Switching DC power supply behavior questions

[Maletrain]

I usually test the output of things I get before I use them, and I have been seeing some things I don't understand from some switching DC power supplies.

After finding some outputs I did not expect on some new things, and before asking this question, I went back and tested one that I have used for a long time and trust with my Power Cab, which is its 14 volt 1.44 amp NCE power supply that came with my Power Cab.  It puts out 14.03 DC volts open circuit, and I just recently tested it on AC and found 2.05 volts AC, periodically flashing to 19.7 volts AC. It does that with an open circuit or when powering the Power Cab and a PSXX breaker and a Pro Cab, which is about 0.15 amp draw.  Since I have been using this for years, it seems like this is an acceptable result.

Comparing that to a different, older 1.11 amp NCE power supply that came to us as a donation, I got open circuit readings of 13.43 volts DC and the same 2 volts AC periodically flashing to 19.6 volts.  So, similar behavior.

I started all of this when I bought this https://www.ebay.com/itm/176576945795 and checked its outputs at the various voltage levels.  At 12 volts, the output is 12.04 volts DC with 2 volts AC, but no periodic flashing to a higher AC reading.  At lower voltage settings, the AC component seems to decrease somewhat proportionately, but it is not linear.

So, is this 2 volts AC superimposed on the DC a normal thing for switching DC power supplies?  Are "DC only" circuits OK with that?

FWIW, these voltage measurements were made with a "Micronta 22-171A" multimeter that I got so long ago I have no idea where it came from.  But, it has been pretty reliable and accurate for my various uses.  Even on DCC, it reads 14.78 volts where my RRampMeter reads 13.9 volts - so not calibrated for true RMS at that frequency, but not way out of the ballpark, either.

[jagged ben]

You're not actually seeing AC voltage.  There's just something about the switched DC that the AC measuring circuitry is picking up and misinterpreting.  Possibly some high-frequency characteristic of the DC waveform that resonates with the meter circutry designed to expect the neighborhood of 50-60 Hz.  But who knows what algorithm the Micronta is using to analyze the AC waveform and why it's confused.  Probably not worth trying to figure out and not worth putting any stock it what it says on the AC setting.

[jagged ben]

DC loads should be fine with switched DC power supplies with rare exceptions.  Like, if it's driving a sensor (such as an RR Cirkits OIB-8) you may need to add filter capacitors and/or use software debounce to get the sensor to ignore the high frequency variations and be 'on' whenever the power is applied.  Or see the probably-usually-overthinking-it discussions of DC motors and some power supplies with voltage 'spikes'.   But something like Power Cab, if it's not complaining, is doing just fine.  Probably has enough filter capacitors and other stuff between the power supply and any important components or outputs that it just doesn't care.  With LEDs if you can't see it it doesn't really matter.

[MK]

Micronta is the line carried exclusively by Radio Schack.  Like Kobalt by Lowes, Pittsburg by Harbor Freight, Craftsmen by Sears, etc.

[peteski]

I agree that there is no way that there is a 2VAC ripple on the power supply's output (or 19V spikes). That woudl cause the the devide using power from that supply to not be very happy.  The meter is likely the culprit.

TBH, I don't recall ever trying to use a multimeter to measure AC voltage at the DC output from a switching power supply.

Yes, that is a RadioShack meter.  I might have the same meter, but I can't find the part #.  Mine has a 2-color case. Gray on the front and black on the back, and 2 rotary knobs for selecting mode and range.

[Maletrain]

Yes, it is the potential for the output to "cause the the device using power from that supply to not be very happy," that caused me to post the question.

Does anybody have an NCE PowerCab switching wall wart and a scope?  It would be interesting to see what it actually puts out.

[Point353]

Yes, that is a RadioShack meter.

http://www.oddmix.com/elec/elt_rs_22171a_micronta_dmm.html

[peteski]

http://www.oddmix.com/elec/elt_rs_22171a_micronta_dmm.html

Yeah, I have that one too.  Cool thing about this meter is that it has 3 ³/₄ digit resolution instead the standard 3 ¹/₂ digit.

If I have some time this weekend I'll see if I can check my Power Cab wall-wart.

[mmagliaro]

99.9% sure it's just an anomaly of the meter.  Unless it's 60 Hz AC, the "AC voltage" readings on multimeters are not reliable uinless you spend the bucks to get a "True RMS" one, and even those can be fooled by erratic signals or high frequency noise.

Can you put a scope on the output?  That's the way to know for sure if there's really a couple lf volts of AC coming out of there.

[Maletrain]

I don't have a scope, so I was hoping that somebody who does could check the output of some switching power supplies.

I had a strange event a while back when I put an NCE amp meter and an RRampmeter in series with my PowerCab track output to see how their current readings compared.  When I put a light bulb that draws about an amp on the tracks for a load, took some readings and then disconnected the bulb from the track rails, the Power Cab started flashing its display.  The wall wart usually shuts off before the PowerCab gets to 2 amps and uses its own circuit breaking feature, so I had never seen that happen before.  It kept going until I threw a switch to disconnect the wall wart from the wall outlet.

Trying to figure out how that could have happened, I am wondering if there is some high frequency AC from the switching power supply that resulted in what was effectively a buck-boost circuit with the other stuff down the current path to the rails and that light bulb.  Suddenly disconnecting the 1 amp load is what set off the whole event. So, a suddenly collapsing current seems to have been the trigger.

Inquiring minds want to know - but maybe not enough to risk a PowerCab in further experiments.

[peteski]

I'll do the analysis if I have time this weekend. I'm was busy today and spending most of the day at a model car show on Sunday taking  photos of the models.  I did a quick test last night and there is no appreciable AC signal on my Power Cab wall wart.  While my wall-wart might not be identical to yours, I don't think there would be much difference.

The voltmeter you're using is auto-ranging, and the range switching speed is rather sluggish.  I think that might be part of false readings.  The meter could be getting "confused".

[Maletrain]

Good thought about the auto ranging feature. 

So, I tried the same meter on a stack of 10 D cell alkaline batteries attached to nothing electronic. The result was the same as with the switching power supply: 14 volts DC and 2.04 volts AC periodically jumping briefly to 19.something and back.

So, that is a problem with my meter.  There is definitely no AC component in my stack of D cells.

For kicks, I also put my RRampMeter on the switching power supply.  In one polarity, the meter would not light up, which is normal for DC.  In the other, it showed 13.4 volts, with both the DCC and AC LEDs lighted, which means DC is present. But, without a battery in the RRampMeter, it would not read a low voltage AC ripple, and I have no idea what it would do with a really high frequency AC component if that came through the switching supply, anyway.
 
So, I am still interested in seeing how smooth the output of a switching power supply really is, if somebody will put one on a scope.

And, I am going to have to get another meter to use for checking for AC in the outputs from DC power supplies.  It may not be so easy to find a meter to do that.  I tried a digital manual-ranging meter and also a manual ranging meter with the old needle meter, and both showed me about 15 volts DC but about 30 volts AC continuously on that stack of 10 D cells. 

My Fluke Model 36 did show proper DC voltage, and on both the D cell battery stack and the NCE switching supply, when in AC mode, starts with a voltage reading of around 2 volts AC, but that diminishes over a period of many seconds down to 0.1 or 0.0 if I wait long enough.  But, that Fluke is not a "pocket meter", so I was hoping for something smaller that will read zero volts AC on batteries.

I am learning to trust nothing until I test it myself.

[peteski]

And, I am going to have to get another meter to use for checking for AC in the outputs from DC power supplies.  It may not be so easy to find a meter to do that.  I tried a digital manual-ranging meter and also a manual ranging meter with the old needle meter, and both showed me about 15 volts DC but about 30 volts AC continuously on that stack of 10 D cells. 

I really don't think measurement AC component of a DC power supply (switching or analog) is really needed or useful (unless there is some sort of an issue). I don't even recall that in my 40+ years of electronics experience I  have ever done that just for kicks (when everything works properly). And if I did, I used a scope.

As I mentioned did a quick test of my Power Cab wall-wart using my scope and the AC component on the DC output was around 200mW peak-to-peak riding on the 13.8V DC voltage output.  As I mentioned I will take some photos of the scope's screen in the next few days and post them (time allowing).

[mmagliaro]

If it can be any switching power supply, I've got a bunch of laptop power bricks I can scope.  Those are all switchers.  I'll put one on the scope and take a look.

[Maletrain]

I sometimes do have a reason to check for AC on things like train DC power packs.  The club gets a lot of junk in "donations" that we need to check.  We have found a few that don't even completely stop locos when turned to zero speed, and some have actually made the DC locos buzz.  Some  have some AC on the external medal cases.  So, yes, we check.