One time when I was employed at Intel, I sat in a "what do you need" exploration meeting with a bunch of Tektronix engineers. I said I wanted the same thing. They said "Why would you want that?". I explained about ground loops between probes' grounds, inevitably being connected to different places, messing with our measurements.
Later, after retiring; https://www.tek.com/en/products/oscilloscopes/probes/isovu-isolated-probes Not cheap. I believe gen 1 had multiple fibres in the probe cable, one of which carried a laser which was used to power the electronics in the probe head. I've no idea how the gen 2 probes work.
Ordinary folks might afford https://www.batronix.com/shop/measurement/probes/Micsig-DP10013.html, also on ebay https://www.ebay.com/itm/393698322257. Unsure if you could touch these for ~$200, via DIY time / effort.
As an aside, I tried to setup my BT device (switch selectable transmitter / receiver) as an isolated two-channel scope probe, using a PC. Stupid W10 absolutely refused to connect to it as a "Bluetooth A2DP Sink", even after trying the app from the "store", touted on line as being able to do so. As if to tease and torture me, it of course connects just fine as a source - think BT headphones - and I was able to capture a waveform using REW's oscilloscope panel from, you guessed it, the little microphone built into the BT "transmitter" device. What a piece...connects in transmission mode to a BT amplifier just fine, but alas, not to the PC -
I thought it might be fun to get the waveform across a filter inductor with both ends sitting at 300V, displayed on REW's oscilloscope panel, using a very floating probe...with an all digital connection, downstream of the A/D in the probe "head" -
Last edited:
I've used a number of HV differential probes, Tek etc. They generally have crappy CMRR testing circuits with fast transients. Signals I wanted to see were overwhelmed by CM noise. Isolation transformers minimize those issues and shock hazards as long as you follow the "rules". We used to float O'scopes all the time with a single transformer but I prefer using two transformers to guarantee nothing has a galvanic path to neutral or ground. .
There are several ways to probe this kind of stuff. if its your job, get something like a Tek THS720 https://w140.com/tekwiki/wiki/THS720 , which has 2 isolated probe channels and is designed to safely look at dangerous stuff line variable speed motor drives (600V variable frequency stuff) and is still a fine scope.
When surge testing surge protectors the way to look at the line with its up to 3KV surge you use two HV probes and monitor differentially. For lower voltages you can use X100 probes and monitor differentially or get a differential probe suited to that. These have a good reputation: http://www.micsig.com/html/41.html#
When surge testing surge protectors the way to look at the line with its up to 3KV surge you use two HV probes and monitor differentially. For lower voltages you can use X100 probes and monitor differentially or get a differential probe suited to that. These have a good reputation: http://www.micsig.com/html/41.html#
Thanks to the many who have put me on a different tack, that of differential probes. It does seem the obvious way to go, and for a quick one-off job, a fairly high speed op-amp and a small handful of 1MΩ resistors (and preferably a few low pf capacitors) seems to be all that is needed. Do people using such op-amps have a favourite common one? I doubt that in SMPS applications one need bother about frequencies much above a few MHz, but I am ready to be put right.
If you want to measure just the ac voltages, any isolation transformer will work. You do not even need a 1:1 transformer. A 10:1 will be sufficient. This will allow 1x probes to be used.
On the other hand, if you want to measure waveforms, a transformer will not work. The transformer will eat all the harmonics or reduce them to low level that it is not suitable.
If your SMPS is is used in any environment that require equal or more than CAT III, then you have use some sort of Industrial waveform monitors that are specially designed for this application. The transients generated in CAT III or higher may damage the electronic measuring device as higher voltages can creep over the SMD resistors.
Regards.
On the other hand, if you want to measure waveforms, a transformer will not work. The transformer will eat all the harmonics or reduce them to low level that it is not suitable.
If your SMPS is is used in any environment that require equal or more than CAT III, then you have use some sort of Industrial waveform monitors that are specially designed for this application. The transients generated in CAT III or higher may damage the electronic measuring device as higher voltages can creep over the SMD resistors.
Regards.
Bonjour à tous just seeing this
Working as consultant and designed HV, transients protection, SMPS since 1970s.
Never needed any of these isolation probes or secondary side transformers
Simply used Variac variable transformers followed by mains 1:1 ratings 5 KVA to run the DUT.
Even in bridge topologies, I have had no issues to probe and debug drivers and power switches.
More useful than just probing voltages is the switch and driver currents
we used Pearson 411 toroids wideband CT and Tektronix current probes P6021, 6022, and Tektronix DC current probes and amplifiersP6301, P6303 etc.
Highly recommend the 1980s Tektronix Circuit Concept books on Oscilloscope Probe Circuits
https://www.w140.com/tekwiki/index.php/Concepts_Series
I hope this note can inform some to a better solution
Bon courage
Jon
Just the ramblings of an old retired EE
Working as consultant and designed HV, transients protection, SMPS since 1970s.
Never needed any of these isolation probes or secondary side transformers
Simply used Variac variable transformers followed by mains 1:1 ratings 5 KVA to run the DUT.
Even in bridge topologies, I have had no issues to probe and debug drivers and power switches.
More useful than just probing voltages is the switch and driver currents
we used Pearson 411 toroids wideband CT and Tektronix current probes P6021, 6022, and Tektronix DC current probes and amplifiersP6301, P6303 etc.
Highly recommend the 1980s Tektronix Circuit Concept books on Oscilloscope Probe Circuits
https://www.w140.com/tekwiki/index.php/Concepts_Series
I hope this note can inform some to a better solution
Bon courage
Jon
Just the ramblings of an old retired EE
SMPSU repair is probably the biggest issue of the last 30+ years - there are two simple solutions:
1) Use a mains isolation transformer to isolate the PSU (or equipment) from the mains - but this costs money, and has restrictions.
2) Remove the earth lead from the scope chassis (take the wire out the plug for non-moulded plugs) - this is pretty well universal in the electronics service trade, and I (among the many) would consider having a grounded scope a serious health hazard. Basically it's all down to understanding exactly what you're doing, exactly why you're doing it, and what the point of the grounds is in the first place - grounding only provides a level of protection under certain circumstances, under other circumstances it makes it much more dangerous.
If you can arrange it, working in a completely ground free environment is probably the safest solution.
1) Use a mains isolation transformer to isolate the PSU (or equipment) from the mains - but this costs money, and has restrictions.
2) Remove the earth lead from the scope chassis (take the wire out the plug for non-moulded plugs) - this is pretty well universal in the electronics service trade, and I (among the many) would consider having a grounded scope a serious health hazard. Basically it's all down to understanding exactly what you're doing, exactly why you're doing it, and what the point of the grounds is in the first place - grounding only provides a level of protection under certain circumstances, under other circumstances it makes it much more dangerous.
If you can arrange it, working in a completely ground free environment is probably the safest solution.
You are probably right, but there are issues: one is political correctness, and lifting ground is highly uncorrect. Another issue is that class I devices rely on a good PE connection for safety. This is theory only, because class I devices are made to the same standards as class II, except they have metal parts accessible.
I am certainly not going to advise anyone to lift the ground of a class I equipment, but it is true that having well earthed devices like an Oscope can become a hazard, because if you touch the GND and some live circuit, it can be much more dangerous than touching a live connection and something that is not earth-protected.
The best way to reconcile those aspects is to use an isolation tranformer: it is not 100% safe, because you can still find yourself in unsafe situations, and you need to be alert and aware of dangerous conditions, but at least you have some level of safety without compromising the basic safety regulations.
That is the option I have adopted for tens of years now: all my lab supplies have an isolated circuit, with its own voltage control/meter and ammeter. Nevertheless, I always stay on-guard when I use the isolation feature.
Unlike cats, we only have one life, thus take care
I am certainly not going to advise anyone to lift the ground of a class I equipment, but it is true that having well earthed devices like an Oscope can become a hazard, because if you touch the GND and some live circuit, it can be much more dangerous than touching a live connection and something that is not earth-protected.
The best way to reconcile those aspects is to use an isolation tranformer: it is not 100% safe, because you can still find yourself in unsafe situations, and you need to be alert and aware of dangerous conditions, but at least you have some level of safety without compromising the basic safety regulations.
That is the option I have adopted for tens of years now: all my lab supplies have an isolated circuit, with its own voltage control/meter and ammeter. Nevertheless, I always stay on-guard when I use the isolation feature.
Unlike cats, we only have one life, thus take care
Let me point out a basic reality check- when troubleshooting a line connected switching supply you WILL be dealing with dangerous voltages, with or without an isolation transformer. You need suitable probes and a good understanding of what you are looking at. Most likely there will be exposed metal on PCB's heat sinks screws or ??? that are at potentially lethal voltages. its not a place to take any shortcuts.
If you do this on a regular basis you need (not just want) a HV differential probe or a scope like a Tek THS720 with isolated input circuits (they can handle 300+ volts between the ground connection and the instrument).
In any case deal with the circuitry like its a petri disk of smallpox intending to kill you. Take no chances or shortcuts.
If you do this on a regular basis you need (not just want) a HV differential probe or a scope like a Tek THS720 with isolated input circuits (they can handle 300+ volts between the ground connection and the instrument).
In any case deal with the circuitry like its a petri disk of smallpox intending to kill you. Take no chances or shortcuts.
If you have a scope, what is the problem in getting an isolation transformer. Neither the scope nor the isolation transformer are meant for one time use.
(if total isolation is required, use a battery operated scope, disconnected from the charger) Some shortcuts are not good when handling lethal voltages. Just because someone did it, does not mean this has to be a practice in hiding. All current safety rules, be it traffic, mining, electrical, electronic have been developed after numerous incidents as a method to reduce accidents. Grounded equipment sets you, the person to the same equ-potential as the ground and prevents a shock hazard. Regards.
Tektronix believes the opposite. One time I decided to use a Tek battery powered bench scope for making a particular measurement where ground loops were an issue. We had lab budget, so I got approvals and bought one. First thing that comes up when you turn it on, is this big on-screen warning not to use the scope without it being connected to earth ground. Even when running on battery, apparently. You have to press a button acknowledging the message, before the regular scope screen comes up. I even had a comm bus opto-isolator to cut the ground connection back to the PC...
We asked Tek about it, described the application as measuring 1-2 Vdc - and they would not back down. Corporate safety got involved and it killed the project - at least that approach to it. The scope then just rotted in a cabinet far beyond when the battery expired.
This is why I like my Fluke 123, it's old, but it's a DMM with an oscilloscope screen. There's more modern versions of that idea available.