I have noticed in the forum a number of people advising the connection of ground (negative power) of all their ESCs to the APM. In some cases some people are also connecting multiple 5 volt lines to the APM in the hope of providing redundancy of the 5 volt line. In my opinion both these practices can lead to intermittant issues that could cause your copter to occassionly behave erratically, and at worse damage your APM (as connecting multiple ESC grounds to the APM can create ground loops). Ground loops will create a noisy and unstable ground for the APM, which is the last thing you want with sensitive gyros and accels. The random issues that ground loops/bad power can lead people to incorrectly assume that there is something buggy with the software, causing unnecessary work for the developers and testers.
Ideally a seperate, single BEC (5 volt regulator) should be used to power the APM as the 5 volt line from an ESC that is also powering a motor will get very hot, causing the 5 volt line to become noisy and possibly brown-out/shut-down in flight. The BEC should not get so hot that it is uncomfortable for you to keep your finger on it for more than a couple of minutes with APM+GPS+Xbee etc connected. If so, a BEC with a larger heatsink or a switch-mode BEC should be used. In addition using a seperate BEC also ensures that the APM does not share a noisy ground return of an ESC.
I think a lot of the intermittant and unexplainable issues people have had can be resolved by ensuring the power supply and powering wiring topology is sound to begin with.
Does anyone know who contact about having documentation updated to reflect the above?
That looks like a star to me!
That said, I wouldn't build it that way (twisting together so many wires in one solder joint) because it's too easy to get a bad connection (electrically or mechanically) to one of the wires, which could cause a crash.
- To me it looks like you have a dry joint to the 'top' wire in both cases, though that might be an artefact of the photo.
Using a small "PDB" circuit board is a good idea because you can examine the solder connection to all the wires individually to make sure they are good.
I did not read this entire thread, so this might already have been covered.
Using separated ground for any R/C controlled unit like a ESC is a BIG NO-NO!
Remember that the ESC is controlled by a 0-5v digital PWM pulse coming from the APM board. If you separate the ESC ground from the APM you remove the zero (0V) signal reference for the PWM signal. It might still continue to work, but it might just as well stop working at any time.
Edit: Just to clarify. The signal cable ground and the power ground are common on most ESC units, so separating the power ground and still interconnecting the signal ground wires (trough the APM unit) has no effect other then that you now only use the thin signal wire (bad) to make a common ground.
so what miss in my scheme? I cant understand.. :-(
My concern is that you are actually creating ground loops with your scheme. As I've posted before, using the power ground wires of the ESC is a bad thing. Why, because every amp that motor draws raises the voltage above ground thus your scheme will introduce tons of noise. You have a very good idea for powering the servos but again, you are floating them while using the power ground as a signal ground. You cannot "float" anything in the system as we use single ended (referenced to ground) signaling.
Basically, exactly what John was saying but I disagree on the point of only utilizing the power ground of the ESC and not also attaching the signal input ground to the same star ground point.
IMO, the best thing you can do is star every single ground together, even the ESC input wires. Next, a lot of issues could be resolved by replacing heavier wires and grounds between BEC (ESC built in or otherwise). The thin wires stock are part of the issue.
For the most part, your scheme is good, we just think the ground wires need tied together as well. The +5V is fine (IMHO).
I'm getting a bit lost in this discussion, but I just wanted to throw a few things out there:
I think one problem with the discussion is that we are discussing electrical power wiring requirements only with a concern for the electrical noise. Unfortunately, especially with quads, we also have to worry about magnetic interference on the Magnetometer. I fear the design requirements to eliminate electrical and magnetic interference might be mutually exclusive. To eliminate mag interference, we need the power wires to always run in a twisted pair. That power distribution board might cause a big problem.
Also, Torg's image looks not to bad, but I strongly feel that the servo ground wires MUST pass through the APM. This is because the signal from the APM to the servo is based on the ground level in the APM, which may not be the same as the ground level on that ESC. This is particularly a problem if you have an Xbee. I used exactly the setup he is suggesting, and had horrible problems. Putting the servo ground through the APM ground rail made it 90% better.
I fear the design requirements to eliminate electrical and magnetic interference might be mutually exclusive
I think that is NOT a valid conclusion. There is a big difference in currents and thus the generated fields. The grounds being tied together should reduce any stray noise and have little to no effect from a magenetic standpoint. Again, the magnetic problem is caused by high DC currents, however the electrical noise is coming from lack of good current return paths (we are talking less than 1 AMP) dealing with voltage sensitive APM connections. Ensuring a solid grounding plan should not generate strong magenetic fields and if anything would help shield and reduce them. By far, the worst single connection from the magnetic intereference standpoint is the battery to the power distro board. By ensuring those wires are the farthest away from the APM and the compass, you will seriously reduce issues.
I honestly believe the worst builds (the ones with all the problems posting here all the time) are those who try to make their own power distro systems, and not use a proven distro board. The worst one has built 2 quads using a rig of wires and a vero strip board (basically trash), constantly complaining it flips on first takeoff every time. The common component in the builds was the trash wiring and then they wonder why it doesn't fly.
Keep high current battery leads as far from the APM as possible, Have a solid star ground system so all grounds do not result in loops to sensors or the APM, have a solid 5 volt power plan even if you need more than one BEC, and finally, keep the wiring neat. The distro boards are not expensive, but all bets are off when people come up with their home brewed solutions. I think not using a proven board and thus a solid ground and power plan is the problem. The board takes care of that.
Well, one of the mutually-exclusive requirements is that the power system needs large wires to reduce voltage drop, particularly on the ground. Reducing magnetic interference is best done with the small wire possible, to have the shortest center-center distance between positive and negative. In fact, ideally the power cabling would be coaxial.
I also fear that the power distro board can be a cause of interference. It's really hard to model, I'm not really sure, but my gut feel is that those boards cause problems. They turn the current into a big ring, and the Mag is usually right in the center of it. Maybe not, but it's worth looking into.
Thanks for the discussions, so far!
What about connecting two LIPos to get longer flight time?
I have my esc`s connected to APM with just the ground and signal and a separate ubec, I am using a PDB to which the feed wires are twisted to cancel out any EMI, I have my PDB below my APM and have tested for any significant magnetic field and found no issues and have had zero problems..
Yes Lars connecting two lipos together in parallel is not an issue, you can buy 1 to 2 plug adapters for this purpose, be sure the wires are thick enough to handle the amps required, or you can always make your own if you feel confident enough to do so..
I don´t like to connect batteries in parallel. The battery with lowest voltage will steal charge from the other one. This I say as a physicist. I would prefer to connect 3 ESCs to one LiPo and the other 3 to the second LiPo.
However, after an interesting discussion with an experienced I was convinced to follow your advice - as long as both LiPos are fully charged and with the same capacity and condition. A 2-to-1 plug adapter is now in the equipment bag and will be used as soon as possible.
Thank you for the advice.
Lars: As an electronics engineer (and not a chemist), I'd say you've opened up an interesting sidebar discussion, if not something warranting its own thread.
Not to play the "mine is bigger than yours" card, but I've run 36 10S (yes, that's 360 total cells) in a parallel array for years, with a bus-type power distribution for both charging and discharging. As you say, as long as all the cells are the same rating (and are otherwise decent quality), this should never be a problem, provided you balance the cells (insure the internal cells in the stack are connected in parallel with the same position cells in the other stacks or employ some means to make sure they all reach the same charging voltage). It's okay to use lithium ion-based secondary batteries this way; DO NOT ever, ever charge NiMH packs in parallel, as disaster will result.
First off, as it relates to anything in power distribution, current is not your friend. Everywhere there are large currents flowing there are also larger line losses (I^2*R, so it's exponential) and larger inductive/electromagnetic stray effects. The best theoretical solution would be to put the two LiPo cells in series, which for the same power, would cut the current in half. When I first started looking at hobby-class ESCs, I thought "they" (the R/C industry at large) were crazy for making 200A MOSFET controllers capable of only 6S (or about 25volts peak), when the entire EV industry is pushing voltages well into the triple digits and using IGBT devices instead of MOSFETs. This is precisely due to efficiency and weight (less current = thinner wire) concerns. Upon closer examination, perhaps "they" are only moderately deranged. I shall offer the conjecture that there are three reasons the voltages in R/C hobbyware are kept lower as a general rule:
#1 - Safety: The surge energy capabilities of LiPoly packs are astonishingly lethal. Normally, your body's resistance keeps lower voltages (<24V) at bay. Higher voltages will naturally lead to more fault current flowing where it shouldn't if there's a problem.
#2 - Cost: High current MOSFETs are cheap as dirt these days, but not so with IGBT or high Vds MOSFETs. The chance of avalanche voltage destruction goes up with higher rail voltages too, as do the costs of the capacitors to keep inductive ringing in check.
#3 - Reliability: Which relates to #2, but specifically to the batteries, the more cells you put in series, the more risk you have of one cell making the whole pack worthless. It boils down to a yield issue for the manufacturers as well.
So, run your LiPoly cells of the same rating in parallel all you like. I'd say that's a far more widely accepted "best practice" than trying to create two power buses.
Hello everyone, sorry for my absenteeism of these days,
I hope I understand the part about the grounds that can not be separated.
My project went ahead anyway and I decided that the servos will be completely separate from APM.
But I have the inability to use a power board because my frame has a central hole of about 20cm.
the frame is like that:
What do you think about this solution?