TECHNET Archives

December 2015

TechNet@IPC.ORG

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Louis Hart <[log in to unmask]>
Reply To:
TechNet E-Mail Forum <[log in to unmask]>, Louis Hart <[log in to unmask]>
Date:
Tue, 8 Dec 2015 13:53:26 +0000
Content-Type:
text/plain
Parts/Attachments:
text/plain (1 lines)
Ahne, interesting comments you make.  It's been a long time, but I worked on 'corona' or, better terminology I think, 'partial discharges'.  In decreasing order of durability were inorganic materials, naturally occurring organic materials, synthetic organic materials.  The company making the partial discharge tester I used is not longer in that business.  The field was technically challenging, since you were trying to study the lifetime of materials which were imperfect in varying degrees. Louis Hart

-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Ahne Oosterhof
Sent: Monday, December 07, 2015 10:32 PM
To: [log in to unmask]
Subject: Re: [TN] z-axis voltage separation

In June there was a thread about high voltages and breakdown which I came across as I was "cleaning" my Outlook file, it was getting too big.

Many words jumped out: creepage, breakdown, spacings, dielectric constant, corona, conductor shapes, environment (altitude, humidity), safety requirements, safety factors.
It is difficult to come up with a 'hard and fast' rule for a designer to follow. Which is why there are so many different rules, the most famous one: it depends. Even if you find rules, try to find out under what conditions were they established or were they work. 
In my experience two important factors are dielectric constant and corona. First think of flux lines / field density. Trouble starts were you have most of them closely spaced. That is why you don't want high voltages between two sharp points. Those points concentrate the field very nicely and allow arcs more easily. 
When you use two different materials with different dielectric constants, and let's say they are the same thickness, the field is highest across the material with the highest dielectric constant. And that is where the breakdown starts, after which all the voltage now exists across the other layer if it can handle it (for a while).
Almost always in high voltage situations breakdown will be preceded by corona through or over the material. To detect it you can listen for it with a detector that hears ultra sound. Or you can put a resistance of a few hundred kOhm in the ground lead of the device under test and connect a scope across it. Corona will show up as small voltage spikes across that resistor. (Don't increase the voltage till it sparks, the scope may not like it.) A little bit of corona can easily go undetected but over the long run it will "eat" your insulation leading to breakdown.
Applying a safety factor is a real 'it depends'. Most so called rules do have a safety factor included, but if the rule was established at sea level, it does not really work in Denver at a few thousand feet of altitude. So test at all conditions were the design will be used and then I would apply a safety factor of at least two.

Good luck and goodbye, 
Ahne
 


-----Original Message-----
From: TechNet [mailto:[log in to unmask]] On Behalf Of Jack Olson
Sent: Tuesday, June 24, 2014 7:47 AM
To: [log in to unmask]
Subject: Re: [TN] z-axis voltage separation

oops... I missed part of the relevant quote during my "copy-and-paste".
sorry.

For the so called *functional insulation*, UL 60950-1 permits to use separation distances lesser than the specified in their charts. They just have to withstand the electric strength test (casually called *Hipot*) per
Par.5.2.2 Table 5B. In other words, where only functional insulation is required, you don't need to meet any specific clearance between PC traces for as long as there will be no electric breakdown between them at the prescribed test voltage. The latter generally is several times greater than actual working voltage between separated traces. Unfortunately, there is no clear information in the literature on what is actual breakdown voltage between the conductors and how to design a PCB to pass a specific hipot.


On Tue, Jun 24, 2014 at 9:31 AM, Robert Kondner <[log in to unmask]>
wrote:

> Jack,
>
>  You are correct, folks keep feeding us creepage and clearance info. 
> There are many technicians out there that will repeat what they know 
> even if it answers a different question. We just need to wade through 
> the noise and look for real data.
>
>  Look for the terms "Reinforced Insulation" where they talk about 
> minimum thicknesses but that still does not answer our questions directly.
>
>  Best I can tell we need to deal with "Insulation Resistance". A 
> temperature and humidity soak followed by a resistance check at a 
> certain voltage.
>
>  If I find anything I will let you know but I bet someone on this list 
> does know, we just need to hear from them.
>
> Bob K.
>
> -----Original Message-----
> From: TechNet [mailto:[log in to unmask]] On Behalf Of Jack Olson
> Sent: Tuesday, June 24, 2014 10:04 AM
> To: [log in to unmask]
> Subject: Re: [TN] z-axis voltage separation
>
> Thank You for the reference. I didn't read it word for word, but it is 
> primarily dealing with the same subject as IPC Table 6-1 (lead to lead 
> clearance, or any conductive to conductive, creepage across surfaces, 
> around grooves and corners, over protrusions, through air, etc.) in 
> other words, CREEPAGE and CLEARANCE.
>
> I'm asking about only one factor (layer to layer through a dielectric, 
> inside a PCB layer structure, masked and conformally coated, no 
> exposed
> anything) the answers from both TechNet and the Designer Council fall 
> into two major categories:
> 1) those that recommend about 0.2 mm spacing
> 2) those that recommend about 3.2 mm spacing
>
> Such a huge difference, and respectable opinions on both sides. This 
> seems like an area that could use some work. And if anyone like an IPC 
> committee takes it on, maybe we could revise Table 6-1 while we're at it.
> 0.1 mm for 100V,
> 0.2 mm for 101V?
> seriously?
> c'mon...
>
> thanks for supporting the fact that I'm not crazy for asking, Jack
>
>


______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] ______________________________________________________________________

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

______________________________________________________________________
This email has been scanned by the Symantec Email Security.cloud service.
For more information please contact helpdesk at x2960 or [log in to unmask] 
______________________________________________________________________

ATOM RSS1 RSS2