Yesterday I had my third fly with ardupilot on my X8 and is anded badly.
In about 3 seconds after start engine stopped and I lost control totally.
Due to small roll and pitch angle my X8 kept flying and crashed in a tree in 300 meters.
Later at home I discovered that connector on a power module on the battery side came off from the wire. Terminals inside the connector had almost no trace of solder. On the wire solder attended and was highly oxidized!!!!
This means that the soldering kept on heat shrink tubing, and water has entered inside accident was only a matter of time and vibration, which is what happened.
I also cut tube from another wire and soldering quality on it also leave much to be desired.
Fortunately the aircraft be restored by the end of the week I will raise it again in the air (It amazingly strong!).
However, I will now have to check all the areas of factory soldering.
Comments
I find solderimg is somewhat of an art form, and you definitiely have to dial in your own settings for the equipment used. I spent a bit of time soldering my power distribution board and Xt60 connector before I was happy with the lead free solder joints.
R, I will certainly defer to Greg as the expert...
I've had great success using traditional 40/60 solder at 800 degrees... with proper tinning, I find that the solder connections are made really fast, so there's minimal heat moving away from the solder joint. The solder joints look nice as well.
Having seen some of your heli stuff, I'm thinking that future current sensors might be made in these options:
[ ] low
[ ] high
[ ] insane
[ ] R_Lefebvre
:-)
Had exactly the same experience, same wire, same connector, but only on one out of 4 that I purchased.
Of course I removed the shrink wrap from all connections and checked them, but only this one had not been done correctly.
They failed to tin the connector
Measuring small DC current changes using inductive pickup is hard as you only get the differential when a change happens. Inductive pickup only really works for AC supplies.
I also use leaded solder as its much easier to work with. I'm not doing this stuff 24/7 and only a small amount. That said lead is not good. I was listening to radio the other day about some research that has linked the drop in violent to crime to dropping of lead in gas. Lead apparently effects the brain centre for violent behavior especially in kids, if you are a borderline case, it can just tip the scale. And it takes about 20 years or so to show up in the stats.
Greg, interesting info there. Are you saying that I really should be soldering at such high temps? I have a WES51, so I can set the temp, but I had read somewhere to use "the lowest temp that does the job", so I typically have been working around 4-500°F, unless I'm doing a big solder needing lots of heat so I crank it up. You're saying I should be using 800? I do have to say, I have not been super impressed with the solder appearance at the lower temps, but though I was doing the right thing.
I have a 250W gun I use for my battery connectors.
The 3DR PM can be improved easily I think. The connector between the PM and the ESC is unecessary. I would remove it, and solder wires between the ESC and PM directly. Now a couple of failure points have been removed. The 6pos connector... sure, it's an extra connector, but at least it's locking. Overall I bet this connector is more reliable than powering the board via a servo cable.
But how come nobody makes a current sensor that clamps around the wire instead?
I agree that voltage measurement is perfectly adequate. I have experienced that a new battery had a high voltage tell very near empty and then dropped off sharply, but after the battery had run 1 or 2 cycles it was pretty linear already.
Voltage measurement does not introduce a single point of failure, and it does not give you a false status if you accidentally start with a uncharged battery.
Back in the early space program days (showing my age) NASA did extensive research into electronic failures. Under higher G and vibrations during rockwt launches the extra weight (Blob Soldering) and oxidisation from poor preparation were major contributors lower reliability. As a result NASA developed a high reliability soldering program which included inter alia mechanical and chemical cleaning of all components being soldered including the solder itself. Attention to creating solder filets rather than large blobs were crtical in the process. Large amounts of solder do not equal better reliiability qu ite the opposite.
Looking at the photos posted shows all these bad elements. Personally I never buy pre soldered components (no disrespect to 3DR team) I prefer to take time with each joint / connector. But thats me. For the XT60 connectors pre-filet the solder in the connector and use thermal not mechanical tools to strip your wires. I would also recommend both mechnical and chemical cleaning of all wires and connectors even if they appear new. This will help the lead free or even lead solder flow better.
Yes, just voltage, that's it.
I've yet to ever have a LiPo drop capacity before it drops voltage. Voltage always predicts the capacity. Sometimes the voltage drops lower than it should be for the capacity, due to high loads, so you get a false reading of low battery. But this is still safe. You land, and then check the battery and see that there was a lot more left. Make note of it, and re-engineer your flight proceedure (go to a lower voltage).
I don't like the idea of my primary power source passing through an extra 12 solder points. If we had a current sensor that was the type that clamped around, a wire, then I'd use that.
But cost/benefit on this one... doesn't work out for me.
Not to mention, nobody makes a current monitor that can handle the Amps I need to handle. Nor for multiple input, multiple output power supplies.
R, do you measure battery voltage?
This is why I don't use current consumption on any of my aircraft. It's one more HUGE thing that can go wrong. I relentlessly try to eliminate all potential single point of failures, and this is a big one!
-
1
-
2
-
3
of 3 Next