Shop OBEX P1 Docs P2 Docs Learn Events
Docking Station — Parallax Forums

Docking Station

xanaduxanadu Posts: 3,347
edited 2012-05-18 12:55 in Robotics
I'm looking for advice on building a docking station. My first hurdle is somewhat elementary. My robot runs on a 12V 35Ah SLA battery. I want to build some kind of dock connector with skid plates (like a Roomba) for it to dock with.

My concern is if I connect my robot into a docking station and charge the battery should I be disconnecting the other electronics from the battery?

The charger is 14.9V @0.75A max loaded, it would be directly attached to the 12V battery. The 12V battery is also attached to a switching power supply with a rated input of up to 28V and also my MCU which has an input rated up to 28V. I would like these devices to be on while the robot is docked and charging.

Given those specs, my little trickle charger shouldn't cause any issues right? If the onboard electronics can handle up to 48V and they're in parallel with the battery which is 14.3V max?

Arcing won't be an issue because the dock will turn on the charger after the contacts have been made, not before. But is there something I should put between them anyway?
1024 x 768 - 146K

Comments

  • tobdectobdec Posts: 267
    edited 2012-05-15 18:11
    Somebody else may have another opinion...but I've never had any issues with a trickle charger hooked up while other things are running. Just fuse it and you should be fine....I just recently did exactly this with a small robot that resides and serves the front counter of my shop so you should be fine.
  • al1970al1970 Posts: 64
    edited 2012-05-16 00:05
    If your specs. are right; you are going to burn out your charger in secs.! To slow charge the battery, it will take 3.5 amps for about 14 hours. How are you going to do this with a "charger is 14.9V @0.75A max loaded" charger? You would need a charger that puts out about 7 amps as it's max load..
    xanadu wrote: »
    I'm looking for advice on building a docking station. My first hurdle is somewhat elementary. My robot runs on a 12V 35Ah SLA battery. I want to build some kind of dock connector with skid plates (like a Roomba) for it to dock with.

    My concern is if I connect my robot into a docking station and charge the battery should I be disconnecting the other electronics from the battery?

    The charger is 14.9V @0.75A max loaded, it would be directly attached to the 12V battery. The 12V battery is also attached to a switching power supply with a rated input of up to 28V and also my MCU which has an input rated up to 28V. I would like these devices to be on while the robot is docked and charging.

    Given those specs, my little trickle charger shouldn't cause any issues right? If the onboard electronics can handle up to 48V and they're in parallel with the battery which is 14.3V max?

    Arcing won't be an issue because the dock will turn on the charger after the contacts have been made, not before. But is there something I should put between them anyway?
  • tobdectobdec Posts: 267
    edited 2012-05-16 05:51
    How would that burn the charger out? Albeit it may take ages to charge i've never seen a charger burn out from simply being a trickle charger....
  • ercoerco Posts: 20,255
    edited 2012-05-16 15:44
    xanadu: Your intended method sounds quite reasonable to be. That's a huge SLA battery you have there, which will act as a large voltage buffer when your charger is connected. Worst failure mode case, you blow a fuse to your battery and the charger's 14.9V is fed directly to your circuit board. As long as it can handle that voltage using reasonable voltage regulation and filtering, you're good to go IMHO.
  • xanaduxanadu Posts: 3,347
    edited 2012-05-16 16:53
    Thanks Erco! Happy to get some more opinions. I have not built any circuits like this. Circuits like this meaning the kind with capability to smash through a wall in my condo. Even worse one could explode when I'm not home, and I miss all the fun.

    al1970- It's 1A I looked it up, it's a good quality charger and yes it will be working hard but it's fed from a 15w solar panel so it doesn't have much to begin with.

    tobdec- Got any docking station secrets to divulge? The next part for me is the connector, I don't think the skid plate is going to work out. I was thinking about running two dual mounted banana plugs, female set on the bot and male set on the dock. Then running the robot into them. Seems like there's got to be better ways than that.
  • W9GFOW9GFO Posts: 4,010
    edited 2012-05-16 16:57
    An automotive 12v plug and socket might work well. Take apart the plug and squeeze the contacts a little so that not so much force is required to make the connection. Many of them have fuses built in too.
  • tobdectobdec Posts: 267
    edited 2012-05-16 17:07
    Erco: Thanks for saying what i meant...im one of those guys that has great logic and theory in my head...but just can't find the words to express my thoughts...if only we could video tape the mind's eye!

    xanadu: I do not have a pic up yet but this is my current project with a docking station http://forums.parallax.com/showthread.php?139970-Business-card-presenting-bot now just imagine that with slot car contacts on top and a 3rd one for my PIR sensor input with a garage type structure overhead and reed switches to tell it when its docked correctly. Albeit my bot is much smaller than yours by the sound of your battery. However I do know of a design that sounds like its just the ticket here....funny story my gf found this project while searching my name on google lol I have the same first name. http://letsmakerobots.com/node/7025 In this gentleman's design he used what looks to be battery terminal springs as contacts and a shaped delrin cutting board material as a guide...couple this with a few reed switches to ensure proper alignment and sensing and you may be ready to rock my friend!
  • ercoerco Posts: 20,255
    edited 2012-05-16 17:20
    You'll want heavy spring loaded contacts for good charging. Wiping contacts are self-cleaning. I've made several skid plate chargers (inspired by Heathkit's Hero 2000 auto-dock charger) and they work pretty well. Much depends on your navigation system and how reliably you can locate and orient your bot to the charger.
  • xanaduxanadu Posts: 3,347
    edited 2012-05-16 17:40
    W9GFO wrote: »
    An automotive 12v plug and socket might work well. Take apart the plug and squeeze the contacts a little so that not so much force is required to make the connection. Many of them have fuses built in too.

    Yes, that's great it would and the 12V socket is bigger making the alignment easier as well.

    That has me thinking about sockets, especially wall sockets, and the fact they're in every room. Maybe I can use a small wall-wart style charger on a fold out arm, and drive it into a wall socket with some PING guidance. Then for my solar docking station use an inverter and a wall outlet and a heavy weight in the bottom. That would solve a lot of problems and work elsewhere. I like it!

    Thanks for the ideas :)
  • tobdectobdec Posts: 267
    edited 2012-05-16 17:42
    BTW xanadu....I want more pics of your bot...it looks big and full of awesomeness!!!!
  • xanaduxanadu Posts: 3,347
    edited 2012-05-16 17:55
    erco wrote: »
    You'll want heavy spring loaded contacts for good charging. Wiping contacts are self-cleaning. I've made several skid plate chargers (inspired by Heathkit's Hero 2000 auto-dock charger) and they work pretty well. Much depends on your navigation system and how reliably you can locate and orient your bot to the charger.

    I don't think I could custom fabricate anything along those lines. I have some nice designs but no means to build them. The robot is big enough that it should plug into a wall outlet. That solves a lot of problems, but now I need to figure out how to do it for less than it would cost me to fabricate the skid plate approach.
  • xanaduxanadu Posts: 3,347
    edited 2012-05-16 18:23
    Thanks! Here's another post about it - http://forums.parallax.com/showthread.php?139902-some-rocky-first-tests

    This is loosely what I want it to do, but articulating and also on the DC side will be a relay so that the charger pulls no power from the wall outlet until it's seated.

    Docking and undocking. You can see the charge light on the charger light up. I probably should annotate the video. It's raw footage of the first time I tried. The long pause before the hand wave in the middle is because the motors were in neutral and I had to run to my bedroom and fix that.
  • ercoerco Posts: 20,255
    edited 2012-05-16 19:57
    Docking to 120VAC presents many challenges, but it's definitely possible. Looks like you're well on your way. But don't dismiss the skidplate being beyond your fabrication skills, I like what I saw in your video. You can do it any way you like. I'd stick with lower DC voltages. Here's an early 8VDC dock I made for a Scribbler: http://www.youtube.com/watch?v=S-nv303EMAE

    With a battery that big, you probably won't need to recharge all that often.
  • tobdectobdec Posts: 267
    edited 2012-05-16 20:09
    Wow im in dumb gear today.....ive been following your build for a while now lol. xanadu: As erco said don't say you cant do the skid plate thing......you've came this far you can certainly make a docking station. I just came up with an idea while talking to my boss about our own project. What if you made a sling with contacts with a cone shape mounted on a rubber hose x with strong rare eartch magnets mounted in them to dock in mid air by simply connecting to the male or female corresponding charger mounted at the right height? So you can very easily either by eye or sensor array dock with very little fuss and lots of room for error...best part itl be a snap to set up with almost no custom cutting or milling.....do you get what im saying? If not I will gladly draw something up in either autoCAD...or most likely mspaint out of laziness paint lol.
  • al1970al1970 Posts: 64
    edited 2012-05-16 23:49
    If it is rated for 1 amp max; that means you can only run it at about 1/2 amp for long time use. You are planning on charging a 35 amp battery and run the robot at the same time. What is your robot using on stand by; 100 ma? That means it would take you about 120 hours to charge up the battery.

    "al1970- It's 1A I looked it up, it's a good quality charger and yes it will be working hard but it's fed from a 15w solar panel so it doesn't have much to begin with."
  • xanaduxanadu Posts: 3,347
    edited 2012-05-17 17:28
    Erco, Thanks. Your docking station is really nice, great project. I'm leaning away from the docking station more because I think this particular robot should be able to plug into any outlet. But, I'm not quitting and will build one for a smaller bot for sure.

    Tobdec, I'm not sure what you mean, a pic would be awesome. So I don't hijack my own thread I'm going to post some of the design ideas I had here as well.

    Al1970, I should have wrote 1 amp continuous, instead of max and 0.75A is the highest I've seen it go while charging. I have no idea what the max rating is, I'm sure that it's at least 25% higher. Sorry for confusion it just follows me wherever I go.
  • ZootZoot Posts: 2,227
    edited 2012-05-17 20:23
    One of my larger 'bots "self docks" to charge and unhooks when he's done. Chris Schur's work was incredibly helpful when I developed my charger/nest schema:

    http://schursastrophotography.com/robotics/dockinglogicfinal.html

    My 'bot is much larger, so the full-frontal contact plate scheme wouldn't work, but the hoops for charging contacts are great. For my nest, I used brass strip metal -- it conducts pretty well and maintains it's "spring". I use two IR detectors with about a 6.5" inch "vane" between them -- both detectors see the nest IR beacon only when the nest is dead ahead. Actually, it's within 6 degrees or so. See attached for a simple diagram I used to estimate accuracy.

    Given that we have a number of 'bots out and about here, all the 'bots -- and this big 'bot's nest -- use an IR code, sort of like Sony SIRCS. This lets them send simple messages, as well as prevents IR "pollution", clobbering IR obstacle detectors, etc. The big 'bot sends an IR message out when he's hungry asking the beacon to turn on. When the beacon receives that particular IR message, it starts sending an "I am here" message -- once the big 'bot picks up the docking station's IR messages, it can use it's two detectors to home in on the nest.

    The 'bot knows it's charging with simple zener diode and 1M or so resistor -- with no charging current on the front contacts, it's a logic low, otherwise high. This let's the robot stop, of course, but also send a brief IR message to the nest to tell it to "shut up".

    Both the nest and the 'bot are protected with fuses right by the contacts and hefty diodes to prevent possible short circuits (w/o blowing a fuse) on the front contacts. I wired up my own charger -- basically a 1.5 amp constant current charger. A recycled printer table top power supply provides juice (the main diodes and the current limiting regulator help drop the 16.5 volts from the printer supply down to a 13.8 volt or so charge current.

    I'm not sure what Erco meant about an inrush of current to the battery/on-board supply if the fuse blows -- if the fuse from the charging contact to the 'bot's battery blows, then the charger is effectively disconnected from the 'bot circuitry.
    1024 x 768 - 105K
    1024 x 768 - 99K
    1024 x 768 - 82K
    1024 x 768 - 88K
    1024 x 768 - 84K
    692 x 756 - 42K
    1024 x 768 - 68K
    1020 x 770 - 10K
    1024 x 513 - 34K
  • tobdectobdec Posts: 267
    edited 2012-05-17 21:43
    Zoot: Don't get me wrong thats an awesome idea...but it just seems like a short waiting to happen with it all open on the floor like that.

    xanadu: I know my pic is cheesy as hell but let me know if you understand this or not

    cone_charger.jpg
    1024 x 246 - 22K
  • ZootZoot Posts: 2,227
    edited 2012-05-17 22:16
    No, shorts are not likely -- there are 6A diodes on both the nest's contacts and the 'bot's contacts. Plus both sides are fused. After I built these platforms, I tested by shorting out the contacts with a screw driver. The diodes didn't even get that warm, no sparks, no problems. *If* the diodes failed, the fuse would blow, as it should. In any case, the 'bot has excellent obstacle avoidance and has never actually bumped it's contacts, but if it did, no harm would come.

    Commercial 'bots like the Rovio, which has charging plates on the bottom, are similarly protected.
  • photomankcphotomankc Posts: 943
    edited 2012-05-18 06:49
    I guess I don't get how the diode would prevent the short circuit on the charger side? I can see it on the bot side since it would prevent the reverse flow of current out of the battery back across the contacts but the charger side has allow current in that direction to operate. I'm guessing that is what the .82 ohm resistor is there for, to limit the maximum current, but I'm not understanding the signifcance of the diodes?
  • ercoerco Posts: 20,255
    edited 2012-05-18 08:25
    @Zoot: Nice bot & charger! Have you shared details of that previously? I'm interested.
  • xanaduxanadu Posts: 3,347
    edited 2012-05-18 10:25
    Zoot that's awesome, thanks for sharing :) The docking station design looks very solid. That is something that I would build, the springy copper strips are perfect for something like this. The link is very informative, I'm not going to task myself with an auto-docking station right now but the dock will be designed with that in mind. You have any videos of your bot in action? Two pings, IR and encoders it must know its way around well.

    Tobdec I see what you mean with the cone and will definitely take that into consideration. The cone shape gives a lot of surface area and should be easy to line up. Also the female cone side should offer a lot of short circuit protection and I'd put it on the robot side.


    I have decided that for my original hurdle, what to put between the battery and charger, will be this;

    Batt+>fuse>diode>N.O.relay> |dock| <diode<fuse<charger+ w/thermal and overload prot
    Batt->fuse>diode>N.O.relay> |dock| <diode<charger- w/thermal and overload prot

    My reasoning for the relay is I do not want the charger attached to the battery (yet). Since the charging amps are low it's not going to take a big relay to do this. This way if there are any docking issues it won't matter because the circuit will never complete until there is a solid capture of the docking power.

    The other relay concern is, I'd rather not have any exposed contacts. So the relay would allow for exposed contacts with the least mechanical process involved in protecting them. I'm trying to design this robot with harsh environments in mind.

    Thanks again guys! I'm going to throw together something this afternoon :)
  • photomankcphotomankc Posts: 943
    edited 2012-05-18 11:31
    erco wrote: »
    @Zoot: Nice bot & charger! Have you shared details of that previously? I'm interested.

    Same here. I was going to go with a more complicated cone&probe but that looks much easier and simple to construct.
  • tobdectobdec Posts: 267
    edited 2012-05-18 11:59
    Let us know how it goes xanadu...and pics please im excited to see how this turns out...bc I too will be building my first "large" robot in a few months for my company and HAVE to get it right.
  • ZootZoot Posts: 2,227
    edited 2012-05-18 12:43
    Sorry, I have not had time to document the project nor shoot video, but I'm going to try to find the time over the next week or so. Pieces of the project have showed up here and there in the forums over the past few years, though.
  • ercoerco Posts: 20,255
    edited 2012-05-18 12:55
    The Octobot Survivor uses those same parallel charging contacts as Zoot. Worth a peek; full construction details at http://www.mondotronics.com/PDFs/3-918_OctoBot_v125.pdf
Sign In or Register to comment.