EPC Fets and Bondit

[John Larkin]

The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great.
Just put a small drop over the fets and hit it for a few seconds with
the blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we
have a serious UV gun here, which I can try) but the standard kit
works great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

I've also wondered the same sort of thing about 2-part epoxies. Do
they have to be mixed down to molecular level?



--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[John S]

On 3/10/2020 2:20 PM, John Larkin wrote:

On Tue, 10 Mar 2020 12:14:57 -0700, John Larkin
jlarkin@highland_atwork_technology.com> wrote:


The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great.
Just put a small drop over the fets and hit it for a few seconds with
the blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we
have a serious UV gun here, which I can try) but the standard kit
works great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

I've also wondered the same sort of thing about 2-part epoxies. Do
they have to be mixed down to molecular level?

Sorry, Bondic not Bondit.

The stuff at the bottom of the vias is hard.

Does it release easily with alcohol like hot glue does?

 

[John Larkin]

On Tue, 10 Mar 2020 12:14:57 -0700, John Larkin
<jlarkin@highland_atwork_technology.com> wrote:

The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great.
Just put a small drop over the fets and hit it for a few seconds with
the blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we
have a serious UV gun here, which I can try) but the standard kit
works great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

I've also wondered the same sort of thing about 2-part epoxies. Do
they have to be mixed down to molecular level?

Sorry, Bondic not Bondit.

The stuff at the bottom of the vias is hard.

--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Clifford Heath]

On 11/3/20 6:14 am, John Larkin wrote:

There are probably more commercial UV epoxies and illuminators (we
....
I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

Dentists use a lot of UV-cure materials. For cosmetic work, they get
applied in multiple thin layers, cured one at a time. I believe this is
so folk don't end up with uncured resins in their mouth - so it's is
quite likely a problem.

On the other hand, anywhere you can't get UV is somewhere you can't see,
so if there's uncured resin there, you'd only know by cutting into it.

I've also wondered the same sort of thing about 2-part epoxies. Do
they have to be mixed down to molecular level?

Improperly mixed (or wrong ratio!) epoxies form a microscopic mesh
containing uncured resin components. If the ratios are close, the result
is almost as strong as properly cured epoxy, but professionals still
take extreme care with proper weighing and mixing of parts.

I learned this from a neighbour who spent his whole career in this
industry. His techniques are being used in Melbourne to make airfoils
for Boeing jets, and his technique&design for making refrigerated
shipping containers is the industry standard. Vale Bill Maloney. He was
a lovely chap.

CH

 

[Phil Hobbs]

On 2020-03-10 15:14, John Larkin wrote:

The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great.
Just put a small drop over the fets and hit it for a few seconds with
the blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we
have a serious UV gun here, which I can try) but the standard kit
works great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

Not very fast, if ever. In optical assembly it's common to tack parts
together by partly curing a few areas, so that if there's some screwup
you can get them apart again.

I've also wondered the same sort of thing about 2-part epoxies. Do
they have to be mixed down to molecular level?

IIRC it depends on the catalysis mechanism. If the catalyst is a small
molecule, it eventually diffuses around pretty well. OTOH it isn't at
all difficult to get mixing on molecular scales. The exponential effect
of folding does a very good job even for epoxy putty.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[glen walpert]

On Tue, 10 Mar 2020 18:34:07 -0400, Phil Hobbs wrote:

On 2020-03-10 15:14, John Larkin wrote:

The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great. Just
put a small drop over the fets and hit it for a few seconds with the
blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we have
a serious UV gun here, which I can try) but the standard kit works
great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

Not very fast, if ever. In optical assembly it's common to tack parts
together by partly curing a few areas, so that if there's some screwup
you can get them apart again.

While that may be true of UV cure acrylic adhesives, there are UV cure
adhesives designed to cure in unexposed areas such as under components,
for example:

https://www.panacol.com/adhesive-glue/led-curing-adhesive
https://www.panacol.com/products/adhesive/vitralit

Unlike the Bondic site https://notaglue.com/ they provide specs on
uncured viscosity and cured modulus, and have a wide range of properties
available. Bondic won't even tell you that ~365nm is optimal for curing
- they claim it is proprietary!

There are other mfgrs of UV cure adhesives designed for electronic
component encapsulation which provide good specs too, some of which seem
to be readily available in small quantities from the usual suspects

BTW I have found that half a bottle of UV cure optical adhesive will cure
in the bottle if left in a dark refrigerator for a decade .

I've also wondered the same sort of thing about 2-part epoxies. Do they
have to be mixed down to molecular level?

IIRC it depends on the catalysis mechanism. If the catalyst is a small
molecule, it eventually diffuses around pretty well. OTOH it isn't at
all difficult to get mixing on molecular scales. The exponential effect
of folding does a very good job even for epoxy putty.

Most Epoxy manufactures claim that under mixing is the most common cause
of failure to achieve rated properties, and many recommend mixing until
the mix visually appears to be complete, then continue mixing in the same
manner for the same time again, mixing twice as long as appears
necessary. The catalyst is rarely a small molecule in epoxy, it won't
diffuse far and is not purely a catalyst responsible for initiating
polymerization but also an essential part of the cured co-polymer resin.

Interestingly many of the UV cure adhesives tailored specifically to glob-
top encapsulation and BGA underfill (UV + anreobic cure) are one part
epoxies, not acrylics.

 

[Phil Hobbs]

On 2020-03-12 12:31, glen walpert wrote:

On Tue, 10 Mar 2020 18:34:07 -0400, Phil Hobbs wrote:

On 2020-03-10 15:14, John Larkin wrote:

The tiny EPC GaN fets, especially the smallest 4-ball ones, are very
fragile. If you nudge one in handling a board, the part can fracture,
or the balls can break away from the board or from the fet itself.

Someone here recommended Bondit, so I got the kit. It works great. Just
put a small drop over the fets and hit it for a few seconds with the
blue LED. It hardens nicely before it has time to spread out. The
viscoscity is just right.

https://www.dropbox.com/s/qw7rkexm0lled0p/GaN_Bondit_3.jpg?raw=1

There are probably more commercial UV epoxies and illuminators (we have
a serious UV gun here, which I can try) but the standard kit works
great and will protect hundreds of little fets.

It filled the vias but not enough to make a convex bump on the bottom,
which matters to us since we will use a pogo fixture for testing.

I wonder: what happens to places, like under the fets or inside the
vias, that don't get illuminated much? Does the UV cure effect spread
out into the dark regions?

Not very fast, if ever. In optical assembly it's common to tack parts
together by partly curing a few areas, so that if there's some screwup
you can get them apart again.

While that may be true of UV cure acrylic adhesives, there are UV cure
adhesives designed to cure in unexposed areas such as under components,
for example:

https://www.panacol.com/adhesive-glue/led-curing-adhesive
https://www.panacol.com/products/adhesive/vitralit

Unlike the Bondic site https://notaglue.com/ they provide specs on
uncured viscosity and cured modulus, and have a wide range of properties
available. Bondic won't even tell you that ~365nm is optimal for curing
- they claim it is proprietary!

There are other mfgrs of UV cure adhesives designed for electronic
component encapsulation which provide good specs too, some of which seem
to be readily available in small quantities from the usual suspects

BTW I have found that half a bottle of UV cure optical adhesive will cure
in the bottle if left in a dark refrigerator for a decade .

I've also wondered the same sort of thing about 2-part epoxies. Do they
have to be mixed down to molecular level?

IIRC it depends on the catalysis mechanism. If the catalyst is a small
molecule, it eventually diffuses around pretty well. OTOH it isn't at
all difficult to get mixing on molecular scales. The exponential effect
of folding does a very good job even for epoxy putty.

Most Epoxy manufactures claim that under mixing is the most common cause
of failure to achieve rated properties, and many recommend mixing until
the mix visually appears to be complete, then continue mixing in the same
manner for the same time again, mixing twice as long as appears
necessary. The catalyst is rarely a small molecule in epoxy, it won't
diffuse far and is not purely a catalyst responsible for initiating
polymerization but also an essential part of the cured co-polymer resin.

Interestingly many of the UV cure adhesives tailored specifically to glob-
top encapsulation and BGA underfill (UV + anreobic cure) are one part
epoxies, not acrylics.

Yeah, I'm mostly familiar with Norland 65 and 85, which I standardized
on long ago.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net