0
\$\begingroup\$

I'm in my infancy as regards electronics, but will soon be acquiring an O-scope as part of my learning plan. I'm at the stage (and will be for some time) where I'm building simple things on a breadboard, using a variable DC bench power supply.

After a great deal of reading on the subject of 'scopes and grounding, I have concluded that the safest approach to 'scoping a DUT is to have it on an isolation transformer (but retain grounding at the 'scope).

I can't understand any way in which I can use the two (variable DC bench power supply AND the isolation transformer) at the same time.

I'm guessing the ISO is strictly meant to be used with AC-powered devices? If that's so, then how would one "float" a DC-powered device?

Thanks all in advance for any advice you can shed on this.

\$\endgroup\$
2
  • 1
    \$\begingroup\$ This video is a must watch for scope newbies: EEVblog #279 - How NOT To Blow Up Your Oscilloscope! youtube.com/watch?v=xaELqAo4kkQ \$\endgroup\$
    – tlfong01
    Commented May 18, 2022 at 1:52
  • \$\begingroup\$ Your DC bench power supply already has a floating output. Unless you ground it yourself, that is. \$\endgroup\$
    – Justme
    Commented May 19, 2022 at 4:15

5 Answers 5

4
\$\begingroup\$

Rarely do I find isolation is required when working on microcontroller projects on the desk. However it helps to understand where other earthing sources are. Eg via USB from a desktop computer. If powering your circuit via a laptop on battery, then that is isolated (unless you have another monitor or other separately powered device connected). If the laptop is powered by the ac adapter it might earth the laptop but more likely it will float but due to the ‘Y’ capacitors in the ac adapter, this will leak half mains voltage but at less than 1mA into your circuit. This can give you a tingle or damage your electronics. It also causes a large AC waveform to be displayed on the scope.

To avoid potential problems I separately earth the laptop via the USB connection in order to shunt the leakage.

Also note that just about any switch mode power supply that does not have a mains earth will exhibit the same leakage. The old iron transformers don’t have the leakage issue.

There’s plenty of questions regarding leakage and Y caps here and on the interwebs as it surprises many people. It is easily measured with a multimeter or a scope. Once you’re aware of it then you can take precautions.

\$\endgroup\$
6
  • \$\begingroup\$ Nice answer. I generally agree about the MCU projects and about the possible questions with class 2 power and modern laptops. I loved this: "The old iron transformers don’t have the leakage issue." Agreed! Heavy iron (200 kg transformer) is always good medicine. ;) Every home should have at least two. :) And as a rule if you can move it, it's too small. You should need a decent tractor or fork lift to move it. I use a JD 4230 that lifts 1500 kg using the front loader. Bigger and heavier is always better. \$\endgroup\$
    – jonk
    Commented May 18, 2022 at 8:04
  • \$\begingroup\$ Sorry. That's a JD 4320, not a JD 4230. Sheesh. To repent, here's me transporting an old hot tub. \$\endgroup\$
    – jonk
    Commented May 18, 2022 at 8:15
  • \$\begingroup\$ Don't know why jonk called it a "nice answer" It is totally non-responsive as regards my question. I want to know how/when to employ an isolation transformer and/or bench DC power supply. I am not at all interested in batteries or USB power. Thanks anyway for giving it a go, \$\endgroup\$ Commented May 18, 2022 at 13:11
  • \$\begingroup\$ @DrSmokeWrench Your input is taken. The video that 'tlfong01' posted up in comments does then cover pretty much everything I imagine you would need to have. Is there something from that video that fails to cover something you want? If so, I can't at this time imagine what. \$\endgroup\$
    – jonk
    Commented May 18, 2022 at 18:03
  • \$\begingroup\$ Well, my question concerns the interplay of two devices: the DC power supply and the isolation transformer. Do I plug the power supply into the transformer thereby defeating it's internal grounding? Do I not use them together? \$\endgroup\$ Commented May 18, 2022 at 22:26
1
\$\begingroup\$

Some dc power suppliers have outputs that are isolated from the line, others will have dc - bonded to earth ground (like a pc power supply) and some have switches or selector bars to choose one or the other.

When it comes to bench top DC work usually the voltages are low enough to not be considered hazardous, so isolation isn’t the biggest safety concern.

\$\endgroup\$
1
  • \$\begingroup\$ It's not about the low DC voltages... isolation is about isolating those low voltages from e.g. mains voltages. \$\endgroup\$
    – TypeIA
    Commented May 19, 2022 at 8:02
1
\$\begingroup\$

You have some valid concerns. Grounding will change as you get into RF etc but for normal work I use earth ground supplied by the third prong on the plugs. The DC stuff I ground my scope to the power supply "-" as that is generally my reference. That will automatically ground the "-" to earth with a line powered scope. I have the isolation transformer but use it mainly when working with mains voltages. I do not know what voltages you are working with so if you are using an isolated supply be sure it is rated at a high enough isolation voltage for what you are working with.

While you are waiting on your scope spend some time with your multimeter and ohm out your power supplies so you know what is happening. It is best to do this with them disconnected as you will want to probe the line plug. My power supplies have a shorting bar on front so I can elect to have them grounded or floating. This grounding thing applies to many but not all bench instruments. Your best tool will be your grounding strap.

If you are going to design and work on devices designed for the mains use 24VAC to do as much prototyping etc as possible. Doing it this way you do not have to worry about floating your scope and or getting a nasty zap.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ If you have any doubts (and you should if untested) check the difference between various "ground" points in your system. If you do analog work, this precaution is necessary from the sensor to the power supply. Even for digital circuits, differential grounds can route current where you don't want it; generally this is due to poor layout. If you ever feel "fuzz", carefully get rid of it! If you have a problem with these things, ask on here (or somewhere else); and include a circuit diagram and picture. Nine out of ten times, this type of problem is physical layout/implementation. \$\endgroup\$
    – rrogers
    Commented May 18, 2022 at 0:42
0
\$\begingroup\$

Let me share my experience working with breadboards, with DC and even AC voltage supplies (AC through a function generator).

I just connect the ground of the oscilloscope probe to the breadboard ground rail (that's where I also connect the negative terminal of the power supply), and then use the probe on any place in the breadboard to find the voltage.

I never do any kind of grounding on the oscilloscope itself. I simply connect the ground connection on the probe to the negative rail of the breadboard to arbitrarily setup a zero voltage point.

I've used voltages up to 18 V (9V and -9V negative) and have never had an issue.

I'm not an expert, but it just might be that the considerations about the transformer etc. are not at all necessarily for typical hobby work - I built audio amplifiers and also AM radio receivers, and never had an issue, even as the power consumption of the speaker (which is like an 8 ohm resistor) is quite significant.

\$\endgroup\$
6
  • \$\begingroup\$ your answer gets closer to the point (information I can both understand and use) -- but I believe I made clear that I want to "float" the device under test, NOT the 'scope. While your breadboard plan will work, it falls far short of a solution when I graduate to, say, working on an old integrated amp. I'll give it one last try -- given that I want to "float" the thing that I'm working on (to avoid possible damage to the 'scope) I've concluded that I need to use an ISO. transformer. It seems to me that that is only a solution if I want AC power, correct? \$\endgroup\$ Commented May 18, 2022 at 13:27
  • \$\begingroup\$ (continued due to text constraints) When I'm using a DC bench power supply, (almost all the units I've investigated are grounded via their plug) how do I isolate the "DUT"? Thanks for you answer. \$\endgroup\$ Commented May 18, 2022 at 13:28
  • \$\begingroup\$ You might be more advanced than me - but to my mind: if you're adding a chip to a breadboard (as opposed to discrete components), you'd want to power the chip through your supply (at it's VCC pin). But even if you're extending an already made PCB, I don't think you need isolate your power supply either. I believe you just need to make sure that the stage you're adding doesn't load the existing PCB. Example: an amp outputs a signal, you want amplify again. If the output is channeled through a high impedance path to the next stage, your power supply won't load the original stage. \$\endgroup\$
    – Daniel
    Commented May 19, 2022 at 2:41
  • \$\begingroup\$ Usually the base of a transistor is a high impedance path, and circuits that provide shielding between stages are aptly called buffers. But you can see the point simply with resistors: a 9V source connected to an earlier stage through a 22k resistors will hardly touch it. I highly recommend by the way the Falstad Circuit Simulator - a wonderful tool for learning & experiments. \$\endgroup\$
    – Daniel
    Commented May 19, 2022 at 2:44
  • \$\begingroup\$ Also adding a capacitor between stages block DC current and in effect isolating your DC power supply. \$\endgroup\$
    – Daniel
    Commented May 19, 2022 at 4:20
0
\$\begingroup\$

Making voltage measurements is a differential measurement. You use the scope to measure the difference between the scope ground and the probe. However, most scope ground/reference leads are also connected to the scope's AC ground pin connection (Earth ground.) When you connect that reference clip to the circuit, be aware you are connecting that node in the circuit to Earth. That could be a problem and may damage your circuit if that node has a voltage referenced to Earth other than zero volts. This can crop up with full bridge rectifiers or FET bridges for example.

You can avoid tying your circuit to the reference ground by making a differential measurement with two scope probes and setting the scope to subtract one channel from the other. Another option is a specialized differential scope probe which which has a + and - input, both of which are isolated from the Earth ground.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.