1
\$\begingroup\$

I was just testing this since I blew up a Raspberry Pi when hot connecting its ground to another ground (or minus) of another power source.

Now I tried connecting the rectified output ground of which the source is a variac with a switching power supply ground (or minus). Both devices are in a socket that has a ground pin but the variac is not connected with it. Having some ~45 volts on the rectified (with 400uF cap) variac output and the switching power supply set at around ~12 volts, I connected the grounds and something blew up in the switching power supply (it still seems to work although I could smell something in there burned).
I read some questions on this matter here and I still wonder what I did wrong. I connected these 'grounds' together before and it worked fine (driving some large capacitive load with the variac using an RPi output pin for control on a large transistor.

  1. Is it necessary for both power supplies to start at 0 V when connecting their ground and turning any voltage knobs?
  2. What may have been blown inside my power supply that it is still functioning? My guess is some output filter cap that briefly took some voltage from when they were still floating compared to each other?
  3. What I also saw is a huge (~120V) AC (50Hz) difference between the 2 negative terminals before connecting them. I guess this is normal and will not produce any current after connecting.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
10
  • 7
    \$\begingroup\$ Voltage difference causes current flow. If you measured 120VAC between two wires and assumed that connecting them to each other was fine, I think that you should practice on some battery powered stuff first before playing with non-isolated line-connected equipment. \$\endgroup\$
    – vir
    Commented Nov 15, 2022 at 20:33
  • 1
    \$\begingroup\$ "Ground" is seldom an absolute potential, and when there's more than one conductive path between one piece of equipment and another, their grounds can be at slightly different potentials. Let those touch, and there will be a current that could interfere with either or both devices, or even cause damage. en.wikipedia.org/wiki/Ground_loop_(electricity) \$\endgroup\$
    – Solomon Slow
    Commented Nov 15, 2022 at 20:43
  • 1
    \$\begingroup\$ Re, "...Variac..." A Variac is an autotransformer. It's output is not isolated from the mains. You want to be extra careful with those. \$\endgroup\$
    – Solomon Slow
    Commented Nov 15, 2022 at 20:47
  • 4
    \$\begingroup\$ I guess this is normal and will not produce any current after connecting. .... totally the dangerous way of thinking .... every part of the circuit you presented is deadly \$\endgroup\$
    – jsotola
    Commented Nov 15, 2022 at 21:14
  • 2
    \$\begingroup\$ there is only a diode separating the DC portion of the circuit from the powerline ... it will drop the powerline voltage by 0.6 V on alternate cycles ... touching a grounded device with one hand and touching the circuit with the other hand will apply full powerline voltage across your body ....that is why you should not be working with a circuit that does not provide full isolation from power line voltages ... cannot really produce any current is foolish thinking \$\endgroup\$
    – jsotola
    Commented Nov 15, 2022 at 21:58

3 Answers 3

Reset to default
3
\$\begingroup\$

Based on the diagram, the variac does not seem to isolate and output is dangerous to connect anywhere since it is mains referenced.

The node marked with question mark will peak at -325V in reference to earth or any earthed device like ground of desktop PC (assuming 230VAC mains).

It is hazardous to touch and should not be connected to other devices.

Yes, if it did not explode at first, the neutral and live might have been swapped in regards to the picture but it's still wrong.

\$\endgroup\$
2
  • \$\begingroup\$ When working with sensitive equipment that should have their ground included, is the right way to do this only when everything is switched of or at least turned to 0 volt? Otherwise there may be a brief current to 'unfloat' them with respect to each other, right? \$\endgroup\$
    – Niels
    Commented Nov 15, 2022 at 22:17
  • \$\begingroup\$ Right now we are looking at a variac with output happily able to electrocute people from the supposedly marked "ground" terminal. This will blow up anything when turned on, even if connections were made de-energized. If you connect e.g an earthed PC to an earthed monitor, they are safe to connect as they both are already at same potential. Same can not be said about equipment with 2-prong mains input. \$\endgroup\$
    – Justme
    Commented Nov 15, 2022 at 22:24
3
\$\begingroup\$

According to all the very helpful reactions I tend to assume the following is considered safe in regards to connecting the two negative terminals (I'm not touching anything in these experimental setups). If anybody still has concerns, please let me know.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
2
  • 1
    \$\begingroup\$ This ^^^^^^ Get an isolation transformer \$\endgroup\$
    – Kyle B
    Commented Nov 16, 2022 at 3:45
  • 1
    \$\begingroup\$ I did and it is all working as I thought it would :) I put it before the variac though, as it can handle 'only' 3 A, so at that side, the voltage is highest \$\endgroup\$
    – Niels
    Commented Nov 26, 2022 at 15:10
2
\$\begingroup\$

It is only safe to connect to ground (not neutral and not hot) on an AC system, even if it is through a rectifier. it appears you are connecting to 'hot' (some kind of translation difference)

You need a step down transformer that is power limited to make it safe.

Another thing you might want to consider is optoisolators to do any switching between circuits.

\$\endgroup\$
9
  • \$\begingroup\$ But every simple wall adaptor with a transformer and rectifier in it works also fine in this setup. I can connect there negative poles together... So is the rectified variac different? \$\endgroup\$
    – Niels
    Commented Nov 15, 2022 at 19:58
  • \$\begingroup\$ I may be talking about negative terminals then, indeed. Not ground. One negative terminal from the rectified variac and one negative terminal from the switching power supply. \$\endgroup\$
    – Niels
    Commented Nov 15, 2022 at 19:59
  • 1
    \$\begingroup\$ @Niels - The "simple wall adapter" such as a typical USB charger is fully isolated from the AC line. So when you connect the grounds between two of them it is just fine and safe ... and low voltage. Your variac is NOT isolated from the line. \$\endgroup\$
    – Michael Karas
    Commented Nov 15, 2022 at 20:16
  • 1
    \$\begingroup\$ @Niels "What I also saw is a huge (~120V) AC (50Hz) difference between the 2 negative terminals before connecting them." And yet.... you persisted. You're lucky all you did was blow up your Raspberry Pi, this sort of setup can be deadly. An auto-transformer such as a variac is not isolated and shouldn't be used as a 'power supply'. They should generally be used with an isolation transformer, either a properly rated one before the variac, or the variac feeding a transformer and not directly into a rectifier. \$\endgroup\$
    – GodJihyo
    Commented Nov 15, 2022 at 20:51
  • 2
    \$\begingroup\$ When you see 125V, but you think it's a "phantom voltage" (which DOES occur, especially when using a high-impedance digital meter)... Just connect 10k or 100k across that, see if the voltage goes away. If it does, then you're right and it's a meter trick. If it doesn't, watch out.... That voltage is "real" \$\endgroup\$
    – Kyle B
    Commented Nov 16, 2022 at 4:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.