Transformerless, Hot Chassis, Isolation transformer - the last word?

[atmars]

Every once in a while I run into a transformerless amp safety (or, lack thereof)  and return to the net to nail down the details yet find that there is a lot about the topic that no one seems sure of.

 I understand the basic warning transformerless guitar amps can only be used safely with and isolation transformer between the amp and the line  voltage.

But, I would like to know some details from those who know what they are talking about (no guesses please).

1). What are the variants on the hot chassis design? I don't believe i have ever seen one that has the line voltage connected directly to the chassis (my understanding this is how the "all american 5 radios were designed") but have seem some where there is no voltage on the chassis no matter which way the plug is inserted in the socket. Are there some basic variants? If so, do they all have the same level of danger?

2). What is the mechanism for each/all that makes them dangerous? Some examples: Literally  hot chassis? Internal fault (say, tube shorting internally)? Voltage difference between said amp and another device (say, PA grounded or un-grounded)?

3). How does the isolation transformer protect the user (I believe I understand this, but i have read a number of explanations, so let's hear it from the experts)? The more details the better.

4). Finally, what is the correct configuration for the isolation transformer? How should it be wired regarding grounds to that device or the amp?

Many thanks in advance!

[eleventeen]

When you listen to an All-American 5 radio, you (1) turn it on; (2) turn the volume to where you want it, (3) adjust the channel select to dial in your station. It is not just possible but incredibly likely you will never, ever do anything different with the thing in 20+ years of ownership and use and if the knobs that you use to make those adjustments are insulating plastic, then the chances of getting shocked are just about zero. Most AA5 radios have zero exposed metal parts. None.


It is absolutely different on a hot-chassis amp because every piece of metal on your guitar is connected to the chassis via the ground connection in your guitar cable. It *may* be that the ground on the input jack is lifted off the chassis via an insulating washer. It *may* be that the one side of the AC input is connected to the chassis via the so-called death cap. Both could be true. The point is, you really do not know whether one or both of those is/are in place, and, if it's just the cap (which most of them are/were) then your life might depend upon that cap not leaking.


The mechanism of the thing being dangerous is a hot/neutral swap wiring error in the location you might be playing in or might one day bring the amp to. The entire picture is worsened by the general mechanics of holding your guitar with one hand and reaching, with the other hand, to some other piece of gear. If there is a voltage difference, it goes right across your heart. Unless you are Frankenstein, that's probably not good for you.


The correct config removes the >>DC<< connection from one side of the line to the chassis. That's the idea of the cap. It removes a DC connection, it blocks DC. Until it doesn't. My isolation transformer is a pair of identical filament transformers, connected secondary to secondary. The 120 VAC output does not know and does not care which wire on the 120 VAC input is ground or neutral.

[PRR]

> I don't believe i have ever seen one that has the line voltage connected directly to the chassis

That hazard was never designed-in on Guitar Amps. Radios, yes.

However many guitar amps use an R-C network between user-chassis (particularly guitar jack) and line to drain buzz. The caps often fail SHORT. And none of them are as young/fresh as they used to be.

> have seem some where there is no voltage on the chassis no matter which way the plug is inserted in the socket.

Show me one.

> i have read a number of explanations, so let's hear it from the experts

Experts are over-rated. Find out what a transformer does.

[atmars]

Thanks for the answers so far.


Maybe a should clarify what the objective is here: although i personally would like clarification on some if these questions, in general i was hoping for definitive answers from the smartest here so there is a resource that can be pointed from other places on the web. There is seems to be much confusion elsewhere and when it comes to devices that are often described is "widow makers" and "deathtraps", that i was hoping to try to set straight as many people as possible that "are pretty sure" they know what they are talking about as the stakes are seemingly so high.


For example, i recently crossed paths with and individual who posts often and knowledgeably elsewhere, was working on one of these amps and added a grounded three prong plug. When i inquired about an isolation transformer, his response was "he couldn't see how there could be a shock hazard when the safety ground was in place".Maybe he is correct, but he didn't sound too sure.


So regarding this for example:


"i have read a number of explanations, so let's hear it from the experts:


Experts are over-rated. Find out what a transformer does."


I think I do understand what a transformer does, but I worry I ( and others) actually don't. Or, don't to the degree we should especially when this level of safety is being considered.

My understanding is the isolation transformer eliminates a  direct connection to the mains power supply as there is no physical connection between the two coils. That system eliminates the ground path that makes you a possible fault between  the earth or other pieces of equipment. So, if that is indeed correct, I am interested in the nuts and bolts of how this works and maybe more importantly how it can be accidentally defeated. I just saw a video where the individual shows a  ground connection that goes from the device on the isolated side of the transformer to the earth ground on the other side, stating that there is no potential between that ground and the isolated hot and neutral. Yet another video show how to construct an isolation transformer using back to back transformers, but does not include a ground in the diagrams. How do those two explanations square?


I have read enough accident reports about people who actually thought they were being safe, but because they did not understand the system deeply enough, they inadvertently defeat the safety system and there is a disaster.


So if anyone who feels qualified wants to respond to the questions in detail, maybe we can distill things down to a document that explains the topic thoroughly and does't leave out any pertinent details, that would be great. Or, alternatively, maybe someone can point to an another resource  out there that they believe to be comprehensive and thorough?

[eleventeen]

With respect, I think you are just blazingly overthinking this.


"My understanding is the isolation transformer eliminates a  direct connection to the mains power supply as there is no physical connection between the two coils. "


Correct. Period, (should be) end of subject, although IMO it would clarify your thinking by saying "DC connection" or "zero DC volt connection" between the two coils.


When you say "mains power supply" you are saying all of 1: the voltage potential between the hot and neutral 2: the "groundiness" of the neutral, and 3: the groundiness of the mains ground, the integrity of that ground as you find it at the wall socket. It is very possible, indeed common, that the neutral as measured at a socket will be a fraction of a volt or a volt or so different from load center ground. That is because socket-neutral might differ from load center ground by the voltage drop occurring over some dozens of feet of wire that connects the load center to the individual socket you may be using. Most modern circuits use the third ground wire in the 3-wire romex-type cable that connects the load center with your indiv outlet. Some circuits use the steel conduit for the same function and there, one might expect a bit more volts difference between socket neutral and load center ground. But we are talking low single digits volts, hopefully not even 10 volts.




"That system (eg; the use of a transformer) eliminates the ground path that makes you a possible fault between  the earth or other pieces of equipment. So, if that is indeed correct, I am interested in the nuts and bolts of how this works and maybe more importantly how it can be accidentally defeated.


It is hard to imagine how this could be "accidentally" defeated. One would have to deliberately work at it. It could be "unknowingly" defeated by using an autotransformer such as a variac. Many/most variacs connect one wire of the output circuit to one wire of the input circuit. They thus have the ability to "pass on" a socket-wiring fault. It is absolutely well known that variacs (which are commonly used during troubleshooting electronic stuff) ARE NOT isolation transformers. This is why.




I just saw a video where the individual shows a  ground connection that goes from the device on the isolated side of the transformer to the earth ground on the other side, stating that there is no potential between that ground and the isolated hot and neutral. Yet another video show how to construct an isolation transformer using back to back transformers, but does not include a ground in the diagrams. How do those two explanations square?


I don't see the conflict. With a transformer (or two, back to back) there is only an *AC* connection from primary to secondary. No DC connection. If the device you are working on has a chassis ground and you plug into a properly grounded outlet, if there is a hot-neutral reversal in the "house" wiring (the exact, precise, and only thing we are talking about, the thing that can shock you) then the fuse should blow before you get the opportunity to get shocked, as a massively lower-resistance and thus higher current path will exist between hot and ground. That's what blows fuses.

[PRR]

> "he couldn't see how there could be a shock hazard when the safety ground was in place".

In place on the _amp_. Who knows how the wall outlet is wired??

In my old kitchen, the third holes didn't connect to anything. Also the white/blacks were "wrong" on 60% of the outlets. It was 1959 all over again. Get the plug the wrong way, you are HOLDing the Live Wire through a R-C network which is 50+ years old.

> only an *AC* connection from primary to secondary.

Actually there is no electric connection. Utility electricity is converted to magnetic energy and back to electric energy. The secondary is "floating". Safe as a 120V battery with neither terminal connected to the world. You can touch the world (dirt, concrete, radiators) and one side of the secondary and NO current can flow. (Real transformers will have <1000pFd and >10Meg stray leakage; better transformers are used in operating rooms where they get under your skin next to vital organs.)

> people who actually thought they were being safe, but because they did not understand the system deeply enough, they inadvertently defeat the safety system and there is a disaster.

Right. And there are SO many ways to mis-understand, I think any "understandable" explanation would do no good. Just make the author liable to the deceased heirs.

[terminalgs]

When C1 fails closed, which guitar player do you want to be?