Reading -564mv on power supply (Owon SP3051) when turned on and output is off

[aaruff]

I recently purchased an owon SP3051 and tested the output voltage when the supply is turned on and the output is turned off and I get a reading of -564mv (as shown below), instead of zero volts. I was told this could be due to the probe leads acting as antenna and the terminal capacitor and diode acting as an AM receiver. I have moved the power supply to areas and I get a constant negative voltage of about -564mv. I have also connected my multimeter to the power supply using the shortest possible wires to limit the RF that the probes could be picking up and the voltage is the same. When I power off the power supply the capacitor (and diode) connected to the terminals discharges to zero volts.

Are there any methods for confirming if this is RF interference or possibly something else? I placed an order for a cheap software defined radio in the hopes of determining if there is some RF causing the issue. I'm fairly new to electronics so I am stumped as to how to remedy the problem. Any advice is greatly appreciated.

[bob91343]

Why is this a problem?  My guess is that it's a normal reading perhaps a diode forward voltage.

In many supplies like this, there is a diode across the output terminals for protection.  You might remove that diode to see what happens.

[Vovk_Z]

What is an input impedance of that Fluke? What if there will be connected something with a smaller resistance - 1 MOhm, or 100 k resistor?

[bob91343]

In other words, the output would probably go slightly negative and the protection diode conducts a little to prevent that.  This is not a precision voltage source; it's just a utility power supply and you shouldn't complain about a few millivolts error.

[aaruff]

Why is this a problem?  My guess is that it's a normal reading perhaps a diode forward voltage.

In many supplies like this, there is a diode across the output terminals for protection.  You might remove that diode to see what happens.

I guess my concern is that the reverse voltage may cause an issue with the components, though unlikely. I'm not complaining, instead I am trying to understand if this is normal. I have seen many reviews where a measurements across the terminals when output is off is at or near zero volts. Is -600mv discrepancy small, or does it hardly negatively affect circuitry? I felt like it wasn't since it's more than half a volt.  At what price point can one expect zero volts when the output is off?

What is an input impedance of that Fluke? What if there will be connected something with a smaller resistance - 1 MOhm, or 100 k resistor?

The fluke has a 10MOhm impedance. All of my other meters also have a 10MOhm internal resistance so I'll have to do a bit of googling to figure out a way to do the measurement bypassing the internal impedance.

[TimFox]

You don't want to bypass the internal impedance.
Just connect a medium resistance (say, 10 k ohms) across the power supply binding posts after turning it off.

[aaruff]

You don't want to bypass the internal impedance.
Just connect a medium resistance (say, 10 k ohms) across the power supply binding posts after turning it off.

I connected a 10k Ohm resistor to the negative and positive terminal of the power supply, then plugged in the multimeter into the power supply terminals for measurement (as in the photo below). The measured voltage was still - 563mv. I'm still perplexed.

Why would the output have a tendency to go negative when the output is off? I thought the output button would work like a single throw single pole switch. Is there a reason why it's not?

Thanks!

[bob91343]

Perhaps dielectric absorption of the output filter capacitor.

[TimFox]

What does the output On/Off switch actually do, physically?

[m k]

How it goes with the green jack?

[radiolistener]

My KORAD KA3005D also has -0.0288 Volts on the output when output is off. Interesting that when output is on with 0.0 Volts setting it doesn't have such negative Voltage.

[Damianos]

It seems that the output on-off is not a switch (relay contact) but an "active zero" operation. Check what it gives at the output when it is set for zero volts, with the output on and off and with and without a load.

Did this device used as the manual may suggest? For example: did this used with a battery connected and the output off or a low voltage setting? In this case, if it is not intended for this use, something inside maybe is cooked.

Study the manual and whatever information (such as application note or anything) is given to the user, there may be something helpful.

PS: I found a manual but it is very poor, it not contain even a specifications table!

[aaruff]

When the voltage and current is set to zero I measure -564mv across the terminals. Yet, when I set the voltage to zero and increase the amperage to 5mA I get a reading of -2mV. A voltage of 10mv and 5mA results in 15mV. I think I will just add a switch to my breadboard every time I use the power supply and turn the output on first, then toggle the switch, so my circuit isn't reversed biased by half a volt when output is off.

Thanks for all your help.

[Vovk_Z]

Just don't set current to zero. There isn't much sense to do that. That's not the real case for using a supply, that's why designers don't bothered to do something with this exact situation (V=0 and I=0 simultaneously).

[Marco]

The power supply seems to be setting voltage and current to zero as an alternative to having a proper output relay for turning the output off.

So Owon did it wrong twice. I assume they have a bias current to a negative supply rails for stability reasons and at 0V/0A and output off it stops counteracting that current.

[bob91343]

It's also possible that there is an internal adjustment to set the zero point.

[Marco]

It's not a voltage offset error, it disappears once he increases the current a bit.

[TimFox]

In general, for a feedback-stabilized supply to go to zero, the circuit must be capable of going slightly negative.
Often, this is done with a reasonably-high resistor to a suitable negative supply.
If the supply turns "off" the output by setting the programming on both voltage and current to zero, I agree that that is an inappropriate design.
The current through the current shunt may have to supply a small amount of current towards the negative supply to reach zero output voltage, and turning the current control to zero may stop this current totally.

[Damianos]

... ...
If the supply turns "off" the output by setting the programming on both voltage and current to zero, I agree that that is an inappropriate design.
... ...

This is the "stupid" point! They set also the current to zero, the current sensing takes over the control, by reading its own offset and drives the output to negative voltage. The result is the forward voltage of the reverse diode at the output, that is heating without a reason. If the output diode is a "normal" rectifier type, then it seems that the reverse voltage is of a source with a weak current capability, but if it is a Schottky type then it is a "stronger" source.

@aaruff:
Notify the seller or the manufacturer (or both) for the results of your tests. There may be a solution of software update or unit exchange (if it is under warranty).

[fischerd]

I also have purchased 2 Owon SP3051 supplies.  Have measured the terminal voltage, unloaded, with the output ON; voltage pretty much equal with the voltage setpoint.  Have loaded the output with a 0.5W, 220 ohm resistor.  Measured terminal voltage matches the displayed terminal voltage which matches the setpoint voltage, provided the setpoint current is above the actual current.  When turning the output OFF, both supplies put out a negative voltage; if the 220ohm resistor is across the output, the voltage is -0.381V, if the resistor is removed, the output is about -0.567V.  Based on this, the Thevenin equivalent resistance is about 100ohms.  So, no, the output voltage is not noise pickup; the output impedance of the "pickup" is too low to be a true pickup.  As others have noted, the output is not switched via a relay, but via an active circuit, which isn't that well designed, hence the -0.5V residual voltage.  Should this voltage be here? No.  Could it have been better designed?  Yes; a relay would have done the trick.  Am I going to loose sleep over this?  No; the -0.5V can hardly turn on a diode (silicon), so whatever circuitry is powered by this supply, will see -0.5V and I won't even hear a big yawn coming from it....  If I spent 4 times the price on another unit, would it have had a smaller residual output voltage?  Maybe, probably.... Do I care?  Yes, but I won't loose any sleep over this.... ;-)