I recently purchased 3 channel lab dc power supply and I can't figure out one thing yet. There are three terminals on the output - + - and ground. When I test in an oscilloscope by connecting + and - I get a given voltage. But when I connect for example positive voltage and ground I get some AC signal.

Like this one: enter image description here

I compared it with my another static 12v AC to DC converter (Doepfer A-100 DIY kit PSU) which has 3 terminals +12v / -12v / and Ground. And when I connect for example +12v and ground it gave me +12v and +12v and -12v obviously gave 24 linear voltage.

So my question is it possible to get the similar behaviour on the lab DC supply? So common ground will stay on 0v. I tried to connect +24v and -24v on a breadboard to get 0v but in this case power supply switching to the current mode. Maybe I should use different mode for it as it has Parallel track, Series track and Channel track (I'm currently using Independent Output mode).

Doepfer A-100 DIY kit PSU Images: enter image description here

enter image description here

DC Lab supply: Owon ODP3033 enter image description here

Thank you.

  • 2
    \$\begingroup\$ Put a small wire between "-" and "GND" on your power supply. Currently it's "floating"... i.e. it has no voltage reference to earth ground until you give it one. The AC you see is what's basically "in the air" (inductive pickup from the wires around you). Most power supplys will include a small bar to make this connection as it's a very common thing to do. Your o-scope is already at earth ground. Here's a pic of a PS that has this bar installed... onelectrontech.com/wp-content/uploads/2020/08/… \$\endgroup\$
    – Kyle B
    Nov 24, 2022 at 0:58
  • \$\begingroup\$ Thanks @KyleB this is really nice tip! But in this case ground became -12v (or any voltage that set on -) and not 0v. For me it's really important to get 0v as ground reference. \$\endgroup\$
    – Kirill
    Nov 24, 2022 at 1:12
  • \$\begingroup\$ The Owon user manual should describe how to get the device into so-called plus-minus mode which does the connections internally and enables you with an easy UI to change the voltages. \$\endgroup\$
    – Justme
    Nov 24, 2022 at 5:46

1 Answer 1


Perhaps this will clarify. But before starting I will comment on the concept of "zero volts". The desire to have an absolute zero volts is understandable but cannot be found. The connection of zero volts to the ground in electrical distribution systems is purely for safety reasons. But if the voltages in a cell phone are to be measured, using the earth as a ground produces meaningless results but they are voltages none the less. So another point within the cell phone becomes the zero volts for the purpose of describing the the local voltages. Usually the negative terminal of the battery is chosen.

Zero volts should be considered as an arbitrary reference. Where the black wire of the voltmeter is placed becomes zero volts for the duration of the measurement.

Trying to force ground (earth ) to be zero volts leads to the question in the Original Post.

Let me model the Doepfer A-100 and the Owon ODP3033 as several batteries. I can deduce that the ground symbol shown in one picture of the Doepfer A-100 is not actually connected to ground because you indicated that you measured 24V from -12 to +12. So the ground symbol is being used to indicate the common point to be used as a reference zero, not actually ground.

The circuit in figure 1 shows how two batteries can be connected to represent the Doepfer A-100 voltage sources. The negative of one is connected to the positive of the other and is labeled COMMON. This node is represented by the ground symbol in the picture.

Voltmeter VM1 uses COMMON as zero volts.

VM2 uses NEG as zero volts measuring COMMON as +12V.

VM3 also uses NEG as zero volts measuring POS as +24V.

VM4 uses POS as zero volts measuring COMMON as -12V.


simulate this circuit – Schematic created using CircuitLab

The circuit in figure 2 shows how three batteries can be used to represent the Owon ODP3033 voltage sources. Notice two things:

  1. The voltage sources are floating, completely isolated from anything else. The red terminal (POS1) is more positive than the black terminal (NEG1).
  2. The green terminal is the chassis connection that is ultimately connected to ground through the power wiring.

You may have been expecting the chassis terminal to be a common terminal as in the Doepfer A-100.

This is not the case. It is the chassis connection (to earth) and nothing more.


simulate this circuit

Figure 3 demonstrates how to configure the bench supply to represent the Doepfer A-100. Notice that the Chassis (ground) is not connected to COMMON in order to preserve the floating nature of the Doepfer A-100. The thick red line is a short wire between the two terminals.

You can use the series tracking feature to do the same thing. In this case the red wire is not required as the connection is made by the tracking feature. Read the manual to understand how the feature works.

Finally the ac signal that you measure from the red terminal to chassis is meaningless. If you want an earth connection to a particular case, then connect the green terminal appropriately. The "ground" lead of the oscilloscope is exactly earth ground through the ac mains. With the scope and power supply both plugged into mains a dc path between the scope ground lead and the green terminals can be measured.

Hope this helps

  • \$\begingroup\$ Hello @russellh, thank you so much for such detailed reply! It's really clarify a lot of things for me. And illustrations are really makes sense. I tried to use series mode according to manual and made some measurements like so: I set mode to series, voltage to 12v, connected oscilloscope ground to CH1 positive connector (as it's should be 0v common) And CH1 negative gives -6v which is looks correct. But when I checking voltage on CH2 positive returns around -9v, which is odd. Does it mean that I measuring it incorrectly, or something wrong with the unit? Thanks in advance. \$\endgroup\$
    – Kirill
    Nov 24, 2022 at 10:55
  • \$\begingroup\$ Quick update: I tested your suggestion with wire between Channel 1 + and Channel 2 -, and it's working as expected! \$\endgroup\$
    – Kirill
    Nov 24, 2022 at 11:05

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.