Timothy
Thanks for the info. Firstly, I would say a well-cured solder mask would have a
BETTER punch-through resistance than laminate, because it is homogeneous.
Secondly, are the figures you quote for the products at around the glass
transition temperature, at which the electrical characteristics of the resins
run down a very steep hill? IMHO, even at room temp, putting enormous voltage
gradients onto epoxy resins is hazardous. Nearly all epoxies contain some sodium
chloride (resulting from the prepolymerisation reaction between the
epichlorohydrin and the sodium hydroxide, to release the epoxy group for
cross-linking). Applying large voltage gradients will dissociate these molecules
and cause the ions to migrate slowly through the resin matrix, especially with
absorbed humidity present. Removing the excitation will tend to cause
re-association but often on a time scale of hours or even days. I have observed
this phenomenon with FR-4 with voltage gradients as low as 100 V/mm although
not, of course, to a dangerous level. Once you have free ions beetling around,
the problem may become unpredictable. To restrict these to a minimum, it
requires a good resin with a minimum of dissociable molecules (may not be the
cheapest) and one which is cured into the largest molecules, to restrict
physically the mobility of the ions. I maintain however what I said before that
you should never allow the test conditions to approach the Tg by anything closer
than about 30°C if you are electrically stressing the resin. Incidentally,
increasing the temperature of the resin will increase the ion mobility linearly
up to approaching Tg and, above Tg, there is little restriction to ion mobility.
I don't claim to be an expert on resin science, although I have some knowledge,
gained partially from work which was conducted by Prof Kausch (Department of
Polymers of the Swiss Federal Institute of Technology, Lausanne) umpteen years
ago. One of his students did his dissertation on the electrical characteristics
of epoxies and some other polymers, I think about 1989, and this kind of
phenomenon, using mainly commercial FR-4s. Only linear polymers, of those tried,
were exempt from ion migration phenomena. As I supplied some of the
instrumentation he used for low voltage insulation resistance, I was closely
involved with the project and was named as one of the adjudicating experts. Prof
Kausch (and his assistant Dr Buégelin) said that the dissertation was one of
best he had seen in all his career as an academic because it revealed some new
phenomena which were not in the literature, as well as being well researched.
Unfortunately, the student in question did not continue in this field of work
after his graduation with honours, at which he also received a special prize.
Since then, I've always looked on much of the specification of thermosetting
resins with a pinch of dissociable sodium chloride taken out of the resins. We
still have much to learn about commercial laminates in this field.
Best regards
Brian
Timothy Reeves wrote:
> I'm not sure that's unacceptable...soldermask is expected to have a
> dielectric strength of 500V/mil (per IPC-SM-840C), which equates to almost
> 20,000 V/mm. The IPC-4101 requires FR-4 to withstand 29,400V/mm for FR-4
> <0.78 mm and 40,000 V when >= 0.78 mm.
>
> Timothy Reeves
> ECD Circuit Board Division
> 13626 South Freeman Road
> Mulino, OR 97042
> [log in to unmask]
> (503) 829-9108 (800) 228-8198 FAX (503) 829-5482
>
> > ----------
> > From: Brian Ellis
> > Sent: Friday, December 10, 1999 8:51 AM
> > Subject: Re: WG: dielectric II (arghh! sorry!)
> >
> > Wanner
> >
> > I'm not an expert in the matter but your message inspires me to say three
> > things.
> > 1. 500 V over 65 micrometres gives just one helluva voltage gradient,
> > about 8 000 V/mm,
> > if I'm not mistaken. This is at the limit of the acceptable IMHO, even at
> > ambient.
> > 2. At 120°C, you are approaching the Tg of the laminate, where the
> > electrical
> > characteristics of the resin go haywire. I believe that Tg is not a sharp
> > threshold,
> > where you have one phase below and one phase above, but a gradual change
> > over a range
> > of 20 - 30°C. Furthermore, the range can be modified by the degree of cure
> > (i.e. the MW
> > of the crosslinked resin) and the humidity content (i.e., it is known that
> > the Tg can
> > drop by 20° when the resin is close to saturation, generally between 2 and
> > 5% by weight
> > of the resin.
> > 3. A laminate is heterogeneous. Pressing a thin single prepreg to give a
> > 65 um layer
> > will give precious little resin above and below the silane-coated glass
> > fibres.
> >
> > Under these circumstances, combined, I think you are approaching the
> > characteristics of
> > an arc welder rather than those of a test method. Pragmatically, I think
> > you should not
> > test or function at higher than 30 - 40°C under the Tg when you have high
> > voltage
> > gradients in heterogeneous insulants and that you should ensure that your
> > supplier
> > presses your boards to a full cure, using only first-rate materials that
> > have been
> > tested for a high MW polymerisation.
> >
> > My 2 centimes worth.
> >
> > Brian
> > -
> > Brian Ellis
> > Protonique SA
> > PO Box 78
> > CH-1032 Romanel-sur-Lausanne, Switzerland
> > Voice: +41 21-648 23 34 Fax: +41 21-648 24 11
> > E-mail: [log in to unmask]
> > URL: Technical and consultancy divisions:
> > http://www.protonique.com
> > Web services division:
> > http://www.protonique.com/webserv
> >
> >
> > Wanner Bernhard wrote:
> >
> > > Hi Techies, sorrysorrysorry for the mismail before! Ok, now the finished
> > > mail:
> > >
> > > I have to analyze failures of a subassemblies of our subcontractor, a
> > > well-known european hi-tech pcb-manufacturer. The failing subassemblies
> > I've
> > > to analyze, contains transformer windings which are implemented as 14
> > and
> > > 16-layer pcb's. The winding shows dielectric breakdown between different
> > > winding layers (and sometimes between different windings on the same
> > layer).
> > > This breakdown happens after some hours run-in at 120°C and 500VDC
> > (500VDC
> > > as test
> > > voltage, failures also at lower tension at at AC). Before them, all pcb
> > has
> > > passed successful a hi-voltage test at ambient temperature.
> > > The diel. strength for the closest distance of 65mm prepreg should be
> > around
> > > 2.5kV, or a little bit less because IPC-2221, § 4.1.2, not 500V!
> > > May be any loss of the electric strenght of a laminate/prepreg is a
> > function
> > > of the temperature, the hi-temp aging, and material defects (the
> > occuring
> > > pcb temperature of 120°C is in accordance with IPC-2222, § 4.3. But this
> > > must be in a specified manner, isn't it?
> > >
> > > Okydky (so weit - so gut), now the pcb-manufacturer's statement is that
> > we
> > > can not use the max. electrical strenght at max. temperature. And the
> > > pcb-manufacturer is not able to define a derating of the el. strenght at
> > a
> > > given temperature. Hard to believe ...
> > >
> > > So, my question is: means IPC I can use the maximal specified electrical
> > > strenght at maximal specified temperature, or not?
> > >
> > > (If not, I would like to append some further questions..)
> > >
> > > As I have reviewed some IPC-standards, there are some "little"
> > limitations
> > > in view of limitation of Temprature, electrical strenght, dielectric
> > > thickness, maximum operating temperature. But none of this comment means
> > we
> > > could not use the electrical strenght at max. allowed operating
> > temperature:
> > >
> > > Designstandards
> > >
> > > IPC-2221, Generic Standard on Printed Board Design
> > > § 4.1.2 Material Selection for Electrical Properties, refers to
> > > Table 4-1, Typical Properties of Common Dielectric Materials. Thereafter
> > for
> > > FR-4 the electric strength amounts 39.4 x 103 V/mm2 and notice notice
> > says :
> > > "The stated electrical strength values are commonly evaluated under test
> > > conditions with a 0.125mm core laminate thickness. These values should
> > not
> > > be considered linear for high voltage designs with a minimum dielectric
> > > separation, i.e., less than 0.09mm."
> > > § 6.3 Electrical Clearance, refers to Table 6-1. Thereafter,
> > the
> > > minimum spacing for internal conductors (B1) at 301-500V (peak) amounts
> > > 0.2mm.
> > > §10.1.2 Electrical Clearance, "Clearances are applicable for all
> > > levels of design complexity (A, B, C) and performances classes (1, 2, 3)
> > > (...)".
> > >
> > > IPC-2222, Sectional Design Standard for Rigrid Organic Printed Boards
> > > § 4.3 Laminate Materials, "The values in Tabele 4-1 are based
> > on
> > > long term aging tests by UL (...) Hot spot temperatures shall not exceed
> > the
> > > temperatures specified in Table 4-1 (...)"Table 4-1: "Clad Laminate
> > Maximum
> > > Operating Temperatures", thereafter for FR-4 and 0.1 mm dielectric
> > thickness
> > > the max. allowed temperature would be 120°C. For FR4 / 0.4mm diel.
> > thickn.
> > > 130°C.
> > > § 4.3 Measurement of Dielectric Thickness, " (...) Thickness
> > by
> > > microsection (...) taken at the closest point between metal cladings."
> > > § 4.3.2 Dielectric Thickness/Spacing, " (..) If the minimum
> > > dielectric spacing and the number of reinforcing layers are not
> > specified,
> > > the minimum dielectric spacing is 0.09mm and the number of reinforcing
> > > layers may be selected by the supplier."
> > > § 4.3.4.3 Glass style, "A variety of glass cloth styles
> > are
> > > available for prepregs (see IPC-EG-140). The glass cloth selection is
> > > dependent upon dielectric thickness and tolerance required, circuit
> > filling
> > > needs, and electrical requirements of the dielectric."
> > > § 4.3.7 Laminate Material, explains the Material Code
> > Designation
> > > and refers to Figure 4-2 for the recommended material selection process.
> > >
> > > Thanks for clarifications and all statements!
> > > Bernhard
> > >
> > > ##############################################################
> > > TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
> > > ##############################################################
> > > To subscribe/unsubscribe, send a message to [log in to unmask] with
> > following text in
> > > the body:
> > > To subscribe: SUBSCRIBE TECHNET <your full name>
> > > To unsubscribe: SIGNOFF TECHNET
> > > ##############################################################
> > > Please visit IPC web site (http://www.ipc.org/html/forum.htm) for
> > additional
> > > information.
> > > If you need assistance - contact Gayatri Sardeshpande at [log in to unmask]
> > or
> > > 847-509-9700 ext.5365
> > > ##############################################################
> >
> >
> >
> > -
> >
--
Brian Ellis
Protonique SA
PO Box 78
CH-1032 Romanel-sur-Lausanne, Switzerland
Voice: +41 21-648 23 34 Fax: +41 21-648 24 11
E-mail: [log in to unmask]
URL: Technical and consultancy divisions:
http://www.protonique.com
Web services division:
http://www.protonique.com/webserv
##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe: SUBSCRIBE TECHNET <your full name>
To unsubscribe: SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Gayatri Sardeshpande at [log in to unmask] or
847-509-9700 ext.5365
##############################################################
|
