Arc gaps on circuit boards?

[larrymoencurly]

On old circuit boards I'd regularly see traces with little pointy areas to act as arc gaps, usually at the AC input or where low voltage signals connected to the outside world. Why do I no longer see those arc gaps very much?

[tom66]

Re: Arc gaps on circuit boards?

I see them quite regularly on modern PCBs. It's a cheaper way of reducing the arc which occurs due to the line filters.

[DaveB1972]

Re: Arc gaps on circuit boards?

Can you post a picture of what you mean? I cant figure out exactly what you guys are talking about.

[Mad_Professor]

Re: Arc gaps on circuit boards?

Can you post a picture of what you mean? I cant figure out exactly what you guys are talking about.

Same here, I never even heard of an arc gap.

[shovenose]

Re: Arc gaps on circuit boards?

I third that - no clue what's being discussed.

[mariushm]

Re: Arc gaps on circuit boards?

See first picture in this page, circled in blue: http://www.eevblog.com/forum/beginne...s-about-a-pcb/

Follow the posts below for explanations..

As to why not seen anymore these days... i couldn't tell you a specific reason.

[shovenose]

Re: Arc gaps on circuit boards?

Ah, now I get it. Maybe because power is much better these days?

[larrymoencurly]

Re: Arc gaps on circuit boards?

Here's what I mean, the little pointy areas on the copper traces, between the input and ground traces. The screw connectors are for the wall buttons of a 20-year-old Stanley garage door opener. I've seen similar pointy areas on the circuit board of a power supply from an original IBM PC, on the 120VAC side between ground and hot and between ground and neutral.

[Mad_Professor]

Re: Arc gaps on circuit boards?

So what exactly is their purpose?

[larrymoencurly]

Re: Arc gaps on circuit boards?

So what exactly is their purpose?

Surge protection?

[redwire]

Re: Arc gaps on circuit boards?

Usually I see them designed in for ESD protection. Some regulatory tests simulate human-body static electricity pulses i.e. 100pF+1k5 resistor at up to 25kV. In Canada static electricity is so bad in winter you can easily generate that.
Some instrumentation/RTU's where lightning can reach their inputs, will use spark gaps but their breakdown voltage is wishy washy and they etch away quickly, so not great for recurring spikes or big spikes.
The spark-gap is a free way to deal with a low-energy HV spike. I think a MOV works better but costs more.

[smason]

Re: Arc gaps on circuit boards?

In Canada static electricity is so bad in winter you can easily generate that.

Yup, and here in Alberta in the winter a single cat and a wool carpet can power a small town.

[goontron]

Re: Arc gaps on circuit boards?

yup, and here in alberta in the winter a single cat and a wool carpet can power a small town.

lol!

[severach]

Re: Arc gaps on circuit boards?

But the cats all go to sleep at the same time so you need idle base load power at the ready. The cost of free cat power + idle base load power is no less than the cost of running the base load all the time and letting the cats sleep the remaining hour.

[Uranium-235]

Re: Arc gaps on circuit boards?

could be tin whiskers. Might not see them much anymore cause of new mixtures of solder

[larrymoencurly]

Re: Arc gaps on circuit boards?

could be tin whiskers. Might not see them much anymore cause of new mixtures of solder

In the 1980s IBM PC PSU I saw the pinty areas were untinned copper, and the board in the picture is about 20 years old.

[mertsok]

Re: Arc gaps on circuit boards?

An excellent way to protect inputs of electronics components from the very high voltages of ESD and EMI spikes is to use spark gaps. These can be built into your circuit board as parts of the PC board etch.

[redwire]

Re: Arc gaps on circuit boards?

I don't find arc gaps that great... you'll still end up with a several kV spike which is too much for the components downstream.
One trick I do is ESD gun the board in a dark room. You can see where the arc is, Murphy's law takes effect and the spark gap can do nothing but instead the arc happens on unexpected pcb traces.
Example is a 0.5mm spark gap gives theoretical 3kV breakdown (with no soldermask). Okay, now that it's from 10kV to 3kV - then what? It can still hop across a resistor and take out the semi's.

In the old days, spark gaps were on CRT sockets to stop arcing and corona from destroying the video amplifiers, later they used spark-gap capacitors.
Nowadays, I think spark gaps are used only to pass IEC-61000 regulatory ESD tests.

For lightning you have to use MOV's or gas-tube arrestors, because the spike has so much energy.

[japlytic]

Re: Arc gaps on circuit boards?

I don't normally include printed arc gaps in circuit diagrams, unless they are in series with a component.