Society of Robots - Robot Forum
Electronics => Electronics => Topic started by: Asellith on April 06, 2009, 01:31:21 PM
-
Ok I live in Florida so this may sound strange but I just got this project dumped in my lap and need some info so I thought I would ask the brain trust here at SOR.
I need to detect if snow has built up inside a satellite dish.
My FM radio station has translators all around the country. They are all feed by satellite and some are on mountains and it costs us $500 bucks or more to send a guy up the mountain if they can get up there. Just to have them call and say oh the dish is full of snow. heaters cost $5000 for these dishes so no use melting the snow or complicating the issue with more equipment that could break. My first idea was to put a light sensor at the bottom to detect light or not but that will not work at night. So my next idea was to use a range finder and detect the distance from the LNB to the dish. When that changes it will send me an alarm.
Does anyone have any experience with detecting snow with sensors? IR doesn't sound like a good idea because it's outside. So I was leaning towards ultrasonics. Also need a way to detect heavy rain if possible.
Thanks
-
a momentary switch that when pushed down with a bit of force sends signal? but ultrasonic is an idea because you can waterproof them. or a weight switch maybe? do you have dimensions of the dish? that would help tonnes. i mean are we talking a small home dish or massive omfg alienzz are comin dish?
-
Maybe two bare ends of a wire, and if there is snow covering them they will conduct some electricity.
-
Maybe two bare ends of a wire, and if there is snow covering them they will conduct some electricity.
snow conducts? oh yea its water...smarty pants. you could have that as a snow sensor combined with a weight switch to see how much there is?
-
The dish is 1.8m in diameter. So not a dish network wanna be dish but not a C-band transmitter dish which is closer to the "aliens are coming" size
-
I would think that there would be an industry "best practice" for this. Perhaps you should find a P.E. in the RF field to consult with. I would suspect that the problems are solved often on a case-by-case basis where some cases the dish is covered by tarp, other cases the equipment can detect the degradation of the signal.
If I were approached with this problem, then first thing I would assume is that "greater minds than myself" have already found approaches to resolving the issue.
ETA: Some dishes can be "greased" with a slippery/oily spray. Other dishes benefit from a "knocker" that whacks the dish, or a vibrating motor.
-
Did anyone google?
http://www.howmuchsnow.com/ping/ (http://www.howmuchsnow.com/ping/)
http://www.google.com/search?hl=en&q=snow+depth+sensor&btnG=Search (http://www.google.com/search?hl=en&q=snow+depth+sensor&btnG=Search)
-
I don't want to fix the problem to much cost. They are only down for a bit when it happens. And these are small stations only reaching a few thousand people anyway. The problem I am thinking I will have with ultrasonics is shooting the beam into a parabolic dish might give me weird readings. I was looking at the Maxbotics sensors but I may have to get a dumber sensor and roll my own code for it.
The big issue I want to prevent is sending someone to the site to just tell me there is snow in the dish.
-
Broadcast really loud music. :-P
The least expensive solution might be a tarp cover or another physical solution. Electronic solutions are fun to design, but need to survive the environment and that might likely break your cost budget.
-
a camera? maybe you could just hook one of those up allowing you to see the dish? so your saying you dont need to clear the crap outta the dish just know there is snow in there?
-
I don't want to fix the problem to much cost. They are only down for a bit when it happens. And these are small stations only reaching a few thousand people anyway. The problem I am thinking I will have with ultrasonics is shooting the beam into a parabolic dish might give me weird readings. I was looking at the Maxbotics sensors but I may have to get a dumber sensor and roll my own code for it.
The big issue I want to prevent is sending someone to the site to just tell me there is snow in the dish.
If you feel you have to use electronics to detect the snow, then put a modulated IR emitter and a detector across the likely snow area, and detect the loss of the IR signal as snow or heavy rain. A beam-break sensor. The detection of an object (snow) in the way could trip a signal letting you know of the obstruction or it could trip something that clears the snow such as a vibrating motor.
-
Perhaps use pressure/force sensors to detect changes in weight an pressure on the disk? Something like these (though perhaps you should go for some bigger ones): http://www.robotshop.ca/force-sensors.html (http://www.robotshop.ca/force-sensors.html)
You would also have to ensure that they are water resistant, though I don't see that being too much of an issue (maybe some heat-shrink or some plastic coverings).
-
gear motions idea is good, you could put a few of the to see how much snow is on the plate and where it is. and gearmotion, great avatar!
-
gear motions idea is good, you could put a few of the to see how much snow is on the plate and where it is. and gearmotion, great avatar!
Thanks man! It took a while before I found and altered what I wanted.
I'm just thinking that any sonar concepts could be hard to implement out in the cold wet environment. But a visible/IR setup could be sealed behind plastic. And would be an on/off detect/not detect, rather than an analog value of How Much that would need a thresahold.
-
I also agree with GearMotion.
sometimes simpler is better
-
And what if a leaf gets over the sensor?
-
Just Building on gearmotions idea a bit, you could have a cheap laser pointer on one side hooked up to a dc power supply not batteries, and a phototransistor on the other hooked up to a digital port on a microcontroller. Then when the light is blocked for longer than 30 seconds or so, it could send out a signal to whoever needs to know, or take care of it on it's own.
And what if a leaf gets over the sensor?
Maybe, you could hook up a camera to send a snapshot of what's going on to whoever would take care of it rather than just saying it's blocked. Then whoever receives the picture can tell wheter or not it's a false alarm.
-
or use the laser pointer to melt the snow :-P
-
actually it just came to me the best option but probably to complicated/expensive would be an array of capacitive sensors along the outside of the dish. When snow stayed in the dish it would change the capacitance for an extended period of time.
Keep in mind that the sensor doesn't need to know right away. Like if I find out 3 or 4 hours later that snow is starting to build in the dish it's not that big a deal. I don't even care as long as my signal gets through. I can check signal strength on the receiver. Also this is a remote location so feedback will be in issue. Actually its like 20 or so remote locations. Possible 42 if its cheap enough and they decide to put it at every station. All I have to communicate with the site is a cellular modem with some digital/analog inputs and outputs as well as USB/Ethernet capabilities but those require a lot of work here to get something setup to monitor those inputs.
I think I got a sensor somewhere. I'll dig it up and take it to work tomorrow and start experimenting with ultrasonics and my local dishes. not sure how to simulate snow but worth a shot.
-
Maybe, you could hook up a camera to send a snapshot of what's going on to whoever would take care of it rather than just saying it's blocked. Then whoever receives the picture can tell wheter or not it's a false alarm.
If you're gonna do that then you might as well just put a surveillance camera above the dish, or just have a camera take a picture every 2 hours or so.
I'm not experienced in this field, but I think you could also make it tell you if the picture starts to change, aka, the pixels that should be gray(the dish) turn white(the snow)
-
Hi,
A large "bowl" or similar form, made out of eg. polyethylene would make a nice cover and can be made in a shape that hinders snow buildup. I have used Tupperware(TM) for a similar purpose on short range (around 20+ km) microwave links with great success, when I was in the TV broadcasting business. Just make sure the chosen type of plastic doesn't attenuate the needed frequencies.
If you really do want an electronic solution, you still need to KISS it.
The only long term viable solution is IMO a capacitive sensor. It can be made in so many ways and can be tailored to the purpose, physically as well as electrically, and there's no beams that can be blocked with a leaf, nor any US-sensors that would probably be quite error prone in your app. considering temperature cycling, humidity cycling etc. (US sensors are electromechanical devices after all).
Do you have the facilities, at the location of the dish, to relay the output of whatever sensor back to civilisation?
-
Not sure if this was suggested yet because im short on time ATM but what about using a some sort of emitter detector pair to detect an obstacle? Have them on opposite sides of the dish maybe near the apex of the curvature? If theyre small enough I dont think theyll affect the dishes performance... maybe even use a small <=5mW laser pointed at it when temps get below like 40°F or something and when the laser isnt detected when its on then theres something on the dish?
Just a thought
hope I helped at all...
-
I wrote a program in VB.NET a while back that could capture an image from a webcam, and compare it to the previous image to see the percent that had changed. If the change was over a certain threshold, the image was saved with a timestamp. Perhaps I could modify it to post the image to a website or something. Just an idea though...
-HyperNerd
-
Wow time flys I got the message that this post has not been used in 24 days. I can't beleive I have been working on this project for almost a month. Anyway new direction..... LASERS!!!!!
So a weather proof laser and a weather proof light sensor should let me determine if snow has covered the dish. I can place the sensor at the bottom of the dish and attach the pointer to the LNB support arms. This would then determine if something was covering the sensor. If I mount the sensor on a small pole I can even determine a few inches of depth. Simple clean and easy.
One Problem....
I found weather proof lasers easy enough but weatherproof light sensors are proving difficult any ideas?
-
I found weather proof lasers easy enough but weatherproof light sensors are proving difficult any ideas?
Mount it in a small, transparent plastic box ?
Chelmi.
-
What if you just coated the non air-tight spaces with Silicon Rubber or something similar? You can get it at most hardware stores I beleive... a plastic box as Chelmi suggested could work but im wondering if it would collect too much more ambient light or deflect too much light away?
-
What if you just coated the non air-tight spaces with Silicon Rubber or something similar? You can get it at most hardware stores I beleive... a plastic box as Chelmi suggested could work but im wondering if it would collect too much more ambient light or deflect too much light away?
If I understand correctly he is using the light detector with a laser beam, so it should be fine.
-
cling wrap? i would say to trial a few things and just use whatever works best..
you could try a plexiglass box?
-
Any luck?
Actually its like 20 or so remote locations. Possible 42 if its cheap enough and they decide to put it at every station. All I have to communicate with the site is a cellular modem with some digital/analog inputs and outputs as well as USB/Ethernet capabilities but those require a lot of work here to get something setup to monitor those inputs.
That many locations, definitely needs to be cheap and mass-producible . . . test it out on a single dish, and when you are happy, do it for all of them.
I'm thinking sharp IR is the way to go. Not hard to test at least . . . Snow absorbs IR, so not sure . . .
Otherwise I like Soerens idea. The Navy does that, covering up radar dishes with bowls. For example, those ball thingies on this ship are covered up dishes:
http://aycu20.webshots.com/image/47099/2003529587801442487_rs.jpg (http://aycu20.webshots.com/image/47099/2003529587801442487_rs.jpg)
(http://aycu20.webshots.com/image/47099/2003529587801442487_rs.jpg)
-
whats up with the picture?
all i got from it was a bunch of the navy standing on a ship ;D
-
whats up with the picture?
all i got from it was a bunch of the navy standing on a ship ;D
The forum auto-shrink image feature broke during the last software upgrade . . . you'll have to use the scroll bar to see the rest of the image for now ;D
-
It's only part of the picture - Follow the link, it shows the picture in whole.
-
It's only part of the picture - Follow the link, it shows the picture in whole.
what? what link?
-
It's only part of the picture - Follow the link, it shows the picture in whole.
what? what link?
Otherwise I like Soerens idea. The Navy does that, covering up radar dishes with bowls. For example, those ball thingies on this ship are covered up dishes:
http://aycu20.webshots.com/image/47099/2003529587801442487_rs.jpg (http://aycu20.webshots.com/image/47099/2003529587801442487_rs.jpg)
-
Hmmm, Im confused with the picture?
-
admin was showing that the navy cover they're dishes with bowls do
prevent... damage?
i think i get it now, thanks guys.
-
I see :) that actually would be a great way to do it, this would keep the snow off of them in general!!
-
does anyone know if it being closed interferes with the signal quality?
-
Hi,
It won't interfere if the material is carefully selected - a metal dome would be a big no-no of course :)
-
If it interfered with the signal strength im pretty sure the Navy wouldnt use it :P Heck im sure if you just Saran wrapped the thing youd be good to go lol
-
If it interfered with the signal strength im pretty sure the Navy wouldnt use it :P Heck im sure if you just Saran wrapped the thing youd be good to go lol
Or like I suggested on the first page - tarp. I've seen many dishes covered in tarp like a drum skin stretched on a frame. Again - A physical solution to prevent snow accumulation might just be the easiest.
-
back to the original problem.
Roombas have a sensor that detects heavy particulates. meaning that if a lot of crumbs hits a sensor then it concentrates on cleaning that area more.
A similar sensor could detect rain and possibly snow drops hitting it. I would imigine its just a simple piezo type sensor with a program that detects high/fast rise peaks...
-
I would see what I could do as far as making something, and then testing it, but unfortunately it is the summer(springish), and there is no snow so I cant :(
-
They make dome tarps designed to do this but they cost around $500 to buy per site. Thats is way out of the budget. The budget for this project is like under $30 because it will cost me about $200 to have the tech install it and ship it. Let alone my time building the device. We just can't justify spending to much money on these sites as some bring in like $1500 a year in contributions and thats barely enough to cover the rent at the site.
-
piezo discs can cost like $1 each...
-
1) Would the current generated by a piezo being struck by a rain drop be enough to read by a cheaply available MCU?
2) I thought piezos only reacted to the intial change in pressure on their surface so like if you pressed down on one with 5 newtons of force and held it w/ 5N of force for like 10 seconds, wouldnt their only be a current from the time it takes the ~0 newton forces on it to reach 5 newtons(such as a fraction of a second if you were pressing on it with your finger)?
If my second question is indeed correct then that method would be entirely ineffective because you wouldnt be able to tell if there was anything currently on the piezo, you would only be able to tell if something was enacting a force on the piezo at that point in time... and even if you had some sort of memmory for it so that when you woke up in the morning to look at the display it said "something hit the piezo" or w/e think of how annoying that would be every time a large gust of wind came by and pushed on the piezo. If its sensitive enough to detect the force of a rain drop, a moderate gust of wind would surely be able to set the thing off.
-
Hi,
As you assume, a piezo only gives a signal when hit (actually, when flexing). The signal looks like a dampened sine, since a hit will vibrate it like that.