Paradise Robotics Forum
Robotic Lawnmowers => Lawnbott => Topic started by: moti on August 23, 2012, 03:45:01 PM
-
I've read on this forum about using multiple perimeter wire loops, and switching between them to direct the robot into different zones. I plan on implementing such a scheme. Since I use an ISY99i home automation controller and Insteon devices inside inside, I could use the Insteon EZIO40 to switch the zone daily or based on any external criteria.
However, it would be nice to control more than just the selection of the zone. For example, it would be great to be able to tell the robot not to go out at all on days when it's pouring rain - or even the day after a rainstorm - since the grass is soaked. The built in rain sensor sometimes brings the robot back, but usually doesn't stop it from going out initially. Manually pressing pause to keep it from going out is somewhat low-tech.
Sending signals via the remote is limited - and even if I could mimic the frequencies with some other transmitter, the range of transmission is limited. Also, the codes are not discreet - sending a "pause" just toggles the state, so you'd have to know what the current state was, and this is not very accurate.
In the case of not going out in the rain - if the behavior of the robot was such that it wouldn't go out if the perimeter wire was non-existant, then I could use one of the EZIO40 relay positions with no wire attached to simulate an empty zone. Then, if the automation controller detected bad weather (the ISY 99I controller has a feature to obtain and react to local weather information), it could switch to the empty zone.
Also, it would be nice if the robot could message or alert that it was stuck. If the robot had some device onboard that would xmit its position, then we could poll it periodically - if the position hadn't changed in some period of time, we could alert.
Has anyone successfully controlled the robot in some other way? If so, please share.
-
If your Insteon EZ1040 can control a wireless DC powered unit that could run off the bot's battery and does a On/Off function then consider having it short out the rain sensors. If the Bot still comes out make sure that you have a very small area also selected that would send him back after traveling 10 or 15 feet. The area would have to catch him whether he was to follow the wire to an area or just start mowing.
You could use it to stop him from mowing the day after a rain storm.
Any time you want him to not mow just short out the rain sensors. Heck, that would even allow you to limit his time in a given area during hot days, etc.
-
I have recently found that my 3250 will not back out if the electric to the charger and transmitter is shut off... i don't know if that will hold true for the 3550. When I get it back from the repair shop will test it.. If that hold true you could easily just control the power to the charger.
-
Since the bot would still be in the charger with the charger off, wouldn't that put a load across the battery, which in turn run the battery down?
-
This is a great discovery. It's stormy here today, so I can't test this - but I'll investigate tomorrow.
-
Has anyone done more experiments this year?
-
Hi Guys - I have successfully controlled the robot remotely similarly to what you have described Moti. There are other posts on this forum which go into detail what I did, but I basically put a wire harness in the bot connected to the buttons on the back and hooked up an RF transceiver with the wires running to it. By sending an RF signal to the transceiver I was able to simulate pressing any button on the back of the robot.
I automated it by using x10 - I had an x10 program on my PC which I could control from my iPad or iPhone which triggered a signal to be sent to an RF transmitter - I had 4 transmitters - one for each button on the robot (on, off, charge, pause). The transmitter would send a signal to the robot, the transceiver picked it up, closed the loop on the wires simulating pressing the button and it worked great. I had a fairly good antenna on the robot to pick up the signals and I was able to control the robot anywhere in the yard.
I did put an outdoor camera close to the charging station so I could see what state the robot was in when it was charging (paused/unpaused). I was able to see this on my i-devices as well.
A couple of challenges with this:
1. The RF receiver required 12v power. There were two options - hook into the robot battery and put a device in the middle to regulate the voltage down to 12V or put in a separate batttery. i tried both and settled on trying to use the robot battery power since a manual recharge of a separate battery got to be a pain. The challenge was that I didn't have a cutoff mechanism to tell the RF receiver to turn off if the bot's lithium battery got too low. One time the robot got stuck and I wasn't there to put it back in the charging base and the RF pull on the batteries pulled the charge on the bot's battery below 24v and ruined it! Expensive mistake!!!!! Lithium batteries do not like going below a certain voltage!!
2. Figuring out what state the robot is in and did it actually receive the signal and execute it. I was able to see the robot when it was in the charging base but if I was away from home and wanted to control the bot through my phone (which did work - I could connect anywhere and submit x10 commands) but when it was out doing it's thing I couldn't track it as well. So ........
I also put a camera on the robot with the same setup. It was an 8v camera and I used a regulator to convert the power from the robot down to 8v. I was able to do the same thing as above and remotely turn the camera on and off from my i-devices. It actually worked very well - I took the plastic off of the light hole in front and mounted the camera out that hole. I could now see what the robot was doing easily.
The challenge here was the same - I had an extra receiver to control the camera which drew power and those regulators - they draw a small amount of power as well. So they just added to the battery rundown issue if the robot got stuck and couldn't recharge in time.
SO - the solution to these issues is either a voltage cutoff device of some type to disconnect all of these devices from the robot battery to avoid battery failure or a separate battery (non-lithium) which doesn't have the cutoff requirement and a way to recharge that battery easily with little to no manual intervention.
If someone has any great ideas about how to fix these challenges I am all ears. I would love to put this stuff back in operation because it is pretty cool. I still have the setup in place but I removed all of the extra devices from the robot because I didn't want to fry any more batteries :).
Does anyone know of a small voltage cutoff device that could be used?
-
Don't know if they make a switch that if you send a current to it the switch will turn On then when it receives a second shot of current it turns Off. If they do then you could use it to turn the power on or off after it recharges. Like a push on, push off switch. Not automated but you could control the on/off times of the charging period using your x10 receiver?? Just a though. Something like this:
http://www.discovercircuits.com/H-Corner/on-off-switch.htm
.
-
Check this out:
http://www.newmarpower.com/DC_Power_Accessories/DC_Power_Accessories.html (http://www.newmarpower.com/DC_Power_Accessories/DC_Power_Accessories.html)
This LVD device would work because you can adjust the cutoff and reconnect voltages. Only issue is the size of it - 5.25" x 5.25" by 3.5". Not sure if it fits in the robot ..........
-
If you can find a box that is water proof that would hold the unit you could mount it on the carrying handle. May have to delete the alarm on the handle if it has one. OR how about just attaching a metal plate to the top cover such that it could hold the new box. Have it extended back far enough that you can still get to the control panel. Maybe have it rotate out of the way when using the panel.
If you get all of this working make up a how-to and parts list with wiring diagram for all of us. Maybe Robotlady could sell it for you.....
-
I've thought about this some more and the issue I still see is that even the LVD I posted the link to above has a current draw even when the load is disconnected. So, if the robot got stuck for a period of time and wasn't brought back to the charging base you could still run into a battery failure. The current draw disconnected is only 50mA (milliamp), so it is not huge, but it is still a concern.
The best case scenario would be a battery with an onboard PCM board that shuts everything down when the battery gets too low. These boards pull current in the microamp level disconnected so would be much safer.
The issue is that the Lawnbott battery does not have a PCM built in and relies on the software for overcharge and discharge management (the charger may protect against overcharge - not positive on that one).
A potential solution but more expensive is to put a different battery in the robot (LiFePO4 32V for example or 29.6V Li-Polymer) that has a PCM built in. I think I could run a separate charger and connect it to the charging plates of the base and leave the other wires as-is for the signal . . . . That way the battery itself would cut off discharge when it needs to, the new charger would be designed for the new battery type, and the signal would still go through the wire. I think the existing battery for the 3250 is between 6 and 8 amp hours (Ah) from what I remember. I'll take a look around for a battery close to this Ah with the right size to fit in the battery compartment and see what the price is.
With that said, last time I researched this, I could not find a battery that would work and I thought about building my own. If I only had the time . . . . . . . . .
-
Maybe mount a small solar panel on the back by the faceplate and charge a separate 12V battery pack with that? Don't know how much current you need, sorry if I missed that.
-
Hi, we are carrying some new products that can help you with your projects:
Arduino (easy-to-program controller)
Raspberry Pi (low cost single board computer)
Many electronic parts
Mini solar panels
The new web site is under development. Meanwhile, call and we will walk you through what we have.
RobotLady
-
I just bought an LiFePo4 battery that is 25.6V and 4.5Ah. It should work in the robot - the max charge on it is 29.4V and it is rated at a higher discharge amp level than the stock battery. I got this one just for testing purposes because it was cheaper than buying one that is 10Ah. I am going to test putting it in the robot and running charge wires out to it from a charger designed for this battery. The battery has a PCM on it that is supposed to protect from over discharge below a certain level.
So, I will hook up all of the stuff I had on the robot to control it remotely including the camera and see if the battery holds up. What I'll do is run the robot until the battery is low and it wants to charge and then I'll leave it out of the charger for a bit and let the additional components pull more from the battery. Hopefully the PCM will auto cutoff when the cells get too low.
If this works, I will think about providing a parts list and instructions. Who knows, maybe I'll sell it as a kit.
-
Update - I got the new battery and put it in the robot yesterday and it worked fine! I haven't run new charging wires to the charging base yet so I put a piece of wood in the entrance to the base so the robot can't enter to recharge. Since the new battery is LiFePo4 I can't let it charge with the base charger. Although now that I think about it, it would probably be ok, but I am not willing to risk blowing up or flaming the robot :D.
It actually ran for a good 3 hrs even with a lower amp hour rating than the stock battery. I have put all of the extra hardware back in the robot (camera, voltage regulators, 2 RF receivers, and the wire harness to connect to the buttons on the back panel), but I haven't connected them up yet.
Last night I brought the robot into the garage to recharge the battery and let it charge overnight. This morning I checked the voltage display on the robot panel and it reads 30.01 which is a little higher than the standard but should be fine. My guess is that when I run new charging wires to the robot base and let the robot do its thing, that the software will send the robot out when the charge level reaches 29.3V or somewhere close to that.
I have got to do something to organize all the extra wires in the robot - it is a mess in there!
Next step is I will run wires out to the base and disconnect the base charging connectors on the charge plates and make sure the robot behaves normally with this setup. Then I will reconnect all of the extra hardware to the new battery and see how the new battery reacts with the extra stuff pulling power constantly. Hopefully the battery will automatically shut down when it reaches its cutoff voltage due to the PCM.
Fun stuff!
Paul
-
This afternoon I ran new wires out to the charging base, disconnected the stock wires from the charging plates and connected my new wires to the plates. I connected the new LiFePo4 charger, let the robot go into the base and it worked great. The robot is now charging with the new battery in the base.
I am curious to see when the robot leaves the base - I am guessing when the charge level is 29v or above?
This new battery can only charge with a max of 1.2amps so the battery charge time is longer than the stock battery. Once I test all of the components and everything works I will likely get a higher amp/hr battery and a charger that can charge at a faster rate.
One thing to note is that the current LiFePO4 battery is 4.5amp/hrs and it is basically the same size as the stock lithium battery which is 6.9amp/hrs. So I will likely be modifying the inside of the robot slightly to accommodate a slightly bigger battery pack.
Paul
-
The robot has been working fine all weekend.
I let the robot work until it was ready to charge on Saturday and then I hooked up the RF receivers and the camera on the robot and left it all on overnight in my garage - the goal here was to run the battery all the way down so that the battery protection circuit would kick in and protect the battery and not let the cells go below a safe voltage. I checked the battery in the morning and none of the devices were receiving any power in the robot - that was good.
I then put the robot back in the base and held my breath as I turned it on .......... And it started right up normally - the battery voltage was much lower than normal for this battery, but it charged fine and kept on going.
So - everything appears to work great with the new battery. Next step is to reconnect everything and test the camera and make sure the receiver is working and that I can control the robot from my phone and iPad.
Is anyone interested in trying this? If so I'll take some time and put together a parts list and a manual.
Paul
-
Paul has taken this in a different direction than I did : - )
I think everyone using this forum would agree that Zuchetti should have made this platform more open - as it would be nice to tweak both the hardware software. While I used to be a tinkerer, I seem to have much less energy to apply in that direction now. Also, since the Lawnbott is expensive, technological, and proprietary, I want to make sure I don't get into any warranty issues if something does need to be replaced (maybe RobotLady could weigh in on what does and does not void any warranty). I would love to be able to control the robot remotely with a PC or more powerful remote, and I'd really like to get sensor data back... maybe after the warranty has run out. For now I limit my hacking to the more basic components, and to external methods to influence the robot.
In that regard, and as a follow-up to my initial post:
Switched wire loops:
I have implemented two wire loops, and thy are connected through a DPDT toggle switch that I flip daily or whenever one loop needs more attention. My property is 1 acre with hills - but given the house, driveway, and mulched beds, the robot mows about half of that. I initially planned (and tried some of) a single loop with 4 zones - but the 2 loop system offers more advantages (which I could outline in a separate topic). As I first mentioned, I plan to control the switching of the loops with an Insteon relay switch, which I will locate in a weatherproof box at the point where I plug in the robot's charger.
Location tracking:
The robot DOES get stuck, and I don't always check to see if it has returned at the end of the day - so this is where technology could help. There are a new group of tiny stick-on devices (Button TrackR, Tile, Twine) on KickStarter that look promising. Button TrackR is the size of a quarter, uses a coin-cell battery, and uses bluetooth to communicate with an iphone app. Some of the gadgets use wifi instead of bluetooth. I'm hoping to leverage one of these devices.
Rain sensing:
I still don't have any great solution for rain. The on-board rain sensor works, but the options are limited. In a deluge, the robot still goes out, and then comes back. I'd like to keep it from going out at all in the rain, and even keep it from going out when the grass is still wet, since all the wet grass clippings make more of a mess - they get compacted and the robot needs to be cleaned more often and more thoroughly. Even if I could come up with some sensor that measures rain, grass moisture, etc and gives a yes/no if it's ok to mow, I still could not limit the robot from going out without wiring into it with some additional receiver. It might be ok to toggle power on the charger/transmitter - even if the robot was still powered on and in the charger - since the voltage drop would not be significant over the one or two days that the robot was held back. I could test the idea of switching to "no loop". More work is needed!
Moti
-
Moti,
You could add a very small circular loop that the robot could follow on it's way out and back when it rains. Just change the type of switch you are now using to allow for the new loop.
To tell when my Jr. gets back into his charger I attached a wire with a flag, made out of silver metal duct tape, to the top edge of his charger house. When Jr. enters he hits the wire causing the wire go horizontal and back vertical when he exits. All I have to do is look out the window and check the flag. A quick check each day confirms that everything is okay.
toolbelt