Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
You can't achieve a very good vacuum at the bottom of a deep mould that
is already filled with a liquid potting compound. Gravity acts on the
liquid, causing the pressure at the bottom of the liquid to be higher
than that at the top of the liquid.
whit3rd <whit3rd@gmail.com> wrote in
news:87c37d58-86ae-4210-a7e6-2677f51f6b29@googlegroups.com:
On Friday, November 1, 2019 at 4:14:54 AM UTC-7,
DecadentLinux...@decadence.org wrote:
I would not recommend an epoxy unless you plan on zero
serviceablity. Use an RTV so you can service it if it fails.
Huh? RTV is 'room temperature vulcanizing' material, it sets up
just as irreversibly as epoxy does. How is that
service-friendly?
Well, son... it ain't yer tube of window caulk. ;-)
RTV for HV potting is 100% serviceable. It shears easily.
Epoxy, on the other hand cures usually quite rigid.
Remember "Stycast"?
A 2-stage rotary pump like the one I use can very easily achieve 0.01On Sunday, November 3, 2019 at 9:09:08 PM UTC-5, Chris Jones wrote:
On 02/11/2019 12:01, Rick C wrote:
On Friday, November 1, 2019 at 8:32:39 PM UTC-4, Bill Sloman wrote:
On Saturday, November 2, 2019 at 11:16:01 AM UTC+11, Rick C wrote:
On Friday, November 1, 2019 at 7:56:30 PM UTC-4, Bill Sloman wrote:
On Saturday, November 2, 2019 at 1:34:44 AM UTC+11, Rick C wrote:
On Friday, November 1, 2019 at 7:47:28 AM UTC-4, Bill Sloman wrote:
On Friday, November 1, 2019 at 6:18:13 PM UTC+11, Rick C wrote:
On Friday, November 1, 2019 at 2:00:50 AM UTC-4, Bill Sloman wrote:
On Friday, November 1, 2019 at 8:43:09 AM UTC+11, Jim Horton wrote:
On 10/31/19 5:33 PM, Rick C wrote:
I don't know a lot about epoxy potting, but I know you don't want a lot of vacuum. If you pull too much vacuum the epoxy will boil and you'll end up with foam as your encapsulant. I know, I did this.
I know, that was one of my concerns. That was one reason I asked about
vacuuming just the epoxy premixing container ahead of time. As I said,
because the transformers are already well potted for the autos, I don't
even think vacuum will be necessary but thought I would throw the
question out there.
I'd evacuate container the before I let in the liquid expoxy encapsulent. If it frothed at the time, that wouldn't matter, as long as you got in enough of the liquid to fill the container. If the process stalled before that happened (which seems unlikely - there isn't going to be much gas in those bubbles if you degassed the liquid before you poured it in - you can use a bit of atmospheric pressure to squeeze in the last of the liquid.
You've got to do this while it's still liquid and more or less free-flowing. Once it starts setting, atmospheric pressure will flatten any bubbles that remain.
Bubbles of atmospheric pressure air are much harder to get rid of.
If it froths, it will expand to many times it regular volume and spill out onto the vacuum chamber. Frothing is bad.
That's why you out-gas the liquid encapsulant in a deep beaker, and makes sure that the froth doesn't go over the top of the beaker ... and only after you have done that do you pour the outgassed liquid into box that contains the gear you want to encapsulate.
Doing all of that under vacuum (or a vacuum of sorts - you can't get the residual pressure spectacularly low) requires a proper vacuum potting setup, which isn't all that complicated, but not all that easy to improvise either - being able to tilt a beaker and get it to pour from the other side of the the vacuum wall requires a vacuum-greased rotating joint.
Why would you even suggest something so complicated?
Because that was the way it was done at places I've seen doing it.
All you need to do is to fill the mold with the device, pour in the potting compound and then put it in the vacuum chamber. Apply the partial vacuum and make sure the potting compound doesn't foam by limiting the vacuum.
Getting the dissolved air out of the potting compound before it goes into the mold lets you do it when there's a lot more exposed surface for the bubbles to come out of, and can speed up the process a lot.
This is not a production run, no one cares if it takes a while. The process you describe is overly complex for a one-off.
It was a prototyping set-up. It was just complex enough to avoid inconvenient and time-consuming problems. People want prototypes as soon as possible, and they don't like hanging around when they don't have to.
To prevent foaming it would be good to have a pressure gauge attached and test the encapsulant to see at what point it foams. Then you can monitor the vacuum to make sure foaming doesn't happen.
You are imagining that the encapsulant has a high vapour pressure component that boils at a specific pressure. Dissolved air doesn't work that way. Any water in the encapsulant will boil at about 20 torr if the encapsulant is at room temperature, but there shouldn't be any.
I never mentioned any numbers.
Surprise, surprise.
I'm not talking about foam from the air. I'm talking about the encapsulant forming enough vapor that the entire stinking mess expands out of the container and spills into the vacuum chamber making a huge mess that then has to be cleaned up so you can start over. This isn't imagination. This is experience.
You've had experience of doing it wrong, and you want to reject the experience of people who had done it often enough to do it better.
I didn't invent that set-up - I got to see it in use and more or less understood why it was set up the way it was, which you seem to be having a problem with.
The only problem is that you insist on making it many steps when those extra steps add nothing of benefit. Now that we have gotten to the root of the issue you stop talking about the issue and attack the person.
Bottom line is there is no need to pump the epoxy first and then again when adding it to the mold.
You can't achieve a very good vacuum at the bottom of a deep mould that
is already filled with a liquid potting compound. Gravity acts on the
liquid, causing the pressure at the bottom of the liquid to be higher
than that at the top of the liquid.
That is true for various values of "deep". In this case you
will never notice the delta from the top to the bottom of a
less than 1 foot deep mold. Materials like this are the same
density as water to a first approximation. Water is 32 feet > to 1 atm. So 1 foot of mold will only cause 1/32 of an atm
difference in pressure.
The size of the remaining bubbles will depend on what the "multipleSurface tension might also hinder the escape of gas from crevices in a
device that is already submerged.
When the volume of trapped air expands my multiple factors, it will
help get the air out. Any remaining bubbles will be very tiny when
the vacuum is released. Since these air bubbles will exist whether > or not you pre-vacuate the epoxy, it has no impact on the result.
Putting the two components of the encapsulant under vacuum, stirringDepending on the purpose of the potting process you may or may not care
about small bubbles remaining, but some people do need to care.
How can you do better in getting rid of them?https://www.youtube.com/watch?v=8GTvWfwFktU
If the vacuum is poor, it will not get the smaller bubbles out. If theI have never needed to be extremely careful about eliminating bubbles so
I have never built anything to pour encapsulant under vacuum, but I do
degas the encapsulant in a separate beaker before pouring, simply
because it does tend to froth and overflow unless that step is carried
out in a beaker able to contain many times the volume of the liquid
encapsulant.
It won't froth if you don't draw too much vacuum.
On 04/11/2019 16:27, Rick C wrote:
On Sunday, November 3, 2019 at 9:09:08 PM UTC-5, Chris Jones wrote:
On 02/11/2019 12:01, Rick C wrote:
Bottom line is there is no need to pump the epoxy first and then again when adding it to the mold.
You can't achieve a very good vacuum at the bottom of a deep mould that
is already filled with a liquid potting compound. Gravity acts on the
liquid, causing the pressure at the bottom of the liquid to be higher
than that at the top of the liquid.
That is true for various values of "deep". In this case you
will never notice the delta from the top to the bottom of a
less than 1 foot deep mold. Materials like this are the same
density as water to a first approximation. Water is 32 feet > to 1 atm.. So 1 foot of mold will only cause 1/32 of an atm
difference in pressure.
A 2-stage rotary pump like the one I use can very easily achieve 0.01
Torr (though perhaps not with the resin present if it is at all
volatile), which is 1/76000 Atm, so 1/32 is comparatively a poor vacuum.
If there is a certain volume of bubble that is large enough to be
buoyant enough to rise to the surface and pop in a reasonable time (i.e.
before the goo cures), then one would want to make as many bubbles as
possible expand sufficiently under vacuum that they exceed this volume
and rise and pop. Roughly, halving the pressure doubles the volume of
all of the existing bubbles, (though I suspect that there may be some
limitation due to surface tension). If at one atmosphere, a bubble is 32
times smaller than the volume at which it would rise and pop in a
reasonable time, then lowering the pressure to 1/32 Atm ought to do the
trick. If the bubble was smaller to begin with, then it is my
expectation that this level of vacuum would not be good enough to remove
that bubble.
Surface tension might also hinder the escape of gas from crevices in a
device that is already submerged.
When the volume of trapped air expands my multiple factors, it will
help get the air out. Any remaining bubbles will be very tiny when
the vacuum is released. Since these air bubbles will exist whether > or not you pre-vacuate the epoxy, it has no impact on the result.
The size of the remaining bubbles will depend on what the "multiple
factors" are, i.e. how good the vacuum is, at the location of the
bubble. If the spaces between for example turns of a coil, are already
at 1e-3 Torr when you pour encapsulant over them, any bubbles that were
not already suspended in the encapsulant should be very small indeed
after the chamber is vented to atmospheric pressure. I suspect that the
result might not be so good if the pouring is done at atmospheric
pressure and you are then trying to get existing bubbles out from
between the turns of a coil.
Commercial equipment seems to do the dispensing under vacuum:
https://youtu.be/ZWkzDv10J6A?t=140
Depending on the purpose of the potting process you may or may not care
about small bubbles remaining, but some people do need to care.
How can you do better in getting rid of them?https://www.youtube.com/watch?v=8GTvWfwFktU
Putting the two components of the encapsulant under vacuum, stirring
under vacuum then mixing under vacuum, then pouring the mixed
encapsulant under vacuum onto the dry assembly needing potting.
I have never needed to be extremely careful about eliminating bubbles so
I have never built anything to pour encapsulant under vacuum, but I do
degas the encapsulant in a separate beaker before pouring, simply
because it does tend to froth and overflow unless that step is carried
out in a beaker able to contain many times the volume of the liquid
encapsulant.
It won't froth if you don't draw too much vacuum.
If the vacuum is poor, it will not get the smaller bubbles out. If the
vacuum is good, but applied very gradually then that approach might
work, but I find a deep, wide beaker (but filled to only about 1/2 inch
depth) gets the job done quickly without requiring so much time and
attention to avoiding spills.
Perhaps. The commercial units seem to mention 5mbar, and perhaps that isOn Monday, November 4, 2019 at 7:49:59 AM UTC-5, Chris Jones wrote:
On 04/11/2019 16:27, Rick C wrote:
On Sunday, November 3, 2019 at 9:09:08 PM UTC-5, Chris Jones wrote:
On 02/11/2019 12:01, Rick C wrote:
Bottom line is there is no need to pump the epoxy first and then again when adding it to the mold.
You can't achieve a very good vacuum at the bottom of a deep mould that
is already filled with a liquid potting compound. Gravity acts on the
liquid, causing the pressure at the bottom of the liquid to be higher
than that at the top of the liquid.
That is true for various values of "deep". In this case you
will never notice the delta from the top to the bottom of a
less than 1 foot deep mold. Materials like this are the same
density as water to a first approximation. Water is 32 feet > to 1 atm. So 1 foot of mold will only cause 1/32 of an atm
difference in pressure.
A 2-stage rotary pump like the one I use can very easily achieve 0.01
Torr (though perhaps not with the resin present if it is at all
volatile), which is 1/76000 Atm, so 1/32 is comparatively a poor vacuum.
There is your first mistake, thinking there is any real advantage to the better vacuum. The issue is not how low the vacuum can go, it is the difference in pressure to 1 atm. So what you call a poor vacuum will work pretty much as well as your good vacuum in this case.
For the silicone that I have used, I would guess about 3mm-6mm, or sayIf there is a certain volume of bubble that is large enough to be
buoyant enough to rise to the surface and pop in a reasonable time (i.e.
before the goo cures), then one would want to make as many bubbles as
possible expand sufficiently under vacuum that they exceed this volume
and rise and pop. Roughly, halving the pressure doubles the volume of
all of the existing bubbles, (though I suspect that there may be some
limitation due to surface tension). If at one atmosphere, a bubble is 32
times smaller than the volume at which it would rise and pop in a
reasonable time, then lowering the pressure to 1/32 Atm ought to do the
trick. If the bubble was smaller to begin with, then it is my
expectation that this level of vacuum would not be good enough to remove
that bubble.
Yes, well thought out. So what is the bubble size of this threshold?
Surface tension might also hinder the escape of gas from crevices in a
device that is already submerged.
When the volume of trapped air expands my multiple factors, it will
help get the air out. Any remaining bubbles will be very tiny when
the vacuum is released. Since these air bubbles will exist whether
or not you pre-vacuate the epoxy, it has no impact on the result.
The size of the remaining bubbles will depend on what the "multiple
factors" are, i.e. how good the vacuum is, at the location of the
bubble. If the spaces between for example turns of a coil, are already
at 1e-3 Torr when you pour encapsulant over them, any bubbles that were
not already suspended in the encapsulant should be very small indeed
after the chamber is vented to atmospheric pressure. I suspect that the
result might not be so good if the pouring is done at atmospheric
pressure and you are then trying to get existing bubbles out from
between the turns of a coil.
The context is the OP's setup where he is encapsulating an encapsulated device.
Commercial equipment seems to do the dispensing under vacuum:
https://youtu.be/ZWkzDv10J6A?t=140
Depending on the purpose of the potting process you may or may not care
about small bubbles remaining, but some people do need to care.
How can you do better in getting rid of them?https://www.youtube.com/watch?v=8GTvWfwFktU
Putting the two components of the encapsulant under vacuum, stirring
under vacuum then mixing under vacuum, then pouring the mixed
encapsulant under vacuum onto the dry assembly needing potting.
I have never needed to be extremely careful about eliminating bubbles so
I have never built anything to pour encapsulant under vacuum, but I do
degas the encapsulant in a separate beaker before pouring, simply
because it does tend to froth and overflow unless that step is carried
out in a beaker able to contain many times the volume of the liquid
encapsulant.
It won't froth if you don't draw too much vacuum.
If the vacuum is poor, it will not get the smaller bubbles out. If the
vacuum is good, but applied very gradually then that approach might
work, but I find a deep, wide beaker (but filled to only about 1/2 inch
depth) gets the job done quickly without requiring so much time and
attention to avoiding spills.
If the vacuum is "good" it won't get the smallest bubbles out. You have raised many theoretical questions and answered none. This is a practical application where the theory is too complex to be of any use other than a general guide. Then you have missed the lesson I learned by trying this myself many years ago.
I'd like to find a really good book or article on the subject.