Isolation Transformers and Safety Considerartions

Thread Starter

SamR

Joined Mar 19, 2019
5,031
I keep seeing Safety from electrocution being brought up as a reason for using an isolation transformer but not why. I certainly understand the ability to block an incoming DC component but not the safety issue. I am assuming the idea is that mains can deliver more current than the transformer in a fault? But I would think that even the faulted current of an isolation transformer is potentially deadly?
 

ericgibbs

Joined Jan 29, 2010
18,767
hi Sam.
In the UK, we use a 3 wire system, Line, Neutral and Earth.
At the generator source the Neutral is connected to Earth/Ground, so the Line wire is 'Hot' with respect to the Neutral and Earth.

So if you are standing on the Ground [Earth] and touch the Line you could be electrocuted.
The Neutral is only a few volts above Earth so touching the Neutral is not usually a problem.

When using an Isolation transformer, neither the of the two secondary wires are live 'hot' relative to Earth, so touching either wire should not give you a shock.
You would have to touch both secondary wires to get a shock.

E
 

LesJones

Joined Jan 8, 2017
4,174
As the secondary of an isolation transformer is floating you could touch either wire BUT NOT BOTH without getting a shock. The main situation where it is useful is working on the primary side of switch mode power supplies (Or very old live chassis TV sets.). The first thing (After filtering) on the input of a switch mode power supply is a rectifier (Either a bridge or voltage doubler.) which means that neither the negative or positive output of the rectifier is near ground potential when fed from normal mains as one leg of the mains is at near ground potential. If it is fed from an isolation transformer you can safely connect the ground lead from the scope to the negative. (You could connect it to the positive but the negative is normally used as the reference for signals.) If you connected the scope earth to the negative and the switch mode power supply with it fed directly from normal mains every negative half cycle of the mains would be shorted out via the diodes in the rectifier.

Les.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
On Work sites in the UK, which is a 230v country, A 110v CT secondary transformer is used, the CT is earth grounded so any danger of electrical shock on any work site is reduced to 55v to GND and considered non-dangerous.
In industrial systems in N.A. any control transformer with a 120v secondary is generally referenced to earth ground by connecting one side of the secondary to GND establishing a neutral conductor for safety reasons.
Max.
 

WBahn

Joined Mar 31, 2012
29,979
I keep seeing Safety from electrocution being brought up as a reason for using an isolation transformer but not why. I certainly understand the ability to block an incoming DC component but not the safety issue. I am assuming the idea is that mains can deliver more current than the transformer in a fault? But I would think that even the faulted current of an isolation transformer is potentially deadly?
An isolation transformer addresses some concerns, not all. In fact, it can introduce safety concerns that NOT using it would have addressed. That's because their are multiple fault scenarios and a technique that makes one safer can make another less safe. So by understanding the various scenarios and assessing which ones are more important to mitigate in your situation, you can choose an appropriate technique and be aware of the additional risks you are assuming as a result and act accordingly.

Consider an electrical system that has no ground reference. You use this to power some device and a fault develops that connects one of the power leads to the device chassis. Now, when you touch the chassis you are in contact with one of the power leads. But you don't feel anything because, while you might be grounded, the power lines are not. But, if normal operation of the device put you in contact with the other power line as well, now you get zapped. But what if the chassis of the device was connected to Earth ground and so to was the side of the power line that normal operation puts you in contact with? Now as soon as that fault develops that connects the other line to the chassis, a large current flows (through the device -- not sitting their looking innocent waiting for you to come along and complete the connection) and the breaker trips. That's one of the scenarios in which NOT having an isolated supply increases safety. Other scenarios include when a fault in the distribution transformer results in primary line voltage (tens of kilovolts) getting connected directly to your device. Without the ground it will be sitting their waiting to bite you, but with the ground it results in a large current flow that trips breakers or blows fuses.

Now consider the case where you are poking around inside a device trying to troubleshoot it. If the supply is grounded, then all you have to do is come into brief contact with a single point connected to the hot line and you can be shocked. With no grounding, you have to connect to both supply lines in order to get shocked. So it is safer for this scenario, but you have sacrificed protection against the other scenarios to achieve that.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
As the secondary of an isolation transformer is floating you could touch either wire BUT NOT BOTH without getting a shock.
This is what I don't understand... I understand that IF I am not grounded I cannot be shocked, but if I am not perfectly isolated from ground any high potential will use me as a path to ground. I know about the guys working on live KV aerial distribution systems but then there are lots of "accidents" involving high V xfmrs and capacitors. Standing on a rubber mat does not guarantee safety but sure doesn't hurt sort of thing.

I also know that I blow a lot of fuses on my 120/12/6 CT XFMR secondary when I make a mistake placing the ground on my Oscope probe. Which is also why I use a GFI on it's primary.
 

WBahn

Joined Mar 31, 2012
29,979
This is what I don't understand... I understand that IF I am not grounded I cannot be shocked, but if I am not perfectly isolated from ground any high potential will use me as a path to ground. I know about the guys working on live KV aerial distribution systems but then there are lots of "accidents" involving high V xfmrs and capacitors. Standing on a rubber mat does not guarantee safety but sure doesn't hurt sort of thing.

I also know that I blow a lot of fuses on my 120/12/6 CT XFMR secondary when I make a mistake placing the ground on my Oscope probe. Which is also why I use a GFI on it's primary.
The notion that if you are perfectly isolated from ground you can't be shocked is probably part of the problem. If you are floating in mid air and put your hands across the power lines you will get shocked, whether the power lines are ground-referenced or not. Imagine a 10 kV battery -- if you touch the two terminals of the battery at the same time, current is going to flow through you. But if one terminal of the battery is driven into the ground, now you only have to touch the other terminal and if you also have a good enough path to ground from other other point of your body, current will flow through you.

The grounds of most o-scope probes are connected to the safety ground wire in the scope's power cord. This is why you can't measure the voltage between arbitrary points, such as across a resistor, with most scopes and instead have to use two probes and the scope's math functionality.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
The grounds of most o-scope probes are connected to the safety ground wire in the scope's power cord
I figured that out the first time the breadboarded wires started smoking before I fused the xfmr.
 
Last edited:

MaxHeadRoom

Joined Jul 18, 2013
28,619
Which is also why standard 3Ph motor starters have a 120V secondary Xfmr for motor control power instead of full motor line voltage.
The tenancy now is to go 24vdc control circuitry, this has been a standard in Europe for many decades now.
Max.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
24vdc control circuitry
Really... Well, things do change. All motor controls that I did were 120VAC excluding PLC/DCS and some sensors/interlocks controlled by it. Used a lot of octal base relays, timers, etc. and some of those were going to lower V coils for interfacing with DCS or PLC but still used 120V contact outputs to the starter controls. One problem would have been all the explosion-proof stuff we had to use in 90% of the plant and then the maintenance Engineer would have a fit having to stock multiple brand items in the storeroom for his shift electricians. He mantra was "one brand only" no having multiple brand types to stock and train his people on maintaining which is a valid point. Most "New" stuff sneaked in on packaged control systems by outside Vendors. We were primarily Foxboro DCS, A-B SLC, SquareD, and Westinghouse substation breakers. I was fitting some VFDs into the non-explosion proof MCC compartments and fighting to do that.
 

MrAl

Joined Jun 17, 2014
11,396
I keep seeing Safety from electrocution being brought up as a reason for using an isolation transformer but not why. I certainly understand the ability to block an incoming DC component but not the safety issue. I am assuming the idea is that mains can deliver more current than the transformer in a fault? But I would think that even the faulted current of an isolation transformer is potentially deadly?
Do you mean because of the interwinding capacitance?
Yes i suppose some current can flow but it must be small because the windings probably have to be physically isolated from either other from some minimum distance. For wall warts each winding must be on the core not on top of the other winding and that allows it to get the higher class rating for protection against shocks. I would suspect that an isolation transformer should have the same requirements. You can check though by looking at it and see if the windings are one on top of the other or side by side.
Some transformers have the winding one on top of the other but also have a metal guard between windings that gets grounded. I dont think wall warts can be constructed like that though.
 

MaxHeadRoom

Joined Jul 18, 2013
28,619
Really... Well, things do change.
I started retrofitting old industrial relay control panels here back in the 80's and converting to PLC, I made everything 24vdc control where possible, as I mentioned in another post, I got some push back on spec'ing in DC solenoids, AC had been the norm in spite of a higher failure rate with the AC variety, evidenced by a large inventory of spare solenoid coils.
Coming from a UK background where 24vdc versions were typically used in most European machinery, I was aware of the lower failure rate.
Max.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
Do you mean because of the interwinding capacitance?
No just in general as most of my experience with transformers has been with 480VAC secondary or higher power distribution substations and the 1200A breakers to large Motor Control Centers. Makes sense from a consumer safety standpoint that wall warts are wound separate and didn't know that. Thanks. No, just trying to learn about transformers used in electronics and from what the others have input here I think I better understand the statement that "using a transformer to isolate from mains current is safer". Not Idiot-Proof, but I do see how it is safer now if you still observe the basic safeguards. The use of 120V CT secondaries in England for construction was new and I can definitely that as an improvement in safety over 240 mains. Most of what I am dealing with now are 120/24/12/6 etc or just 1:1 lo voltage isolators.
 

MrAl

Joined Jun 17, 2014
11,396
No just in general as most of my experience with transformers has been with 480VAC secondary or higher power distribution substations and the 1200A breakers to large Motor Control Centers. Makes sense from a consumer safety standpoint that wall warts are wound separate and didn't know that. Thanks. No, just trying to learn about transformers used in electronics and from what the others have input here I think I better understand the statement that "using a transformer to isolate from mains current is safer". Not Idiot-Proof, but I do see how it is safer now if you still observe the basic safeguards. The use of 120V CT secondaries in England for construction was new and I can definitely that as an improvement in safety over 240 mains. Most of what I am dealing with now are 120/24/12/6 etc or just 1:1 lo voltage isolators.
Oh yeah the spaced windings also introduces extra leakage inductance which helps filter the DC in those DC wall warts.

But i kinda wish we had 230 or 240vac here because the wiring can then deliver more power to any load. At 120 and 20 amps the max power is of course 2400 watts but with 240 and 20 amps the max power is 4800 watts which is significantly more. Instead we have 120vac and when the line dips it can go as low as 80vac which is problematic in the summer months.
Most homes have a split phase system where you can get 230vac if you go across two phases but most equipment is made for 120vac anyway. If we want to 230vac that would change too.
 

Thread Starter

SamR

Joined Mar 19, 2019
5,031
When I built my house I had a few "dedicated" single outlet duplex receptacles w/#12AWG installed thinking I may at some time add a 2KW linear to my ham rigs and could swap one to 240 easy. A couple more kids put that on hold for 20+ years and still haven't taken back over that room. Not a big deal to add a circuit to most houses and the panel is a snap. Most electricians can do it in an afternoon on the side after work and don't charge you out the wazoo for it. I did wire the house with all #12 and 20A breakers except the 240 utility/appliance ckts and haven't hit the limit anywhere yet thank goodness. GA Power very reliable now and replaced the old feed through here with much heavier 3 phased distribution since there are some small industries way on down and more residential construction along the coast here, so no more browning out. Just the regular tree and hurricane outages.
 

MisterBill2

Joined Jan 23, 2018
18,179
On Work sites in the UK, which is a 230v country, A 110v CT secondary transformer is used, the CT is earth grounded so any danger of electrical shock on any work site is reduced to 55v to GND and considered non-dangerous.
In industrial systems in N.A. any control transformer with a 120v secondary is generally referenced to earth ground by connecting one side of the secondary to GND establishing a neutral conductor for safety reasons.
Max.
In all of the industrial controls that I ever designed we NEVER grounded either side of the 120 volt controls transformer output.
 

MisterBill2

Joined Jan 23, 2018
18,179
If a circuit has no earth connection of any kind then a person at "earth potential" will not get a serious shock if they connect to any point of that system. THAT is the safety benefit of an isolation transformer. "Earthing" any part of that circuit removes that bit of safety. I routinely have this conversation with folks who believe that grounding one side of a portable generator output somehow increases safety, although they are not able to explain how. So there is a mix of superstition and facts and the result is often actions taken without adequate consideration of reality.
 
Top