on a line crew, but dang if I can make an amp that doesn't hum way too loud.
power line grounding and amp grounding could not be more different! I just have to make sure everything is connected, an amp actually cares where and how its connected....
Yes on the Joules. Its important to know what's actually happening with safety items, not just what the manufacturer tells you. I regularly take things apart at work to see how and where they fail.
We used to put 1A fuses in suspect underground cables, to see if they were actually bad. (Bang test) The thinking was that blowing a 1A fuse couldn't hurt, till the engineer showed me how we were actually putting 10,000A at 7200V for a short time, and were damaging very hard to replace circuits. We now use specialized testing equipment.
I wasn't feeling like things were hostile, this forum is amazingly helpful! I realize that I know just enough about electricity to understand that I have tons more to learn.
Most of us linemen are cautious about being too confident and forward outside of work. We have to work and understand things in such a way that we stay alive. Its common for us to argue and fight when someone has different ideas because I'm the one that gets killed for following or misunderstanding someone else's advice.
I also realize I'm a tradesman in the midst of engineers. Our engineer is the smartest man on earth! So I have serious respect for what many people on this forum know and understand!!!
I teach a transformer theory class, and a grounding class at work in addition to working on a line crew, but dang if I can make an amp that doesn't hum way too loud.
power line grounding and amp grounding could not be more different! I just have to make sure everything is connected, an amp actually cares where and how its connected....
Yes on the Joules. Its important to know what's actually happening with safety items, not just what the manufacturer tells you. I regularly take things apart at work to see how and where they fail.
We used to put 1A fuses in suspect underground cables, to see if they were actually bad. (Bang test) The thinking was that blowing a 1A fuse couldn't hurt, till the engineer showed me how we were actually putting 10,000A at 7200V for a short time, and were damaging very hard to replace circuits. We now use specialized testing equipment.
Rod Elliott has a new page dedicated to mains safety including guitar amps & death caps -
Electrical Safety Requirements
Electrical Safety Requirements
An amplifier that is properly grounded for safety will almost always have a ground loop the instant it is connected to another piece of equipment that is also properly grounded. You can’t get rid of the ground loop without bending up the ground pin somewhere. The only thing you can do is arrange things such that the ground loop doesn’t matter. It is not always obvious how, with the exception of balanced interconnects which is basically cheating (because neither conductor for the input signal is ground).
I've often wished that electric guitar pickups and wiring was balanced.
Wikipedia says the XLR connector already existed in the 1940s, so clearly audio engineers of the time already knew about the problems that go with small audio signals and single-ended cabling and circuitry.
Sadly, the pioneers of electric guitar were not engineers, but tinkerers with little to no technical education, and so the chance was missed. I'm sure the high cost of valve amplification at the time didn't help.
But imagine if Harry DeArmond or George Beauchamp had thought to centre-tap their guitar pickup coils, and bring out both the centre tap and the two ends to the guitar connector. We might all have been spared the next century of struggles with hum, buzz, radio interference, and so on. (And when Leo Fender later copied DeArmond's pickup, he would presumably also have copied the centre-tap arrangement.)
The opportunity surfaced again with Seth Lover's invention of the humbucker. The centre-tap was there for the taking, since the two coils in each humbucker had to be connected in series anyway. And balanced audio cabling in professional audio was well established by this time.
But Mr. Lover appears to also have been primarily self-taught in electronics, a technician and not an engineer, and may not have been aware of the advantages or even existence of balanced audio interconnects. At any rate, the opportunity was missed yet again. It would not have helped that existing guitar amplifiers all had single-ended inputs, and it would cost a minimum of one extra triode to process the balanced input signal.
-Gnobuddy
Surely you have an ohmmeter? So you can show that "normal" is more like 10K-100K. On 120V, that's hardly dangerous (why Edison worked with 100V; fewer deaths).
As you say, everybody is different. I knew a guy could not feel straight 117V AC. His skin resistance was close to 1Meg.
But then we have the worst cases. George comes off the surfboard saturated in seawater and sweat, stands barefoot on concrete, grabs the guitar.
Also, and often key: with continued current the 10K skin resistance breaks-down. Cells get leaky. Body juice soaks the dry skin cells. Skin resistance can fall below 1K faster than a person can understand the danger. Now we are into lethal shocks.
GFI *only* detects current.
It has line voltage across it but does not monitor it. We could say it simply assumes all the 120V may be going to a person. But the real danger is the nerves in the heart. The heart is a sack of salt-water and would be a low-low resistance, much-much less than skin resistance. From "outside" we can't know how much voltage is dropped in the skin and how much is dropped in the heart. It is reasonable to monitor just the current.
Since line voltage is usually 120V, yes in a sense it does monitor Watts. And the time-curve does approximate Watt-Seconds (which is Joules). However it will cut-off at the same current (there's a Zener reference) at any line voltage where it will "work". On the up-side, maybe 180V before a dropping resistor smokes and it goes dead. On the down-side maybe 60V before the circuit starves or the trip-coil gets too weak.
PRR
Ive often wondered why my multimeter doesn't jive with the standards, but lack the info to answer why.
IEEE Std 1048-2003 lists it as 1k but says it's currently under consideration.
That's also the standard listed in osha 1910.269 app c
They use that number to determine the required resistance of an equipotential grounding system, to make sure current across your body is low during a fault.
1000 ohms is an approximation that started in the 50s doing straw man tests.
If you read carefully though, you are absolutely correct. It's hard to tell skin resistance from body or internal resistance, and moisture makes a world of difference!
I've measured hand to hand, and gotten similar results as you did. I just figured that my $100 fluke was not the device they used to make the standard.
I can say from experience though, that whatever my real resistance is makes me pretty jumpy when I have a hole in my gloves hooking up house services in a transformer. [emoji3]
What is super cool though is being 15ft from an energized 345kv line and having arcs across your steel toe boots, and drawing 4in arcs from the steel tower to your fingers as a result if induction. Looks like your a wizard, but feels like hornets!![emoji298]
Ive often wondered why my multimeter doesn't jive with the standards, but lack the info to answer why.
IEEE Std 1048-2003 lists it as 1k but says it's currently under consideration.
That's also the standard listed in osha 1910.269 app c
They use that number to determine the required resistance of an equipotential grounding system, to make sure current across your body is low during a fault.
1000 ohms is an approximation that started in the 50s doing straw man tests.
If you read carefully though, you are absolutely correct. It's hard to tell skin resistance from body or internal resistance, and moisture makes a world of difference!
I've measured hand to hand, and gotten similar results as you did. I just figured that my $100 fluke was not the device they used to make the standard.
I can say from experience though, that whatever my real resistance is makes me pretty jumpy when I have a hole in my gloves hooking up house services in a transformer. [emoji3]
What is super cool though is being 15ft from an energized 345kv line and having arcs across your steel toe boots, and drawing 4in arcs from the steel tower to your fingers as a result if induction. Looks like your a wizard, but feels like hornets!![emoji298]
The best example of the effectiveness of balanced audio I've seen is in the Hammond organ. The cable from the organ console to tone cabinet carried 110VAC and balanced audio alongside each other in the same cable. Worked like a charm. Granted the audio was pretty high-voltage, low-impedance, but still.
Impressive, right?
Ever plug one of those cheap microphones sold for karaoke into a P.A. with better bass response than the karaoke machine? They hum like crazy, because they're usually single-ended. It's easy to see why balanced microphone cables were already in use early in the history of studio audio.
A couple of years ago I needed a vocal mic at short notice, so I bought a $6 dynamic mic at the local dollar store while waiting for my mail-order hand-held condenser mic to arrive. The cheap mic sounded better than the ubiquitous Shure SM58 - but it hummed at 60 Hz, though it had the usual XLR connector. When I opened it up, there was no internal transformer, and the XLR was wired to deliver a single unbalanced signal!
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.