There is a difference between "isolation" and "floating".
Power transformers provide "isolation".
A circuit connection to earth ground makes it not "floating". Remove the earth connection (assuming that nothing else is connected to LINE, NEURAL, and EARTH) and the circuit is "floating". (Note that some circuits might still have a connection to chassis via a resistor, capacitor, or power line filter.)

 

There is a difference between "isolation" and "floating".

Thanks for spotting the problem in my thoughts. That's what I meant, floating.

 

sounds like you meant using 2 to 3 prong adapter without connecting GND. that is a common way to abuse it and create safety hazard.

 

Small TVs used to be manufactured without a power transformer in order to save cost. The power cord had a 2-pin polarized plug. One prong was wider that the other. If your house wiring was incorrect, the TV chassis would be at LINE voltage.

We rotated the CRT yolk 90° and repurposed a cheap TV into a simple histograph driven by a DEC PDP-15 using only interrupts. In order to prevent inadvertently connecting the TV incorrectly, we used a 1:1 isolation transformer.

 

sounds like you meant using 2 to 3 prong adapter without connecting GND. that is a common way to abuse it and create safety hazard.
View attachment 373131

I believe use of "cheater" plugs are now prohibited and illegal.

 

maybe, don't know... how would that be enforced since they are readily available (sold on every corner) and most people don't even know what the ring terminal on the side is for. and even those that do, may think that running extra wire is - inconvenient.

 

That's because they can't guarantee an isolated circuit? it's not hard to put a circuit that detects high voltage to the earth and then shows an alert message.

I don't know what you mean by guaranteeing an isolated circuit.

Earth ground protects people from electric shock by providing a low impedance path through earth ground to bypass them. Floating devices that have earth ground defeats the protection.

Earth ground and neutral are connected at the power panel, but neutral carries current and is at a higher potential (due to voltage drops) than earth ground which doesn't (under normal circumstances).

When I was a technician (but before I learned that electrical wiring used different color conventions than electronics), I was trying to measure AC voltage on a suicide cord. I connected the scope ground clip to the black wire and that fried the scope probe lead. I managed to disconnect before the breaker tripped, but it protected me from getting shocked. Then I became aware that black was hot in building wiring; not ground...

 

You'd think that we could get this right, across the globe.

 

I don't know what you mean by guaranteeing an isolated circuit.

Earth ground protects people from electric shock by providing a low impedance path through earth ground to bypass them. Floating devices that have earth ground defeats the protection.

Earth ground and neutral are connected at the power panel, but neutral carries current and is at a higher potential (due to voltage drops) than earth ground which doesn't (under normal circumstances).

When I was a technician (but before I learned that electrical wiring used different color conventions than electronics), I was trying to measure AC voltage on a suicide cord. I connected the scope ground clip to the black wire and that fried the scope probe lead. I managed to disconnect before the breaker tripped, but it protected me from getting shocked. Then I became aware that black was hot in building wiring; not ground...

I agree with what you said, I knew it. I was just not expressing myself in the best way. I meant, why don't float every probe of the scope? Just because it's cheaper? But then if you need to do it on the fly, you have to spend a lot of money buying floating probes?! That's not the best thing I have in mind when they say we're in 2026.

 

I meant, why don't float every probe of the scope? Just because it's cheaper?

Most electrical equipment tries to comply with electrical codes. Equipment with a metal chassis ties circuit ground to earth ground for safety.

But then if you need to do it on the fly, you have to spend a lot of money buying floating probes?! That's not the best thing I have in mind when they say we're in 2026.

If you know what you're doing, you can float your scope. I can count the number of times I've done that (in 50 years) on one hand.

Instead of floating the scope, if I want to make a measurement across something that isn't grounded, I use 2 probes with the A-B (A+B with B inverted) feature on my scope.

 

I agree with what you said, I knew it. I was just not expressing myself in the best way. I meant, why don't float every probe of the scope? Just because it's cheaper? But then if you need to do it on the fly, you have to spend a lot of money buying floating probes?! That's not the best thing I have in mind when they say we're in 2026.

In order to float the scope probe they would have to make each channel a differential input. I don't know if any scope has that feature. There is a coax connector that has two signals and a shield, much like TRS connectors.

AI's answer:
Oscilloscopes with Built‑In Differential Inputs
Several oscilloscope models — from budget USB units to high‑end lab instruments — offer built‑in differential inputs, allowing you to measure the voltage between two points without referencing either to ground
Notable examples:

• PicoScope 4444 – A high‑resolution USB oscilloscope with 4 fully differential, high‑impedance inputs (20 MHz bandwidth, 12/14‑bit resolution). Each channel can be paired with intelligent probes for voltages from mV to ±50 V, or high‑voltage CAT III probes up to 1000 V. Designed for non‑ground‑referenced measurements such as CAN bus, balanced audio, and power systems
• Tektronix oscilloscopes with high‑voltage differential probes – While Tektronix’s scopes are typically single‑ended, they are compatible with high‑voltage differential probes (e.g., P5200A, P5202A, P5205A) that provide floating, isolated measurement up to ±640 V (200×) or ±1300 V (50×) with bandwidths up to 100 MHz
• These probes are used with supported Tektronix models such as DPO/DSA/MSO70000, TDS3000, TDS5000, and others, requiring the correct adapter or external power supply.
• Other manufacturers – Many USB and digital oscilloscopes (e.g., Rigol, Siglent) offer differential input channels or differential probe interfaces, but you must check the specific model’s datasheet for “floating” or “differential” input capability

 

Most electrical equipment tries to comply with electrical codes. Equipment with a metal chassis ties circuit ground to earth ground for safety.
If you know what you're doing, you can float your scope. I can count the number of times I've done that (in 50 years) on one hand.

Instead of floating the scope, if I want to make a measurement across something that isn't grounded, I use 2 probes with the A-B (A+B with B inverted) feature on my scope.

Yeah, it works, but as I have 2 probes in my scope, I end up with only one differential probe. If the probes were floating, I'd had 2. I can't measure 2 points that doesn't share a common wire. That has troubled me a lot of times in a couple of years.

 

Yeah, it works, but as I have 2 probes in my scope, I end up with only one differential probe. If the probes were floating, I'd had 2. I can't measure 2 points that doesn't share a common wire. That has troubled me a lot of times in a couple of years.

If this is something you need to do frequently, buy better equipment.

I have multiple Tektronix 7704A scopes. Each can probe two signals differentially by utilizing two 7A26 vertical amplifiers. You can make the same measurements with 2 battery operated scopes. I hesitate to tell anyone to float their scope because that can lead to dangerous situations. I don't like to think about what could have happened to me if I had floated the scope when I was trying to measure line voltage on that suicide cord. If the scope had been floating, line voltage would have been present on the ground lead and the chassis, presenting a shock/electrocution hazzard.

 

If this is something you need to do frequently, buy better equipment.

Buying equipment to replace something that should have in the previous is something I dislike because it makes you sell your current one and invest more on the next. It's very likely to lose money. Things are so expensive in my country that even companies with a large income (2 million dollars + a year) doesn't want to spend much on these equipment.
I know this is the company fault, but it's a struggle to deal with so used scopes that barely works.

 

Buying equipment to replace something that should have in the previous is something I dislike because it makes you sell your current one and invest more on the next. It's very likely to lose money. Things are so expensive in my country that even companies with a large income (2 million dollars + a year) doesn't want to spend much on these equipment.
I know this is the company fault, but it's a struggle to deal with so used scopes that barely works.

I don't know what to say. I don't know of any reputable company that wouldn't pay for equipment that they needed.

 

Buying equipment to replace something that should have in the previous is something I dislike because it makes you sell your current one and invest more on the next. It's very likely to lose money. Things are so expensive in my country that even companies with a large income (2 million dollars + a year) doesn't want to spend much on these equipment.
I know this is the company fault, but it's a struggle to deal with so used scopes that barely works.

You don't have to do that. All my oscilloscopes are conventional 2-channel with single ended inputs.
I cannot recall the one or two times that I needed to make a differential measurement, for example, viewing the current pulse on a high side resistor. In other words, it is a very rare occasion. For that, you can use the difference method already in your oscilloscope.