What exactly killed my big seagull

[r6dan]

If you see what I meanPosted by Geoff Smith 1 on 23/03/2011 14:05:16:

Hiya Tim, just a question from a newby. What does IYSWIM mean, I've seen it crop up a few times and have now got very inquisitive. Cheers mate.

 

 

Geoff

If you see what I mean

[Geoff Smith 1]

Posted by r6dan on 23/03/2011 14:16:16:
If you see what I meanPosted by Geoff Smith 1 on 23/03/2011 14:05:16:

Hiya Tim, just a question from a newby. What does IYSWIM mean, I've seen it crop up a few times and have now got very inquisitive. Cheers mate.

 

 

Geoff

If you see what I mean

Thanks Dan, another mystery solved. Cheers

 

Geoff

[Simon Chaddock]

In the world of probability would a separate independent radio battery actually improve the overall system reliability?

I think the answer is no.

 

As Tim has suggested some sort of automatic back up system would improve things but the real problem is you don't know in advance what and where the failure is actually going to be.

 

Maybe an automatic airframe parachute system is the ultimate answer like this.

[Chris Channon]

Hi all, so was the crash because of the tinned but not soldered joint as per BEB ?

Thanks

Chris.

[Tim Mackey]

Well all I can say with certainty is that it WAS soldered immediately after tinning - there is no way I ever start to solder something and don't actually complete it.

I guess we will never know for sure why it came unsoldered - but it did.

Dry joint, air-pocket, whatever ...these things do happen unfortunately.

I shall just be even more careful and also more brutal with testing all joints physically after completion.

[Tim Hooper]

An interesting investigation, young Timbo! Very objective and thorough.

 

I have to confess that I didn't understand much of the equations and arithmetic, but the end result seems very conclusive.

 

I'm now wondering whether to use a Ubec in my latest project, or to hedge my bets and fit an old-school rx pack.

 

Hmmmmm...

 

tim

[kiwi g]

An old school flight pack or not , the bad connection can still exist.

we have seen that for years .

 

End of the day .. It was bad luck , not bad flying , or bad building ...damn shame that pc9 is a great plane and flyer.

[kiwi g]

hiccup sorry

Edited By kiwi g on 23/03/2011 19:44:26

[Peter Beeney]

Tim,

It is a great sorrow to read about the demise of not one but three models in one day. That’s a very abysmal story indeed! Certainly I can well appreciate the feeling of helplessness as the model crashes out of control, that’s happened to me on more than one or three occasions!

Regarding the PC9, after looking at the evidence, I think I would have to agree with BEB and say that I don’t think that the fact that the connector heating up was entirely the cause of the wire becoming disconnected from the banana plug. I don’t have a ready explanation, but, for me, there would still be quite a few unanswered questions. Although I’m afraid to say I don’t fully understand BEB’s little calculations anyway, so I’m simply trying to consider the practicalities of it all. I think I would be inclined to try a few experiments in terms of trying to establish exactly how hot the cables and connections became with these amounts of load on; and measuring a few individual volt drops whilst I’m there.

I’ve used used these 4 mm banana plugs and sockets in the past, for perhaps considerably more current than 100 amps, but I have say straight away this was in relatively very short bursts. I never saw any evidence that anything ever became heated at all, other than the smoothing capacitors on the input to the ESC; but that’s another story. My problem was the cells. Nicads were the batteries then, and if you drained a pack of say 16 by 1700mAh Sanyo’s at anything up to about 150A , they tended to very rapidly run out of steam, so to speak. One or two of the other lads I knew were giving them even more wellie! I had two cell that went open circuit at different times, I cut the top off of one to find the strap from the positive plate to the centre contact had completely burnt away. I repaired it with a piece of wire, wrapped some tape round it and used it as a glow driver for a short while.

There is one tiny point that I’m confused about. I’ve never seen an Eagle Tree logger, unfortunately they don’t play with a Mac, so this has to be an assumption on my part, the fact that it doesn’t have it’s own power supply. It would appear to use the supply from the main battery, I suspect the demand is quite light.
So, now the battery negative leg has become disconnected, thus everything instantly fails, including the supply to the logger. This makes me wonder how it still appears to maintain a recording? I appreciate the running details are in the memory, and don’t require a supply, but I’d expect the recording to cease when the power failed. Of course, if it does have it’s own independent supply then that explains it. Or, maybe there is probably another perfectly simple explanation which I’m unaware of; and I’ll kick myself for not thinking enough!

Hope you manage to resolve it all for the next one!

PB

[Vecchio Austriaco]

Tim, you said AWG12 - this means approx 4mm2

If you start from an ambient temperature of 20 degC (inside fuselage) you will reach a conductor temperature of 180grdC at a constant current of 67A. At this condition the cable has a temperature of 167degC at the cable surface. Looking at your diagram you were more or less in this region if we consider you had 95A peak and sometimes also much less. In my opinion the cable is on its limits. But normal tinn/lead solder (if you didn't use any low melt alloy) should still be solid at 180degC

The temperatures I calculated are the final temperatures the cable would reach in continous operation - after 25seconds you are below these values.

 

If you build the model again I sugest to use 6mm2 (AWG10). This would allow 86A continous without going above 180degC. As you don't fly at full throttle all the time this should be sufficient. Also I suggest a 2 cell lipo, 1500 or 1800mAh combined with an electronic voltage regulator/switch as for instance the digiswitch from Power box systems. This what we use on our Sebart Angel50. Problemfree and it shows you also the battery voltage level (3 color led) I suggest such a solution for all larger electrics.

 

I know - after you had your problem everybody knows it better - but I think we just want to help and inform others for the future

Edited By Vecchio Austriaco on 23/03/2011 21:07:49

full of spelling errors even before I go to the bar...

Edited By Vecchio Austriaco on 23/03/2011 21:09:50

[Biggles' Elder Brother - Moderator]

Interesting VA - I didn't do the analysis for the wire - only for the connector itself!

 

BEB

[Tim Mackey]

Peter B said

 

"So, now the battery negative leg has become disconnected, thus everything instantly fails, including the supply to the logger. This makes me wonder how it still appears to maintain a recording? I appreciate the running details are in the memory, and don’t require a supply, but I’d expect the recording to cease when the power failed. Of course, if it does have it’s own independent supply then that explains it. Or, maybe there is probably another perfectly simple explanation which I’m unaware of; and I’ll kick myself for not thinking enough! "

 

Thanks for the sympathy Pete

However if you check my report you will see that the battery leads did NOT become disconnected - it was the ESC input leads, which come form the OUTPUT of the logger.

[Tim Mackey]

Thanks VA - I will use heavier gauge cable on any re-build. However, I should point out that the cables already fitted to the battery are thinner than 12 gauge, and the ESC input leads look positively weedy !

Surely the leads supplied already fitted to such components should be capable of carrying the potential currents for which they are intended - in the case of the battery 150Amps, and the ESC 100A constant, 130A burst. In that case my oversized cables should have been more than sufficient, but, I will as you suggest "go larger"

AS to the switching /power supply I already have a couple of models which use the powerbox digiswitch fed by 2s lipo - and very fine units they are too.

[Tom Wright 2]

Tim Would i be correct in saying that if the data logger had not been installed ,you would still have a nice undamaged PC9 ?

 

TW2.

[Peter Beeney]

Tim,

Yes, thanks, obviously I should have read what I was looking at a bit more carefully, it was the simple answer after all!
However, from looking again at the photos it seems as though those cables are very much the standard size ones? Presumably they are as supplied? I’ve never particularly heard of any cables giving very much in the way of trouble, more usually the device it’s connected to. Although 12 AWG doesn’t really sound very inspiring, I have to admit.

If I can find some bits handy I’ll cobble together a load frame and try to apply a little ‘warming up’ to some cables and connectors and see what happens in practical terms.

PB

 

[Kozmyk]

Vecchio that's Very interesting.

Could you show the calculations that you used?

What you've come up with may be the closest thing to a reliable method of calculating cable size requirements for electric flight modelling.

Tables of values published on the net are nearly always referring to long cable runs or AC house wiring which just leaves us having to guesstimate.

We could have a reliable table of cable current capacities for modellers then.

[Tim Mackey]

Posted by tom wright 2 on 23/03/2011 22:20:51:

Tim Would i be correct in saying that if the data logger had not been installed ,you would still have a nice undamaged PC9 ?

 

TW2.

Not really - we know the cause was down to a soldered connection "letting go", I fit the connectors on the entire power-train, so even without the logger in series, the culprit connection would have still been the ESC negative input connector - so power to the ESC would have shut off exactly the same. The difference is, that without the data, I would be left wondering what happened.

1) Battery failed?

2) ESC failed?

3) Signal failed?

4) Pilot failed?

5) Connector failed?

6) Receiver failed ?

7) UBEC failed?

8) Cable failed?

etc ...you get the idea

[Tim Mackey]

Peter - yes the cables were as fitted to the ESC input as bought.

The slightly heavier ones were also standard as fitted to the logger output.

[Tim Mackey]

Kozmyk - Surely I am not alone in thinking that any cable fitted to a battery /ESC etc as purchased should be fit for purpose, and capable of carrying the potential currents .

IE the battery is 5A30C ...thats 150A.

[Tom Wright 2]

Tim

Now i am confused ...if it had a u bec fed from the batt side ,how come the model lost signal?

 

TH2

[Kozmyk]

Posted by Tim Mackey - Administrator on 23/03/2011 22:56:08:

Kozmyk - Surely I am not alone in thinking that any cable fitted to a battery /ESC etc as purchased should be fit for purpose, and capable of carrying the potential currents .

IE the battery is 5A30C ...thats 150A.

 

Tim I'd certainly Hope so but that's leaving a lot to trust isn't it?

With the considerably higher C ratings we're seeing these days the situation could easily become marginal.

For instance, where can you see online a table that gives a comparable current rating to that which we are asking of our battery leads?

I realise that the insulation is rated to a higher temp than ordinary cable (typically 200°C) and that allows us more current than the tables say but there is then another danger in that 60/40 solder melts at 183°C.

 

This is a side issue to your crash investigation, sorry for that, but Vecchio's post inspired the question "What's the REAL ampacity of our cables?". Perhaps a separate thread is in order.

Edited By Kozmyk on 23/03/2011 23:31:49

[Tim Mackey]

Posted by tom wright 2 on 23/03/2011 22:59:54:

Tim

Now i am confused ...if it had a u bec fed from the batt side ,how come the model lost signal?

 

TH2

 

It is nothing to do with signal. The UBEC is hardwired into the ESC input leads, so that plugging in the battery ( in this case via the logger ) to the ESC also powers up the UBEC.

Conversely, unplugging the battery at the end of flight ( or in this case mid flight! ) removes power to the UBEC also.

[Tim Mackey]

Hi Kozmyk - yup separate thread me thinks over in electric flight general topics - good topic

[Tom Wright 2]

Tim.

Would it then be fair to say that if the ubec power pick up point was at the deans and no data logger was fitted the model could have been recovered in dead stick mode?

This may be a simple or obvious comment but to my mind does raise significant  questions about installation reliability.

TW2.

 

 

Edited By tom wright 2 on 24/03/2011 00:02:34

[Tim Mackey]

I dont use deans Tom.

Its simply impossible to make these kind of assumptions - a soldered joint failed, and that joint could easily have been anywhere in the main power supply chain. Whether the logger was fitted or not, and despite where the UBEC connects, that joint failed.

It was suggested earlier that had a separate battery been fitted, maybe the model could have been saved - maybe, but certainly not guaranteed at all.