S800 alternative props

[DennyR]

Interesting, I just read on RCG that you have the old GPS, and that might explain why you have not had FOD.

With 5.12 I had two models shut down and drop from the sky without the FOD both had the early GPS. First flight after upgrading from 5.08. One had 250 flights on it without any problems. This was not firmware it was a software problem....

Interestingly the Naza GPS looks like the bad WKM GPS I wonder if that is also problematic.

 

[m4rc0nd35]

I wonder why DJI chose not to include data logging into the systems design? Sure, data logging capacity would cost some processor capacity and take some programming effort, but the engineers at DJI seems to be skilled people so they could do it.

To me, choosing not to provide data logging may seem a little arrogant; like implying "no need for it - we only release error free software" and "flight tuning our system is SO easy, thus flight data analysis is never required".

I am not saying this is the DJI stand (I was just thinking loud), but then again they never commented on the absent logging feature... Maybe in next generation MC? :encouragement:

On a side note, I look forward to the day our MR systems come with fault tolerant software (at the extent possible in this application).
/ Tomas

DJI can make a module with external SD card to record the data.

 

[DennyR]

This has been very costly for DJI so I hope they learn and make a data logger and a simulator that is not a part of Ace one or some other expensive kit.

 

[BorisS]

Skyped with Sidney yesterday conceding the S800 and the GPS developments.

He gave me some additional tips for the S800. He also had the white flashes when having the s800 in idle and he says its due to the RX mainly. Best position under the Center plate were the carbon shields the IMU and MC from the RX. In case of the futaba one whisker straight the there one down to get a 90 angle between them.

More interesting is his point to use conductive grease on the contacts of the arms. These contacts in the frame that meet with the contact of the arm. have little springs behind them and life time is increased if those are greased. Important not to over do it with the grease just a bit ! Just gotta find a product of high quality for that.

Boris

(edit) interesting ( or at least for me ) was also the story how they test the S800. The actually hook it up to a regulated power supply and fly it like that. 1 hour in the air take it down check things and take it up again. This procedure than for 70 hours in one go. S800 with the new GPS issue flipped randomly ( after 20 hours or 30 hours etc) they couldn't determine when or why what actually triggered the FW bug to cause the flip !

 

[jes1111]

More interesting is his point to use conductive grease on the contacts of the arms. These contacts in the frame that meet with the contact of the arm. have little springs behind them and life time is increased if those are greased. Important not to over do it with the grease just a bit ! Just gotta find a product of high quality for that.

It would be a lot safer if he specified a suitable lubricant - some lubricants will attack certain plastics, so you really need to get the right one! Also, if you lubricate a switch contact (which this is effectively) then you should definitely NEVER connect/disconnect the arms with power applied - arcing will break down some lubricants into a non-conductive (and highly adhesive) film, thus actually making the contact worse. In general, lubricating conductive contact surfaces (especially when they carry high currents) is a "non-trivial" subject!

An alternative that might be a better idea (assuming the contacts are not gold-plated) - regular cleaning of the contacting surfaces with a fibreglass brush followed by thorough removal of all residue with isopropyl alcohol.

 

[BorisS]

It would be a lot safer if he specified a suitable lubricant - some lubricants will attack certain plastics, so you really need to get the right one! Also, if you lubricate a switch contact (which this is effectively) then you should definitely NEVER connect/disconnect the arms with power applied - arcing will break down some lubricants into a non-conductive (and highly adhesive) film, thus actually making the contact worse. In general, lubricating conductive contact surfaces (especially when they carry high currents) is a "non-trivial" subject!

An alternative that might be a better idea (assuming the contacts are not gold-plated) - regular cleaning of the contacting surfaces with a fibreglass brush followed by thorough removal of all residue with isopropyl alcohol.

Disconnecting/connecting while power applied i would never do, never treated the arms as hot swappable and hope noone understood them in this way. Cleaning is a way no question although its a bit..ch getting to the contacts, but the springs no way . It would prevent dirt to get behind the contacts and to affect the springs behind them eventually ! It will try the lub thing if i find a product that i feel comfortable with !

Boris

(edit) hope people understand this correctly there is no issue with the contacts, just thoughts on how to improve and extend the life time of the setup !

 

[DennyR]

Disconnecting/connecting while power applied i would never do, never treated the arms as hot swappable and hope noone understood them in this way. Cleaning is a way no question although its a bit..ch getting to the contacts, but the springs no way . It would prevent dirt to get behind the contacts and to affect the springs behind them eventually ! It will try the lub thing if i find a product that i feel comfortable with !

Boris

Boris
I would just clean them with a small amount of alcohol. Products like Contact 400 for example must not be used anywhere near a PC board The highly conductive film will short everything including your processors. It is very difficult to remove once it soaks in. I would be more worried about the video downlink than the Rx. I have my Spectrum stuff mounted high on the top plate and the 1.3 Tx is under with no line of sight to my receivers also up there is the telemetry Tx.

I have no white flashes, ever... As you can see this is a well used model that has undergone a lot of changes.

View attachment 7150View attachment 7151

 

[jes1111]

Disconnecting/connecting while power applied i would never do, never treated the arms as hot swappable and hope noone understood them in this way. Cleaning is a way no question although its a bit..ch getting to the contacts, but the springs no way . It would prevent dirt to get behind the contacts and to affect the springs behind them eventually ! It will try the lub thing if i find a product that i feel comfortable with !

Boris

Here's a quote from this informative page:

HOW CONTACT LUBRICANTS WORK
Lubricants in large quantities act as insulators. So it's not surprising that some people mistakenly think that lubricants used on switch contacts need to be electrically conductive. Curiously, there is virtually no difference in contact resistance between lubricated and unlubricated contacts.

To work on switches, lubricants must not interfere with metal-to-metal contact. A switch's contacts may appear smooth, but under a microscope, they would resemble a landscape of tiny peaks (asperities) and valleys. Current flows only where the peaks touch. So, the actual contact area is considerably smaller than the apparent contact area, sometimes less than 1% of the apparent area. The normal force on the contact is distributed across these asperities, so the pressure is high, typically hundreds of pounds per square inch. This pressure easily forces lubricants out of the contact zone and lets the contacts make metal-to-metal connections.

I'd wait for DJI to verify they've conducted (at least) one of their 70-hour tests on this idea before proceeding

 

[BorisS]

thanks guys, points taken

I just hope the GPS thing really resolves it. Just skeptical since since the number of flipped S800 out runs the number of new GPS none s800 setup that flipped ( did any ? ) by far ! Really hope they got it right !

 

[BorisS]

A friend asked me to film a similar event on the 29th oct. Really don't want to be embarrassed if the thing flips or hurts someone.

Boris

 

[DennyR]

I think that even when you have the right GPS unit. There is no point in taking public risks unless you are serious about doing this to earn money. A lot of testing first and a ops. routine that is as near risk free as you can get it. I think I did about a hundred flights with different set-ups before going public with the s800. I do hope that DJI will pour the money they made selling the many S800's into making reliability the number one objective. There is still some way to go. Too many variables for the less experienced to get wrong. We see far too many bells and whistles that are not needed, Basics right first. What is the point of having autonomous waypoint functionality when the dam thing cant even hold a reliable hover 10 feet away. Having made the statement that there is nothing wrong with the S800 airframe would suggest that the vibration levels are of insignificant importance. I have to disagree.

We are all on a collision course. where accident reports will meet with greater legislation and only the few will be allowed to operate outside of model club flying sites. So far I think we have all been very lucky that a major incident has not happened. Not only DJI have put us at risk. The market is awash with bad stuff that is sold for the hobby market that gets used at the pro end.

 

[DennyR]

East is least and west is best. OK you don't have to understand (1/2 ro V2) to get in the air but if you want to navigate from where you are and come back again. then you do need to know that the Magnetic North is not actually at the North Pole. Perhaps when DJI carry out their improvements they will ADD the offset adjustment to the Assistanant/Autopilot. Then we don't have to physically tweak the compass unit in the general direction. In fact the airframes own magnetic deviation could be programmed in as well if you can find an area free from magnetic disturbance to do the swing.....

If we fly near the limit of visual range then I think it is a good idea if the model tracks true. FWIW...

 

[ChrisViperM]

Has actually someone replaced the WooKong M on a S800 and replaced it with something like a Hooverfly Pro, ZeroUAV YS-X6 or similar....? Would be interesting to see if this combination works. At least the problems could then be related to software (WooKong) or hardware (S800). Zenmuse could still be used with a WooKong IMU.


Chris

 

[DennyR]

As we try to make the S800 even better with different props, the improved stiffness of the arms is also well worth doing. The cheaper solution is to fill the webbing with something easier and cheaper than my original solution. A typical fiberglass repair kit such as Davids P 40 can do the job quite well and is easy to sand an refinish. If you happen to find carbon nano tube fiber to mix in as well then thats even better.

Chris
If VBrain was not so unfriendly to set-up I would be trying that next. Fastest board out there.:tennis:

http://api.ning.com/files/uWrWqV*Fz...FoZm25FCaTfrf5o880dj-/IMG_20120719_171605.jpg

http://www.virtualrobotix.com/profiles/blogs/vbrain-coming

 

[DennyR]

http://www.virtualrobotix.com/profi...-update-on-vbrain-project-benchmark-of-arduco

This is quite interesting

 

[ChrisViperM]

Chris
If VBrain was not so unfriendly to set-up I would be trying that next. Fastest board out there.:tennis:

http://api.ning.com/files/uWrWqV*Fz...FoZm25FCaTfrf5o880dj-/IMG_20120719_171605.jpg

http://www.virtualrobotix.com/profiles/blogs/vbrain-coming

Looks interesting....never heard about it before. Thanks for posting....


Chris

 

[DennyR]

A bit like APM on steriods....Nah, far too much brain wanking.

 

[R_Lefebvre]

East is least and west is best. OK you don't have to understand (1/2 ro V2) to get in the air but if you want to navigate from where you are and come back again. then you do need to know that the Magnetic North is not actually at the North Pole. Perhaps when DJI carry out their improvements they will ADD the offset adjustment to the Assistanant/Autopilot. Then we don't have to physically tweak the compass unit in the general direction. In fact the airframes own magnetic deviation could be programmed in as well if you can find an area free from magnetic disturbance to do the swing.....

If we fly near the limit of visual range then I think it is a good idea if the model tracks true. FWIW...

They don't do that already?

It's pretty easy to set up a system in the program where it looks at your current GPS location, computes the correct declination for that location (there are look-up tables available), and then goes with that. <cough>

http://www.virtualrobotix.com/profil...mark-of-arduco

This is quite interesting

Yes, it is. But all he's doing is running Arducopter with the CPU idling 95% of the time. What's the point? Just run an APM.

I am very interested in his boards too, but at this point, it's not "plug and play" and I'm out of time. So I stick with what's "easy".

 

[DennyR]

Rob

No declination setting. They suggest that you turn the unit. No link to the look-up tables. Also very slow to acquire.

How reliable is AMP for helis BTW. Looks like MR version is now working good.

 

[R_Lefebvre]

Rob

No declination setting. They suggest that you turn the unit. No link to the look-up tables. Also very slow to acquire.

How reliable is AMP for helis BTW. Looks like MR version is now working good.

That's really surprising that you can't at least set it. That's like a single byte in a parameter. So you have to clock the GPS unit? That's silly. Slow to acquire GPS lock? How long? We're just a few seconds now with the battery backup GPS. DJI is really going to have to step up their game.

Yes, the MR version is working quite well. 2.8 is pretty awesome. While I was beta testing, I blew the blades off my quad because I was having too much fun. Never did that before. Quad landed on pavement upsidedown, smashed all the electronics. So I'm down for a while.

The SRH stuff lags behind still. Mostly a case of the simple fact they are more expensive and fragile, so less people willing to try out some open source thing. It also slows me down too, it's expensive to have even a little mishap. Also, SRH is like the redheaded stepchild. The code is mostly designed for MR's, and I basically have to comb through the whole thing and find what works or doesn't for SRH's.

But anyway, it's working fine now generally. I have about 20 flights with no vibration induced problems. The key thing is the bearings. Gotta have good bearings. Soon as they start to go, you're in trouble. And torque tubes can also be a problem, one solution is to use two bearings in the tube. So the heli hovers great, super easy, even in wind. And it's very easy to fly in stabilize mode. I've had it up to about 110 km/h trying to see what it'll do (that's about it at a low-ish headspeed). If you haven't seen this, you should, especially the on-board stuff which is in the second half, using a hard-mounted camera.

I'm now working on waypoint mission flying. I've got it up to 10 m/s (36 km/h) and that's going OK. The problem I'm running into is that at some point it starts flying less like a MR, and more like an airplane. But I'm slowly clearing away problems.

I'm also trying to get rate/acro working better.

Anyway, I honestly wouldn't recommend anybody go ahead and stick one on a big expensive heli just yet. But if you've got a "squadron hack" and want to have some fun, the more the merrier.

Oh, here's what happens when you try to run missions too fast, hit 70km/h here, and the vertical component of the airspeed is too great.