calculating minimum air gap

[cowboyengineer]

Mechanical engineer here, so sorry if this is a an easy question, or in the wrong forum. I need to determine the minimum air gap between two voltage carrying terminals, in order to prevent arcing. The terminals are parallel, with no protruding shapes in the direction of the other terminal. The circuit voltage is 14v dc with very low current Assume the nominal air condition is low humidity and high temperature (>130 degrees C). Hope this is enough information. Thanks

 

[IRstuff]

Is it really 14 Vdc?
Is this on a board?
Why are there no design standards for you to use?

Can't imagine being able to get terminals close enough to cause arcing without raising other concerns, e.g., the substrate bending and causing the terminals to short out altogether.

TTFN

 

[cowboyengineer]

This is not on a board, but terminals on a connector for a small coil. Yes there are design standards, our design group specifies that the gap must be greater than 0.6mm. The point is that I want to check the calculation myself, the design group is 6000 miles away, and I don't realy trust the design group based on previous circumstances. I just want to verify the potential for failure due to a problem with our supplier. Thanks

 

[IRstuff]

If it's only 14 V, then I don't see a problem. Typical IC bond lead spacing can be less than 0.2 mm at 30 V.

TTFN

 

[cowboyengineer]

Thanks for your help, but I have many opinions on what is ok or not. What I'm looking for is a calculation to determine the theoretical limit. Thanks

 

[IRstuff]

Take the dielectric breakdown voltage of air (3E6 V/m) and multiply by the gap

TTFN

 

[nbucska]

As an old engineer I have learned:
Calculate what has to be calculated,
Don't calculate what don't have to be --or can't be --
calculated
and know the difference
--- OR ---
spend unlimited money and time.

<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[nbucska]

I am an EE, IR is one, too, and hi is right except:
you specify >130C but not the highest temp. IRstuff's rule is not valid for much higher (e.g. 500 C )...

To prove the previous rule of thumb:

A yound engineer once specified a $1500. -2V precision pover supply -- I changed the design to use a 10K 1/4W resistor to -15V which is less than a dime.











<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[ScottyUK]

Unless this is going to be housed in a hermetic package, the theoretical limit is just that. As soon as you start to introduce contaminants either into the air or as surface deposits the practical minimum separation will become much larger. Most people design their products for the real world - I'm curious why the interest in the theoretical minimum if you're building a product for manufacture, or does 0.6mm just feel 'too small'?


----------------------------------

One day my ship will come in.
But with my luck, I'll be at the airport!

 

[IRstuff]

That's really almost a whole separate issue.

When simply considering likelihood of failure, there are significantly more and more probable possibilities than worrying about the 14V delta causing an arc across a relatively huge gap.

TTFN

 

[cowboyengineer]

Thanks for all your replies, and I agree completely wtih nubucska's opinion (don't calculate if you don't need). I understand the mitagating factors that were brought up by the others too. This request is derived from a potential quality problem were are experiencing. Our managment [not engineers]is up in arms with concern, demanding an array of ridiculous testing, recreation testing, etc., and I'm simply trying to validate my opinion that we are making a mountain out of a mole hill. Through my experince and emperical knowledge of the product, I believe our management concerns are unwarranted. Calculating the theoretical limit (which we are far from approaching in actual use) just gives me an idea of under what conditions should we be concerned. Thanks

 

[electricpete]

I agree with Scotty's comment. If you use that theoretical limit based on air breakdown as discussed above: 14V / (30E3*volts/inch) you come up with something like 0.0005 inch. That is a ridiculously small number and likely much smaller than the dust particles around. I hope you would never contemplate assigning any practical significance to that number.

Does this coil ever switch? If so (in absence of surge supression) there may be a voltage transient may be orders of magnitude higher than 14v.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

While we're talking theoretical - You might want to google on Paschen's law. It predicts a minimum voltage below which arcing will not occur regardless of how high the field (how close the terminals). For air under some idealized conditions that Paschen minimum is in the neighborhood 300 volts.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Whoops. My calculation 30kv/inch was wrong. Mixed my units. 30kv/cm or 75kv/inch

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

http://home.earthlink.net/~jimlux/hv/paschen.htm

identifies the minimum "sparking" potential for air is 327 volts.

Once again what it is saying is that at your voltage 14V (or any voltage below 320volts) in theory you could put your terminals one Angstrom apart and there will be no arcing.

Revised calculation 14v / [75kv/inch] is even more
outrageous 0.0002 inches.

Once again I am not proposing to use any of the above "theory" as a guide. Exactly the opposite.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[nbucska]

Hi Epete:
I think if the gap is VERY small, the current can just tunnel through. There is a gap between the Cu atoms,
too...



<nbucska@pcperipherals DOT com> subj: eng-tips
read FAQ240-1032

 

[CarlPugh]

Someone said that life is non linear.
This is especially true for breakdown voltages.
A rule of thumb is if the voltage is doubled, the gap is tripled.

Reference Data for Radio Engineers fifth edition page41-1 has Spark-gap breakdown voltages.
The lowest chart goes is 1000 volt peak For needle gap it is 0.025 inches at 25C, 29.9 inches of mercury.
The multiplication factor for 60C is 0.50
So breakdown for 0.025 inch gap at 60C is 500 volt peak.

Page 41-2 states "The breakdown voltage is approximately proportional to pressure and inversely proportional to absolute (degrees Kelvin) temperature"

If you are actually operating at 120C, I would suggest making some tests.

 

[electricpete]

Yes Nick I agree. One Angstrom was an exageration for illustration. The statement remains true - arcing in air does not occur below the threshhold voltage regardless of gap length under whatever idealized conditions assumed by Paschen's law.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[IRstuff]

Sorry to disappoint, but I've had a 270 Vdc supply arcing across the outside of a badly designed PCB. Gap was about 10 mils



TTFN

 

[electricpete]

OK, let me get this straight. When I say "idealized conditions", you think I'm suggesting that is applies to every real world situation? Read it again and I hope you'll agree it's obvious that's not what I'm suggesting.

But since it is the subject of this thread, why don't you tell us the factors that contributed to your arcing?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.