Nasty incident

[PB]

I've recently had a nasty incident in my garden with a new EP model under test.

I had just tested the model with a Wattmeter and tacho and all was well, drawing 36A at full throttle from a 3s 4000 Lipo connected to a new 40A ESC driving a 35-48/55A motor. This set up gave me 7600 rpm with a 12 x 6 prop - perfect - ready for the flying field maiden tomorrow.

I then went to carry out my standard disconnect routine. I put the Tx on the ground, still switched on of course, with throttle stick fully down. I went to move to the model to disconnect the battery but I stumbled and my automatic reaction was to put my hand out as I went down. My extended hand hit the Tx and switched it off.

Before I could recover my stance the motor in the model powered up to full rpm and the model careered across my garden and smashed into my swing seat trashing one wing. I was shocked and astonished that this new ESC had put the motor on full power on loss of signal.

When I recovered my compoure and stopped shaking I carried out some tests of some spare ESCs and motors using a spare Rx. One other ESC also put the motor on full power when I switch off the Tx, but two other ESCs did not. Also, the ESC which put the motor on full power did so whatever the position of the throttle stick, that is, if it was on say half throttle and I switched off the TX this ESC put the motor on full power. The two ESC which didn't put the motor of full power when I switched off simply cut power to the motor if it was running when I switched the Tx off.

Now OK, I have never EVER switched off my Tx before disconnecting the battery from my EP models - it's my unfailing disconnect routine to leave the Txd on - so this was something completey new to me. Surely it is dangerous for an ESC to put full power to a motor on loss of signal, so are these two ESC of mine that do so faulty?

Thoughs would be greatly appreciated.

[jrman]

Sounds like you hadn't set the failsafe correctly.

[WolstonFlyer]

What make and model of TX do you have.

I can turn my DX8 on and off and nothing will happen (I have done it by accident a few times) because the RX bind is done with the throttle off and this sets the failsafe.

Edited By WolstonFlyer on 18/07/2013 16:48:48

[stevejet66]

Yep you failed to set the failsafe, Rebind it with the throttle closed, Reconnect everything, power the motor up on slow revs then turn the tranny off, If the motor stops your failsafe is now set!

[Peter Beeney]

PB - I’ve not checked this out yet, but it might just be tied up with the ‘failsafe’ settings? Assuming 2.4 GHz? If on loss of signal the receiver continues to send a signal to the ESC it’s maybe possible that the motor will run. But if the receiver power is lost for whatever reason, but there is still power to the ESC, then I think the motor will stop/not run regardless. It needs to be checked out, I guess.

I have to say, I aways encourage the use of a restraint, under any circumstances, and I have to say that if I were doing any such checks (with a propeller on!) the model would stay in said restraint until every thing was disconnected. Models have run amok before, usually when the operator least expects it!!

Very pleased to see you had no personal harm done, at least.

PB

[PB]

The Tx is a DX6i. Yes of course you're absolutely right, but that's so unlikely it never even occurred to me. I'm completely mystified. To the best of my knowledge I have never bound any of my EP setups with the throttle stick fully up. I mean to do so is just unthinkable, but it must be the case because I've rebound those two ESCs and they're now OK.

I have no idea how that came to be, but it's a lesson learned - at a cost.

Thanks for the input.

[Biggles' Elder Brother - Moderator]

Object lesson in why setting the failsafe first - straight after binding and before you put a prop on it - is absolutely essential. Never just trust a live electric model - even the sweetest of them can become demons!

BEB

[Steve Hargreaves - Moderator]

Ooo nasty indeed PB. Good job nobody was injured.....

What Rx are you using??

As the others have suggested it does sound like a failsafe issue....if the failsafe was set to go to full throttle on loss of signal then it did what it was supposed to do I guess.

I'm sure you are sad at the damage to the model but count your blessings...it could have been much worse had your fleshy bits been in the way.

It wouldn't have saved you in this instance but one trick I have recently started using is a free mix on a switch to negatively mix the throttle channel back into itself. When the switch is flicked this effectively disables the throttle control and prevents any sudden start ups if the throttle stick is accidentally knocked; whilst carrying the model to the strip for example. Check your radio manual and see if your set will allow you to do the same.

[Andy48]

The other thing I would do is fit a safety plug to your model. This provides an extra level of safety protection. I always stand astride the fuselage to bend down to insert and remove the safety plug.

[WolstonFlyer]

He DX6i does not allow you to mix throttle on its self like the DX7 and 8 do.

One common mod on the 6 is to replace the throttle cut push button with a toggle switch.

[Biggles' Elder Brother - Moderator]

The other point here of course is that the model was obviously not restrained. We recently brought in a new club rule stating that all models - including electric - must be restrained when "live". This followed a recommendation by BMFA to that effect. If I'm watt meter testing or just "running up" at home I still use the restraint - Mrs BEB moans about the holes in the lawn - but tough, at least I've still got all the bits I should have!

BEB

[PB]

Thanks Steve, I'll look at that. This incident has got me quite worried to be honest because somehow, it seems, I've managed to set two of my ESCs to 'fail dangerous' and I simply have no idea how that could possibly have happened. What procedure did I do to get them in that state? I have no idea.

The only time I have switched my Tx on with the throttle stick fully up is when I want to put an ESC into programming mode. That seems to be the industry standard method of entering ESC programming modes. So OK, I've switched the Tx on with the throttle stick fully up and the ESC goes into it's programme mode. Fine, but there's no bind plug in the Rx at the time so how can the Rx pick up that throttle setting and reset the failsafe? It doesn't make sense.

It's just a mystery I'm going to have to live with I guess but you're right BEB, I'll never trust a live EP model again after this one.

[Erfolg]

BEB

I do wonder how many of these rules are written by electric fliers.

Many electric models require the model to be on its back, to arm and disarm. It would be laughable to see a 3m model with a separate nose pod using a restraint. The use of restraints either will not work or be an encumbrance with many arrangements. A clearly IC derived rule.

I have also turned of my Tx with the motor running and off. With the motor running, they have continued. When stationary, they stay stationary. With the Rx failsafe set they power to low throttle. I mainly use HK Blue Series, although I do have a few others.

A proper solution is not just to prop your TX any old way, to use only Esc that do not go to full power. To arm and disarm the model from the rear.

One of my clubs requires you to disarm on landing, the other when you get into the pits. I prefer on landing, though the issue then is, if someone decides to land, although this is the same as retrieving a model, just a few seconds longer, which is probably age and dexterity related anyway.

[Pete B]

Posted by PB on 18/07/2013 17:54:33:

The only time I have switched my Tx on with the throttle stick fully up is when I want to put an ESC into programming mode. That seems to be the industry standard method of entering ESC programming modes.

I'm not sure that may have affected the programming of the failsafe, PB, but the recognised method which I use when I want to set the throttle limits or programme the ESC is to switch on the Tx with the throttle in the closed position, then open it to full, before connecting the Rx to a power source.

Once it has beeped to recognise the throttle position, close the throttle, or leave it open if you wish to go into the ESC programming mode.

Once you've set up your model, it's always advisable to re-bind it so it memorises the failsafe position.

Pete

[GrahamC]

What you have [probably] done is reversed the throttle channel after the bind....

When the plane is all set up, the last job should be to rebind the transmitter to the receiver.

[Erfolg]

The rebind issue seems to be a Spektrum issue, as I have seen a few problems at one club where I am a member.

I have not seen this issue with Futaba on the 8fg or 6ex (which I have). In fact i switch of my Tx to test the failsafe, then back on for control.

I have once powered up the Rx without the Tx being on, nothing happened at all, other than the failsafe positions being found. For some reason i do not like doing this, therefore only done once. I was conscious that I was near the propeller if things did not go well and I would have to disarm the model. It seems unreasonably risky, for no apparent gain.

[Andy48]

Posted by Erfolg on 18/07/2013 18:14:46:

Many electric models require the model to be on its back, to arm and disarm.

Not if you fit a safety plug.

[Biggles' Elder Brother - Moderator]

Restraining electric models is not just an IC derived rule Erflog - it's recommened sensible best practice.

Bear in mind in the case of smaller models - which is what I suspect you have in mind - simply holding the model in your hand, or an assistant holding it, would constitute a reasonable and fully acceptable, restraint. What in my view is not reasonable or acceptable is leaving a live model with no restraint on it what so ever - unless of course its on a strip and about to either taxi or take off!

BEB

[GrahamC]

Posted by Erfolg on 18/07/2013 18:58:57:

The rebind issue seems to be a Spektrum issue, as I have seen a few problems at one club where I am a member.

I have not seen this issue with Futaba on the 8fg or 6ex (which I have). In fact i switch of my Tx to test the failsafe, then back on for control.

I have seen the issue on a Futaba radio, but being a Spektrum user,I couldn't tell you which one. The problem occurs on radios where the failsafe is set to the throttle position indicated on the TX when the bind occurs. If you subsequently reverse the throttle channel, the receiver doesn't know that you have done it... So failsafe is now full throttle.

I've seen it at least three times at my club. On one occasion, a model shot through the pits... Fortunately missing everyone... And that was a Futaba radio too.

This is to my mind a very significant issue as it seems that there are a significant but small proportion of models out there which will turn into unguided missiles if the transmitter is turned off by accident.

I think that the BMFA advice is now that the correct operation of failsafe be checked before every flight. I would have thought that on arriving at the field might have been more realistic.

At the risk of starting a very boring debate, I wonder if it should be routine to ask members flying a maiden flight to demonstrate that the failsafe opporates correctly before the first flight. .

Edited By GrahamC on 18/07/2013 22:12:59

[Eifion Herbert]

The only way I can think that would happen is when you first set up the model and bind the receiver and then find the throttle channel needs reversing, so you reverse it. Now while the Tx is on everything is fine but when radio link is lost the receiver retruns to the fail-safe positions that were set when it was bound, which in this case would be high throttle. This is why with spektrum kit the proceudre is bind receiver, set up servo reversing, throws etc, then last of all rebind to set failsafe.

[Erfolg]

Graham

The sequence of setting, is more to do with human failings than the equipment having a fundamental issue, as you describe with Futaba. Even worse, the failsafe was not tested in private, at home, before subjecting the public to the vagaries of the model and its equipment.

I am absolutely shocked at the level of incidents as you describe them. I can honestly say, at my principal club, which is 100% electric, we have never had one incident as you describe. Are we perfect, OMG no, far from it.

Yet I am beginning to get the feeling that many are seeing the Failsafe as a major step change in safety. I do not. In all my life I have only seen two flyaways, both in the days of 27, when much of the equipment was a long way from 100% reliable. Yet I do advocate the usage, even if the improvement is only statistically 0.01 better, nothing, I judge it is worth doing.

The main obstacle to my safety is often myself, on that basis I do all my testing very carefully, as I know the main problem is me,

[Peter Beeney]

PB - As it so happens, something similar has happened to me recently. I was playing with something else to do with electric motors, and I must admit I was really surprised when the motor went to full speed on switching the transmitter off. Unfortunately I didn’t stop and investigate it there and then, but at the time it did occur to me that the throttle control would not have been reversed, also I think that generally speaking it’s most unlikely that I or anyone else is going to run the binding procedure with the throttle stick in the up position. If I can reproduce the situation I will endeavour to establish the circumstances that allows this to happen. With the default servo direction being normal with respect to Spektrum, and this being the default of most ESC units, (it seems to me, at least), there is not much necessity to reverse this control anyway.

With regard to the restraint, admittedly the motor on a large powered glider with a folding prop and lying on the ground might not in fact be able to turn too freely if it starts up, but many models have undercarriages these days, and are reasonably powerful, such as the trainer types around, the Riot, for instance. In my experience models don’t start unexpectedly when they are being powered up, they start unexpectedly when they are standing all on their own, and no one is expecting it; and again, in my experience of other electrical faults, this is the the result of two different situations coming together at the same time, such as a model accidentally being left powered up, and someone perhaps then turning the wrong transmitter on, and/or having the wrong model model memory. I know these things should not happen, but with the best will in the world they sometimes do, and at least a restraint will prevent the model from haring off into the back of someone’s leg. The last-ditch failsafe situation, perhaps?

In my view safety is to some extent self discipline, and so adopting a procedure that appears to be be safe and then adhering to it I would think is sensible idea, it’s what I do. It soon becomes easy, like wearing the seat belt, or taking the handbrake off. I know that these can be missed, too, but not very often; and hopefully when that other ‘situation’ is not lurking close by to cause a major mishap, either. If you consider that an electric model can start without warning and treat it accordingly then you shouldn’t go far wrong. Probably 99% of models are never going to start under these circumstances, but just the 1% could be a personal disaster. I’d tend to think that anything that involves some sort of input from a human being has the potential at some point, sooner or later, to be mis-operated.

PB

[PB]

That's sound advice Pete B thanks, I'll do that from now on. I don't really want to drag this on, but thought I'd report that I have now carried out lots of tests with a motor clamped in a vice in my garage with battery, ESC and receiver connected and I have tried every which way to get the setup into 'fail dangerous' mode 'accidentally', but haven't been able to do it . I even tried binding the Tx as normal then leaving the bind plug in while trying to get the ESC into programming mode - perhaps that's what I did - but that doesn't work because the Rx needs to bind first.

If I could identify a procedure which makes this condition happen then I'd be prepared to admit that I made a mistake of some kind, but I can't and that's worrying. It's shaken my confidence a bit, and I'll be ultra cautious from now on.

Thanks for all the input........it's probably run its course now.

Edited By PB on 19/07/2013 08:51:23

[GrahamC]

One possibility is that the model memory being used had previously been used for an IC model which required the throttle channel to be reversed, and the memory had not been wiped before setting up the model?

One observation I would make is that as a relative newcommer, I've never flown without a computer TX. One of the incorrectly set failsafes I have expereinced was my own mistake, and I got bit by a WOT 4 foam-e, and I've learn't from my mistake.

All the other cases were more experienced modelers than me. I think it may be the case that in the past, failsafes were set up in different ways, and that some flyers simply havn't realised that [most] failsafe positions are stored in the RX with no reference to what might have subsequently changed in the TX. That requires a different way of thinking.

Anyway - you may never get to the bottom of it, but the lesson learned perhaps is that a live electric model might not do what you expect it to do, and so should always be treated with the utmost respect.

[Andy48]

Posted by Peter Beeney on 18/07/2013 22:49:13:

In my view safety is to some extent self discipline, and so adopting a procedure that appears to be be safe and then adhering to it I would think is sensible idea, it’s what I do. It soon becomes easy, like wearing the seat belt, or taking the handbrake off. I know that these can be missed, too, but not very often; and hopefully when that other ‘situation’ is not lurking close by to cause a major mishap, either. If you consider that an electric model can start without warning and treat it accordingly then you shouldn’t go far wrong. Probably 99% of models are never going to start under these circumstances, but just the 1% could be a personal disaster.

I will repeat again, if you fit a safety plug correctly wired in to electric models then the model simply cannot start without warning even with a battery connected. I take my plane onto the flying field, switch on the transmitter, stand astride the model behind the wings, insert the safety plug with the transmitter on the ground, stand back and then go through the pre-flight checks.

On landing, I remove the safety plug in the same manner and carry the model, now completely safe off the flying field.

All my models have a safety plug, and once that is inserted, it sends a clear mental message that the plane is armed.