Connecting separate systems to a common ground

[yanir]

Hi, I would like some advice on how to do the following:
System A: 110vac input half wave rectified and 5v Zener to microcontroller.

System B: 3.3v micro controller powered from ac adapter (110vac to 5vdc, 5vdc to 3.3v).

I want to use nfets to short some pins of system A micro to ground by system B.

Full disclosure: I already attempted this and killed system A when I clipped my oscilloscope ground clip to ground. ( I connected gnd on system A to ground on system B which worked fine until my ground clip entered the picture. Was connecting these grounds together completely wrong?

Thanks 


Sent from my iPhone using Tapatalk 2

[vk6zgo]

Hi, I would like done advice on how to do the following:
System A: 110vac input half wave rectified and 5v Zener to microcontroller.

System B: 3.3v micro controller powered from ac adapter (110vac to 5vdc, 5vdc to 3.3v).

I want to use nfets to short some pins of system A micro to ground by system B.

Full disclosure: I already attempted this and killed system A when I clipped my oscilloscope ground clip to ground. ( I connected gnd on system A to ground on system B which worked fine until my ground clip entered the picture. Was connecting these grounds together completely wrong?

Yes,if you are using 110vac direct from the Mains,as neither side of this is really at the same Earth potential as your Osciiloscope ground clip.

The Mains Neutral & Earth are connected together where the supply line enters your building.
If we lived in a perfect world,the Neutral & Earth connections to your power socket would be connected back to this point with ideal connections with zero resistance.

In the real world,both connections possess resistance.
Normally,no current flows in the Earth circuit,but it does flow in the Neutral line,causing a voltage drop along that line,so that the potential with respect to  the point where Earth & Neutral are linked,& between the Earth & Neutral  pins of your power socket may be several volts.

That is the "Not so bad" scenario,where the Neutral line ends up being your "ground" in System A.

The "nightmare" scenario is when you get the Active & Neutral "back to front",so that the Active line becomes your "ground".

In this case,when you  connect the grounds,they both become  110vac 'Hot" with respect to Neutral,& to any Earthed object,including the case of your Oscilloscope!
This will not be at all healthy if you touch the "A" or "B" ground.

In this case,you will get a big Zap! if you place the 'scope ground clip on one or both of those "grounds"

[Rufus]

That is the "Not so bad" scenario,where the Neutral line ends up being your "ground" in System A.

Until you break the neutral connection, even momentarily when plugging in or out and your ground becomes live through your power supply. I would say "as bad", just not so often.

[vk6zgo]

I'm used to switched power points.

In any case.It is a bit different to a light bulb or a transformer primary.
If the power supply is still intact,the current is limited by that supply.

You'd still get a bite,though,& it definitely not recommended!

[yanir]

Thanks for the feedback. In this situation System A is not mains earth referenced. See attached for an example schematic.

System B is also not mains earth referenced. I used an isolated adapter for system B. I connected System B's ground to System A's gnd at the VSS pin of the micro in the attachment. I got into trouble when I clipped the ground clip of the probe to a USB connector ground shield on System B (which was obviously System B ground). Still having trouble seeing how I had a difference in potential from GND to mains earth.

I just watched Daves 'How not to blowup your Oscilliscope video (have been meaning to for sometime, better late then never...). As usefull as this video is, it doesn't totally apply to this scenario since none of my devices were reference to mains earth. What happened?

BTW The Zener in system A appears to be shorted now (but not a fuse inline with AC L), I'm hoping that replacing the zener will revive it.

[madires]

Why is system A is not mains earth referenced? Isn't neutral connected to earth at your mains panel, pole transformer or substation?

[yanir]

Why is system A is not mains earth referenced? Isn't neutral connected to earth at your mains panel, pole transformer or substation?

Ah! Thank you. I get it.

[vk6zgo]

Thanks for the feedback. In this situation System A is not mains earth referenced. See attached for an example schematic.

System B is also not mains earth referenced. I used an isolated adapter for system B. I connected System B's ground to System A's gnd at the VSS pin of the micro in the attachment. I got into trouble when I clipped the ground clip of the probe to a USB connector ground shield on System B (which was obviously System B ground). Still having trouble seeing how I had a difference in potential from GND to mains earth.

I just watched Daves 'How not to blowup your Oscilliscope video (have been meaning to for sometime, better late then never...). As usefull as this video is, it doesn't totally apply to this scenario since none of my devices were reference to mains earth. What happened?

BTW The Zener in system A appears to be shorted now (but not a fuse inline with AC L), I'm hoping that replacing the zener will revive it.

That is a seriously horrible circuit!

You have used what I described as the "nightmare scenario".

Your "ground" at Vss is connected to the Active line via D2 & R.
If you then short Vss to Mains Earth,both those components are connected in series across 110vac .

D2 will almost certainly be destroyed,& may go open or short circuit.

In the latter case,it is "the gift that keeps on giving",as when the Earth is removed,D1 is now fed raw AC,& will almost certainly go open circuit,so high voltage AC will appear across C1,destroying it.& the MCU.

Luckily,R will limit the current to some extent,but all the damage will already have been done.

[yanir]

That is a seriously horrible circuit!

I know, and I found two app notes (one from microchip and one from ST ) that explain how to do it! It's common in cheap household appliances.

Appnotes attached. I took the schematic I posted above from the ST appnote. Microchip's example at least has Neutral connected to ground.

Getting an isolated powersupply for the lab...

[vk6zgo]

Looking at the circuit again without my initial recoil of horror,R,at 18k should have allowed you to get away with shorting Vss to Earth,as it should restrict the current through D2 to 6.11mA rms,or 8.64mA Peak.

As you seem to have killed the Zener,& my suggestion in my previous posting seems to be the only way that would happen,it does seem that D2 has failed,--maybe its reverse voltage rating was too low.

All that said,the circuit would work just as well with the "ground" on the neutral side--what the designer was thinking of is beyond me!

These sort of circuits are acceptable embedded inside a double insulated  device hard wired to the mains,but are an "accident looking for somewhere to happen" on the experimenter's bench.