Revs' Talon Tri (re)Build Thread

[richardg6paj]

How do you explain David Windestål';s design then? His arms are v e r y  l o n g!!!

I don';t have to explain anything, I said it';s just my opinion. Just can';t see the point in having very long arms.

Richard

[richardg6paj]

Revs, I just noticed one of your pictures you are using a digital servo. I read recently somewhere that someone was having difficulty using a digital servo for a gimbal with the APM and couldn';t get smooth movement, something to do with matching the APM';s output frequency with that of the servo. So changed to a high quality analogue servo and all was good.
Worth trying?

Richard

[flybywire]

How do you explain David Windestål';s design then? His arms are v e r y  l o n g!!!

Er, David is David, the normal laws of physics need not apply   

[Revs]

Revs, I just noticed one of your pictures you are using a digital servo. I read recently somewhere that someone was having difficulty using a digital servo for a gimbal with the APM and couldn';t get smooth movement, something to do with matching the APM';s output frequency with that of the servo. So changed to a high quality analogue servo and all was good.
Worth trying?

Richard

That';s interesting Richard, cheers.

After posting last night I was thinking about this, so I had a read up on the servo refresh rate for APM. It seems there should be an option in the advanced config called MOT_SV_SPEED to set it between 50, 150 or 250hz, but in the version I';m using (3.1.5) it doesn';t seem to be there. As far as I can tell the default is 50hz but this servo I';m using can take up to 250hz. At 50hz the movement is very steppy, not smooth. I bet this is the issue.

I remember with the KK that if the servo rate was set to ';low'; it was lumpy, but set to ';high'; it was nice and smooth.

I think the next step is to either confirm that the option can';t be adjusted. If it can I';ll change it and try again. If not I';ll do what you suggest and try a good MG analogue servo.

Thanks again bud

[guest325]

Could well be the problem, I only use analogue servos with my Tricopters but I use decent quality ones.

[richardg6paj]

MOT_SV_SPEED to set it between 50, 150 or 250hz, but in the version I';m using (3.1.5) it doesn';t seem to be there.

I think the settings are
Servo speed (MOT_SV_SPEED)
Servo update speed in hz
◾Values: 50, 125, 250
But if I were you, go for an Analogue Servo, at least just to prove a point.

Richard

[Revs]

Yeah that';s right, but as I said it';s not there in the advanced config. I';m gonna get 3.2 installed tonight to see if it';s the same with the software.

Any recommendations on a good MG analogue servo?

[richardg6paj]

Yeah that';s right, but as I said it';s not there in the advanced config. I';m gonna get 3.2 installed tonight to see if it';s the same with the software.

Any recommendations on a good MG analogue servo?

I';m trying this on my mini 250 tricopter:-
http://www.ebay.co.uk/itm/121298398309
I don';t know what size you';re after but they do the larger one I';m sure.

Richard

[richardg6paj]

This maybe what you';re suffering from.
https://www.youtube.com/watch?v=GNI7TaK5POc
You could add a little bit of friction to the Yaw Mechanism.

Richard

[Revs]

I';m trying this on my mini 250 tricopter:-
http://www.ebay.co.uk/itm/121298398309
I don';t know what size you';re after but they do the larger one I';m sure.

Richard

Gonna try that one myself. Spec seems fine. That one is out of stock though bud. Try this.. 331138640823

[guest325]

I usually use MG16R';s but have recently been trying 9gm Turnigy MG Analogue (very much the same as the Tower pro';s) -not sure about durability yet.
The other thing is that I';ve only ever ran my Tri';s on KK2 FC';s which as you can appreciate are not the same animal as what you are using - I like them and have no intention (yet) of trying anything else.

[Revs]

I';ve ordered a TowerPro MG90S, Darrell, so we';ll see how that goes.

Also, Ive just installed 3.2, but there';s still no MOT_SV_SPEED option. I';ll join up at DIYdrones and see if a dev can help out. I';d like to see if I can get the Savox servo working well as it';s a very nice servo.

[nbpicklesno2]

Going back a bit. David Windestål'; booms on the earlier versions were 45cm to 50cms long but they gradually got shorter in the later versions. The current V3 has 33cm booms. The earlier versions were (obviously) using earlier software and hardware and were not as sensitive to vibration.

You could try a servo tester to find if 50, 125 or 250hz works best. On the Tower MG90s it';s 50hz.

[guest325]

It';s one of the reasons for my liking kk2 fc';s they will tolerate a reasonable amount of vibration before they start playing up,  what I';ve found without exception is that the higher the resolution of the fc (ie 32 bit as opposed to 8 bit) in my case they are considerably less tolerant of vibration, but that said they do have the obvious advantage of much greater processing power.
I';m soon going to try some of the 32 bit fc';s to see what makes them tick; I have a Acronaze32 mini naze32 and a cc3d to try out, should be interesting to see how they perform compared to a properly set up kk2!

[nbpicklesno2]

I';d love to see something more scientific on FCs as I really am tired of reading that one FC is "more locked in", "more dialed in" or some other nonsense, than any other particular board.

I like to think that I can tell the difference between my kk2 and naze but can I really? Or do I just think one flies differently than the other. Is it just down to PIDs and not necessarily processing power or other differences in hardware/software?

I haven';t got a clue. I like to think I';m clever but really, am I?

Of course I am.

Got ya. You thought I was being humble.

[guest325]

I like to think that I can tell the difference between my kk2 and naze but can I really? Or do I just think one flies differently than the other. Is it just down to PIDs and not necessarily processing power or other differences in hardware/software?

That';s exactly the way I think, to me if a job needs doing the simplest answer is invariably the best, if my Tri / Quad or whatever flies well on a basic fc why bother with something more complicated - more to go wrong,  less proven, etc.
This is one of my hobby horses - GPS functions,  to me used as a failsafe is wrong because they are NOT 100% reliable,  doesn';t matter who makes them or the fc it interfaces with not 100% reliable therefore not a viable failsafe (by definition).
I set my failsafe at 45% throttle so if anything goes t1ts up I turn my tx off and let it descend gently to terra firma! Never fails!

[flybywire]

Thankfully both the CAA & bmfa disagree with you over the gps reliability aspect Darrell.
Of course nothing can be 100% reliable, i';ll settle for 99.9!

[richardg6paj]

GPS functions,  to me used as a failsafe is wrong because they are NOT 100% reliable,  doesn';t matter who makes them or the fc it interfaces with not 100% reliable therefore not a viable failsafe (by definition).
I set my failsafe at 45% throttle so if anything goes t1ts up I turn my tx off and let it descend gently to terra firma! Never fails!

Sorry Darrell, I have to disagree with you on that one. It';s not the system that';s unreliable or inaccurate, I think it';s the way it has been implemented, it is far too complicated for the majority of people, the manuals are poor or even non existent. Confirmation of gps reception is usually done with sequence of flashing lights which are mixed with other warning lights. There should be a nice LCD display to show actual fix and number of satellites available this would make the system more understandable. Hopefully, as things progress they will get better. At the moment if the pilot understands the system, it works incredibly well and works every time. It';s just a matter of educating people.

Richard

[guest325]

Sorry Darrell, I have to disagree with you on that one. It';s not the system that';s unreliable or inaccurate, I think it';s the way it has been implemented, it is far too complicated for the majority of people, the manuals are poor or even non existent. Confirmation of gps reception is usually done with sequence of flashing lights which are mixed with other warning lights. There should be a nice LCD display to show actual fix and number of satellites available this would make the system more understandable. Hopefully, as things progress they will get better. At the moment if the pilot understands the system, it works incredibly well and works every time. It';s just a matter of educating people.
Richard

Maybe you';re right but to me a failsafe should be idiot proof, if it';s not then by definition it can';t fail safe reliably. Don';t get me wrong on this one, I';ve got nothing against use of GPS - I';ve got one and had one before but will not use them as a Fail safe - it';s just my opinion, I really did expect to take flack for having this opinion but when you';re involved in FMEA';s on a regular basis it does help to see what';s what.

[flybywire]

Well, the ';debate'; if you can call it that will go on and on and on of course, but I';m pretty happy with the results of all my radio loss failsafes, all have come home, and I have either let it land itself, or simply cancelled the failsafe and carried on (if the reason for the FS was obvious, i.e dissapearing behind a hill or such land feature or heavy wooded area maybe.
Some folk like Darrell will not trust it, and some folk like me will, simple as that.  I am only interested in the fact that it has now been approved by the relevant authorities, so I am not breaking the law using it, so in this I am satisfied.  Of course, as Richard points out, many a slip twix cup & lip, so if you don';t know how to set it up, or you don';t give it a chance to settle in and get accurate coordinates, then it';s a chocolate teapot.

BTW: I may be guilty of making a few contentious remarks, occasionally, but please believe me when I say that I am never glib when it comes to my model building or setups of said models.  I am extremely careful when flying anything, and I always attempt to comprehend the limitations of what is of course ';hobby grade'; kit, and take the comments other make seriously (though I may not agree with them per-se).
I';ve invested shed loads of cash in my multi';s, BYM the last thing I want is for one of them to wave goodbye and start a new life in a neighboring county!

[guest325]

Some folk like Darrell will not trust it, and some folk like me will, simple as that.  I am only interested in the fact that it has now been approved by the relevant authorities, so I am not breaking the law using it, so in this I am satisfied.  Of course, as Richard points out, many a slip twix cup & lip, so if you don';t know how to set it up, or you don';t give it a chance to settle in and get accurate coordinates, then it';s a chocolate teapot.

You hit it on the head there, it';s a matter of what you are comfortable with - I see RTL as a great feature of the GPS system and have used it quite a lot as a party trick when I first started flying my 450 at our club, you have also hit it on the head as regards setting it up correctly and allowing adequate time for full satellite lock - ignorance of these factors as you say makes it no better than a chocolate tea pot. Safety critical devices should under these circumstances not allow the model to fly - ie unless there is a full satellite lock the model should not be able to take off UNLESS there is an over-ride in which case the pilot must accept the responsibility for doing so. If this was how the system could be made to work then I would have no problem with it being used as a failsafe.

[Revs]

While I';ve been waiting for the servo to turn up I';ve been looking at the logs. Everything looks good 

When hovering and the rear end isn';t vibrating the IMU says it';s well within limits. All three axis are 0.1-0.2g.

Battery consumption in a hover is 8.9A with the MultiStar 4000mA 10c battery (320g), so that works out at 20 min flight time with a 75% battery drain. So the MS battery is working as intended. Very pleased with that. Although that';s without the GoPro, but that';s only light.

The analogue servo is due before the end of the week so hopefully over the weekend I';ll be able to get that in and it';ll sort the yaw out.

[richardg6paj]

ie unless there is a full satellite lock the model should not be able to take off UNLESS there is an over-ride in which case the pilot must accept the responsibility for doing so. If this was how the system could be made to work then I would have no problem with it being used as a failsafe.

Exactly, now you';re talking Darrell. And there is no reason why that shouldn';t happen. But, I still think that the "Home" co-ordinates should at least be able to be checked along side another GPS receiver instead of relying on flashing lights. After all the Home position is crucial and one digit out can make all the difference.

Now apologies to Revs for taking over his Thread! I';ve finished with that particular subject for now.

Revs, did you ever find anymore regarding configuring your digital servo? I guess you didn';t.

Richard

[Revs]

Nothing definitive, no. Not had a reply at DiyDrones. Fingers crossed this new one does the job.

As for this GPS stuff, the way I look at it is; as long as you';re not being silly, like flying in thick cloud or fog, it';s there if you need it. But it';s not good practice to rely on it. I';ve never had an issue, but my mate took my old 350QX out in the fog we had a few days back (against my recommendation) and after a few minutes it decided it was in the wrong place, banked left at 45* and plowed in to a parked up wagon trailer. QX didn';t win and is now waiting for bits.

[flybywire]

Nothing definitive, no. Not had a reply at DiyDrones. Fingers crossed this new one does the job.

As for this GPS stuff, the way I look at it is; as long as you';re not being silly, like flying in thick cloud or fog, it';s there if you need it. But it';s not good practice to rely on it. I';ve never had an issue, but my mate took my old 350QX out in the fog we had a few days back (against my recommendation) and after a few minutes it decided it was in the wrong place, banked left at 45* and plowed in to a parked up wagon trailer. QX didn';t win and is now waiting for bits.

Are you saying this was down to a GPS issue?
I know we all use anecdotes from time to time, but unless we stick to facts, that';s where all the nonsense starts to creep in from the Luddites, and GPS haters.

[Revs]

Are you saying this was down to a GPS issue?
I know we all use anecdotes from time to time, but unless we stick to facts, that';s where all the nonsense starts to creep in from the Luddites, and GPS haters.

It would seem so, but there are no logs to confirm that. I know someone else it happened to on the same day, same area, same weather conditions, with a cheap Chinese quad. It was doing a RTL when it changed direction, traveled about 75m away and landed on a mill roof. Too him 2 days to find it, but he dried it out and it';s still working.

Moral of the story; never use GPS in fog. But it says that in the manual so he can';t blame the copter.

[guest325]

Moral of the story; never use GPS in fog. But it says that in the manual so he can';t blame the copter.

Heavy cloud can give the same problem,  I have experienced it myself - it';s all too easy to disrupt tbings for my liking.

[flybywire]

Hmm, ok, probably best to leave it there then, as i say it is anecdotal to say the least.
I flew into light fog/cloud last week, and had no issues.  I had a geo fence set, which i had forgotten about, which worked fine, but i would never fly in thick fog, so will never know if it would be an issue with my particular gps setups or not.
One thing i';m content for others to find out for themselves!

[richardg6paj]

Fog and mist and general cloud cover doesn';t block gps signals. Water does block gps signals, so possibly large droplets of rain would block or reduce the signal strength. but defo not fog.

Richard

[Revs]

It was particularly think fog here the other day when this happened. I was driving over the Pennines on the same day and at points visibility was down to about 2m and I almost couldn';t see to drive.

Everything I read suggests that fog shouldn';t be an issue, but it';s unlikely that two copters having similar issues within an hour of each other can be a coincidence. Of course my buddies could be BSing me and blaming a dumb-thumb on GPS 

Anyway, that';s enough GPS chat for one thread