This topic has been discussed before, I just can't seem to find the thread. In any case, I've heard of people adding switches between the batteries and main board to minimize the arc spark - wondering what you all think of that idea and if anyone has come up with any better solutions. Argument against this seems to be that it adds another potential point of failure. Argument for this seems to be it greatly minimizes potential damage and with a beefy enough switch it's a non-issue. I get sparks nearly every time and the tips of my connectors are charred but there is plenty of clean contact beyond the tips and I have yet to experience problems. But that spark does bug me and I wonder about the long term effects.
I think we should be able to come up with a solution. I'm no EE, but I think if you're running dual batteries, you could plug in a small battery to one power lead, then plug in the large LiPo, then replace the small one with another large one. I think the sparking could be eliminated then. Maybe. Anybody care to chime in?
I think the questions you need to ponder are: 1. How much current is flowing to cause that initial arcing? 2. The switch might prevent arcing at the LiPo connector -- but would it eliminate internal arcing in the switch? And the biggy: Why do you want to eliminate the arcing? Andy.
1. No idea but I get arcing with both smaller and larger batteries. 2. Also no idea but my guess is that arcing in the switch would potentially minimize and reduce the amount of the spark because of the elimination of human-fumbling-of-connectors. The biggy: I don't know how relevant an issue this is. Is it? Can I continue to create sparks every time I connect a battery for the next 20 years on a daily basis with no risk of damaging the components or do I compromise the integrity of the parts when this arcing happens? I think it's an important enough issue to find out the answer to. Dan
Dan plug them in faster. He who goes slowly gets larger arching. Be bold. You will still get a spark but much smaller. Lipo connections have been sparking in RC use since day one. If you see melted plastic then replace the connectors.
Some connectors handle the spark better than others. The EC series tolerates them well because the arc affects only the tips and doesn't tarnish the primary mating surface. On something like a power pole or Deans the arc slowly pits/tarnishes the primary mating surface which I've seen fail over time as resistance slowly increases. I'd prefer no sparks myself but in this kind of a power critical, high current application I'd rather have good connectors that can take the sparking than putting a switch in line. nick
Good information, Nick. I hadn't realized that the arcing affected the mating surface on the Powerpoles. Thanks Andy.
I learned that the hard way, Andy. As the arcs pit and slowly eat away at more of the surface it creates tiny bumps and ridges that prevent the contacts from mating properly. It can get to the point where the resistance and resulting heat is high enough to melt the enclosure and cause complete failure of the connection. Some postmortem pics of such an event:
http://www.abb-conversations.com/2012/11/what-do-arc-faults-and-air-bags-have-in-common/ Its a common problem in the Electrical Engineering Industrial world. Snuffing out the arc with an air blast or sulphur hexaflouride is not an option for an Multirotor. Someone claimed that connecting the positive leads first before connection the negative leads and almost eliminate the arc..... or was it the other way around.....? Speed of connection also helps. Using a switch!!!! The switch will have to be very heavy or it will degrade quickly into a low grade fuse or the switch contacts will weld together.
The switch would probably need to be rated for 150 - 200 (or more) Amps otherwise it will become a one-time fuse. Andy.
There are a few issues here.. let me try to break them up. Firstly, capacitors charge up very fast, which means a large instantaneous current spike on connection. When a connector is initially plugged in, there is very little contact area - so the amount of amps is over a very small surface area, the voltages and amps you're dealing with are many times higher than you need for arc welding, and this is essentially what happens when you plug the connectors in. Once an arc forms, it can carry virtually infinite amounts of current limited only by the voltage drop (inverse resistance). The heat and energy of this arc is rather enormous, hence vaporising your connector - just like welding as I said before. The melted plastic on the photo above is caused by a completely separate issue - the resistance of the connector is simply too high, so there is a lot of energy lost on it - creating waste heat, which causes the connectors to get hot and your plastic to melt. This could be a related issue, if the constant arcing has caused the connectors to oxidise or just removed enough material to give a poor connection. I'd guess that its the crimp however, they don't look like they are going to carry 200+A of a copter very well. Furthermore, a large octocopter will pull 400A on a full power takeoff.. and possibly be using 200A on hover. This is a *lot* of current. Take a look under the bonnet of your car, check the size of the cables going from the battery to the engine. These cables do not see this many amps. Now we know those primary issues, we can get onto why switches are a *terrible* idea. Firstly, switches have contacts in them. You won't see it happen, but there *will* be arcs inside the switch. These can fuse the switch on, or just erode the switches contacts causing a bad connection or high resistance. High resistance means high heat, which means things melt. Things melting when you are flying is bad. Intermittent power when flying is bad. You now have a possibly poor conductor, that doesnt quite make contact the way it did when it was new because it has been worn away.. then your copter isnt perfectly balanced (as we can see by the extreme use of anti-vibration systems in the copter world) so the switch is vibrating, possibly with a very high amp load. The worn conductor plates are now bouncing around on each other a little bit, giving your intermittent power - it might be at a frequency such that you wont notice, you're just not getting the power you should. If you are pulling 200A when this happens, guess what - lots more arcs every time the switch vibrates open and then shut. This all adds up to give switches an exponentially increasing failure chance. It might last a month, it might last a year - the end result is always going to be the same, a copter crashing back to earth. On top of this, a 400A (ideally 800A for safety factor) rated switch is going to be enormous. As in seriously huge and heavy. So what is the solution? The manual way is to have a resistor in parallel with your main connector, on its own wire. You plug this in first, then your main wire. This restricts the flow of energy going into the caps to a sensible level (its also much much better for your capacitors and all the wires and tracks going to them to do this!). Think of a battery as a fire hydrant, it *can* supply an incredible flow of water, and sometimes you want that. Its really not appropriate for watering the new seedlings you put in the garden though, it will dig them up and blast them across the street leaving a huge crater. What you need to water your seedlings is a flow restriction - thats the resistor. This lets a trickle of flow go to your seedlings, then when you need to put the fire out you plug the hose all the way in (the main connector.) Some connectors have this concept built in, like HobbyKing's new XT90 anti-spark. There is an initial ring of metal that is on a resistor which lets things charge up before the main unrestricted terminals touch. The other option is a smart electronic solution, built into the power board or a separate connector. These have some high amp fets, a micro-controller (or analogue circuitry) and a big resistor. The resistor is always on, allowing small amounts of current to flow and charge up the board - then the micro or circuitry will turn on the FETs allowing full current flow.. there are not a lot of these out there, I make my own for clients. There are two approaches to it, monitor the current and then supply the main power route when the current demand drops (the best route, but more expensive) or switch on the FET after a period of time. FETs are solid state, so there is very little that can go wrong from a mechanical standpoint unlike a switch, and when a FET does fail, it fails in the ON position which is quite nice. Multiple fets can be ganged together to provide additional redundancy. The main issue is they create heat - even a very good high amp fet might have resistance of 0.001-0.002 ohms, it doesn't sound like a lot but when you start talking hundreds of amps it adds up to quite a lot of heat - this is primary why (good) speed controllers get hot. Cheap speed controllers like Castle Creations have other issues causing heat as well. So... to sum up... its a tough problem and there isnt really an ideal solution. Switches however are the very very least ideal.
Mark..... I hear you..... I do Arc Flash studies on a regular basis as part of my Industrial Electrical Engineering work. Many electricians have been severely burnt by Arc flash. http://etap.com/arc-flash-analysis/arc-flash-analysis-software.htm Thanks for sharing the information on PDB equipped with FETs etc. I often wondering what the heck was going on with that!
Thanks for taking the time to write that post, Mark -- and to Graydon too. I think the net is that we live with the arcs on our LiPo connectors and be ready to replace them when they get too pitted. The good news is that the EC5's seem to pit at the leading edges of the contact surface, but do their heavy current work with different parts of the conductive surfaces. Andy.
The AS150 connectors mentioned here: http://forum.freeflysystems.com/index.php?threads/shock.963/page-2#post-53167 seem to be the best solution for now, or the anti-spark XT90's. 7MM bullets are super nice though, considering the hundreds of amps we pull on heavy lift rigs. With those connectors, the key will be to plug them in slowly, or pause with just the tip in, then continue after a second or two.
Mark: I forgot to ask -- are you really seeing sustained loads pulling 400 Amps for takeoff and 200 Amps for hover? That seems a tad high from my experience -- or are these impulse loads? Andy
Depends on the size of the copter, amusingly the bigger copters run lower amps. Those really are mostly max amp numbers and not sustained, but the system still needs to deal with it. 30A/motor max on a big copter, 10-20A hover... but something a little smaller will be 60A max and 20-30A hover on 8 motors. Obviously all depends on payload conditions, altitude etc. I dont personally have any large copters, i'm purely on the technical side designing and building things - I'm speaking of client copters here
If anyone tries these connectors, please report your experience. If they're all they claim to be, I'll probably convert all my existing connectors.
Nick: Curious about what kind of current are you running through those 45amp Powerpoles? I've been using PP for over 2 years, with no problems. I still use the 45amp on my CS6. For my two larger birds I'm using the 75amp ones. They are massive, but only the tip sparks, and pits a bit. The contact area is fine. I actually contacted Anderson about using the 45amp on a 60 amp constant (hover) draw, and they "suggested" I don't. I'm pulling 50-60 amps constant, with about 100 amps Max, but seldom over 85.