Venus GPS - encapsulated in epoxy....

GPS/GNSS Development Boards
1

My quest to create a submersible/waterproof gps has me wondering where I’ve gone wrong:

I bought both the Venus gps and the recommended Vtorch antenna with SMA connector (VTGPSIA-3). My plan was to encapsulate both GPS and antenna in a block of, well, encapsulant and have a piece of glass for the antenna to look out into the sky.

I attached the antenna to a small (2" x 2") piece of glass by applying epoxy around the edges of the antenna. I tested the GPS/antenna and everything worked 100%.

I then put the glass/antenna plus Venus gps into a small container with the glass firmly attached to the bottom of the container with wire leads from the gps extending up above the container. I then poured an encapsulant into the container.

I ended up with a block of encapsulant with wire leads out one side and the other side with the antenna facing out the glass ‘window’.

Everything looked good. No encapsulant got between the antenna and glass. All the leads are sticking out of the encapsulant just fine. BUT - the unit used to be really fast to pick up a signal. Now it can take minutes or won’t connect at all!

The Venus gps can still be programmed and sends out a signal. But it only sportatically picks up signal.

My questions:

  • it looks like the base of the antenna (radome part) might be touching the SMA connector, could the is be the problem?

  • would encapsulant somehow get inside the antenna and cause the problem?

  • is there a better way to make a gps/antenna waterproof (submersible to say 10 metres?)

Any/all feedback appreciated.

http://i754.photobucket.com/albums/xx18 … /Block.jpg

2

I was reading a topic on this yesterday, and it had doubts as of to the waterproofness of the epoxy (eventually the epoxy can detach from wire leads and allow a possible entry point). As I can’t exactly help you in regards to the epoxy, a possible solution pointed out was to simply stick the thing in a [OtterBox. Should do the trick, and they are relatively cheap.](http://www.otterbox.com/waterproof-cases/)

3

FYI, the GPS Satellite signals probably won’t go through more than a foot of water. So you wouldn’t get any data while diving anyway. Also silicon rubber adhesive/sealant works better as a waterproofer, since it is more flexible that epoxy.

4

By encasing the antenna in epoxy; I suspect you’ve screwed up its properties. The antenna isn’t just that center bit of metal - its the whole ceramic block.

–David Carne

5

hello lowbattery

Unlike many other GPS you will find, the skytraq is not shielded. When you use an amplified antenna so close to it there is a chance of getting into an amlification loop (dont know how better to explain it). Best thing will be to use a passive patch antenna instead of active antenna and remove the bias inductor on the board. If you have to have to use the amplified antenna then do two things that should improve quality

  1. place the antenna furter away from the GPS board

  2. Make sure that none of the epoxy actually goes into the rear circuit of the amplifier stage of that antenna. This is usually a saw filter design and expect certain amount of air around it to work perfectly at the required frequency.

cant think of any other thing that would effect your results other than properties of the epoxy. Try changing to another brand and see if that makes a differece (if all else fails)

6

The dielectric value of epoxy & glass is quite different to air (which the ceramic antenna is designed to operate in). Thus, the antenna is probably being “de-tuned”, as well as having some of the signal blocked.

I would think a better option would be to spray the electronics with a “conformal coating” and then mount the assembly in a plastic box with gasket.

7

yes, detuning can be a problem,

if you have access to a network analyser ( university lab) then you can look at the characerisics of the antenna when encapsulated and in free air. if there is shift in frequency then usually it is possible (by luck only) to us a grinder and grind away the corners of the metalization on the top of antenna until it is detuned by the same amount in the opposite direction. Now when you encapsulate it, the epoxy will bring it back into tuning.

8

It won’t just be the antanna that’s the problem - Encapsulating any sort of RF device is a really bad idea. Even UHF tx/rx modules don’t like it. The encasulant’s dielectric properties mean that things like the stray capacitance of tracks and between coil windings will be different from their carefully-designed values in air.

If you really must encapsulate ,you need to create a pocket inside for the module, or use one with a sealed (or selable) screening can, so the compound never touches critical parts of the PCB or components

9

There are a number of good points raised here, but “test list” would be:

-Proximity of antenna to board

-SMA connector touching ground plane of antenna

And lastly something I know a little bit about…

I would have a good hard look at the material that you potted the unit with. There are potting compounds designed for just this task that are non conductive and can be applied directly to boards with no effect. They are available under the Hysol and Loctite brands. If you want the “really good stuff” have a look for Raychems S1125 potting compound. It the best available in our experience.

The suggestion of an otter box is also a good one. Depending on your application and requirements for weather proof qualities there are a number of good waterproof connectors available that could satisfy your requirements. Tell me a bit more here and maybe I can help.

10

Another Venus gps and Vtorch antenna should be arriving any day. I’m still open to suggestions on how to encapsulate this setup.

I want this gps to be able to withstand being submerged down to a depth of 10 metres so I don’t feel too secure about an Otterbox with connections drilled through it. I used a proper encapsulant (MC Chemicals 832C) that is non-conductive so I don’t think the properties of the encapsulant are an issue (although where the encapsulant flows into is a separate issue).

I used plain two-part 5-minute epoxy to adhere the antenna to the glass (see original pic) before pouring in the encapsulant. The antenna worked just fine glued to the glass (so glass itself wasn’t an issue). I also think this prevented any encapsulant from entering the antenna but I can’t be absolutely certain.

Next time:

  • antenna will be farther away from gps, antenna cable will be directed away from gps

  • try to prevent ingress of encapsulant into antenna

  • encapsulate just the antenna and test before encapsulating combined antenna/gps

  • an alternative would be to create a ‘bubble’ around the antenna before encapsulating (although I’m not quite sure how I’d accomplish this)

I appreciate all the suggestions.

11

lowbattery:
Another Venus gps and Vtorch antenna should be arriving any day. I’m still open to suggestions on how to encapsulate this setup.

I want this gps to be able to withstand being submerged down to a depth of 10 metres so I don’t feel too secure about an Otterbox with connections drilled through it.

But do you expect it to pick up any satellite signals at 10M depth? I highly doubt it will even work at 1M, as the signals are very weak, and impure water is an excellent conductor.

Mike

12

Its not conductivity thats a problem - its the addition of a whole bunch of a dielectric that will change the antenna tuning.

Also, as many have said, it won’t work at any depth.

–David Carne

13

busonerd:
Its not conductivity thats a problem - its the addition of a whole bunch of a dielectric that will change the antenna tuning.

Also, as many have said, it won’t work at any depth.

–David Carne

But isn’t the reason why you can’t (effectively) receive radio signals under water due to it being a conductor and ‘grounding’ out the signal? I wasn’t even going so far as addressing the dielectric properties of the potting compound, mainly because in this case I think its a moot point, since you won’t be able to receive the signal anyway.

Mike

14

Sorry for the confusion; my comments on dielectric applied to the epoxy; and not to the seawater.

I was just trying to say that the epoxy doesn’t really matter anyways, since as you said; seawater kills the signal.

–David Carne

15

I assumed the application was some kind of submersible sensor that dives to a certain depth and returns to the surface to get a GPS fix. Some marine animal tracking systems use GPS, so it is possible.

16

Just as MichaelN said, this is for getting a surface signal but being able to withstand depth.

If I take the advice of keekay_tk and don’t let:

" epoxy … into the rear circuit of the amplifier stage of that antenna "

I assume that I still need to deal with de-tuning. Is this correct?

If simply having encapsulant surround the sided and bottom of the antenna is going to de-tune it, what size of ‘air pocket’ would I need around the antenna to prevent de-tuning?

17

sarantel claim that thier antenna is quite capable when potted (or doesnt get as badly effected as other antennas). Also it comes with a raydome so it can be potted up to the start of raydome palstic and still be waterproof. Try that one of possible.

18

lowbattery:
If I take the advice of keekay_tk and don’t let:

" epoxy … into the rear circuit of the amplifier stage of that antenna "

I assume that I still need to deal with de-tuning. Is this correct?

If simply having encapsulant surround the sided and bottom of the antenna is going to de-tune it, what size of ‘air pocket’ would I need around the antenna to prevent de-tuning?

Yes, you almost certainly need to deal with detuning and how much you need to account for it depends on the antenna design and encapsulant.

Before burying another receiver alive, I recommend a design cycle something like this:

Test the antenna and receiver board in the open air and see how close you can bring them together before self jamming sets in. Don’t use a soft metric like “works ok” or time to acquire because there are too many variables that impact those numbers. Get the system tracking and monitor the SNR (look at the GPGSV sentence if you’re getting NMEA data) as you bring the antenna near the receiver. Unless physical design constraints force the issue, you don’t want to sacrifice any SNR to self jamming, so you don’t want to see those numbers dip at all. Even coming up from depth, you’re going to have water and salt on the antenna cover that will impact your signal strength.

It might be worth shielding that Venus board by putting it in a metal can that you ground firmly (not a weak ground near digital components, try to find where the cable shield connects). Do the same SNR testing with and without the can to make sure you’ve given enough clearance around the components. It’s not just the antenna that will detune-- everything in the RF signal path is susceptible to detuning, including the board traces. Eventually you can put the whole metal can in the encapsulant. Inside and outside that can are nearly different worlds.

You don’t have to worry about detuning the antenna cable- it’s shielded.

Look for a simple way to water proof the antenna alone-- it doesn’t need to be visually transparent. Can you get away with just shrink wrapping the thing with a very, very thin layer of plastic? What about some form of heat shrink tubing? If you can water proof just the antenna element, and then mount the sealed antenna on a metal plate that can act as a ground plane, you won’t have to worry so much about what’s behind the plate. Just keep as much air around the other sides of the antenna as possible.

Do you need to worry about air pockets rupturing your encapsulant? If you can get away with it, you’ll want to leave some air around the receiver board too-- either inside the shield, or if you’ve chosen not to shield it then think about putting it in a small plastic box inside the potting material. That has the dual benefit of helping keep the circuits tuned, and giving you the ability to cut the board out of the potting if things go wrong again and you need to start over.

Once you’ve gotten everything working, without the encapsulant, and you can physically arrange the parts, then try potting and see how it goes.

[/list]

19

just remembered another thing about this. Taoglas used to do a kit of several detuned antennas. These were detuned in the opposite direction, so when you pot it or put it into a box or something, then the antenna comes back into tuning. You may source such a kit and pot all of them exactly alike and see which one works best for you.

ok… found the links …

http://www.taoglas.com/docs/GPD25%20dev … %20kit.pdf

and

http://www.maxrad.com/product_overview_ … _num=11214

20

I’m curious, why not a peice of ABS pipe with tight gasket seals? You can theoretically create something twice the size of the board and antenna, instead of this brick of inactive material twenty time the volume, 40 times the mass of the active unit?