EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: pascal_sweden on October 10, 2014, 10:09:44 am
-
Best practice: Wanted to check what the best practice is: using an isolation transformer to power the device under test, or using an isolated differential probe while both device under test and scope are connected to ground.
Isolation transformer: Where can I find a cheap, but reliable isolation transformer? Which manufacturer is a good A-brand?
Isolated differential probe: Where can I find isolated differential probes to connect to Rigol scope? (DS1000/2000 series) Does Rigol offer also isolated differential probes? Or which manufacturer provides this? It's for my Rigol scope.
-
The differential probe is considered the safer approach under the strict exception that you place the probes before powering on, and not moving them when on for devices that contain over 50V,
does the differential probe have to be isolated, most will handle high voltages, and there differential nature means you can watch a 200mV signal between 2 nodes at 350V from what i have experienced,
-
Best practice: Wanted to check what the best practice is: using an isolation transformer to power the device under test, or using an isolated differential probe while both device under test and scope are connected to ground.
Isolation transformer: Where can I find a cheap, but reliable isolation transformer? Which manufacturer is a good A-brand?
Isolated differential probe: Where can I find isolated differential probes to connect to Rigol scope? (DS1000/2000 series) Does Rigol offer also isolated differential probes? Or which manufacturer provides this? It's for my Rigol scope.
That rather depends on what you want to test; the isolation voltage, differential voltage, bandwidth, source impedance can all affect the answer.
Consider an alternative: avoid the issue entirely. Have a sensor that is connected to the device under test, an isolation mechanism (of which there are several) with its input connected to the sensor output, and connect the scope to the output of the isolation mechanism.
-
i just bought an isolation transformer myself on ebay.
my limited research shows it protects your cro (ground loop), but also you if you touch something bad.
pintek dp25 seems to be cheapest isolation probe. should work on you rigol.
alternative is to get a floating cro such as battery powered hantek. prolly cost more than some probes however.
-
Best practice is isotransformer for the DUT, seperate isotransformer for the scope AND using a diff probe.
But then that was in my previous company mandatory for DUT's that generated kV's and were extremely (lethal) dangerous.
If you want to keep it simple and the DUT is not extremely dangerous than I would choose the diff probe, but be sure you have a proper RCCB installed (check that it switches of with 35mA leakage current through a power resistor).
-
RCCB? =)
-
RCCB? =)
Residual Current Circuit Breaker
-
Consider an alternative: avoid the issue entirely. Have a sensor that is connected to the device under test, an isolation mechanism (of which there are several) with its input connected to the sensor output, and connect the scope to the output of the isolation mechanism.
Duh. That is what a differential probe is for :palm:
Never ever use an isolation transformer. It is a cheap ass antiquated method which has many pitfalls in an uncontrolled environment.
-
depends on what you're measuring.
obviously its not best practice to ground the switching element of a supply so you can take a look at the highside fet source drain waveform..
(--btw, the parasitic capacitance of most isolation transformers is pretty lossy, and it makes a great snubber)
-
i did this setup in a repair lab
Want to stress one point of attention for newbies: if you do this be 100% sure the variac is an isolated one, this means that you can not measure any resistance between any of the outputs to any of the inputs! The wrong kind a autotrafo-variac has a connection between one of the inputs and one of the outputs.
I have had my share of people using autotrafo-variacs and thought they were safe, you can not tell it from the looks unless the manufacturer was so smart to put an earthed socket at the output then you know it is not isolated.
-
Newbies don't recognize terms like variacs =) I asked for a simple, brief description of best practises. But all this feedback just overcomplicate things. How difficult can best practise be? There should be one, brief answer that covers everything :)
-
How difficult can best practise be? There should be one, brief answer that covers everything :)
Sure here you go:
if you are a newbie or you do not know exactly what you are doing (for instance you have to ask these questions) : DO NOT measure on a mains connected device.
-
So I have to use a 5V battery to power my DUT? =)
-
Consider an alternative: avoid the issue entirely. Have a sensor that is connected to the device under test, an isolation mechanism (of which there are several) with its input connected to the sensor output, and connect the scope to the output of the isolation mechanism.
Duh. That is what a differential probe is for :palm:
Clearly I should have explicitly stated that such a sensor would be permanently built-in to the equipment. Sorry if that wasn't evident.
-
If you have to measure in a mains powered device be sure to have a ground fault interrupter (GFI) and use good differential probe. Connect the probe leads before powering the device.
Then again you can do a lot without needing mains power. A very common trick is to supply the primary part of a switching power supply with a bench power supply (and do the same for the secondary side). That way you can check whether the PWM controller and transistors are OK with the circuit running but without any danger.
-
So I have to use a 5V battery to power my DUT? =)
With mains connected I mean unisolated mains connected, if you use a proper lab supply as Nico stated there is a transformer inside the supply that isolates the mains from your device.
-
For clarification: IMHO it is better to use ground referenced mains with a ground fault interrupter than an isolation transformer. If you touch mains you are most likely to create a path to ground so the interrupter safes your life. With 'bench power supply' I mean a low voltage DC power supply.
-
Can you provide a picture of what you mean?
If I use a Philips PE1542 DC power supply, the output voltage between + and - is not referenced to ground, as the - is not connected to ground. Isn't that already enough, or what am I missing here?
What is a ground fault interrupter? I have never heard of this. Do they sell this commercially?
Don't you need a differential probe only when you measure high voltages? For normal applications where your DUT is powered with 5V, coming from a DC power supply, like above, you don't need it right? I am actually not familiar with how a differential probe works. Maybe also here a short introduction can help.
Last but not least. Why don't they make scopes where the - of the input channels is not connected to ground? After all you measure a potential difference on the input, so ground is irrelevant.
-
Best practice: Wanted to check what the best practice is: using an isolation transformer to power the device under test, or using an isolated differential probe while both device under test and scope are connected to ground.
When working with off-line circuits, I usually use both at the same time but often end up using singled ended probing with an isolation transformer because I lack 4 high voltage differential probes and sometimes need better performance than they will provide.
Isolation transformer: Where can I find a cheap, but reliable isolation transformer? Which manufacturer is a good A-brand?
I built mine from two big transformers connected back to back. This gave me two isolated half power (but still high power) 120VAC outputs and oddly enough two low power isolated 120VAC outputs because the secondaries had an extra 120VAC winding.
Isolated differential probe: Where can I find isolated differential probes to connect to Rigol scope? (DS1000/2000 series) Does Rigol offer also isolated differential probes? Or which manufacturer provides this? It's for my Rigol scope.
I would check out Pintek for something new but cheap:
http://www.pintek.com.tw/ (http://www.pintek.com.tw/)
https://www.eevblog.com/forum/testgear/teardown-pintek-dp-25-differentail-probe/ (https://www.eevblog.com/forum/testgear/teardown-pintek-dp-25-differentail-probe/)
I would also consider Testec and PMK.
-
Can you provide a picture of what you mean?
If I use a Philips PE1542 DC power supply, the output voltage between + and - is not referenced to ground, as the - is not connected to ground. Isn't that already enough, or what am I missing here?
This will work great for testing assuming your DC power supply has the required voltage and current capability.
What is a ground fault interrupter? I have never heard of this. Do they sell this commercially?
Ground fault interrupters measure the current in the neural and hot wires and if they do not balance out indicating a fault to earth ground, then they disconnect the circuit. Note that a ground fault interrupter does not actually measure current through a ground conductor.
Don't you need a differential probe only when you measure high voltages? For normal applications where your DUT is powered with 5V, coming from a DC power supply, like above, you don't need it right? I am actually not familiar with how a differential probe works. Maybe also here a short introduction can help.
Even when not making off-line measurements, differential probes can be used in low voltage circuits. Not all measurements are ground referenced and sometimes even the ones that are have large amounts of ground noise which a differential probe can reject.
Last but not least. Why don't they make scopes where the - of the input channels is not connected to ground? After all you measure a potential difference on the input, so ground is irrelevant.
They do make oscilloscopes which are like this and they are very handy but not a replacement for differential probing in all circumstances. The Tektronix TPS series of DSOs are a good example.
-
SIglent has a universal 2 channel probe isolator box. The price in Europe is around 130 euro.
http://siglent.com/ENs/prodcut-fjxx.aspx?fjid=404&id=25&tid=1&T=2 (http://siglent.com/ENs/prodcut-fjxx.aspx?fjid=404&id=25&tid=1&T=2)
(http://siglent.com/Uploadfile/image/20140727/20140727224728_1823.png)
I would strongly recommend to use probes with isolated BNC connectors though. I have no experience with this box but I'm tempted to buy one (and tear down before use).
-
For clarification: IMHO it is better to use ground referenced mains with a ground fault interrupter than an isolation transformer. If you touch mains you are most likely to create a path to ground so the interrupter safes your life. With 'bench power supply' I mean a low voltage DC power supply.
The problem with this is if you touch the neutral which is elevated from ground by the current through it, you can also create a path to ground in the form of a ground loop which the ground fault interrupter will detect. Making this measurement without causing this problem requires a differential probe but the GFI only contributes safety.
-
SIglent has a universal 2 channel probe isolator box. The price in Europe is around 130 euro.
I forgot about these. Tektronix used to make one as well.
They are not complete substitutes for differential probes unless used differentially but provide the functionality that you would find in an oscilloscope which has isolated inputs.
Incidentally, using a differential probe with an isolated input will fail if there is no ground reference for the probe's output because the common mode voltage will drift out of range.
-
I continue to be baffled at the strong interest in these forums for measuring hazardous circuits.
What on earth kind of projects are people working on that they need to do such dangerous measurements?
I'm thinking of getting into thoracic surgery in my basement, just on friends and family. And maybe on small animals also. Do you think it is better to use pre-sterilized supplies and surgical implements, or do you think I should invest in an autoclave to sterilize my own things?
:palm:
-
I continue to be baffled at the strong interest in these forums for measuring hazardous circuits.
What on earth kind of projects are people working on that they need to do such dangerous measurements?
Some of my own poducts/designs generate a couple of hundred volts to meet the specs for the customer. And sometimes you run into less well designed equipment like a lead-acid battery charger with one of the output legs connected to mains directly :scared: I'm still happy I turned the damn thing inside out before attaching my oscilloscope directly. :phew:
-
For your personal safety: never use a Variac. It's there to limit damage to your equipment, not you.
For safety with higher voltage circuits, learn to use one hand only (your right one). Use an insulation transformer for your circuit under test, if you can. A diff probe only isolates the test equipment it's connected to.
-
For your personal safety: never use a Variac. It's there to limit damage to your equipment, not you.
you probably can't imagine why they continue to be manufactured, for well over 100 years, can you.
-
In the past 40 years I have used a variac twice and both occasions where very recently: to test a device at various line voltages.
-
Consider an alternative: avoid the issue entirely. Have a sensor that is connected to the device under test, an isolation mechanism (of which there are several) with its input connected to the sensor output, and connect the scope to the output of the isolation mechanism.
Duh. That is what a differential probe is for :palm:
Never ever use an isolation transformer. It is a cheap ass antiquated method which has many pitfalls in an uncontrolled environment.
Used by tens of thousands of people over tens of years with a vanishingly small number of dangerous incidents.
Of course,that was "used by people with a functioning brain" ;D
-
I continue to be baffled at the strong interest in these forums for measuring hazardous circuits.
What on earth kind of projects are people working on that they need to do such dangerous measurements?
Um, how about repairing used test equipment? The horror, the horror.....
-
I continue to be baffled at the strong interest in these forums for measuring hazardous circuits.
What on earth kind of projects are people working on that they need to do such dangerous measurements?
Um, how about repairing used test equipment? The horror, the horror.....
i'm sorry, my ears are still ringing, what did you say?
-
Richard, it's not uncommon for me to troubleshoot 230 or 460VAC circuits and once in a while to check 600VDC bus voltages. Welcome to industrial controls. Ever look at a Ward Leonard DC motor drive? C16J tubes running at 300VDC at 15 amps! Scarey stuff! I would love to have a high voltage diffrential probe when looking at that stuff. The day I get casual working around that stuff is the day I NEED to retire. Sometimes I can reach in and get my voltage readings safely and other times I just shut the system down and hook up my leads with clips. Older equipment seems to have been designed with less safety in mind for the service engineer it seems. Open contactors all over the place. Even some of the newer machinery is not well thought out. 24V and below is a piece of cake, but it's usually in the same enclosure as the mains voltages with associated transformers and contactors. It's just the way it is working on this stuff.
BillWojo
-
Oh, I have no doubt there are folk with legitimate need to test all sorts of high voltage, mains-connected and otherwise dangerous circuits. I am just surprised that people with such legitimate needs would come to a forum like this and ask what seem to be rather basic questions.
From previous experience in other forums, the people who are asking basic questions are in poor position to conduct such tests safely.
-
Oh, I have no doubt there are folk with legitimate need to test all sorts of high voltage, mains-connected and otherwise dangerous circuits. I am just surprised that people with such legitimate needs would come to a forum like this and ask what seem to be rather basic questions.
From previous experience in other forums, the people who are asking basic questions are in poor position to conduct such tests safely.
Fair point. But everyone has to start somewhere. Is it better to assist someone who is trying to learn or to ridicule them and discourage others from asking questions?
The OP's question was very reasonable IMO and based on the responses a question that even those with experience cannot agree on one simple answer.
-
I am uncomfortable answering broad, generic questions like that because there are so many variables and rather different ways of handling different kinds of situations. I have the nagging feeling that broad, generic responses are borderline irresponsible without adding some amount of guidance in analysis of the situation and help matching the proper solution to specific cases.
And it doesn't help that generic gadgets like "variac" have such a variety of types, some of which are fully isolated, and many (most?) of which are NOT. And with no reliable way for beginners to discern the difference.
It just seems like many of these discussions feature beginners who don't exhibit proper respect for potentially lethal situations and practices. But I guess a couple of previous generations of experimenters and techs managed to survive without advice from the interweb.
-
I continue to be baffled at the strong interest in these forums for measuring hazardous circuits.
What on earth kind of projects are people working on that they need to do such dangerous measurements?
In my case it is usually from repairing or designing off-line switching power supplies.
-
Whenever I work on one of these, I try remind myself how very careful I must be. Using a marked up schematic might just save your life.
And always remember the famous Jim Morrison line "the future's uncertain, and the end is always near".
-
Consider an alternative: avoid the issue entirely. Have a sensor that is connected to the device under test, an isolation mechanism (of which there are several) with its input connected to the sensor output, and connect the scope to the output of the isolation mechanism.
Duh. That is what a differential probe is for :palm:
Never ever use an isolation transformer. It is a cheap ass antiquated method which has many pitfalls in an uncontrolled environment.
Used by tens of thousands of people over tens of years with a vanishingly small number of dangerous incidents.
Note the word 'antiquated'. People used to live in caves for thousands of years. Following your logic a house as we know it now is just as comfortable as a cave :palm: Back in the old days a GFI was expensive or it didn't exist. Nowadays a GFI is a commodity.
-
What about a Philips PM 8940 Isolation Amplifier? Any good?