Bench Switching power Supply Cnecting Pos to Ng for Curent Reading?

[DW1961]

Can someone explain how that works?

You connect the positive to the negative and power the unit on, then dial in your current. How does the PSU prevent a direct short? Also, why not just turn on the power supply and dial in current without connecting P to N? What's the application of being able to connect P to N for a current setting?

Thanks.

[TimFox]

The current limiting (or constant-current stabilization on a precision power supply) regulates (adjusts) the output voltage to control the current through the load connected across the output terminals.
For an arbitrary load, either the voltage setting or the current setting will rule, so both the output voltage and output current cannot exceed the values set by the knobs.
Therefore, it is totally safe to short the output, forcing the output voltage to zero, to set the current on a “proper” constant-current supply.  If the knob is not calibrated, you need to do this to read the current on the meter.  With a finite resistance load, you might not know which of the controls is limiting the output.
Note that some linear power supplies (not test equipment) use “foldback” limiting that reduces the current limit into a short circuit to reduce stress on the pass element.  This has other issues, but makes the unit short-circuit proof.

[DW1961]

The current limiting (or constant-current stabilization on a precision power supply) regulates (adjusts) the output voltage to control the current through the load connected across the output terminals.
For an arbitrary load, either the voltage setting or the current setting will rule, so both the output voltage and output current cannot exceed the values set by the knobs.
Therefore, it is totally safe to short the output, forcing the output voltage to zero, to set the current on a “proper” constant-current supply.  If the knob is not calibrated, you need to do this to read the current on the meter.  With a finite resistance load, you might not know which of the controls is limiting the output.
Note that some linear power supplies (not test equipment) use “foldback” limiting that reduces the current limit into a short circuit to reduce stress on the pass element.  This has other issues, but makes the unit short-circuit proof.

So doing it like that instead of trusting the knob itself, is just a way to get the 'real' current? I guess one could also compare the shorted current with the knob's current to see if they are the same (calibrated)?

[TimFox]

All of my commercial bench power supplies have current-limit or constant-current controls, but none of them has calibrated knobs for current.  They have either two meters or a meter switchable between I and V.  My Keithley current supplies have a calibrated current control (decade switches) and an uncalibrated voltage compliance limit.  My precision voltage supplies have decade switches for voltage and uncalibrated knobs for current limit.
If your modern unit has a meter that can read accurately the current set-point, that doesn’t guarantee the actual current, so it’s good to check.

[DW1961]

All of my commercial bench power supplies have current-limit or constant-current controls, but none of them has calibrated knobs for current.  They have either two meters or a meter switchable between I and V.  My Keithley current supplies have a calibrated current control (decade switches) and an uncalibrated voltage compliance limit.  My precision voltage supplies have decade switches for voltage and uncalibrated knobs for current limit.
If your modern unit has a meter that can read accurately the current set-point, that doesn’t guarantee the actual current, so it’s good to check.

So connecting the P to the N is the way to do that?

[TimFox]

Yes:  connecting the positive output to the negative output through a low-resistance wire or a low-resistance ammeter means that, at any current within the range of the supply, the voltage across the outputs will remain very close to zero, and you can set the current through that wire to the value you want.  When you remove the wire, the output voltage will be the voltage you set and can be measured with a voltmeter.
For sake of discussion, set the current to 0.1 A and the voltage to 5 V.
If you connect a variable (high-power) resistor to the output after that and measure the voltage and current across and through it, at low resistance (R < 50 ohms for this example), the current is 0.1 A, and the voltage is that current times the resistance R.  For R > 50 ohms, the circuit "crosses over" to voltage control and the voltage will be 5 V, and the current through it will be 5 V divided by the resistance R.
In general, for ideal sources, shorting a current source and open-circuiting a voltage source are good, safe things.  If you open-circuit an ideal current source, the voltage goes infinite (bad thing), but for a practical case the voltage goes to the compliance limit, which might be dangerously high.

[DW1961]

Yes:  connecting the positive output to the negative output through a low-resistance wire or a low-resistance ammeter means that, at any current within the range of the supply, the voltage across the outputs will remain very close to zero, and you can set the current through that wire to the value you want.  When you remove the wire, the output voltage will be the voltage you set and can be measured with a voltmeter.
For sake of discussion, set the current to 0.1 A and the voltage to 5 V.
If you connect a variable (high-power) resistor to the output after that and measure the voltage and current across and through it, at low resistance (R < 50 ohms for this example), the current is 0.1 A, and the voltage is that current times the resistance R.  For R > 50 ohms, the circuit "crosses over" to voltage control and the voltage will be 5 V, and the current through it will be 5 V divided by the resistance R.
In general, for ideal sources, shorting a current source and open-circuiting a voltage source are good, safe things.  If you open-circuit an ideal current source, the voltage goes infinite (bad thing), but for a practical case the voltage goes to the compliance limit, which might be dangerously high.

The way they have it in the manual, you don't need to use a resistance element. You just connect the alligator P and N clips together. Concur?

[TimFox]

I said wire or ammeter, but connecting the clips to each other connects the positive and negative outputs through the wires to the clips.  That makes a low-resistance connection.

[DW1961]

I said wire or ammeter, but connecting the clips to each other connects the positive and negative outputs through the wires to the clips.  That makes a low-resistance connection.

OK got it. Makes sense now.