Reliability?

[UAVPilot]

Hello everyone
On this site AutoQuad team asked to give feedback and help make things better, I assume that it mean that by giving feedback to the team, they can further develope the boards and the software, this is exellent idea. I have no clue about the electric schematics of the boards, so I base only what I see in the pictures on this site.
First of all, what Iv been designing similar boards for 10 years, I'd like to focus to the connectors and their reliability. It appears to be, that connectors are attached to the board only by soldering joint and with SMD pins(?) Machines with motors and vibration, solder does not hold for long, specially when there is a stress pointed to it, in this case a cable and the connector itself, included with vibration. If you have to use SMD-connectors(as I can see you have to due to the double side layout), at least add VIAs to the connector pads or very close proximity, this is not easy if using TH-vias and double side connectors. Best way would be 4 layer microvia-board, but price will go a bit higher.
Second thing is the ESD-protection, do you have any methods on board to prevent ESD-strikes entering the board via cables etc. ? Have you tested your board against ESD and what was the results? Did you make test in temperatures ?
Thanks
Juha

[bn999]

Thank you for taking the time to discuss this. All of your points are valid and ultimately should be addressed. But one needs to realized that AQ is an experimental hobby board. The primary design goals were more along the lines of whether it can be made to work at all and at the most economical levels. Considerations such as reliability are luxuries that can be studied in future versions if the design proves successful and there is a demand for more professional commercial applications. I think ESD and temperature considerations would fall into the same categories.

I'm not saying I'd trust my life to the continued proper operation of the hardware, but I do have close to 1000 hours of flight time on AQ hardware with the same sort of SMD design. Thinking back, I don't remember having a single crash caused by hardware or connector failure. They have all been either pilot error or software failure - both are to be expected during rapid development.

[Wingspinner]

I'm rather surprised by UAVPilots line of questioning. For someone who claims to have 10 years of PCB design experience his questions seem to imply ignorance of modern design technique as well as design requirements of different target markets.

Just the fact that this board set only costs $400 and is shipped as a bare board would be an obvious tipoff to even the inexperienced that this is targeted towards the hobby, non-mission critical market segment. Someone with a lot of experience should know that the answer to this question would depend on the target market. However, even the low-end commercial UAV segment requirements are not much different than what autoquad provides. Only once you move to the very high end commercial (systems costing many tens of thousands of dollars, or military, or space are things like connectors and additional ESD protection significantly more robust. Someone experienced in design would know that today's modern chips have robust ESD protection built into the chips (and have for about 30 years - ever since the old CMOS 4000 series).

Anyhow perhaps just an innocent question but seems like an odd question for what is obviously a hobby class product.

Wingspinner

[s_sergiu]

Hi UAVPilot,

I have also long experience in fabricating PCB's. I should agree with Wingspinner.
You are referring at the connectors for PWM out? If yes, as long as board is connected inside one multicopter, I can not see what the problem is. They are SIGNAL connectors, with maximum current passing of 1A(limited by the switching power supply IC) . So if you connect just servo cable to send signal to the motor is not strong enough for this? But what about STM MCU or gyros or any other IC ? Shoud they have also pass-trough pins like DIP IC's otherwise they will come apart from PCB because of vibration?

Cheers,
Sergiu

[kinderkram]

Mikrokopter uses this for ages. And I never heard of any of the pins being torn off through vibrations.



If it's badly soldered and there's some tension on the wire I could think of such a scenario. But it must vibrate heavily to cause a tear off.

[UAVPilot]

[UAVPilot]

[UAVPilot]

[s_sergiu]

Hi UAVPilot,

You are right to ask about reliability when have attached to copter expensive camera. However, unless somebody will present one small video clip showing in what particular conditions of vibrations ,etc,etc, the connectors will fall apart, I can tell you from my experience, that in last more than 5 years i made many types of autopilots (this is first time when I bought one AP) and the only situations when i crashed the copter were : mechanical problems like propeller broken while in flight, motor coming apart from arm, software errors or pilot errors. I never had such connector problems even after strong crashes!.
On the other hand, if vibration is so high then we had a problem with copter itself. If motor+ propellers are balanced well enough (I use XOAR propellers for example), then vibration should not be a problem at all. Also when you are willing to attach expensive camera, of course you will use at least well balanced propellers so even the camera will deliver higher quality recording .

Cheers,
Sergiu

[UAVPilot]