5
\$\begingroup\$

I am currently building a nixie clock with some high frequency components (GPS) and I am planning on putting the clock in a metal enclosure. The highest voltage used in the clock is approximately 510V (referenced to CLOCK GROUND). There are three different grounding configurations of the metal enclosure I am contemplating. I want to discuss the safety considerations of each and ultimately determine which is the safest / best practice option (or if the correct / safest option is something else and not one I have listed below!). I was unable to find a post which discussed the safety considerations of all of these options.

Below are a list of the 3 grounding options I am considering with accompanying schematic representations for each option. Each schematic has the connection in question highlighted in yellow.

EDITED: Images updated to explicitly show EARTH connection for clarity.

Option 1: Connect the metal enclosure to the CLOCK GROUND


Schematic representation of Option 1, with connection in question highlighted in yellow

Since I will have some SMA connectors mounted to the metal enclosure, if I do not insulate these connectors from the metal enclosure of the clock, they will connect the metal enclosure to CLOCK GROUND. This option will allow the CLOCK TRANSFORMER to act as an isolation transformer and also will prevent large voltage differences from building up between the metal enclosure and the internal electronics of the clock. However, I am concerned that this does not follow the common practice of connecting the metal enclosure of mains powered equipment to EARTH.

Pros:

  • CLOCK TRANSFORMER acts as an isolation transformer.
  • Prevents large voltage differences from building up between the metal enclosure and the internal electronics of the clock.

Cons:

  • Does not follow the common practice of connecting the metal enclosure of mains powered equipment to EARTH. The metal enclosure may build up a large potential difference with respect to EARTH, leading to a shock if someone touches the metal enclosure.

Option 2: Connect the metal enclosure to EARTH


Schematic representation of Option 2, with connection in question highlighted in yellow

If I choose this option, I would insulate the SMA connectors (which are CLOCK GROUND) from the metal enclosure of the clock. This seems to be following the common practice of connecting the metal enclosure of mains powered equipment to EARTH. However, should I be concerned that the internal electronics of the clock may build up large potential differences with respect to the case?

Pros:

  • Follows the common practice of connecting the metal enclosure of mains powered equipment to EARTH.
  • CLOCK TRANSFORMER acts as an isolation transformer.

Cons:

  • Large voltage differences may build up between the metal enclosure and the internal electronics of the clock.

Option 3: Connect the metal enclosure to both EARTH and CLOCK GROUND


Schematic representation of Option 3, with connections in question highlighted in yellow

This option prevents the internals electronics of the clock from building up large voltages with respect to the case and, at the same time, follows the common practice of connecting the metal enclosure of mains powered equipment to EARTH. However, this option eliminates the safety of the CLOCK TRANSFORMER acting as an isolation transformer.

Pros:

  • Follows the common practice of connecting the metal enclosure of mains powered equipment to EARTH.
  • Eliminates the possibility of large voltage differences building up between the enclosure and the internal electronics of the clock.

Cons:

  • CLOCK TRANSFORMER is no longer acting as an isolation transformer. Touching an internal electrical node in the clock (such as CLOCK VDD) could result in a serious electrical shock.

It appears that each option poses some advantages, but still has some safety flaws. I suppose working with mains power is inherently dangerous, however, is there one option that stands out as significantly safer than the other two?

\$\endgroup\$
7
  • 3
    \$\begingroup\$ Can you try it without worrying about the "telephone pole transformer"? \$\endgroup\$ Sep 3 '21 at 19:14
  • 3
    \$\begingroup\$ 1) I hope you realize that the GPS's antenna needs to be outside the metal case otherwise it will not receive anything. 2) The general rule for AC powered equipment with a metal case which can be touched by the used is that the case needs to connected to mains earth, that rules out option 1. Option 2 or 3 are both OK regarding safety. I would opt for the "only one ground" option so that would be option 3. \$\endgroup\$ Sep 3 '21 at 19:48
  • 3
    \$\begingroup\$ Do you have 3-prong AC outlets? If yes, you should show all three for clarity. \$\endgroup\$
    – Mattman944
    Sep 3 '21 at 20:50
  • 1
    \$\begingroup\$ I would argue that the baseline for consideration should be option 3. From there on, you can analyze whether the Earth connection with the Clock Ground is indeed harmful and if so proceed to carefully isolate the Clock Ground more and more, e.g. via ThePhoton's answer. \$\endgroup\$
    – tobalt
    Sep 3 '21 at 21:17
  • \$\begingroup\$ @Bimpelrekkie: you are correct regarding the GPS antenna; the SMA connector is to attach an external GPS antenna. \$\endgroup\$
    – hlove
    Sep 5 '21 at 3:46
20
\$\begingroup\$

Option 4: bond the enclosure to earth, and connect it to the clock ground with a high impedance, maybe 1 megohm.

This prevents large voltages building up between the internal circuit and the enclosure, and at the same time limits the damage that can be done by an electric shock from the internal circuit.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ This makes the most sense to me. High voltage differences within the box imply significant surface static charges at nodes at either end of the HV power supply. But also then the possibility of charge accumulation, too. The metal case must have everywhere an electric field of zero inside the metal shell itself, but any accumulation of charge inside the box will be countered by charge at the inside surface of the box and therefore also then reflected oppositely again at the outside surface of the box. Having something to bleed off any internal circuit charge accumulation is a good idea. +1 \$\endgroup\$
    – jonk
    Sep 3 '21 at 21:33
  • \$\begingroup\$ This makes a lot of sense to me, and I think a very elegant and safe solution! Thank you for the suggestion. I will likely also add a small ceramic capacitor to shunt high frequency noise, as suggested by Tony Stewart EE75 below. \$\endgroup\$
    – hlove
    Sep 5 '21 at 3:54
  • \$\begingroup\$ It is worth putting two or three high-valued resistors in series, just in case one gets short-circuited for some reason. This would be strictly required if the resistors were the only safety device, but in this case it is just a good practice. Also make sure that the resistors have voltage rating for the highest voltage present in the device. \$\endgroup\$
    – jpa
    Sep 5 '21 at 14:01
8
\$\begingroup\$

From a safety standpoint the enclosure should be connected to earth, not to the neutral leg of the power supply. You should have a three prong plug with the enclosure connected to the ground pin. Now anybody touching the enclosure is safe assuming the wall plug is properly grounded. Even if it isn't you need another fault to make the enclosure dangerous. You don't get to count on the neutral wire being at earth ground.

\$\endgroup\$
2
  • \$\begingroup\$ Very good point. I have updated the images to explicitly show the EARTH connection, in addition to the hot and neutral wires. Thanks for the catch! \$\endgroup\$
    – hlove
    Sep 5 '21 at 3:51
  • \$\begingroup\$ Safetywise, either 2 or 3 is fine now. I haven't done RF design, but I was told the ground on the secondary side needed to be connected somewhere. We had DC in and connected it to the incoming ground, not earth. \$\endgroup\$ Sep 5 '21 at 3:57
5
\$\begingroup\$

You can power this with a 9V and a boost regulator, both from the same low voltage low VA transformer over a wide range. All you need is a 12Vdc wall supply input with DC and a ground ESD wire to bleed off any charge on a metallised, plastic case or metal case.

There is no need for a big 1:1 transformer.
You do not need over 500V, only 190 Vdc for std Nixie tubes

enter image description here

The case can be double insulated or earth bonded with AC input or DC input with an ESD ground wire with 1 Mohm R and small ceramic C to shunt RF noise. The Resistor will bleed off static charge in a double insulated case that can be spray coated on a plastic box with static dissipative paint. But if you prefer a metal box , 3 wire earth bonding is must with fusing on power.

\$\endgroup\$
1
  • \$\begingroup\$ Thank you for the input! I omitted most of the information regarding the nixie clock in the original post since I was mostly concerned about the safest way to ground the enclosure. The clock uses the 60Hz mains frequency to keep track of time. Additionally, I am using a Dekatron as an additional display, which is why 510V is needed. As you and The Photon suggested, I am planning to use a ~1M Ohm resistor in parallel with small ceramic cap to connect CLOCK GROUND to EARTH. I also will put a 1A fuse right after the CLOCK TRANSFORMER (in series with the secondary). I appreciate your comments! \$\endgroup\$
    – hlove
    Sep 5 '21 at 4:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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