X8 Skywalker - what goes up, must come down...

[skyscraper]

Perhaps a temperature sensor on the motor  would have helped here... With 100 A going through it it must have got hot!

This one is pretty easy to use.... linear scale 0 - 100 degrees c

http://www.ti.com/lit/ds/symlink/lm35.pdf

regards
Andy

I think the only reason 100A was going through it was because it was already failing and the esc was pumping juice into it to get it spinning. Usual launch power was something like 60A.

Yeah... but my point was that you really need to be notified ASAP that something is wrong, either via OSD or Telemetry on the ground station. You could do analysis of current but a motor that is getting too hot to touch is a good simple indicator (If you are running that hot your magnets will degrade) and easy to diagnose... (in fact sensor goes up to 150 degrees c but IIRC motors shouldnt be allowed over 100 degrees c)

regards
Andy

[markvanhaze]

I see what you mean now. The safe way seems to be to connect UBEC to input side with jp1 open. Use another UBEC on output side if needed for servos and dont use the plus leads (thats 2 cables - not only the red one but also the one beside it) on the power module. I will try that out and let you guys know how it works.

I'm still not convinced, my power module powers the front of my apm which also supplies power to the ezuhf. The sbec powers the servo rail or the back part of my apm. So even if I move the sbec in front of the power module that will still power the back end of the apm, granted, but when the power module gives up the ghost I loose power to the front end of the apm.

That means no GPS, no osd, no ezuhf right?

Rgs
Anthony.

I don't know what the front end of your apm controls, but if its required to stay airborne then yes this solution is useless to you. The ezuhf wouldn't be as essential if your autopilot / RTH could still function. Problem would be watching it circle above you until battery dies  :laugh: still potentially better than going down miles away :+

😂😂😂

The only way for that to work is not to use the power module and power both ends of the apm through the sbec which is how I had it before.

Maybe I could revert back to that and use the power module without the positive lead? Worth a test I suppose.

[UAV Aviation]

I know its an old topic, but since Iam upgrading my X8 I ask the question, why not using separate power sources and circuits for the motor and the avionics??
I heard about burned current sensors before so that's the reason I only have the motor lipo connected to the current sensor and I use separate lipos and becs to power the AP, RX, camera and AV TX.....

[Lola]

Well after the crash markvanhaze had he said about connecting the sbec to the front side of the power module, the side the lipo connects to.

I thought this was a great idea so I did it. I have two sbec's, one powers the front of the apm and receiver, the other powers the minimosd.

The last thing for me to try is to add a third sbec to power the back of the apm but before I do that I will have to remove the positive lead coming from the power module going into the apm.

I'm glad you've brought this topic back up because I had completely forgotten to try that.

I'll keep you posted.

[Lola]

Correction to my last post, getting old lol. The first sbec powers the back of the apm ie the servo rail. The power module powers the front. 😀

[Billy_boy_2010]

I know its an old topic, but since Iam upgrading my X8 I ask the question, why not using separate power sources and circuits for the motor and the avionics??
I heard about burned current sensors before so that's the reason I only have the motor lipo connected to the current sensor and I use separate lipos and becs to power the AP, RX, camera and AV TX.....

I suppose you can.

But I'm not really sure how much it helps. Are you building in more points of failure or more backups and failsafe? 

I suppose, by putting each egg in a different basket, you're allowing the plane to part function in some way regardless of which power source failed. But, with more components in the plane, There are more ffailure points in there in the first place (even if the consequences of a failure are now lower).

Personally- I can never be bothered and I swear by a one battery system. My planes just aren't worth enough to worry about like that. And the added complexity is just more stuff to go wrong. The only thing I do do is run a separate bec to power the RX and the servos- and the bec output is always way above the draw so it's far from stressed.

[UAV Aviation]

I hear what you say, but the critical point in the whole system is the current sensor... so why not take it out of the RX, VTX and AP power supply loop??

[dogzilla]

If you are worried about the current sensor going then get a dual input power switch, something like this : http://www.jetimodel.com/en/katalog/Accessories/BEC-Regulators-Switches/@produkt/DSM-10/

That way you can measure your current as normal and if the current sensor goes out, the switch will failover to the secondary input. This could be done with two 12v regulators, one each in front and behind the current sensor.

That also gives you a redundancy with your 12v supply.

Only downside is having to rely upon the switch but the assumption is that the switch and 2x PSU's is more reliable than an individual PSU on it's own.

For the kind of flying I'm doing at long ranges, I'm really not expecting to recover the airframe no matter which way it fails. So I concentrate on 1) keeping it simple  2) using good quality components and 3) testing the hell out of the components before they go in the plane and also in the air before I commit to taking it far.

The battery, current sensor, AP, and BEC are all pretty much fatal if they go, we can fix most of those things if we really want to. PROPER drones have two AP's and some kind of logic switch in the middle which can flick over if one of the AP's stops responding as it should. I'm sure someone has done that with a pixhawk or apm already.

For me though, complexity is a risk factor, both technically and from human error perspective, so there is definitely something to be said from keeping things as simple as possible.

[skyscraper]

I am pretty sure OP used a current sensor based on a resistor which will overheat with overcurrent. https://store.3drobotics.com/products/apm-power-module-with-xt60-connectors.

MFD ( which you are using?) current sensor is Hall effect AFAIK which doesnt overheat with overcurrent. http://uk.rs-online.com/web/p/product/7248739/?grossPrice=Y&cm_mmc=UK|Shopping-_-Google+PLA-_-Allegro+Microsystems|Hall+Effect+Sensor+ICs-_-7248739&kpid=&kpid=7248739&istCompanyId=f7e7b05b-2daf-4c0e-8825-3633baf8113b&istItemId=xwippixpx&istBid=tzit&gclid=CJvqruvf6ccCFSQHwwod6hoDhw

But yeah as dogzilla says ...  K.I.S.S. all the way.

regards
Andy

[dogzilla]

Yeah that's a good point. As far as I know you can melt hall effect sensors but it requires a vast amount of current to melt the copper, just as much as you'd need to melt a cable. So like 200-300 amps. Your battery would probably explode first.

Perhaps look at getting ESC's which have over-current protection which will shut down if they start drawing too many amps. Lots of them do that. for example they won't burn up if the prop gets stuck and will shut down.

That would also stop your current sensor exploding.