DX8 / E-Flite retracts problems

[John Dimond]

I have a Deuces Wild with E-flite 60/120 trike retracts. As discovered elsewhere on Forum, with a DX8 set up with 'wingtype 2xAil/1x flap', the flaps don't work unless plugged into GEAR on Rx, and with gear into an AUX port. With this set up, things appear to work fine at first - but on repeated test of gear I got intermittent unsynchronised gear cycles and uncommanded servo movements. Horizon have had the DX8 Tx, AR8000 & retracts back for checks, but returned (after about a month) with 'no fault found'. I have very carefully re-assembled the whole set-up - gear first, worked fine, every time. Then I progressively plugged in remainder of servos, and all seemed to work normally. Just when I thought I'd cracked it, and for no apparent reason, the gear started to cycle out of synch' , and various control surface servos started to twitch again. The attached video clip shows problem before it went back to Horizon, but the performance now remains very similar.

 

Has anyone else seen this problem - ANY IDEAS PLEASE????!!!!
http://www.youtube.com/watch?v=uQHJ9NugLcE

[Frank Skilbeck]

To help eliminate the problem, have you tried it with the ailerons on a Y lead and Flaps into flap and gear into gear, if this works you know it's probably something to do with the Tx software/programming.

 

The only other thing I can think of is that your battery pack is a bit weak and after a little use the voltage on the servo signal line is a little low and this affects different servos differently. Are you using a 4 or 5 cell pack.

 

I was helping a club colleague set up a plane with retracts, flaps and ailerons all on Y leads, we had intermittent operation of one flap which we eventually tracked down to a run down Rx pack (this wasn't a Spektrum set BTW)

[John Dimond]

Thanks Frank - previous owner had ailerons on Y lead - I prefer seperate channels for differential etc. I'll try reinstating old set-up to see if it makes a difference - although I feel the DX8 should be able to cope with 2 x Ail set-up. Spektrum are aware of problem - I'm amazed they haven't sorted a fix by now (I don't think they've even raised a service bulletin on the subject - I only stumbled upon 'work round' from another forum post). I've flown the DX8/AR8000 with 2 x ail set-up in two other models (fixed gear) with no problem, so still suspect it'sa something to do with the retracts.

 

I was running off a 3S LiPo via one ESC - still giving 11.0v+ after problems occurred.

[John Payne]

Hi John, an unusual problem. If you are running off one ESC, can it provide the amount of current required for all functions? As I understand it, it isn't necessarily the battery voltage, but the current that the ESC provides that is important. Just a thought, but try it with an rx pack and see it it still happens, even if you have to disconnect the ESC to try it. Cheers, John.

[Pete B]

Forgive my warped sense of humour, John, for I understand your frustration, but there's a good chance that video could go viral amongst modellers...........bordering on the hilarious, I'm afraid - and I've got coffee all over the keyboard........

 

Anyway, on a more serious note, Frank's suggestion to check each function separately is a good starter. I suspect it may just be something as simple as a misbehaving servo somewhere on the model.

 

I would set it up on a fresh memory and just connect the ailerons, rudder and elevator with your present wing type. Once they're functioning correctly, add the flap (and you can try an alternative servo as well) and finally add the gear. If the introduction of any part creates the problem, that will narrow it down.

 

Has the Rx been flashing to indicate a 'brown-out' has occurred when all the servos start cycling? That would indicate the power supply, whether ESC, UBEC or separate pack is unable to cope with the demand.

 

The final option might just be a programming error, so setting up on a clean memory will give you a chance to go through the sequence again.

 

Let us know how it goes.

 

Pete

 

[Frank Skilbeck]

It was the unsynchronised gear that lead me to look at battery as these are probably driven from a single channel via a "tri" lead, so a spurious signal from the Rx should make them all go together.

 

To eliminate the ESC Battery Eliminator Circuit why don't you try a standard Rx battery and leave the throttles unplugged. On my Spektrum receivers the servos can jump when it is powered up hence I'd be checking that an interruption in the power supply isn't causing the Rx to "glitch".

 

[John Dimond]

Thanks chaps - I've removed the ESC/BEC & replaced with a 4 cell 4.5v Rx pack, and all seems to work OK - I can see the Rx light remains constant throughout repeated cycles (previously it would go out for a moment then back on again when one or more gear legs failed to cycle properly).

Conclusion - I think you're right - I need a seperate Rx battery on board. All I need now is to work out how to hitch it up - preferably with externally accessible switch. I'm sure I've seen some info on another thread - I'll get searching..............

[Frank Skilbeck]

You've quite a power draw on the Rx battery there, so I'd personally use a a sub C Rx pack as these can deliver higher currents without dropping too much voltage.

 

As for the hook up, all you need to do is remove the red wire on the ESC's to the Rx (pull the red pin out of the plug and tape it back. A conventional switch on the Rx battery to receiver is all that's required there. The other alternative is to use a separate high power UBEC and this can be connected to your flight pack along with the ESC controllers (just disconnect the ESC +ve wire again)

[Peter Beeney]

Having now read through the thread I would consider, with the greatest respect, of course, that there is yet another solution. It may appear to be a bit complicated, but it would likely give the highest level of security, at least as far as the power supply to the receiver is concerned; but that’s only in my opinion, I would quickly add.
My personal choice would be to always have a battery supply, at least on anything larger than a foamy, this is for a variety of reasons, the most important one being that of safety. If someone loses the main flight pack supply, for whatever reason, and that definitely can happen, then however good a pilot that someone is, with no power to the servos from a dead BEC they stand very little chance of maintaining control. So the model will crash anywhere. Certainly a separate battery supply also won’t guarantee a successful landing, by any means, but at least you have half a fighting chance, as the power is sustained to the rx and servos.

If I assume this is my model, thus making this is a hypothetical case, it’s definitely not any instructions on how to do it, I would simply parallel the two BEC’s together from the two ESC’s, this should easily overcome any current shortfall that I seem to be experiencing when using only one supply. I don’t know what my BEC output is, a guess would be 5V at 2 or 3A. Paralleling these together now gives me 5V at 4 or 6A. I could just leave it at that, but that doesn’t satisfy my requirement for a rx supply in the event of a catastrophic failure; admittedly unlikely, anyway, but to be certain I can also hook up a 4 cell AAA NiMH pack as backup. This has also been done in the past. The little problem here is that the battery is in charge, by some fair measure, at least to start with; however, it will gradually get discharged to some degree. If I take a closer look at the voltages I will see that the open circuit voltage output from the BEC is most likely to be just under 5 volts, perhaps 4.9989V, or say 4.99V. Checking the battery open circuit voltage I might find this is 5. 44V, in a reasonable state of charge. If I now connect this battery in parallel with the two BEC’s this instantly results in the battery taking control and proclaiming that it’s quite capable of supplying all the load, including the 5 volts requirement on the ESC PCB’s. The BEC’s are effectively shut down for the duration. If components are paralleled together, the same voltage is applied to all, for most intents and purposes, anyway.
However, the battery shouldn’t get to much ahead of itself, because this will now gradually discharge under load until the voltage gets down to 4.99 volts, where the BEC’s will then start to take some of the strain. The battery will gradually discharge further, until it’s not really doing anything much, basically now down to 4.99 volts open circuit, so at that stage it’s not really in a good state of charge. This, though, is probably of little consequence, it just means that the battery has to be charged regularly.

With all this in mind a further refinement might be a forward biased diode in the battery circuit, this will then reduce the battery voltage at the receiver to say, 5.33 minus 0.6 = 4.73 volts, slightly below the 4.99 volts delivered by the BEC. The 0.6V is the diode’s forward voltage drop. So now the BEC’s are going to do all the work from the start and the battery will sit there waiting for the call to arms which may never come. This being when the flight battery fails, but in the meantime, if the voltage falls because of any large demands from the servos then it will step in and hold the voltage up to it’s level. If the flight pack did fail, then the battery would only supply 4.73V but because it’s in a good state of charge it will keep this up for quite a while.
One final addition to complete my installation would be an onboard battery voltage monitor, it’s perhaps not a lot of use having this battery in the model if I don’t know what condition it’s in!

I said I’d use triple A cells, I don’t need much capacity, and this helps to keep the weight down. I’m sure that this in general will be seen as unnecessary and too complicated but I’d say it’s very simple to do, and it’s of the highest integrity in terms of power security. I feel there are still too many crashes with models due to electrical problems. One advantage of the added battery system is the fact you can really check everything out, apart from the throttles and motors, without connecting the flight pack(s). Also, for me anyway, that’s another good safety point, now there’s absolutely no excuse for me taking off with the ailerons reversed.

I thought I might describe to myself how I would do all this, but to avoid possibly crashing the forum with a long post, I’ll do a second part.

PB

[Peter Beeney]

PART TWO

With regard to the two BEC’s, I can just slot these into a Y lead, without removing any red wires, and into the throttle channel. Or if I wanted to play around with asymmetrical steering and such, into two channels. The diode, such as a P600D 6A silicon from Maplin for 99 pence, I would insert in the switch harness, the receiver side of the switch, so I can charge the battery……
I’d pop the crimp terminals out of the plug and slide a suitable piece of heatshrink over the wires up to the switch, then fairly close to the switch, out of the way, carefully part the wires and cut through the red. Strip the insulation, for about half an inch. Then I need a servo extension lead, from LMS, cut in half, I need just the socket here, the plug will come in useful somewhere else. Part the wires at the end and cut the red a couple of inches short. I’ll remove the insulation from this and twist together with the red from the switch, then I’d cut most of the wire from the end of the diode away from the silver band, wrap one turn of my twisted wires around this solid wire, a quick touch with a smoking hot iron and it will solder in a flash. I’m not worried about overheating the diode, I’m most unlikely to hurt it. Then carefully bend the diode wire to lay flat against the black, more or less, slide the heatshrink along and shrink down. Then I’d move down the switch wire, about an inch, very carefully remove a small piece of the black insulation by slicing along the wire carefully with a sharp knife, I would not be tempted to going round with the knife to mark the copper strands, that makes a potentially very vulnerable point, tin a small length of this bare wire, cut the black from the socket, the white signal wire can I will just cut slightly shorter, out of the way. Tin the tip of the black wire, touch it against the other tinned wire, wave the iron at it and it’s joined. Then I’d cut the remaining red to length, solder to the diode as before, bend to shape and a piece of heatshrink over this staggered joint will complete the job. I’ve done this a few times, for Y leads, or battery connections to servos etc., I don’t believe in cutting the wires unless I have to.
Now I can very conveniently plug my monitor into this socket, it will give me a true reading of battery voltage at all times.

This makes a lot of reading but now I can forget it. The battery still needs charging, of course, and I haven’t checked, but if there is a low self-discharge 4 cell AAA rx pack available that would be ideal. Also a check with one BEC at the time, to make sure the system hasn't broken down anywhere. Because everything is now working well within it’s limits I can be totally confident it will be ultra reliable.
Also I can add in retrospect, that just the additional battery, without a diode, is still a very good option.

PB

[John Dimond]

Wow! - thanks for that Pete - although I'll need to read it through a few times if it's going to sink in! I have already resolved to power my Rx from a dedicated battery (probably 6.6v 700mA LiFe) I'll make it switchable for convenience/safety, and I shouldn't need to recharge during a flying session (it should long outlast my collection of flight pack LiPos on any given day), but wiil have a spare charged & ready anyway, just in case. The possibility of having the BEC(s) remain 'live' to 'take over' as back-up is particularly appealing - I'll do it if I can.

[Peter Beeney]

Yes, I’m afraid I have to admit it was very long winded, John, it’s just another alternative. Probably easiest to just ignore it. If you are not too familiar with this stuff it really must look like a bit of a rook’s nest, something of a useless tangle. I’ve always found that generally with these sort of bespoke jobby’s the hardest part is getting the bits together!

If I went down the LiFe route, and I would definitely do this, every time nowadays, I would simply fit the batteries and leave the BEC’s still connected, although I think it’s very unlikely indeed that they would ever get fired up. They would simply be permanently shut down by the higher LiFe voltage. A few years back this might have been quite important, just a tiddly smidgen over five volts and you’d zap everything in sight, I believe, I think that’s possibly one reason why regulators are still traditionally just below 5 volts output, perhaps, but today everything is far more flexible. On at least three occasions I’ve back fed three different ESC’s on the BEC output line with 7 volts, for 24 hours, just to see what happened, with no obvious adverse results. However, I don’t know what the long term effect of raising the 5 volts from the BEC to 6.6 volts would be on the ESC’s printed circuit board components, probably very little, but it is an unknown risk. It’s the equivalent of hooking up a 6V, 5 cell nickel pack with an ESC, there is really little to gain. Unless, of course, the BEC output is 6 volts. So there is a choice of connecting or not connecting the BEC’s.

If, indeed, I were flying a twin i/c plane this would only just rely on a single power pack as the radio power supply, and provided I maintained that pack in good order I would consider that easily sufficient, as do just about all the other pilots that fly average sized models.

I would definitely use one of the Hextronic dedicated LiFe / Lipo onboard monitors, though, I think they are very cheap from HobbyCity, make sure it’s switched to the correct scale, (some people seem to get confused!), and then they are brilliant. From the accuracy as well as the visibility angle. As I’ve said before, I’d trust this right down to the red led’s, and the “Stop Flying Now‘ last two red led’s are ultra bright; and, of course, another advantage of the LiFe is that you can charge it at a high rate if necessary, too!

From your video, it would certainly appear that you need a nice solid power supply. From the evidence I’ve seen so far, I certainly think the LiFe’s will be up to it.

Good Luck.

PB

[Frank Skilbeck]

If you LiFe pack has a balancer socket, you could consider one of these http://www.giantcod.co.uk/power-battery-tester-p-406643.html I tested it on a 2s lipo in a small foamy and can confirm that the alarm is loud.

 

Or just use the telemetry on your DX8 to read out the Rx battery voltage.

[John Dimond]

Pete/Frank - I've ordered some LiFe Rx packs - I'll let you know if it does the trick. Frank - I already have some of these monitors - thought they'd be useful because wiring telemetry out to the necelles might be tricky - and yes, I could use it on Rx instead, if I had a spare TM1000 (the one I have is installed in another airframe, don't like to keep swapping over).

 

(I've started another thread on programming these monitor/alarms - perhaps you can help there!)