Author Topic: Power supply ground question  (Read 3372 times)

0 Members and 1 Guest are viewing this topic.

Offline simpson

  • Contributor
  • Posts: 48
  • Country: us
Power supply ground question
« on: January 12, 2016, 04:25:28 pm »
I just bought a used BK Precision 1601 Power Supply in good condition and have downloaded the user manual. I have 2 questions about how the ground  terminal on the front of the unit is used.

This is the what the manual says:

1. I am confused by the highlighted part of instruction #6 in the manual which says “if the [positive/negative] polarity of the device being powered is also to be ground reference ...”.

Huh? So, when would I need to jumper the appropriate output terminals to ground as it says? In other words, when are ground reference voltages needed and when are they not needed?

2. If I am powering, say, an opamp circuit that requires positive and negative supply voltages  plus a ground can I use the 3 appropriate terminals on the power supply for this purpose? If so, would each supply (pos & neg) be considered to be half of what the supply output meter is showing?


Offline SebG

  • Regular Contributor
  • *
  • Posts: 66
  • Country: ca
  • Currently looking for an EE Job
Re: Power supply ground question
« Reply #1 on: January 12, 2016, 04:52:15 pm »
Basically the ground terminal is earth ground (ie. it is connected directly to the ground plug of the wall socket)  This seems like a single voltage power supply so you cannot get a split voltage without a resistor divider driving a voltage follower op-amp for example.  The ground terminal is there to help shield your device being powered from external noise, if it has a metal box or something like that.  The output from the supply is isolated from the mains 120V AC or whatever so you can earth ground reference either the positive or negative voltage if needed, say if you had 2 of those supplies then one can be the positive voltage and the other set up as a negative voltage with the common (for both voltages) being earth ground.

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 1049
  • Country: us
  • Retired, now restoring antique test equipment
Re: Power supply ground question
« Reply #2 on: January 12, 2016, 05:24:48 pm »
Think of the unit as a battery connected between the Pos and Neg terminals of the front panel, with the case connected to the Gnd terminal.  This allows you to ground one of the two terminals, but it is not necessary.  Of course, there is a specified maximum insulation voltage that can be applied between any of the terminals and case ground.

Offline Gyro

  • Super Contributor
  • ***
  • Posts: 5262
  • Country: gb
Re: Power supply ground question
« Reply #3 on: January 12, 2016, 05:33:43 pm »

Edit (accidentally hit the post button):

It's always a good idea to connect the ground terminal to one of the output terminals (almost always the negative in practice) unless you specifically need a floating supply. Even with a linear power supply there is some capacitance between the primary and secondary of the mains transformer that can cause some very low current parasitic voltage to appear on (not across) the output terminals. In an unlucky case, such as a sensitive fet input, this might just be enough to damage your circuit - think of it in the same way as basic ESD protection. You can even use a resistor if worried about ground loops.
« Last Edit: January 12, 2016, 05:43:30 pm by Gyro »

"Victor Meldrew, the Crimson Avenger!"

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 14700
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Power supply ground question
« Reply #4 on: January 13, 2016, 07:05:11 am »
Typical use cases:

1. Powering a breadboard. Probing with oscilloscope might be grounded, but that's a momentary contact, and at the same ground anyway (give or take ground loop, which is only important for sensitive circuits).  Ground the 'common' side of the supply.
2. Testing or operating an audio amplifier.  The signal input will most likely be coming from a grounded system (e.g., computer audio out), and the supplies may be bipolar or floating with respect to ground, preventing either side being directly grounded.  You don't need a ground in this case, or ground can optionally be connected with a third cable (as noted).
3. Powering a floating circuit.  Suppose you are testing a circuit where an isolated section requires power, and which is intentionally elevated from ground (an example might be a gate driver circuit, which is referenced to either the switching node, or the supply negative, and where neither is common to the reset of the circuit).  Direct grounding is impossible in this case.  Connecting a third ground cable to the circuit's ground (wherever that might be) probably wouldn't accomplish much, either.  (The circuit under test should still be grounded somewhere, at least if it's handling more than 42V.)

"Common" is most often the negative side of single supply circuits, or the middle supply of bipolar circuits.  "Common" is just a convention, and doesn't have an absolute definition (because voltage is a difference by definition).  Typically, a circuit is designed so that all its signals are with respect to common, so that power supplies, probes, meters, shields, etc. can all be used relative to common. 

Note that most power supplies, function generators, etc. are not usually very good for use on rapidly varying reference voltages, like high side gate drivers.  There is usually a substantial amount of capacitance, at least due to the capacitance between windings in the power supply transformer, but often including additional filtering (Y type) capacitors, especially for those that use switching power supplies.

Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!

Offline simpson

  • Contributor
  • Posts: 48
  • Country: us
Re: Power supply ground question
« Reply #5 on: January 14, 2016, 04:59:35 pm »
Thanks, all.  That helps.

Offline Jeroen3

  • Super Contributor
  • ***
  • Posts: 3403
  • Country: nl
  • Embedded Engineer
Re: Power supply ground question
« Reply #6 on: January 14, 2016, 07:06:23 pm »
Try to measure AC voltage and current (uA) from any of the output terminals to the ground terminal.
If you do not connect any terminal to ground, this voltage, and as a result the current, will find different ways to ground.
- Your scope.
- Your jtag programmer.
- Your usb logic analyzer.
- You.
You might not want to use your pc/scope as lowest impedance ground path. Sometimes you have to, if the EUT requires to be floating.
The leakage should be incredibly small since it's a bench supply, but other isolated supplies might not have the high quality components your BK Precision enjoys.
Such as used in your PC or Laptop or wall-wart. But the same concept applies, and is often underestimated.
« Last Edit: January 14, 2016, 07:09:07 pm by Jeroen3 »

Offline simpson

  • Contributor
  • Posts: 48
  • Country: us
Re: Power supply ground question
« Reply #7 on: March 08, 2016, 12:21:29 pm »
Forum member w2aew just uploaded a new video on this subject (thanks):


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo