I think I remember this brought up in the forum a couple years ago or more. Unless the technology has changed I think the issue was the time it takes to charge a battery as this is an inefficient method vs direct connection. Edit: I'm speaking of free roaming robots which tend to have larger batteries to charge.

Plus how would the wireless charger know when to stop charging so as to not overcharge a lipo battery... Over charging a lipo is a bad thing... However, maybe (as Troy mentioned) since it charges slowly it "may" be safe?

@Richard R, as he describes it, the robot can monitor the battery level and back away from the charge plate when its done charging.

@Troy, DJ is adamant that the robots should be turned off (when using LiPo's) while being charged, so a different method of determining the charge is complete and turning the robot back on wold be required (this is not the case with Lead Acid batteries or NiMH. A lot of discussion over the years about developing self finding charging docs, but other than those who use Roomba's as a base which have the capability built in, I am not sure if any of us DIY'ers have actually implemented anything yet.

Personally, I am waiting for 2 potential upcoming EZ-B products, the Lidar, and the navigation system using a fixed position camera so that I can solve the autonomous navigation issues first. Then I'll tackle self charging (assuming my job slows down when my current project that is taking all my time wraps up in January as I expect. I have had virtually no time for robots all year. I haven't even opened the box of HDD servos I bought months ago....).

Alan

@Alan, that makes sense on the lipos of course. I didnt think of that.

New products? I will have to search for mentions on it.

Quote:

New products

The Lidar came up in a totally unrelated thread but DJ posted a picture of the work in progress. Can't find it now of course. Was within the past month.

There was more discussion of the fixed mount navigation camera, which I also can't seem to find (searching from the phone is tricky). This was a few months ago.

As with any of these works in progress that DJ mentions, it could be weeks or years before we see product, and some get canceled for various reasons, but it is fun to speculate.

Here's info about the IPS: https://synthiam.com/Tutorials/Help.aspx?id=230

Conversations are happening about the product launch. I'm trying to move a budget around to get the DIY components on the assembly line. It'll happen - it's just tough because Revolution is the money maker and that makes it difficult to convince the people who handle the money :D

Understanding that advancing the technology starts with the DIY users and products, which trickles down to revolution. Wheels are in motion... stay tuned :P

I had been considering putting a crowd fund together to manufacturer the diy items - but that might result in disappointing results because the diy community is so small these days

DJ, I’ve noticed that other established companies such as the UpBoard maker have successfully used crowd funding for new products

I really appreciate your support for the DIY community

Frank

Quote:

I had been considering putting a crowd fund together to manufacturer the diy items - but that might result in disappointing results because the diy community is so small these days

I share this concern, but it would give you a good gauge of the the actual interest (or, like you did with Revolution, you could do pre-orders instead of using a crowd funding site, but then if you don't meet goal, you have to do refunds, so you are essentially carrying a liability between the pre-order and the delivery).

I think the 2 most talked about DIY things that are difficult to achieve without kludging together 3rd party stuff are autonomous navigation and action (solved with IPS and Lidar) and the subject of this thread, power docking to extend run time (requires autonomous navigation, and a lot of other work, and probably harder to retrofit into existing Revolution designs - so more strictly DIY). I think though that adding autonomous navigation and action fits so much more than DIY. It has great educational tie in. I mean, how much better is it to tell your robot to find its own way from one room to another than to do yet another line follower (I almost passed out when you did a line follower ROLI. I mean, an Arduino can do that....). Also fits into companion robots (follow me around, converse with me, pick things up for me, etc), and industrial uses (security, factory stock transport, package delivery, etc. etc. etc.).

Just my 2 cents, for what it is worth (probably about 2 cents).

Alan

I'd like to jump back to the original post for a second. I feel that wireless charging is likely not the ideal solution to make the charging process more seamless. My reason being is that wireless chargers do not put out very much current, if they do it's likely that the transmitting and receiving coil are very large in size (larger than our robots). Wireless charging with a coil of the size pictured in the first post would presumably take an entire night to charge a 7.4V 1300mAh battery.

That being said, I feel like we as a community are in a good place to be able to try a system similar to the Roomba docking system. You see the new Linkman charger only requires a 12V source to power it. If the Linkman charger was to be attached to an ezrobot (or DIY robot) with two connections broken out (+12V, GND) and then have a Glyph or beacon system to auto-locate the dock, one could create a automatic charging system.

A simple momentary cherry switch mounted on the robot could be used to disconnect the power to the robot once it is completely seated into the dock. Of course you would need to take it off the dock when charged or have the dock kick the robot off of itself. I guess that would make the dock a robot as well :).

Anyway, just some ideas. Although I love wireless charging it's just not all that practical for our scale at this moment in time. As mentioned though, I do think automatic charging is possible with a little creativity.

@jeremie, I like this approach

Thanks, Frank

@Jeremie I have to agree, I think I have been watching to many iphone commercials. Why create a complex solution when a docking station would solve the problem.

@jeremie, I’m going to try to implement your idea.

Just to be sure, it’s ok if the version 3 changer is left connected to the battery, just as long I disconnect the 12v power. I was always under the impression that we needed to physically disconnect the charger when the battery was charged up... unplugging power to charger was not sufficient.

Thanks, Frank

My charger appears to be 7.4v 1000mah and the battery is 7.4v 1300mah so we are charging at about .75 C. Depending on number of sensors, active actuators etc if you charge the unit while still turned on with existing charger you may actually find you are not charging the battery but actually causing it to go flat and will kill the LiPo (they don't like to go flat). I assume you could probably charge the battery at 2C without to many issues (although LiPo's are prone to go up in smoke for little to no reason so you may want to put battery in a fireproof bag inside bot) . So you could use a high C charger to do the work but again you want to disconnect charger once battery is charged or you could find your bot goes up in smoke if there is a problem. Maybe safer to stick to the .75 C charge and turn bot off before charging.

So I think a possible process could be Robot self doc's . Robot has a Magnetic switch for on off power. The magnetic switch is activated by an electric magnet on docking station so it cuts power to robot turning robot off when it identifies robot is docked correctly. Next a relay cuts in to connect power to charger, when charger detects battery is charged the same relay disconnects power to charger and then turns off electric magnet turning robot on again.

You could use a second V4 controller that is permanently connected to a power outlet to handle the smarts of the relay, magnetic switch etc. Getting the docking contacts correct and detecting when it is docked will be an interesting task as well.

Good point Frank!

I believe from my previous calculations on a fully charged battery it would take approximately 45 days for the battery to parasitically drain backward through the Linkman charger down to 6V. I'll just a test tonight and get back to you.

@Nink great idea to use reed relays and magnets! Genius!

Unless i'm missing some detail a 7.4v Lipo is a 2 Cell battery. So you need to balanced charger (3 wires) ground and 3.3v for each cell, a 3 cell battery has (4 wires) etc.

DFRobot has a 7.4V lipo with only 2 connectors https://www.dfrobot.com/product-489.html

I asked their support and is also in the comments:

Quote:

Yeah, this lipo battery has a bulit-in protect circuit. Which could protect the battery from the danger like overcharge and overflowing. It will be more safe in the application.

Oh @Nink and @ptp sorry for the misunderstanding guys. The idea is to have the Linkman charger attached to the battery already through the 3 pin balance plug. It would be the 12V barrel jack that would be broken out and attach to the charging base. I wouldn't want to break out the LiPo positive and negative connections at all, that's asking for trouble lol.

@nink Great idea on using reed switch

I’m going to start testing this

Frank

@ptp I have a few of those batteries (One came with my fat shark FPV HMD and 2 with a headplay HMD). Batteries work great on the development kit as voltage and plug are same.

So i missed that detail, that is a good idea, so the robot carries half of the charger ?

Can you post the battery charge link ?

@Nink, can you charge and use at same time ?

@ptp That is a very interesting battery

@jeremie, Has EZ Robot evaluated this type of battery?

Thanks, Frank

Frank, I'm guessing, the battery is the same... the difference is the charge/protection circuit and how far they went to protect/handle the charging process.

I made self charging robot and i added 2 thermistors to monitor 10 x AA NimH batteries, maybe overkill, but, i was so afraid that something could go wrong with my charger, so i added an extra circuit to monitor the temperature.

@PTP Never tried to be honest, but can't see why not as long as you are not overloading the transformer. Lots of people take the balance charger and LiPo's out of the fat shark goggles battery and then use 18650B Li-ion batteries (holds double the charge per gram than LiPo's).

Is anyone using Li-Ions on their EZB's ?

@nink, I’m not able to find any reed switches that handle more that 0.5A

@jeremie, What is the fuse rating inside the Revolution robots?

To be safe, I think I’ll go with the microswitch idea you first mentioned. Small ones are raated at 15A

Frank

@Faengelm No we haven't evaluated that battery in particular. The fuse rating in revolution robots is 20Amps.

@Faengelm how about a reed controlled MOSFET?

@Nink Great idea. This is what makes the forum so great!

I’ll prototype this and let you know how it works out

Regards, Frank

@Faengelm From my measurements last night my battery went from 7.558V down to 7.554V in 12hours. From these calculations it looks like the charger would parasitically drain the battery from full charge to 6V in about 300 days.

I'm sure that it would actually be less days than that but I'm pretty confident that you are ok to leaving an unpowered Linkman charger attached to the battery in your design.

@jeremie, Thanks for researching this

Frank

Honestly in my mind I would recommend making a " dock stand" for each robot. The robot would have a couple contacts on the bottom some where to make a connection to a charger. Plus, it could be a could stand / display for the robots too.

@jstarne1 That’s a good point.

I’m pusuring this with several levels of increasing capabilities:

Level 1 would still require turning off the power switch, but would just allow setting the robot on the stand... making the power connection

Increasing levels would handle the shutting off the power switch and determining the end of the charge cycle to disconnect the power

Frank

@nink Here is an update on my prototyping.

I thought I'd start with the manual stand approach as described by @jstarne1. The first problem is how to make a reliable easy to connect power connection.

I chose a magnetic charging/data Android phone adapter with some micro USB breakout boards

If you decide to try this be sure to order a cable that has data capability as you need the D+ & D- pins joined together for the center tap on the LiPo cells. You need both pins as the magnetic connector does not care about orientation

Of course this approach still requires you to turn off the robot power switch and monitor charging to later disconnect

Now I'm off to work on the suggestion from @nink and @Jeremie

Frank

Very cool love watching the progress. Your magnetic android charger idea seems to work well. Robot only needs to get close enough to the charger to connect. Great work. (looking on ebay for magnetic charging plug)

Here are a couple more Revolution Robots with the quick-connect charger

Frank

Cool I ordered half a dozen micro USB and breakout boards and some magnetic charging cables, some USB female breakout boards and some ballance cables from China. About $20 total should be here in a few weeks. Will try and catch up.

@Nink Just be sure the cables support data transfer instead of just charging

The next step in my project is to add @jeremie’s idea of using another robot to provide the charging station

It is powered by a wall adapter instead of a battery and will contain the version 3 charger for the dockable Revolution robot connected through the magnetic connector

Regards, Frank

Yes I did thanks @Frank I ordered a 2.4 amp charge and "Data Sync" cable. My only concern is the data lines (that it looks like you are also using for power) does not have the correct gauge wire to support the required amperage so they may get a little toasty and burn out. Looks like yours are working fine. Worse case I just grab 4 diodes and make a quick bridge rectifier so the + and - are always correct irrespective of the way you join the connector. This would also free up the data wires to talk to say a circuit or even perhaps just a relay and disconnect power to the robot during charge cycle.

@Nink I need the D+ D- wires because the charger requires 3 wires to charge the battery

This is because there are 2 cells in the battery that must be balanced charged

Also, I don’t think you need to worry about the orientation of the connector switching polarity as these cables are more the the USB C design where they don’t care about orientation

regards Frank

ok understand sorry I forgot about the balance charging. I am using the ballancer on battery I mentioned before so only 1 positive and 1 negative terminal.

Here is a picture of the 2.4 amp cable I purchased. There is 15 cores for data line and 35 for power.

Ah, I forgot you were adding the charger into the robot... I might try that too as you mentioned that frees up some lines for possible power control of the robot

Frank

@Frank 2 months later my magnetic cable turned up. (Just love ordering parts from China).

Works well. I will add them to all my EZ-B great idea! Thanks

@Nink Great news

Frank