8
\$\begingroup\$

I recently received (as a birthday gift) a Rigol DS1052E oscilloscope. I read about oscilloscopes but I never really used one.

So my question is:

  • How can I measure the ripple voltage of a floating linear power supply, using this oscilloscope?

Can this be done with this oscilloscope (or the common mode noise will be too high, since the ground clip of the probe is tied to earth)? To get a fairly accurate reading would I need differential probes? Or can I use the two channels of the scope and add them together?

\$\endgroup\$
9
\$\begingroup\$

You should be able to safely measure the floating linear power supply, as long as you know what you're doing and you're sure that the supply is in fact floating.

So first step is to make sure that the supply is floating. It would be simple to use a multimeter to confirm that there isn't a conductive path between the power supply's rails and ground. Is that is true, you can simply connect oscilloscope probe's ground connector to some point in the circuit. Often this would be the circuit's negative line, but does not need to be.

If the power supply isn't floating (or to say more clearly is grounded), you'd have to connect the oscilloscope's probe ground to the grounded rail of the power supply. That would be usually the negative rail, but could be positive, so to be 100% sure, you'd need to confirm the ground connection with a multimeter.

Do note that once you've connected the ground clip of the probe to some part of the circuit, that part is now ground referenced! That is important, because the ground clip of the other probe is also connected to ground and if you touch another part of the circuit with it, you'll short it to ground, which could have very negative consequences.

Here's a diagram of usual probe connections inside of an oscilloscope:

ground connection

So if you for example connect the ground clip of one probe to negative rail of the supply and the other to positive, you'd have a short-circuit.

Now about the actual measurement itself:

First step would be to see if the probe can handle the voltages and to determine the appropriate probe setting. Usually 10x attenuation is used on probes, since that represents what is usually negligible load on the power supply and provides more bandwidth for the oscilloscope.

After that, connect the ground clip of the probe to the power supply and the probe tip to the point which you want to measure. Some sources recommend that the tested device be powered-off during the connection and that to me looks like a good idea since it minimizes the chances to make a short somewhere where it shouldn't be while connecting the probe. Once you connect the probe, check that the probe is properly connected and isn't touching anything it shouldn't be, like heatsinks (which might be connected to power supply's negative side).

Next, activate the oscilloscope and make sure that the probe attenuation factor is set at the same setting as seen on the probe. Next make sure that the probe coupling setting is correct. It shouldn't be set to ground and it should be set to DC. More about that is in the manual under To Set up the Vertical System.

Next step would be to set the oscilloscope trigger voltage for the connected probe a little bit higher (or lower) than the power supply's nominal voltage. This should make the scope trigger on ripple.

After that, turn on the power supply. You should be able to see a (more or less) flat line representing the supply's output voltage on screen and you may see some interference riding on that voltage.

Next part is a bit more difficult to explain and is a bit more experimental, but once you do it a few times, it will be easy.

The idea is to zoom in on the interference you see. You could try with automatic measurements and see how they work out. In case they don't show what you want to see, I'll explain how to do it manually. The whole story is explained in the horizontal and vertical settings part of the manual. Basically you use the scale knob to zoom in on the wave you see and then you use the position knob to set the wave on the center. I usually first adjust vertical settings, then horizontal and repeat the procedure until I can clearly see the ripple. Once you see it, you can measure the ripple using the graticule or you can use cursors. Cursor use is explained in the example 5 at the end of the manual for the scope and in the To Measure with Cursors section. When you're using graticule, you simply look up how much time or volts each division represents and then multiply the number of occupied divisions by the value you have. Cursor measurement will usually provide you with more precise result.

So far I haven't mentioned the math menu, because there is no need to use it. You definitely need to reference some point in the circuit to the oscilloscopes ground, since the scope does all measurement with respect to ground. If you connect one probe to the positive rail of the power supply and the second to negative and subtract them, you'll gt same result as if you measured against the probe clip's ground.

Do note that in the case of the isolated linear power supply, you can't get ground loop and have noise, since there will be no current going from power supply's ground through the oscilloscope's ground to the main ground, because the PSU itself isn't ground-referenced and there's no closed loop for current to go through.

A bit about AC coupling: As Vorac says, if you set the probe to AC coupling, you'll remove low frequency signals. This includes the DC component of power supply voltage, which will leave you with only the ripple. This way, you can avoid the need to use vertical position controls to bring the noise into view because it will be already centered to zero volts, so you can just zoom in on it.

Another handy thing are the trigger settings. You can also set filtering to triggering circuit, so that it will work on AC, DC, low or high frequencies. AC trigger coupling will remove all signals under 10 Hz from trigger circuit, so slow periodic signals won't interfere with trigger. LF reject will block all signals under 8kHz and HF reject will block all signals above 150 kHz. This can sometimes be useful if you're trying to focus on only one component of the signal and trigger on it.

\$\endgroup\$
  • \$\begingroup\$ +1 Great answer! Maybe a section about AC coupling could be added, to provide an alternative o the tedious zoom. \$\endgroup\$ – Vorac Jul 6 '12 at 12:56
  • \$\begingroup\$ @Vorac I was a bit afraid that it might remove some needed noise, but it only removes everything under 10 Hz here, so it's probably going to be OK. \$\endgroup\$ – AndrejaKo Jul 6 '12 at 13:19
  • \$\begingroup\$ @AndrejaKo Thank you for this great answer. I have one more question: can I rely on the ripple measurements done with this scope (Rigol DS1052E)? I mean isn't the internal noise of this unit too high for ripple measurements? \$\endgroup\$ – Buzai Andras Jul 8 '12 at 12:51
  • \$\begingroup\$ @Buzai Andras Well, that depends on how good reading you want. I haven't actually read much about internal noise problems with this oscilloscope, so I can't tell if it's a big problem or not, but the unit seems to be well-recommended, so it probably wouldn't be a big problem. Do you have any sources mentioning the internal noise problems? Also you need to tell us more about the linear power supply itself. Is it some special low-ripple type? What ripple voltage do you expect? \$\endgroup\$ – AndrejaKo Jul 8 '12 at 16:02
  • \$\begingroup\$ @AndrejaKo In the process of learning electronics I started to design my own power supply. Nothing special, just the standard LM317 type :). So I was thinking of measuring the ripple for it. I would expect of a 10-15mV (peak to peak) ripple. Thank you. \$\endgroup\$ – Buzai Andras Jul 9 '12 at 10:40
1
\$\begingroup\$

If I could refer you to the manual, page 2-92. This is the interface for automatic measurement of {frequency of dominant harmonic, min value, max value, amplitude} among others. I think you need to measure the amplitude peak to peak.

On the other hand, you can "Add, Subtract and Multiply Mathematic Functions" from the Math button (Mathematic Functions == two different channels signals). I think you do not need this, however.

Lastly, a word of waning (though you know this and even mention it in your answer, I would like to stress it). As the ground clip is tied to earth, you should never measure the wall power without an isolation transformer. Your power supply is floating, so this is not a concern this time.

\$\endgroup\$
  • \$\begingroup\$ never measure the wall power without an isolation transformer What? The whole point of the ground connection on the oscilloscope is to make sure that the device is safe to use and should never be circumvented, and especially not for this reason. \$\endgroup\$ – AndrejaKo Jul 6 '12 at 11:26
  • \$\begingroup\$ Well, the zero clip of the measurement probe is grounded. In my layout, wall ground is short circuited to wall zero (bad bad - but many people have it this way). Thus, connecting the clip to a hot wire results in a short. Again, proper wiring requires three-wire power lines, but at many houses this is not the situation. \$\endgroup\$ – Vorac Jul 6 '12 at 11:53
  • 1
    \$\begingroup\$ Since the OP has floating power supply, connecting the grounded probe clip to a positive line shouldn't create a short. Also even if the building's ground isn't zeroed but connected to real ground, you'd still have a short with non-isolated power supply if the probe ground is connected to positive since the zero line of the building power is actually referenced to ground. \$\endgroup\$ – AndrejaKo Jul 6 '12 at 12:09
  • \$\begingroup\$ @AndrejaKo, could you please explain further. The way I see it the clip is shorted to wall zero through the oscilloscope. So touching a live wire creates a short {live wire - zero} and a light show. Good point that building zero is grounded. \$\endgroup\$ – Vorac Jul 6 '12 at 12:21
  • \$\begingroup\$ I may have to bring a picture here, so before I begin drawing, I'd like to confirm that I understood you correctly: You want an explanation why when you connect the probe clip to the positive rail of a floating power supply, you don't get a boom, light-show and a liquid ground clip cable, right? \$\endgroup\$ – AndrejaKo Jul 6 '12 at 12:23

protected by Community Oct 27 '15 at 7:58

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

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