Edge 705 Home Made Rain Protector

Edge 705 Home Made Rain Protector

Postby Jamie T » Sun Oct 18, 2009 4:36 pm

Howdy,

I just recently got a 705 and I already knew about the problem with rain affecting elevation readings by blocking the pressure holes. I went for a ride on the weekend with a lot of climbing and it rained a fair amount and the heights were completely wrong (e.g. I did a 1000m climb and the edge said I did a -10m climb).

Does anyone have a suggestion for some sort of home made remedy to keep the vent holes free of water in rain?

Thanks.
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Re: Edge 705 Home Made Rain Protector

Postby DavidCHall » Mon Oct 19, 2009 2:48 am

Jamie T wrote:I already knew about the problem with rain affecting elevation readings by blocking the pressure holes.


This is new information to me. I assume the 305 would also be affected? Do you have a link to any discussion on this point?

Jamie T wrote:Does anyone have a suggestion for some sort of home made remedy to keep the vent holes free of water in rain?


I'm not comfortable with getting my Edge 705 wet at all: Never mind what Garmin say about it's tolerance for water.

My solution is to put the whole device in a small ziplock bag, and put the whole thing in my pocket: Which has worked well for me.

Intuatively this should influence the barometric readings (because the bag is airtight, and and pressure on the bag would give a pressure spike), but practically I've seen no evidence of strange elevation readings or data.
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby Jamie T » Mon Oct 19, 2009 4:33 pm

Hi David,

Sorry, I can't find the original discussion - I think it was on the old motion based forums. I definitely remember a couple of users saying that when the unit gets wet water blocks the hole between the inside and outside and means that there is no pressure change in the unit when you climb. One of the users said that they just wrapped the unit in a plastic bag.

I'd upload an image of the profile of my ride but I don't know how - the first part of the ride was fine and the elevation profile is correct - the second half was heavy rain and there is little elevation change when there should have been about 1000m or so which just confirmed the other posts I had seen before.
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Re: Edge 705 Home Made Rain Protector

Postby CHnuschti » Tue Oct 20, 2009 10:36 am

Jamie T wrote:Does anyone have a suggestion for some sort of home made remedy to keep the vent holes free of water in rain?

Did never test, but I can imagine that if you attach/glue a piece of somewhat highly transpirative material (as for example used for air cleaners/filters) over and around the holes of the barometric sensor, this could prevent water to obstruct the tiny holes. I can imagine that if you make such a piece big enough, the water would drip down at the edges of the piece following the gravitation, so keeping it away from the holes.

Maybe attaching a small section of a pipe around the holes also could do the job.

regards

Edit: Glueing a female screw might be a quite simple solution:
Image
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Postby Jamie T » Tue Oct 20, 2009 6:56 pm

Thanks for the tip. I might try the nut idea.
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Postby DavidCHall » Tue Oct 20, 2009 11:08 pm

Jamie T wrote:I'd upload an image of the profile of my ride but I don't know how


See the FAQ section...

Jamie T wrote:the second half was heavy rain and there is little elevation change when there should have been about 1000m or so which just confirmed the other posts I had seen before.


Water blocking holes on the barometric sensor may cause the elevation reading to be a few meters out. You might also loose some finer profile detail.

But 1000m is a lot to loose. Even if the barometric sensor failed completely there is still the GPS input, which more than capable of tracking down to a few meters accurately.

Something else is wrong here...
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby CHnuschti » Wed Oct 21, 2009 9:42 am

At least here in the garmin forum, some people confirm altitude issues related to rain: https://forums.garmin.com/showthread.ph ... ight=water
Picture of the threadopener there isn't too dramatic IMO, but seems to show that, assuming the gps altitude sensing would continue to work, altitude still might be relevantly off.

Woud be indeed nice if you could attach your picture Jamie T to get a better clue.

regards
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Postby DavidCHall » Wed Oct 21, 2009 10:07 am

CHnuschti wrote:Woud be indeed nice if you could attach your picture Jamie T to get a better clue.


Yes: Please post an image.

PM me and I will respond with an email address: Mail me your .TCX file and I can confirm your data issue, as well as, if you like, post an elevation image.

BTW: Might also be useful to run elevation correction on this activity and compare elevation data: This will tell another story...
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby Jamie T » Wed Oct 21, 2009 2:30 pm

Hi guys - thanks for the response. I've attached the elevation profile from sporttracks. As you can see the elevation was plotting fine until the 25km mark. Checking against my topomap data I descended down to about 180m, rode down a flatish river valley for about 4km and then climbed back up to an elevation of about 500m and then dropped back down to 30m. I incorrectly stated that I lost 1000m before - I meant to say that I did 1000m of climbing.

It started raining quite heavily at about the 20km mark so I assumed that was the problem. I have been for a couple of rides since and it is plotting fine again.

David I already have your e-mail after you kindly gave me one of your tcx files to play around with. I'll e-mail you the file but don't worry about spending much time on it.

I've also attached the profile after using the elevation correction plugin just for comparison.

Image

Image
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Postby CHnuschti » Wed Oct 21, 2009 4:50 pm

Image

I've made a comparison with SRTM elevation data (blue=original, red=SRTM).
As you can see elevation is way off beginning at approx. 18 km. Judging the picture it also indicates that gps elevation sensing, supposed to be at least able to continue to work once the barometric sensor is obstructed is no help to correct the elevation. The hole activity duration is more than 3 hours, so it also seems not to be a question of "missing time" for the gps elevation sensing to become active.

One can speculate if the jumps in elevation at 23 and 38 km (differing approx. 1:05 hours) are the timepoints where a correction of the elevation over gps sensing occours. You can interprete that the correction here jumps back to the original altitude at the timepoint where the barometric sensor ceased to work correctly.

regards
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Postby DavidCHall » Wed Oct 21, 2009 9:09 pm

CHnuschti wrote:I've made a comparison with SRTM elevation data (blue=original, red=SRTM).


No doubt: something went VERY WRONG with your activity elevation data!!!

Thanks to this excellent analysis we can even see how wrong, and exactly where. (Isn't ST great?)

But now we are left with a vexing question: WHAT happened?

All I have is a few theories: {Supported by a look at the .TCX data}

1. Your Edge 705 has some fatal elevation firmware bug: Are you updated and running the latest version?

2. In the rain / terain your Edge lost satellite contact, and hence altitude information, but maintained 2D accuracy. (Clutching at straws here, this theory just doesn't hold up)

3. The water blocked your barometric pressure holes. (I really doubt this: The altitude loss is so great that the pressure would be considerable, certainly enough to clear a bit of water: Almost enough to clear chewing gum)

4. As you started to desend in the rain your (wet) Edge drew a few drops of water through the pressure holes: This wet the sensor, adding conductivity where it is not welcome, and confused the Edge firmware. (As bizarre as this one sounds I think this is the likely explanation)

Solution: Put your Edge in a ziplock bag...
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby Jamie T » Wed Oct 21, 2009 10:24 pm

DavidCHall wrote:
1. Your Edge 705 has some fatal elevation firmware bug: Are you updated and running the latest version?

Edge is only a month old and it came with 2.9 firmware uploaded. I have done approximately 20 rides with it and elevation has been plotting perfectly.

2. In the rain / terain your Edge lost satellite contact, and hence altitude information, but maintained 2D accuracy. (Clutching at straws here, this theory just doesn't hold up)

Satellite visibility was good for all the track.

3. The water blocked your barometric pressure holes. (I really doubt this: The altitude loss is so great that the pressure would be considerable, certainly enough to clear a bit of water: Almost enough to clear chewing gum)

I don't know enough about it to add anything useful to your item 3.

4. As you started to desend in the rain your (wet) Edge drew a few drops of water through the pressure holes: This wet the sensor, adding conductivity where it is not welcome, and confused the Edge firmware. (As bizarre as this one sounds I think this is the likely explanation)

That is possible.

Solution: Put your Edge in a ziplock bag...


Will do and thanks for the help :)
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Postby hubrat » Thu Oct 22, 2009 5:10 am

I've had a very similar occurrence on one ride a few months ago. Descending in the pouring rain when I got an instant elevation jump which then remained at a constant elevation for the duration of the ride.

Other rides have been fine since. I'm still on 2.6 firmware.

Elevation Correction plugin has corrected that particular ride.
hubrat
 
Posts: 18
Joined: Mon Oct 05, 2009 4:48 am

Postby DavidCHall » Thu Oct 22, 2009 7:20 am

There are a few threads (in various places) that mention altitude / elevation issues with an Edge 705 when wet:

1. I would expect similar issues with any barometric GPS: But don't find mention of this. {Maybe nobody noticed and posted an issue yet?]

2. The "general" conclusion is that the water blocks the holes: If your readings were erratic and perhaps 10M out I might fall for this: The magnitude of the noted errors means this has nothing to do with blocked holes.

3. My conclusion, based on the fact that the common denominator is rain, is that the real issue is something inside the Edge 705 getting wet. I might be wrong, but my conclusion sounds reasonable until somebody has a better theory.

Conclusion: If it's raining, keep your Edge dry!!
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby Jamie T » Thu Oct 22, 2009 1:57 pm

Funnily enough I had the exact same thing happen again in the same place on the same ride last night! Climbed to top of hill and had torrential rain along the top and I then descended back down again. Elevation profile showed that I stayed at 400m when I actually dropped from 400m down to 30m. I checked the back of the edge at the bottom of the hill and all eight pin holes were blocked with a film of water. I got home and dried the holes off and then watched the elevation rapidly drop from 400 back to 30m. Conclusion - water is blocking the holes which is preventing pressure change.
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Postby CHnuschti » Thu Oct 22, 2009 4:53 pm

DavidCHall wrote:3. The water blocked your barometric pressure holes. (I really doubt this: The altitude loss is so great that the pressure would be considerable, certainly enough to clear a bit of water: Almost enough to clear chewing gum)


IMO water blocking the holes, hence preventing a correct measurement of the barometric pressure, is still the most probable cause. I don't think this is an issue of the edge 705 alone, as far as I remember, also in the old motionbased forum this issue was addressed for the edge 305.

Also judging Jamie's records this looks very coherent, elevation record goes wrong together with the stated begin of the rain at km 20.

And hey, this is my professianal field. There are some scientific facts which underline this theory too.
Water has a relatively high surface tension, which among other leads to the effect of the capillary action.
As known drops may hang downwards on a ceiling, this is one of the effects of the surface tension. When raining, drops will flow down to the holes of the barometric sensor on the edge. Once they are there, they are literally sucked into the holes through the capillar action.
I measured the holes of the edge 305, they are around 0.7mm, which gives a capillar height potential of 40mm, so you get the idea of the existing forces.

Now it is in theory true that if you descent some hundreds of meters the atmosferic pressure indeed changes enough to be able to push in (out if ascending) the water from the holes. However, this is theory, if it continues to rain there is a constant supply of water to be sucked into the holes again. The capillar forces also may prevent the water from being removed by the change of the atmosferic pressure alone.

Well, some numbers. 10mm water height (cause the pressure of) = 1 mbar. Atmosperic pressure diminishes by 10 to 12 mbar per 100 m ascent (nominal 1013 mbar at sea level).


Played around a bit with my edge 305, immersing it into 7 cm of water (pressure change is equivalent to approx. -70 m). The numbers were admittedly quite inconsistent, in some cases going even opposite to the expected direction, maybe because the test was indoor without gps, also it's not cristal clear if there is a difference in behaviour if the unit is recording or not. However, one consistent observation was that the elevation (=pressure) does almost not change, once the immersed unit is taken to the air again. Looking at the holes, while it is dry around it, it looks like the water is still kept in the holes. If you blow out and "clear" the holes, a significant change of the elevation occours.

As conclusion, for me the theory of the water obstructing the holes of the edge seems to be the most consistent explanation.

regards


P.S. Jamie, would you mind to send me the tcx of your last ride mentioned. Would like to look at it. Thanks.
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Postby DavidCHall » Thu Oct 22, 2009 11:13 pm

CHnuschti wrote:P.S. Jamie, would you mind to send me the tcx of your last ride mentioned. Would like to look at it. Thanks.


Add me on the cc if you don't mind: This is an interesting academic exercise: It would be nice to find an agreed conclusion.
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby DavidCHall » Thu Oct 22, 2009 11:31 pm

We can agree the a wet Edge with barometric altitude is likely to poor altitude / elevation readings.

The unresolved question is why are the Edge readings wrong when wet?

I'm not very comfortable with my theory that the water somehow affects the barometric sensor output: Perhaps the sensor is sensitive to water? Or the sensor conductivity changes? (Assuming the sensor produces an analogue output) It's a theory, but I'm not comfortable with all aspects.

But I'm equally not fully comfortable with the going theory of water causing a blockage:

At some point this problem causes 500M of elevation differance / error, as shown by the elevation correction analysis produced by CHnuschti: 500M Elevation translates to an air pressure error of 5.9KPa. {See here}

5.9KPa Translates to a column of water 600mm high.

At best the holes are 5mm long. (An earlier post mentioned 0.7mm, I'm just rounding up and allowing for some other length (worst case) than might not be obvious)

Add 45mm for surface tension and capillary effect.

How do we account for the other 550mm? Hence my discomfort with the theory: How does the water block that much pressure? {Unless there are some very small internal holes: In which case the capillary / surface tension forces might be much more significant}
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby Terry Dickerson » Fri Oct 23, 2009 7:53 am

DavidCHall wrote:[color=blue]{Unless there are some very small internal holes: In which case the capillary / surface tension forces might be much more significant}


When I took my 305 apart I seem to remember there was a membrane covering the inside of the holes. This is just from memory (~1 year ago) though.

TERRY
Terry Dickerson
Cambridge, UK
Terry Dickerson
Donated!
Donated!
 
Posts: 49
Joined: Mon Jan 21, 2008 12:46 pm
Location: Cambridge, United Kingdom

Postby texmurphy » Fri Oct 23, 2009 8:22 am

Assume the logic and hardware of the 705's barometric altimeter is similar to the 305's:
The 305 will sometimes, for days on end, show near same elevation for a given turn on location...
...and then on the next day be wrong by 10's or 100's of meters; and
When off by 10's or 100's of meters the 305 can take several hours to recalibrate to the correct altitude.

So a single heavy rain event, where possibly the holes are fully occluded by water or an electrical event happens, that throws off the altimeter calibration, might result in several hours for recalibration.
Some reports indicate that power off/on may clear the altimeter to a better state (MotionBased Forums).
User avatar
texmurphy
Donated!
Donated!
 
Posts: 1871
Joined: Wed Jul 05, 2006 7:38 pm
Location: Maryland, USA

Postby DavidCHall » Fri Oct 23, 2009 8:22 am

Terry Dickerson wrote:there was a membrane covering the inside of the holes.


That might do it....
User avatar
DavidCHall
Donated!
Donated!
 
Posts: 977
Joined: Mon Sep 24, 2007 8:29 am
Location: Cape Town, South Africa

Postby CHnuschti » Fri Oct 23, 2009 10:35 am

Image

Thanks for the file (blue=original, red=SRTM). You seem to have tough rides in your training plan. ;)

Rain began at approx. km 17 according to Jamie. Also here a similar picture like the other shown comparison, the elevation is not able to diminsh according to the ride. Only one clear jump in the elevation can be seen once the record began to be incorrect.
IMO these jumps are interesting since they possibly can give some indications about how the coupled measurement gps+barometric sensor does work. It is still unclear. There were some suggestions there is a correction over the gps elevation measurement at regular intervals, some postings about this guessed every 15mins. On the other hand, once the measurement is messed up, everything seems to go havoc anyway, doubtable if these jumps have to say anything, since you would at least expect the would go in the correct "direction", but not even that is the case



DavidCHall wrote:How do we account for the other 550mm?

As stated, I also think a clearing of the holes through the atmosferic pressure change is thinkable, however this is assuming a single, short event of rain. If the rain continues, there is continous pouring down of water which can obstruct the holes again once they are possibly cleared. Also you have to take into account that at descents like in the 2 examples, the pressure increases, hence you might expect the water in the holes to be pressed in, so it would be still present within the unit.
I also assume the height of the holes to be quite small, 5mm probably is an overestimation, wouldn't be surprised if it is 1-2mm at best. I would not dare to make a quantification of the capillar effect in numbers, but I am confident its effect of the water being "automatically" sucked in occours.

JOKE: so please Jamie next time start when it's ALREADY raining :D and ride straight upwards in order we can analyze also that case. ;)

regards
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Postby Cinderblock » Wed Nov 04, 2009 12:10 pm

The error is quite dramatic.

Here's Columbus Fall Challenge, Saturday, per my 705, for a claimed total of 3414 feet total ascent.

Image

Here's the same ride per RideWithGPS.com, for a claimed total of 8619 feet (reasonably close to what the organizers claim).

Image

It was actually raining pretty much all day, but as you see, it took a couple of hours before the holes got clogged.

For comparison, here's the ride using ST with the Elevation Correction plugin, for a claimed total of 13,148 feet.

Image

The profile is reasonably close, but as much as I'd like to believe it and put it in my training log, nobody claims it's 13,000 feet.
Cinderblock
Donated!
Donated!
 
Posts: 10
Joined: Tue May 20, 2008 6:24 am

Postby Jamie T » Fri Nov 06, 2009 10:42 pm

The elevation correction profile is based on data that has a low precision. You can see all the lots of little ups and downs when you're really doing a steady climb so they add to the total climbing. If you're finicky you can manually smooth out the profile by altering the faulty nodes.
Jamie T
 
Posts: 18
Joined: Sun Aug 16, 2009 8:04 pm

Postby CHnuschti » Mon Nov 16, 2009 2:54 pm

Check out also this post regarding problems with water that might have entered an unit.

regards
CHnuschti
 
Posts: 280
Joined: Tue Feb 13, 2007 3:28 am
Location: Switzerland

Next

Return to Garmin

Who is online

Users browsing this forum: No registered users and 0 guests