Chinese DIY Kits: how to power them without batteries?

[picofarad]

Hello everybody, a bit of time since I wrote last time. 

So, recently I got into those DIY kits from China (always the same: function generator, mini scope, frequency counter, component testers...). Very fun to build and they also works quite decent, at least for an amateur.
I've always powered them using standard 9 volts batteries; now I would like to power them more permanently, so using a power supply. My first though was to use a small wall wart switching power supply, cheap ones from Amazon.
Here's my question: we all know that switching power supplies have floating voltages. In many dc power supplies I've measured 80-90V AC between ground and DC positive lead. Ok, it's a very small current, but I would like to use those kits with my lab equipment, especially my analog scope. And, from what I know, there's always a problem when an oscilloscope referenced to ground probes something that isn't, with several tens of volts AC floating. How can I solve this problem?
I know that using a linear wall wart power supply would solve the problem, but now they seems to be impossible to find!

Thanks for the future answers!

[rhb]

You can build a good linear supply out of scrap using an LM317.  Old transistor radios are a good source of suitable parts.

Another option is ground  through Y capacitors.  Find a suitable box with a 3 wire grounded plug (e.g. old computer PS).  Stick the wall wart inside, connect the wall wart to the power leads and  ground both output  leads using Y capacitors.  The metal foil fail open rather than short, so those are to be preferred over the ceramics.  Add three output terminals on the case.  I'm not entirely sure if you need two Y capacitors.  I need to test that.  One may do.

You can also use an old computer PS if you provide a small load so it doesn't shut down. that will give you +5 +12 and a wimpy -12.

I just recently became aware of the AC voltage on wall warts.  As someone pointed out to me in the FeelTech FY-6600 thread, it's a small capacitor connection to the mains. It just needs similar a capacitor  to ground.

I suggest doing all three.  It's cheap, educational and provides useful tools.

Have Fun!
Reg

[Peabody]

I had no idea this was a problem.  I just measured AC volts (on my multimeter) on the 9V wall wart I use with my Chinese kit scope, and it shows zero volts AC.  How do you go about measuring this 80-90V AC that you're seeing?

It's things like this that make me realize how little experience I actually have.

[ez24]

How about big batteries   ie  18650 ?

[Brumby]

I had no idea this was a problem.  I just measured AC volts (on my multimeter) on the 9V wall wart I use with my Chinese kit scope, and it shows zero volts AC.  How do you go about measuring this 80-90V AC that you're seeing?

Find yourself a properly earthed point - such as the chassis of an appliance that uses a 3 pin mains plug that has the earth pin connected to that chassis.  Place one probe from your meter on this chassis and the other on one of the output pins (low voltage side) of the wall wart (doesn't matter which).  There, you will see the voltage.

The impedance of this voltage source is quite high, so if you use a DMM with the typical 10M input impedance, you will likely get around the voltage mentioned.  If you were to use an older moving coil meter at the classic 20Kohm/V sensitivity, then expect a much lower reading.

[jonovid]

How about big batteries   ie  18650 ?

I use a set of sealed lead acid batteries 6 and 12 volt  then add 2 or more diodes to get were i need to be  in 0.6 volt steps .
did this to stop  ground loops

[picofarad]

I had no idea this was a problem.  I just measured AC volts (on my multimeter) on the 9V wall wart I use with my Chinese kit scope, and it shows zero volts AC.  How do you go about measuring this 80-90V AC that you're seeing?

It's things like this that make me realize how little experience I actually have.

Are you sure you're measuring between mains ground and DC positive terminal?
Having 80 volts AC it's normal for a switching power supply, as usually the current is quite small (1 mA at worst). It's not a problem when you are powering a single device (if you measure AC between DC terminals, you should read 0 volts AC), but it becomes a problem when you want to connect it to other equipment referenced to ground.

You can also use an old computer PS if you provide a small load so it doesn't shut down. that will give you +5 +12 and a wimpy -12.

I just recently became aware of the AC voltage on wall warts.  As someone pointed out to me in the FeelTech FY-6600 thread, it's a small capacitor connection to the mains. It just needs similar a capacitor  to ground.

It's a recent discovery for me too, and I'm getting a EEE degree!
I think I have some Y capacitors salvaged from a old TV, I'll try this fix. Thanks!

[imidis]

For antything that will run on 5v I just used usb and power them via power bank

[rhb]

It's a recent discovery for me too, and I'm getting a EEE degree!
I think I have some Y capacitors salvaged from a old TV, I'll try this fix. Thanks!

Y capacitors are "special" because of the safety factor, so make sure they really are.  Any capacitor of suitable size will have the desired result.  If you're using a wall wart, any ceramic rated at 200 V or better will do.  After all what will you lose if it shorts, a wall wart? It only matters if you use it for something that runs when you're not there  to see the magic smoke come out and put out the fire.

They don't have time to teach you a *lot* of things at the undergraduate level.  If you really want to understand electronics you need an MS and lots of bench time repairing stuff or a PhD.  Better yet, a PhD and lots of bench time like the Signal Path host.  Just the learning the math takes an undergraduate degree.  The good news, bad news of modern life.

[Peabody]

Find yourself a properly earthed point - such as the chassis of an appliance that uses a 3 pin mains plug that has the earth pin connected to that chassis.  Place one probe from your meter on this chassis and the other on one of the output pins (low voltage side) of the wall wart (doesn't matter which).  There, you will see the voltage.

The impedance of this voltage source is quite high, so if you use a DMM with the typical 10M input impedance, you will likely get around the voltage mentioned.  If you were to use an older moving coil meter at the classic 20Kohm/V sensitivity, then expect a much lower reading.

Ok, I get about 35VAC on my 10M digital meter.  My analog meter shows no needle movement on any AC scale.

I guess this is where lack of formal training shows itself.  I don't understand where this AC comes from.  I popped the case on my 9V wall wart, and found pretty much what I expected to find - Mains into FWBR and some capacitors, FSD210 controller, transformer, feedback via an optocoupler.  So what's the source of this mystery voltage?  Is it at mains frequency or switching frequency?  And if the transformer and optocoupler aren't enough isolation, why would a linear supply be any better?  I haven't checked, but take it from the discussion that a linear supply would not have this problem.  Why not?

I think I'm in the opposite situation from the OP.  I'm using the wall wart to power my kit scope.  The devices I've probed so far are either battery powered, or powered by my laptop's USB, with the laptop charger plugged in.  So I guess I should see if the laptop charger also has this AC problem - although unlike the wall wart, it does have a three-prong plug at the mains.  Anyway, I don't understand whether I need to do something different to power the scope, or if using the wall wart is ok despite this ethereal 35VAC.

[grifftech]

You can build a good linear supply out of scrap using an LM317.  Old transistor radios are a good source of suitable parts.

Another option is ground  through Y capacitors.  Find a suitable box with a 3 wire grounded plug (e.g. old computer PS).  Stick the wall wart inside, connect the wall wart to the power leads and  ground both output  leads using Y capacitors.  The metal foil fail open rather than short, so those are to be preferred over the ceramics.  Add three output terminals on the case.  I'm not entirely sure if you need two Y capacitors.  I need to test that.  One may do.

You can also use an old computer PS if you provide a small load so it doesn't shut down. that will give you +5 +12 and a wimpy -12.

I just recently became aware of the AC voltage on wall warts.  As someone pointed out to me in the FeelTech FY-6600 thread, it's a small capacitor connection to the mains. It just needs similar a capacitor  to ground.

I suggest doing all three.  It's cheap, educational and provides useful tools.

Have Fun!
Reg

instead of LM317 use LM78xx if possible

[edavid]

And, from what I know, there's always a problem when an oscilloscope referenced to ground probes something that isn't, with several tens of volts AC floating.

It is usually no problem at all.  Once you connect the ground of the kit to the ground of the scope, which you have to do to make any measurement, there is no longer any voltage, just a tiny current flowing.

However, you can use a 3 wire line cord, and connect the negative side of the wall wart output to ground.  Then you won't have to connect the scope probe ground clip to get rid of the voltage.

[jeroen79]

instead of LM317 use LM78xx if possible

Why?

[grifftech]

instead of LM317 use LM78xx if possible

Why?

no set resistor needed, simpler

[picofarad]

Ok, I get about 35VAC on my 10M digital meter.  My analog meter shows no needle movement on any AC scale.

As I said, the current is quite small. You'll need an analog meter with a 10 µA-like movement to see it.

I guess this is where lack of formal training shows itself.  I don't understand where this AC comes from.  I popped the case on my 9V wall wart, and found pretty much what I expected to find - Mains into FWBR and some capacitors, FSD210 controller, transformer, feedback via an optocoupler.  So what's the source of this mystery voltage?  Is it at mains frequency or switching frequency?  And if the transformer and optocoupler aren't enough isolation, why would a linear supply be any better?  I haven't checked, but take it from the discussion that a linear supply would not have this problem.  Why not?

I think I'm in the opposite situation from the OP.  I'm using the wall wart to power my kit scope.  The devices I've probed so far are either battery powered, or powered by my laptop's USB, with the laptop charger plugged in.  So I guess I should see if the laptop charger also has this AC problem - although unlike the wall wart, it does have a three-prong plug at the mains.  Anyway, I don't understand whether I need to do something different to power the scope, or if using the wall wart is ok despite this ethereal 35VAC.

The topic is quite broad, and as I said it's not even mentioned in a Bachelor Degree course; I leave the explanation to someone who knows more about it. I figured out this "problem" it's mainly on IEC Class II devices, so double insulations and working currents low enough that no earth connection is required.

From what I've seen in my lab, the AC voltage is at mains frequency, with a waveform that looks like a distorted sine. For what concern the AC amplitude, keep in mind that what you're reading with your DMM is RMS voltage, so in your case 35 V rms = 49 V peak (V peak = ?2 * V rms).

[rhb]

instead of LM317 use LM78xx if possible

Why?

no set resistor needed, simpler

And not adjustable!

Come on!  Does it *really* matter if it's 5 parts or 6?  Besides, the obvious thing to do is add LM78xxs for the fixed voltages. so there goes the part count.

[Brumby]

Ok, I get about 35VAC on my 10M digital meter.  My analog meter shows no needle movement on any AC scale.

That's pretty much what you can expect.  The analog meter needs more current, so the voltage drops significantly.  The fact that it dropped to "unmeasurable" is not at all surprising.

I guess this is where lack of formal training shows itself.  I don't understand where this AC comes from.

In a word - capacitance.  Every conductor in the world (and beyond) has a capacitance to every other - it's just that most of the time these are so small they are ignorable.