Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length.

[NFMike]

Right, I'm new to RC, though not to modelling or interference suppression, so bear with me. I have had a trawl for info but not really found anything to help me here.

Long and involved 'story so far':-
Twin prop kit, two MFA 6:1 geared motors, 12V SLA battery, fusegear and ESCs from ACTion (1 combo twin/mixer for propulsion in mode 3, 1 straight esc for bow thruster), Hitec 485 servo for rudder, Giant Shark T4U tx/rx pair (2.4GHz). The tx is in mode 2, the rudder is l/r on the right stick and CH1, throttle is up down on left stick, thruster is l/r on left stick.

When I was installing the rudders and servo I did a trial hook up of the battery, rx, rudder servo and tx. No ESCS or motors. It worked fine except the servo had a more or less continuous twitching - not a lot of motion but clearly audible and visible.
I've now got the rest of the kit installed (but just the bare motors - no propshafts, etc. installed) and have been testing. I did have some problems with motor interference but I think I've got that handled now. What I'm left with is, well, unexplained.

It appears that the twitching is still present and is affecting all the channels. With the throttle at 0 (all motors are stopped) not only does the rudder twitch as before, but the motors and bow thruster do as well. The motors sometimes twitch in opposite directions (the LEDs on the ACTion fuse boards tell this) which could/would be the steering channel via the mixer, but also sometimes in the same direction which must be the throttle channel. The bow thruster LED also shows twitching.

Clearly it isn't motor interference (the first test had no motors or ESCs at all and the twitches are initiated when nothing is moving). I did wonder about wifi interference, so just tried turning off our wifi, but that made no difference. So I'm left with the suspicion that it is the tx/rx - they are extremely budget, so not complaining really. Has anyone else had this?

A different 'problem' is that when I first power up (tx first, then boat) the motors blast off at full speed. I'm not sure if they come to their senses on their own or when I move the throttle stick. I also (deliberately) turned the tx off and again the motors went into full forward, which is slightly worrying should the tx fail while on the pond. I've seen talk of failsafes, etc, but don't know if or how that might relate to this set. Any advice gratefully accepted.

[Netleyned]

Are you setting up under or close to an old fluorescent light?
I have seen the sort of twitching you describe caused by one of those.

When you switch on the TX first ,give the throttle stick a couple of waggles up and down
then centre and switch on the RX. The motors should behave then.I do this with my T4U.

Could be other reasons for it and you will probably get a plethora of 'cures' coming up on here

Ned

[Martin (Admin)]

Yes, I had the "full speed ahead syndrome" with my T4U Tx too but as Netleyned says, this is a 'known fault',
 ( also know as Helicopter fail safe mode ) - just 'full circle' both sticks after power up so the Tx knows you are there
 and didn't switch on by mistake.

 Also, try the radio with the Tx ( always on first! )  a 4.8 Rx battery (4 pencells etc.) and just the rudders servo, see if that's OK.
 If so, try the servo iin all the channels.
 If not try another servo.

  .... and as Netleyned say try it in the dark! 

[NFMike]

Thanks for that tip you two. Waggling before receiving seems to work.

Ned, I have some CFLs I use as work lights and already tried turning them off.

However, I have made some progress - although only sideways at present - but need to bounce it off someone else first.

[malcolmfrary]

The fluorescent light 'should' not be a problem on 2.4GHz.  On an older system, yes, it will give RF interference which will show up on the first channel in the data frame, normally the one that does the steering.
A twitching servo is often a sign of either a broken servo or iffy power.
Failsafes intended for air use do not always help anybody else, and can often do the exact opposite for us.

[scoop]

Sometimes standing too close to the model can cause twitching, walk away (about 10 yds) with everything switched on and see if that stops the twitching. Have you 'earthed' the motor cans to the propshafts ? You should never switch your Tx off with the motors still running as they will go into failsafe, if it's set up wrong it may go to full power instead of to stop, have you also tried servo reversing to stop the twitching ?

Just a few things I've come across and learnt over the years, 9 times out of 10 though the trouble comes from keeping boats in a cold damp shed where corrosion can occur in the delicate electronics found in Rx's and cable connectors etc, a quick squirt with electrical cleaner can sometimes solve this problem.

Regards
Scoop

[dodgy geezer]

I'm not too happy about the initial servo twitch. That has never gone away? Is is only on one channel, or is it associated with that servo? If the servo is dodgy, it may be sending all sorts of interference into the system...

The T4U is cheap, but works well. Cheap, however, means no quality control, and I have had T4Us with quite poor soldering on the wires to the pots. If the connections in the Tx are iffy, this could cause your problem.

Have you used the T4U with other servos and found it to be solid?

In general, you fault-find with R/C by replacing individual units until you can narrow down the fault to one item. Then you usually replace it.. 

[NFMike]

Thanks for the inputs.

Sometimes standing too close to the model can cause twitching, walk away (about 10 yds) with everything switched on and see if that stops the twitching.
I did sort of try that - but only the other side of the room.

Have you 'earthed' the motor cans to the propshafts ?
Not yet, but the shafts are not in the tubes anyway - the test was with free motors.

You should never switch your Tx off with the motors still running as they will go into failsafe, if it's set up wrong it may go to full power instead of to stop,
I was deliberately checking to see what happens if the tx goes off.

have you also tried servo reversing to stop the twitching ?
Not sure how that can help as things would end up backwards.

keeping boats in a cold damp shed
It's all new stuff being assembled in a warm, dry spare room.

The T4U is cheap, but works well. Cheap, however, means no quality control, and I have had T4Us with quite poor soldering on the wires to the pots. If the connections in the Tx are iffy, this could cause your problem.
Yes, I am aware of this and before using the tx/rx I took them apart and remade a number of iffy looking joints (all on the tx in fact). But that's not to say I haven't missed some. (In fact there are a couple of small stickers on the tx board with hand written arrows that appear to point at joints that look remade, so I think there may actually be some testing and remedial action done somewhere.)

Have you used the T4U with other servos and found it to be solid?
I only have the one servo. I got a decent spec Hitec unit, rather than a 'bargain' unit, in the hope that it would be reliable.

Having said all that I have made progress, though it is a very strange tale, to me anyway.

I decided to put the mixer/esc into 'tank' mode, which effectively changes it to a pair of unmixed escs, so I could better see what was going on.  This meant using another r/c channel and also removing the the rudder servo lead from the mixer/esc unit and putting it direct into the rx. Much to my surprise all the twitching stopped!

However, I then found that when 'stopped' the port motor (only) was continuously 'stepping' about twice a second until enough forward/reverse was applied for it to run continuously. Starboard was fine. This didn't seem to be an erratic, interferency thing - it seemed fairly regular, though not clockwork.
To find out if this was being caused by the rx or the esc I swapped the port and stbd r/c leads in the rx module.  Everything then behaved perfectly - no twitching, no stepping of either motor (just the motors were now the wrong way round).  Swapping the rx plugs back again results in the port motor stepping again when it should be stopped. WTF?
As a final test I physically swapped the motor wires so the port esc was powering the starboard motor and vice-versa, just in case it was something about the motor that was causing it.  The stepping remained on the port esc (making the stbd motor move).

All of this suggests to me that the stepping is some subtle interaction between the rx and mixer/esc, and that the twitching was perhaps also caused by it - maybe some sort of feedback loop - but that is beyond my knowledge of these things.

I also did another test by powering off the escs, so I was running the rudder alone as I did a couple of weeks ago. Seems fairly OK - just a tiny bit of jitter, but nothing like the twitching I was getting earlier. I am wondering now if that is what I had two weeks ago - memory can be a terrible liar.

I bounced all this off Dave at ACTion, but the only other suggestion he had was to try a different means of powering the rx module in case it was a volts stability issue. I did that but with no change in symptoms.

I then put everything back into mixer mode, with the rudder servo on the mixer/esc, and the serious twitching returned as before.

Having played on the bench with the mixer control I must confess to being not very convinced - not that the principle of mixing isn't OK, but I just don't like the lack of control over each 'engine'. So using tank mode instead seems attractive and also works around this twitch problem.
To get around the stepping problem I've taken the pragmatic option and reversed the P&S plugs in the rx to get rid of the stepping and then swapped the in and out connections of the esc modules to keep the external fuses and LEDs matching the port and starboard drives. As the miser/esc unit is mounted on it's side anyway the 'wrong way roundness' of these connections is not obvious.
It all seems to work OK now and the final job this evening was to mash up a centreing arm and spring for the 4th tx direction, so both sticks self-centre vertical and horizontal. Oh, I had to reverse one of the channels on the tx as well.

I also retested the tx fail with this setup and now only the port motor goes to full chat; the stbd stops. Not ideal, but only half as bad as both going flat out :)

[scoop]

Just had another thought, if you have a mobile phone is it anywhere near your radio gear ie; in your pocket ?

If so try switching it off then switch on your Tx and Rx

Cheers
Scoop

[dodgy geezer]

Have you used the T4U with other servos and found it to be solid?
I only have the one servo. I got a decent spec Hitec unit, rather than a 'bargain' unit, in the hope that it would be reliable.

Having said all that I have made progress, though it is a very strange tale, to me anyway.

I decided to put the mixer/esc into 'tank' mode, which effectively changes it to a pair of unmixed escs, so I could better see what was going on.  This meant using another r/c channel and also removing the the rudder servo lead from the mixer/esc unit and putting it direct into the rx. Much to my surprise all the twitching stopped!

I can't tell from your comments whether stopping the twitching was associated with removing the servo or not.

Any component in an RC link (including the connectors!) can fail in such a way as to inject interference. You should try assembling your system with different items in order to isolate the one that is at fault. Most people have several alternatives to each component (Servos nowadays can be less than £2 each!). If you do not have alternative components, I am sure someone from a local club will be only too happy to lend you some...

   

[dodgy geezer]

To find out if this was being caused by the rx or the esc I swapped the port and stbd r/c leads in the rx module.  Everything then behaved perfectly - no twitching, no stepping of either motor (just the motors were now the wrong way round).  Swapping the rx plugs back again results in the port motor stepping again when it should be stopped. WTF?
As a final test I physically swapped the motor wires so the port esc was powering the starboard motor and vice-versa, just in case it was something about the motor that was causing it.  The stepping remained on the port esc (making the stbd motor move).

That sort of thing makes me suspect connectors. You can get situations where one socket won't work well with a particular plug. And, of course, you can have breaks inside the insulation of a wire....

[malcolmfrary]

It appears that the twitching is still present and is affecting all the channels. With the throttle at 0 (all motors are stopped) not only does the rudder twitch as before, but the motors and bow thruster do as well. The motors sometimes twitch in opposite directions (the LEDs on the ACTion fuse boards tell this) which could/would be the steering channel via the mixer, but also sometimes in the same direction which must be the throttle channel. The bow thruster LED also shows twitching.

Looking back to the OP, this says power to the analogue bits, or the bits where digital is converted to analogue or vice-versa.  The highest power level is at the RX end of things, the RX could be coughing spurious signals out to whatever is plugged into it, whatever is plugged into it does rely implicitly on stable voltage, not only the pos appearing on the red wire, but also on a solid ground connection via the black, and as has been noted in the past, if an ESC is drawing its motor neg voltage via the control wire, strange things happen, not only to that ESC, but to everything else in the system.
The TX also converts analogue voltages to digital, but at a much lower and less variable power level.  The fact that the TX shows signs of mending also leaves the possibility that it was not as comprehensively mended as possible.  It might have been returned with a reported fault, had that sorted, tested OK and been sent on its way with an undiscovered problem, or had another created by the application of technicians thumb when fixing the lid back on.  The ticking of the port channel indicates that centre off is not quite as 'off' as it might be or again, wandering voltage.  The stick just moves a wiper on a pot - exactly how perfect the alignment of this might be on assembly is anybody's guess.

[NFMike]

... if you have a mobile phone is it anywhere near your radio gear ie; in your pocket ?

No. And it would only interfere for a few seconds every so often anyway, unless I was on a call ... but I'd not be fiddling with the boat then. :)

[NFMike]

I can't tell from your comments whether stopping the twitching was associated with removing the servo or not.

... Most people have several alternatives to each component ...

Stopping the twitching seemed to be associated with taking the servo connector out of the mixer/esc and putting it back on the rx - but I changed the mixer/esc operation at the same time. Shouldn't change two things when fault finding, but it wasn't intended as part of the tests, just a diagnostic aid - the 'cure' was unexpected.

As far as r/c is concerned I'm in the 'most beginners' category - I've got one of each bit.

[Netleyned]

The T4U tx has been known to have some issues of dodgy soldering on the wires to the stick pots so have a look inside and do what I did and resolder all the pot leads and also the leads on the battery holder terminals.

Ned

[NFMike]

As said above the major twitching seems to have been resolved and I have done a workaround for the stepping problem. Regarding the latter I did another quick test this morning (it only involved swapping plugs on the rx).

The 'correct' set-up, if there is such a thing, for tank mode with this equipment is rudder servo in rx ch1, mixer/esc connection S in ch2 and T in ch3. This gives the stepping problem on the port esc.
Swapping S to ch3 and T to ch2 cures the stepping problem.
Changing them again so S is on ch3 and T is on ch4 causes the stepping problem again.
Swapping S to ch4 and T to ch3 cures the stepping problem again.
So this seems to me to perhaps be a timing thing - I'd guess the rx 'serves' ch1 then 2 .. 3 .. etc, in sequence. And something in the mixer circuits in the esc unit is somehow sensitive to this, so that swapping the sequence of the inputs gives a different result.

All very strange.

Still, all I have now is a tiny bit of jitter or dither on the rudder servo when all is still, but that may even be normal. I need to see what people at the lake have.
I do still get a bit of twitching at some motor speeds, but that is clearly interference. I saw a comment elsewhere that these motors (MFA) are very dirty, so I maybe need to do a bit more suppression.
I've had a look without success - does anyone make screened servo leads?

[NFMike]

The T4U tx has been known to have some issues of dodgy soldering on the wires to the stick pots so have a look inside and do what I did and resolder all the pot leads and also the leads on the battery holder terminals.

Ah, I don't think I did those - I'll have a go later.

[malcolmfrary]

I've had a look without success - does anyone make screened servo leads?

Someone , somewhere, maybe, but I've never come across any.  A brave, or inspired, or afflicted man might try making them up using screened stereo P/U lead, but the benefits are doubtful.  You could always look for ferrite beads with a hole big enough to thread the plug through, or smaller ones which would involve removing the plug, doing the threading, and refitting the plug with the pins in the same order.  Finding the real root cause might be simpler.
I do have a preference for setting things up as basic as possible to at least make sure that the parts work as individuals - if I then need to add logical complications I can be fairly confident that I am not trying to solve two or more problems at the same time.  Most fault finding logic presumes that the thing has worked, and there is just the one problem.  Having more than one problem at the same time makes solving them complicated.

[dodgy geezer]

I do still get a bit of twitching at some motor speeds, but that is clearly interference. I saw a comment elsewhere that these motors (MFA) are very dirty, so I maybe need to do a bit more suppression.
I've had a look without success - does anyone make screened servo leads?

If the problem was a dirty motor I don't think screened servo leads would help - they might ward off RF interference, but a dirty motor would probably be putting interference straight down the lines...

I still suspect plug/connector issues, and think that using a different servo would give us some more data..

[NFMike]

Yes, the lack of any screened leads from Googling made me think there wouldn't be any. Screening is a bit of a black art, so it was just a thought.
Funnily enough I was thinking about ferrite beads, but to put on the motor wires. I may still give that a go.

I found an interesting thing while Googling around - a presentation about interference for Ford motors (wipers, etc). Two of the things to avoid were slots in the casing (MFA - tick) and plastic end caps (MFA - tick). Probably a couple of reasons the MFA are 'dirty'. It's unfortunate the rudder servo lead has to pass the motors - and with the low deck in that area it's tricky to get much physical separation.

I've just ordered a dirt cheap servo off ebay for comparison. It's only a 9g job so probably won't drive the rudders - it's just to see what happens with a different servo.

[dodgy geezer]

Yes, the lack of any screened leads from Googling made me think there wouldn't be any. Screening is a bit of a black art, so it was just a thought.
Funnily enough I was thinking about ferrite beads, but to put on the motor wires. I may still give that a go.

I found an interesting thing while Googling around - a presentation about interference for Ford motors (wipers, etc). Two of the things to avoid were slots in the casing (MFA - tick) and plastic end caps (MFA - tick). Probably a couple of reasons the MFA are 'dirty'. It's unfortunate the rudder servo lead has to pass the motors - and with the low deck in that area it's tricky to get much physical separation.

I've just ordered a dirt cheap servo off ebay for comparison. It's only a 9g job so probably won't drive the rudders - it's just to see what happens with a different servo.

There are two ways that a motor may cause interference. Radio Frequency Interference (RFI) direct from sparks from the commutator and pulses direct down the motor power wires. Your Googling has probably found items on RFI, if it's talking about motor shielding.

RFI can be minimised by decent brush maintenance, using capacitor suppressors, and shielding the motor. The typical item to be impacted by RFI is the receiver. But 2.4Ghz does not respond to motor RFI anyway. I don't think there'll be much induced interference in the servo leads from RFI.

Powerline transmitted interference can be minimised by decent brush maintenance, using capacitor suppressors, and putting inductors on power lines. The typical item to be hit by powerline interference is the ESC. If anything, I suspect that this might be your problem. I have had this issue experimenting with cheap ESCs and Taycols.

One of my worst Taycols, with heavily sparking worn brushes, would make 27Mhz impossible to use in its vicinity, even with capacitor suppression. But a Radiolink T4U 2.4Ghz system ran perfectly right next to it. Once I connected an ESC in, however, the powerline interference again made things inoperable until I put inductors on the power lines...

[NFMike]

Yes, this makes some sense.
Did you use wound inductors or just 'beads' on the powerlines? I was on eBay till silly o'clock last night looking at offerings. The catch with wound ones is the current needed to pass for these motors, but they are likely to be better suppressors.

I also have longer leads between ESC and motor than planned originally and that won't be helping. I'll have to see if I can improve that as well.

[essex2visuvesi]

Yes, this makes some sense.
Did you use wound inductors or just 'beads' on the powerlines? I was on eBay till silly o'clock last night looking at offerings. The catch with wound ones is the current needed to pass for these motors, but they are likely to be better suppressors.

I also have longer leads between ESC and motor than planned originally and that won't be helping. I'll have to see if I can improve that as well.

Not sure if it helps but I had issues with my RC heli when I extended the servo lead from the ESC... I ended up using a ferrite ring on the servo lead and its bee trouble free ever since

Does the ESC have a built in BEC as this could be part of the problem

[dodgy geezer]

Yes, this makes some sense.

Oh? Bother, I'll have to try harder....   

Did you use wound inductors or just 'beads' on the powerlines? I was on eBay till silly o'clock last night looking at offerings. The catch with wound ones is the current needed to pass for these motors, but they are likely to be better suppressors.

I also have longer leads between ESC and motor than planned originally and that won't be helping. I'll have to see if I can improve that as well.

I used ferrite rings - about 1/2" diameter. Find a dumped TV or computer power supply - they are full of these components, ready wound...

[essex2visuvesi]

I got mine from a dead PC power supply, unwound the copper wire that was around it and then wound the servo lead round it like this