Motor Failure

[Old Man]

JST connectors are THE crap.

As for FAA standards, we aren't flying full size planes/helicopters. Or even full size 'drones', so they need to sit back and relax. They can simply take a look over the pound to see how every other country is doing it, and has been for a few years. Voila done, save the million of dollars of bull**** 'research'.

None of this even addresses the primary issue: Clueless people with $500-$1000 to spare to buy a off the shelf quad and fly it over densely populated residential area's, into buildings, near airports, etc.

I do not disagree at all.

 

[FerdinandK]

... please do not forget in your considerations, that in risk management, there is only the question "When"/"How often", not "If". So of course some regulations will improve the safety and some wont, and here we are talking about probabilities. Vehicles that fly fail, like every other vehicle, only that in case of flying, failure means crash. If the flight is over populated areas the probability is higher (which can be quantified) to hit someone, than in spare or unpopulated areas. It is also not a matter of size (Hubsan or Airliner), just a matter of damage they can do. So whatever (new) method or technology can provide safety it will fail, out of each failure we can learn and improve, but it will fail again. Even if you think, that you are so professional, and followed all guides and rules, the vehicle will fail, and you should be prepared for the day this happens.

If you need some prove of evidence just look at the news, there have been unsinkable ships and invulnerable nuclear power plants, lost airliners, .... and it always was some "unfortunate series of evens". Not what you have thought about will take you down, it is what you have not thought about (beside the standard errors).

Another rule for flying machines is "weight kills", so what ever clever idea you had on how to do this or that, always have the added weight in mind.

best regards
Ferdinand

 

[Old Man]

Hence the reason pilots license required, that alone is going to weed out many.:tennis:

But place a significant financial hardship upon many that have until now been operating safely. Being rated ASMEL-instructor/airplane I won't have any problem with pilot certification. I do believe everyone that will participate commercially should be made accountable in the aviate-navigate-communicate departments, they should be able to effectively use a Sectional chart too, but is holding a Pilot's certificate really necessary for low level, short range, short duration stuff? Yea, people would have to voluntarily keep their aircraft low and close. I would think new testing and a specific licensing classification would be in order, something that could be handled via written test, flight simulator, and personal demonstration with feet on the ground. OTH, I would also be a bit jealous that people would not have to spend all the time, expense, and effort I did to become a Commercial Pilot. I could probably get over that. Probably.

This stuff is all new so much will need to be viewed from new perspectives. Many of the old ways will carry over but to use all of the old ways will subject far too much to the limited capabilities of the past.

 

[Old Man]

... please do not forget in your considerations, that in risk management, there is only the question "When"/"How often", not "If". So of course some regulations will improve the safety and some wont, and here we are talking about probabilities. Vehicles that fly fail, like every other vehicle, only that in case of flying, failure means crash. If the flight is over populated areas the probability is higher (which can be quantified) to hit someone, than in spare or unpopulated areas. It is also not a matter of size (Hubsan or Airliner), just a matter of damage they can do. So whatever (new) method or technology can provide safety it will fail, out of each failure we can learn and improve, but it will fail again. Even if you think, that you are so professional, and followed all guides and rules, the vehicle will fail, and you should be prepared for the day this happens.

If you need some prove of evidence just look at the news, there have been unsinkable ships and invulnerable nuclear power plants, lost airliners, .... and it always was some "unfortunate series of evens". Not what you have thought about will take you down, it is what you have not thought about (beside the standard errors).

Another rule for flying machines is "weight kills", so what ever clever idea you had on how to do this or that, always have the added weight in mind.

best regards
Ferdinand

As in all things, nothing is ever 100%. You make good points, and as someone that darn near took someone out with a UAV on the wing a long time ago, I could not agree more. The person in question was in the right place at the wrong time

 

[edgarmillennium]

Hello Guys:

Sorry to ask, Im kind of new, can somebody tell me what controller the hex on the video have, that do not yaw when one motor is out?? I changed my Naza lite for the V2 after a prop failure. But probably will be a good upgrade later to have one that do not yaw.

Thanks

 

[baja-king]

Hello Guys:

Sorry to ask, Im kind of new, can somebody tell me what controller the hex on the video have, that do not yaw when one motor is out?? I changed my Naza lite for the V2 after a prop failure. But probably will be a good upgrade later to have one that do not yaw.

Thanks

I'm pretty certain that [MENTION=9737]Quinton[/MENTION] flies with a WKM...

 

[R_Lefebvre]

Of greater issue than a single Herk board or independent ESC's is the power distribution boards we use. Consider that out primary power outputs are all solder joints of one type or another. Solder is fragile and will not indefinately hold up in a high vibration environment. When the solder terminates inside wire insulation is a natural brittle point in the wire and a well noted failure point in all things that have wire solder joints. We need a well engineer re-design of our PDB's that eliminate solder requirements, be they bullet connectors or direct soldering to the board. Some relatively simple mechanical modifications to a much lighter version of a Power Hungry board would do the trick.

Rob L.,

I hope you're listening since you have an M.E. background. I had an idea for this some months ago I gave to a fledgling engineering company but as soon as they drew a Solidworks (v2005) drawing of my design, and afterwards wanted to charge me $5k to have the drawing. I walked away from it. They never bothered to do anything with it themselves.

I agree that the PDB designs on the market are overly large, overly complicated, poorly designed and expensive. I like things simple, cheap, and that "just work".

This is what I have done in the past. I simply used large copper washers from McMaster and directly soldered all the wires. On this one I used acrylic washers (simply cut with a hole saw), on short-run production I used fiberglass washers. The insulating washers compress the wires, and create strain relief so there is no strain on the solder joints. The "PDB" itself can probably handle 50,000 Amps, and cost like $5.

I do prefer solder joints over a bolted joint, as lugs have their own set of problems. And using bullet connectors wouldn't eliminate the problems with solder joint strain.

The crux of the problem, is simply lack of strain relief on the wires. It wouldn't be that hard to do.

Currently I have a quad which simply uses the Hobby Wing Quattro 25A ESC. My last Octo just used 2 of the same. Super simple, easy to wire, and just works. But obviously limited on amperage/multirotor size though.

 

[R_Lefebvre]

Wow, now this is the thing.. I have not been at this long, but after a year and a half I do NOT know who to believe.
I have heard some very smart people say that a solder joint is the only way to go. @FerdinandK

I have learnt during that time not to believe everything I read online.
Now I do not know who to believe, one minute I hear one thing, next minute its the opposite..this drives me crazy, too many experts but who do you listen to?

Surely there are always solder joints at some points, after all how do you connect the bullets?

Here's what I have learned over the years, and this applies to more than just multirotors:

If somebody says "This is the the one, singular answer. The be-all end-all. Listen to me, I'm an expert. Everything else is crap. Do as I say." that usually means they don't really have the answers.

Nothing in life is black and white. And things are even less clear in engineering. Every solution has pros and cons. An expert is somebody who understands this, and can discuss the problem in depth, suggest options, and decide which solution has the best combination of pros with the least cons. These people often sound uncertain about what they are talking about. But they are not. It is because they understand that it is the ANSWER that is uncertain.

In this case, my belief is that soldering is the preferred connection. A good solder joint is lower resistance, and is more mechanically reliable. The problem is that word "good". Large, high power joints can be difficult for unskilled people to do properly. I have a 260W iron for jobs like this. And those copper washers had to be pre-heated with a propane torch first. That is a demonstration of how many amps this thing can take (the electrical conductivity is actually loosely connected to the thermal conductivity). It took about 2 hours to build that. And then, yes, soldered joints absolutely need strain relief. Of course, that is not to say that a crimped or lug connection doesn't also need strain relief. Though many crimp connectors include the strain relief in the connector. Bullets can have some strain relief by way of shrink wrap, though I don't think it's fully adequate in all cases.

 

[Cheshirecat]

If it didn't hold in GPS mode, but was able to in manual mode, then it would seem like the programming or tuning of the WKM was the cause. Because obviously there was physically enough thrust. I'd suspect the Yaw I-term could not grow big enough. But that's just a guess.

Sorry i didn't answer your question, i missed it the first time round.
So no i didn't put it in manual, i didn't want to increase my workload at the time, i have a bunch of pre-terminated ribbon cables on order which will hopefully arrive tomorrow.
when i fit them i intend to change the mixer setup to a non DJI standard as recommended by Andreas Baier, this will make it a double decker quad, at the moment it is set will all 4 upper motors running CCW and all lowers running CW. clearly not able to balance torque in the event of a motor failure as is.
Baja King has run an identical ship with the non standard mixer and Andreas recommends it, and as they both know what they are talking about i shall follow that advice.

 

[maxwelltub]

Hey @Cheshirecat I hope you changed your motor config. I believe I had the same problem as you with a lose micro match connection. Few weeks ago I head a motor rev, but it stopped fairly quickly after it started. Didn't notice any change in flight. Last week I had the whole board crap out. Luckily I was running the stacked quad method and was able to land with no problem. After bringing it down and spooling up a few times on the ground I only had one motor not spin properly. I haven't had time to do a full strip down at the shop but my first guess is the micro match cable. Can you detail your solution, any links to the replacement cables you ended up getting?

 

[Cheshirecat]

Hi

Playing catch up. I ended being able to replicate the problem at will on the test bench, 100% sure the problem was with the Micromatch connector termination.
They are not that easy to terminate without the correct tools (which cost a lot) so i purchased spare pre terminated cables and connectors (E-Bay supplier) and just swapped them all out.
Ran bench tests for a good couple of hours and could not invoke the issue after swapping the cables, since them it has sailed through the RPQS test and had several uneventful flights.
I did contact Andreas about the difficulty of terminating these reliably, who told me he is working on changing things in this area of teh Herkules (no detail though) i have also added some procedures to my operations regarding reaction to a motor failure / loss of Yaw control event. Just need to practice it to the point it becomes an automatic reaction (no substituting hours on the sticks )

 

[maxwelltub]

Glad to hear it! Because I had a whole board go out, if my motors were set to dji specs I would have been f'd! Could you supply a link to the new cable with terminal ends? Thanks for posting this in the first place also!

 

[Amerly]

Well this is a very interesting discussion... I'm running standard mixing with A2 and Herkules III (X8) and was interested in going for the Andreas special mixing... The idea of having lower props (under prop wash from the top ones and spining faster) doing all the torque conteract of the top ones is not very desirable for heading lock and safety (connector loss)... There is many old treads about this mixing or this board on this forum (some knowledgable people not active anymore) but it's good to have fresh news to !

Micromatch connectors gave me hard time to secure on the board and took me a couple of try to install properly on the flat ruban in the first place (with enough clamping of the wires but without breaking this tiny plastic box) ! Really the weak point of that superb BLDC... Good to know that Andreas (ichwillnix)
is rethinking this part... I hope he will be part of this discussion to...

 

[maxwelltub]

I've always done this mixing method, even before I had a Herk board. I'm currently flying all 18"s but I'll probably put 17"s on top. It has way better yaw authority. I used to fly 15" and 16" but when I upgraded my motors I just bought all 18"s. I would highly recommend this mixing setup, especially with the Herk.

 

[Cheshirecat]

Glad to hear it! Because I had a whole board go out, if my motors were set to dji specs I would have been f'd! Could you supply a link to the new cable with terminal ends? Thanks for posting this in the first place also!

Check out e-bay listing number 400593772482
250mm are slightly long but 200 too short (well in my config anyway)
So you lost a Quad side of the Herk??? I did try setting up the motor scheme to the recomendation From Andreas but could not get it to work. so back on DJI standard Mix (wish i was not)

 

[Carapau]

Ive got an X8 with a MTOM of 20kg and it is set in the standard DJI config. I lost a motor on it whilst flying at 20kgs due to a faulty ESC. The first I knew about it was just as I was coming into land as I noticed one of the props was stationary. No issues with yaw authority what so ever. If you are having issues I would suggest that the rig is perhaps underpowered for its load. My rig uses 18" props on the KDE 5215 435kv motors giving it a theoretical total thrust of 59kgs or 47 if you take into account a loss of efficiency due to coax setup of 20%

 

[maxwelltub]

Ya I lost a whole board due to micro match. Also agree with what @Carapau is saying in terms of losing yaw control with one motor. However if you loss 4, all of which spin one way would be a bad situation. Which is why I use the stacked quad configuration. I'm not sure about the 20% loss, although its really hard to put a hard number on it. I've calculated mine at a max of 15%, testing by measuring thrust at 50% throttle.
But in other news @Cheshirecat that listing number didn't work. Do you know the name of the red end piece that terminates the ribbon cable?

 

[Cheshirecat]

Ya I lost a whole board due to micro match. Also agree with what @Carapau is saying in terms of losing yaw control with one motor. However if you loss 4, all of which spin one way would be a bad situation. Which is why I use the stacked quad configuration. I'm not sure about the 20% loss, although its really hard to put a hard number on it. I've calculated mine at a max of 15%, testing by measuring thrust at 50% throttle.
But in other news @Cheshirecat that listing number didn't work. Do you know the name of the red end piece that terminates the ribbon cable?

Must be a regional thing with E-Bay, search Amp Micromatch 8Way 250mm 1483352-3, if there are several in the UK i would be surprised if you cant find them state side.

 

[Cheshirecat]

Ive got an X8 with a MTOM of 20kg and it is set in the standard DJI config. I lost a motor on it whilst flying at 20kgs due to a faulty ESC. The first I knew about it was just as I was coming into land as I noticed one of the props was stationary. No issues with yaw authority what so ever. If you are having issues I would suggest that the rig is perhaps underpowered for its load. My rig uses 18" props on the KDE 5215 435kv motors giving it a theoretical total thrust of 59kgs or 47 if you take into account a loss of efficiency due to coax setup of 20%

Well on paper (at least) mine is 21 kg less 20 % 16.7kg with a tow of 9.5kg is that under powered?? interesting you had no ill effect with the standard DJI Mix. i would still like to configure as a double quad but my WKM just won't have it. No idea why???

 

[maxwelltub]

Yup found some state side, I just didn't know what to search for except 8 way ribbon cable. I was having trouble finding the ones with the right end piece, but with this info I found a supplier locally