Line powered device with no ground questions

[Chris Carlen]

Hi:

Any tips I might want to consider for safety when building a gadget (my
Nixie clock) that will plug into 120-240VAC and use only a two-wire cord?

I have a 5x20mm 250VAC fuse in the hot before the transformer primary.

I even have pads on the PCB if I want to install a thermal fuse which
would cut the line voltage if the transformer core overheats due to an
overload that isn't enough to blow the line fuse, a failure mode which
could start a fire without the thermal fuse, but which is rarely
protected against in consumer electronics.

I am planning to put a 1Meg resistor from the signal GND to the neutral
of the power line. That's so any static discharges get dissipated and
the signal ground can't float with respect to earth, assuming of course
that the wall socket is wired with the correct polarity. But with 1Meg,
even if the polarity is backwards, it can't kill anybody.

But perhaps I should also put a cap from the signal GND to neutral?
Does it have to be a "safety" cap? What exactly is the point of safety
caps anyway? Are they just rated for high surge voltages and able to
tolerate continuous, albeit negligible, dissipation? The only thing I
can see it would do is in the event of a static discharge, it would
prevent a large peak voltage from developing on the secondary side.

I have plenty of creep distance, at least 10mm, from any line traces to
the low voltage traces.

Should I add any other surge protection or safety features?

Comments appreciated.


Good day!

 

[Frank Bemelman]

Hi:

Any tips I might want to consider for safety when building a gadget (my
Nixie clock) that will plug into 120-240VAC and use only a two-wire cord?

I have a 5x20mm 250VAC fuse in the hot before the transformer primary.

I even have pads on the PCB if I want to install a thermal fuse which
would cut the line voltage if the transformer core overheats due to an
overload that isn't enough to blow the line fuse, a failure mode which
could start a fire without the thermal fuse, but which is rarely
protected against in consumer electronics.

I am planning to put a 1Meg resistor from the signal GND to the neutral
of the power line. That's so any static discharges get dissipated and
the signal ground can't float with respect to earth, assuming of course
that the wall socket is wired with the correct polarity. But with 1Meg,
even if the polarity is backwards, it can't kill anybody.

Don't trust the polarity. On the European continent, plugs don't even
have a polarity, you never know what is neutral or hot.

But perhaps I should also put a cap from the signal GND to neutral?
Does it have to be a "safety" cap? What exactly is the point of safety
caps anyway? Are they just rated for high surge voltages and able to
tolerate continuous, albeit negligible, dissipation? The only thing I
can see it would do is in the event of a static discharge, it would
prevent a large peak voltage from developing on the secondary side.

I have plenty of creep distance, at least 10mm, from any line traces to
the low voltage traces.

Should I add any other surge protection or safety features?

Comments appreciated.

A wall wart solves a lot of these problems.

 

[Chris Carlen]

Don't trust the polarity. On the European continent, plugs don't even
have a polarity, you never know what is neutral or hot.

Yes, that would argue for having no connection between the two.

A wall wart solves a lot of these problems.

I don't like those things because they are very cheaply made, and so I
trust my own work better than them. I have a constant paranoia of a
wall wart bursting into flames one night, or while I'm away.

They are also clunky and want to fall out of the wall, forcing me to buy
power strips just for the strings of wall warts under the computer table.

I think for folks wanting to sell their electronics without having to
get certifications, they are great. But I will not be selling my work,
though making it available to anyone who wants to reproduce it.

Thus, I opted for a direct line connection. Oh, also I wanted access to
the AC for time, but I suppose I could have used an AC wall wart.

So tell me, what *should* be done for safety when building two-prong AC
line connected equiptment?

Thanks for the input.


Good day!

 

[N. Thornton]

Hi:

Any tips I might want to consider for safety when building a gadget (my
Nixie clock) that will plug into 120-240VAC and use only a two-wire cord?

Hi Chris. You cant rely on N being N and L being L - swaps are common.
Lots of places L/N arent even identified.

I have a 5x20mm 250VAC fuse in the hot before the transformer primary.

I even have pads on the PCB if I want to install a thermal fuse which
would cut the line voltage if the transformer core overheats due to an
overload that isn't enough to blow the line fuse, a failure mode which
could start a fire without the thermal fuse, but which is rarely
protected against in consumer electronics.

Its standard kit in Oman.... FW thats W.

I am planning to put a 1Meg resistor from the signal GND to the neutral
of the power line. That's so any static discharges get dissipated and
the signal ground can't float with respect to earth, assuming of course
that the wall socket is wired with the correct polarity.

Well, neither neutral nor live is earth, so it wont be at earth. I'd
use 2 Rs in series, and tap from the junction of them. Ensure adequate
Vrating on the Rs: class II in UK requires very high insulation test
V.

But perhaps I should also put a cap from the signal GND to neutral?
Does it have to be a "safety" cap?

Yes, Y rated.

What exactly is the point of safety
caps anyway?

safer Theyre required when youre connecting L to E, or L to user.

Are they just rated for high surge voltages

I believe so, and they're self healing, and more reliable.

and able to
tolerate continuous, albeit negligible, dissipation?

any cap can do that.

The only thing I
can see it would do is in the event of a static discharge, it would
prevent a large peak voltage from developing on the secondary side.

I have plenty of creep distance, at least 10mm, from any line traces to
the low voltage traces.

Should I add any other surge protection or safety features?

I dont know where youre locatde Chris, so I couldnt say. If youre in
Oman you'll need a thermal cutout, if youre in Kuala Lumpur youll need
a method of user electrocution, and if in India youll need to install
bureaucracy and politics. And for Russia you'll need gauzed vents to
protect against water splashes.


Regards, NT

 

[Tim Shoppa]

Any tips I might want to consider for safety when building a gadget (my
Nixie clock) that will plug into 120-240VAC and use only a two-wire cord?

Why have you ruled out a grounding plug and a 3-wire cord?

I am planning to put a 1Meg resistor from the signal GND to the neutral

Big mistake.

Does the clock you're building connect with any other equipment? Is that
other equipment grounded in any way? Does "other equipment" include an
antenna?

Tim.

 

[Chris Carlen]

Why have you ruled out a grounding plug and a 3-wire cord?

Originally because I planned to give one to my folks, who live in an old
house with 2-prong sockets, and some 3-prong sockets that my dad wired
up, and he doesn't know what he's doing.

Also, most consumer electronics is 2-prong.

Big mistake.

Please explain why. I understand that if L and N designations are not
reliable, then this might be an ineffective way to avoid a floating
signal ground, which might not really be a problem anyway.

Does the clock you're building connect with any other equipment? Is that
other equipment grounded in any way? Does "other equipment" include an
antenna?

Well, I have a TV, VCR, DVD player, stereo tuner, amplifier, CD player,
and various other consumer electronics devices all with polarized
2-prong plugs. So what is your point?

And the TV, VCR, and DVD player, all connect to cable or antenna systems.

Let's get specific. I want to know:

1. Is it undesireable to have a floating system? If so why, if not why
not?

2. What are sensible precautions to take with ungrounded equipment to
make it safe?


Thanks for the input.


Good day!

 

[N. Thornton]

2. What are sensible precautions to take with ungrounded equipment to
make it safe?

Just a couple for you:

tie the mains wires together in pairs so that if one connection comes
adrift it wont touch the case.

build it to very high insulation levels. Eg 4kV rather than 1kV.

I think follow those 2 and youve really covered most issues, if not
quite all.


Regards, NT

 

[R.Legg]

Any tips I might want to consider for safety when building a gadget (my

1. Is it undesireable to have a floating system? If so why, if not why
not?

It's more difficult to make, but probably more desirable to achieve,
particularly in simple devices that do not interconnect, mainly
because they can be used safely in all wiring systems.

2. What are sensible precautions to take with ungrounded equipment to
make it safe?

Place it in a non-conductive enclosure, with no external access to the
end-user. This requires carefull design of any adjustment knobs or
switches.

Assume that the isolated circuitry is also hazardous, and design
accordingly.

If you can hipot from line to an enveloping conductive wrap at 3KV,
without breakdown, AFTER performing all standard single fault
abnormals and a 24 hour, high humidity soak, you have demonstrated a
potentially safe two-wire-input-powered device.

This procedure will typically destroy a half-dozen samples in the
process, but all of them will have demonstrated integrity of the
insulation present, despite all inadvertent damage to the internal
contents.

If properly documented and witnessed, this will give you a clear
conscience and reduced liability if one of your parents assassinates
the other by tossing the clock into their bubble-bath.

If you do the tossing, you are still liable to face the consequences.

Nixie tubes, eh? I assume your parents have expressed deep longings
for the good old days, down at the SciFi/adventure film studio
back-lot? Or is it a problem with display visibility?

RL

 

[Tim Shoppa]

Originally because I planned to give one to my folks, who live in an old
house with 2-prong sockets, and some 3-prong sockets that my dad wired
up, and he doesn't know what he's doing.

Also, most consumer electronics is 2-prong.

More on this below

Please explain why.

If he doesn't have ground hooked up right, chances are 50/50 that he
has live and neutral reversed too.

I understand that if L and N designations are not
reliable, then this might be an ineffective way to avoid a floating
signal ground, which might not really be a problem anyway.

But even if neutral *is* neutral, the 1M resistor doesn't help any.

Well, I have a TV, VCR, DVD player, stereo tuner, amplifier, CD player,
and various other consumer electronics devices all with polarized
2-prong plugs. So what is your point?

And the TV, VCR, and DVD player, all connect to cable or antenna systems.

Ungrounded consumer electronics are popular for two reasons:

1. Most TV's have a "live" (well, nominally connected to neutral) chassis.
This is entirely to avoid the cost of a transformer in the power supply.
Even though the chassis is nominally connected to neutral, the design
requires complete insulation between the chassis and the outside of the
cabinet (and things that might get poked into the chassis) as the
assumption is that a good fraction of households have hot and neutral
reversed.

2. Lots of smaller consumer electronics are not grounded because typical
household grounds are so bad that ground loops will introduce 60 Hz hum
in consumer audio systems when components are interconnected. Or
hum bars for video systems. (Even then hum-bar eliminators - really
isolation transformers - are common accessories for cable boxes...)

1. Is it undesireable to have a floating system? If so why, if not why
not?

It doesn't make much difference. And again, my gut feeling is that
connecting "ground" to neutral in any way is a Bad Thing.

2. What are sensible precautions to take with ungrounded equipment to
make it safe?

Insulation. Isolation. If you've got a power transformer in your device
with no hot-wired ground-to-neutral connection then you have isolation
automatically.

Tim.

 

[Chris Carlen]

Place it in a non-conductive enclosure, with no external access to the
end-user. This requires carefull design of any adjustment knobs or
switches.

Assume that the isolated circuitry is also hazardous, and design
accordingly.

But doesn't this contradict the supposed reason why everyone just uses
wall-warts instead of line power? I mean, wall-warts are almost all
2-prong with isolated outputs.

If you can hipot from line to an enveloping conductive wrap at 3KV,
without breakdown, AFTER performing all standard single fault
abnormals and a 24 hour, high humidity soak, you have demonstrated a
potentially safe two-wire-input-powered device.

This procedure will typically destroy a half-dozen samples in the
process, but all of them will have demonstrated integrity of the
insulation present, despite all inadvertent damage to the internal
contents.

If properly documented and witnessed, this will give you a clear
conscience and reduced liability if one of your parents assassinates
the other by tossing the clock into their bubble-bath.

That is always possible.

Nixie tubes, eh? I assume your parents have expressed deep longings
for the good old days, down at the SciFi/adventure film studio
back-lot? Or is it a problem with display visibility?

It's intended as a gift. My grandma taught me that you should give
people things that you like. That way you share yourself with them. So
I gave my grandma a big box of nuts and bolts.

She loved it!

Also, the clock is something that has taken a lot of time and effort.
More suitable for a gift than some consumer gimmick junk.

Good day!

 

[Chris Carlen]

If he doesn't have ground hooked up right, chances are 50/50 that he
has live and neutral reversed too.




But even if neutral *is* neutral, the 1M resistor doesn't help any.

Yes, perhaps I will dump the resistor, but still not sure if I want the
ground.

Ungrounded consumer electronics are popular for two reasons:

1. Most TV's have a "live" (well, nominally connected to neutral) chassis.
This is entirely to avoid the cost of a transformer in the power supply.
Even though the chassis is nominally connected to neutral, the design
requires complete insulation between the chassis and the outside of the
cabinet (and things that might get poked into the chassis) as the
assumption is that a good fraction of households have hot and neutral
reversed.

Still today? Shocking! I see power trannys in almost all other stuff
though.

What is done for the inputs and outputs of TVs then? There must be some
sort of isolation.

Insulation. Isolation. If you've got a power transformer in your device
with no hot-wired ground-to-neutral connection then you have isolation
automatically.

Tim.

Well I do have a power transformer. Most folks are doing Nixie clocks
either to run off line voltage directly for the HV, or to use a wall
wart with a little SMPS booster for the HV. I hate wall warts, so I
opted for a fully isolated line powered transfomer power supply, with a
SMPS booster to get the HV from low voltage DC. A little cumbersome
perhaps, but I also had economies of tiling in mind. It is expensive to
make good 2-layer silkscreened PCBs as a hobbyist, so I wanted to
implement my flyback SMPS DC-DC converter project with the Nixie.

The dreadful possibility is what would happen if some bonehead hooked
the hot to the ground terminal of a socket. I hope my dad wouldn't do
that, but if someone does that, there's little that can help them.

Good day!

 

[Spehro Pefhany]

Still today? Shocking! I see power trannys in almost all other stuff
though.

It's a SMPS. But where's the transformer?
Hint: It operates at approximately 15.734 kHz.

What is done for the inputs and outputs of TVs then? There must be some
sort of isolation.

Yes.


Best regards,
Spehro Pefhany

 

[N. Thornton]

Tim Shoppa wrote:
Hi

or one that really is live, since it is connected to the output of a
full wave bridge rec. Funny how that catches many people out.

What is done for the inputs and outputs of TVs then? There must be some
sort of isolation.

Capacitance for the aerial socket. Most TVs here have no other in or
outs precisely to avoid safety prolems. If you make a TV with a row of
in/out connectors, youre going to need an isolated supply.

When I was doing some live chassis stuff I tried to come up with
fiendish arrangements to keep external connections safe while still
using no PSU isolation: I did figure out ways to do it, but not that
had an approvably high level of safety. They basically relied on
monitoring and a relay.

The dreadful possibility is what would happen if some bonehead hooked
the hot to the ground terminal of a socket. I hope my dad wouldn't do
that, but if someone does that, there's little that can help them.

Non grounded kit will be safe in that case.


Regards, NT