GPS PLUG CONNECTION FIX

[flyNfrank]

I can confirm that v3 new ones have same crappy gps connector - 1 of 2 we just got 'fell' no power from few hundred feet & smashed lots of stuff. so I got to see connector.

my question is why aren't you just replacing it with the proper keyed $ 1.50 one rather than sticking or gluing stuff in it?

http://www.mouser.com/ProductDetail/TE- ... AcshhfI%3d

thank you!

For the record, the Mouser part linked above does include the correct plastic body. However the crimp pins are not user friendly without proper tooling. Tyco, the connector mfg, does make a better pin for us to crimp. Just haven't made the time to detail and test for performance. Your existing GPS terminal pins WILL NOT retain properly into this Tyco tab'd plug.

From the limited testing with the Tyco plug and terminals. The quality is much better by a factor of 10 just based on visual comparison. The tension is noticeably better than the DJI original, but it's not a fair statement since I have pulled the DJI plug more than 100 times. The tested mechanical cycles for the Tyco is about 100 per hour, still doesn't tell me life expectancy based on our application. For what the plugs are originally designed for, the removal/insertion count is expected to be low. I was going to order three varieties of pins, low and high contact and then a high pressure contact.

Just don't want everyone ordering a part and be pissed that it doesn't work... Give me a couple of weeks to detail or help yourself to the same.

I know I sure as heck don't have a problem is telling you RichWest, I appreciate your time and effort you put into this. Obviously my little temporary fix is just that, a temp fix which seems to keep from thrashing out the quad. Thanks for what you do!

 

[Navman]

Hey Frank
Found my P2 had a Very loose GPS connector and did the Hot Glue/zip tie end mod on it and glue on the flytrex connector that was a very good fit, but since I was in there anyways..

Now to open it back up and look for hot spots and yellow on capactors... :geek:

Good info Frank!

 

[flyNfrank]

Hey Frank
Found my P2 had a Very loose GPS connector and did the Hot Glue/zip tie end mod on it and glue on the flytrex connector that was a very good fit, but since I was in there anyways..

Now to open it back up and look for hot spots and yellow on capactors... :geek:

Good info Frank!

The vPlus and V3 both have movement of the plug in the connector on the board. That movement moves in the direction of front to back of the quad. Each time the cover is removed or plug is removed, causes the internally parts of the plug to widen causing it to become loose enough to allow a GPS connection issue.

This particular problem could be the single most related issue to fly-aways. It's great to see those that are wise enough to take 3 to 5 minutes to do the mod. Since the time I discovered this problem and announced it, I have never been approached by anyone saying they did the mod and reported a fly-away or dropping from the sky.

 

[RichWest]

Really the issue is the contact pressure from the individual connectors inside the plastic plug's body. By loading the plug back, your just making a different contact reference on these connectors. Which is a good fix. The connectors have two contact fingers and it appears they are prematurely fatiguing from a variety of things - vibration and/or the act of inserting and removing from the board. It could be as simple as corrosion, the contacts appear only as copper without plating. Either way the connectors are of poor quality. No identifying markings tells me this is an off-off brand.

As an example, pull one of the servo-patch cables going to the TX connection in the bird. These have greater tension when removing and they have only three connectors resisting the force. Two different class of products are being used.

I will be ordering connectors this weekend. Have narrowed the group further after my conversation with Tyco this AM...

 

[mad in nc]

I will be ordering connectors this weekend. Have narrowed the group further after my conversation with Tyco this AM...

If you find the right connector all would be good then to what should have been from the factory.

Pls post the part #, we all will have to buy 100 or 1000 pc's probably which is cheap insurance for the long run to your corrosion/ pressure quick fix I have in place and I'll be better knowing this is the right form/ fit solution...

one more check off on the DJI quality index.

 

[copyhere]

hello,
im new here and have this problem on a brand new phantom 2 vision+ version 3 mine was so bad that after three flights I couldn't get any gps at all upon startup I only would get slow flashing yellow as if it was disconnected completely. DJI wanted me to send it back to them for warranty repair it was only used 3 times. so before i sent it in I found your thread about this and started working on it myself. I found that wiggling the connecter around by hand with the phantom powered on would start to connect if i held it tight in the right spot it would acquire satellites. I was happy, I proceeded to do the tie wrap mod. but it did not put enough pressure to make a good connection. Next I tried removing the female pins out of the connector and tightening the 2 contact points in each one and reinstalling the result was still no good. next i looked at the solder connections on the board that the male portion of the connector is attached to which is visible through the battery compartment. I re-soldered the 4 of them still no good. so I had an Idea I sprayed some deoxit contact cleaner on it worked it on and off a few times and problem solved a solid connection even when I wiggled it around by hand there is no intermittent loss of connection. I reassembled it and test flew it a few times and works perfect with out the tie wrap mod and it is very loose. In conclusion I believe in my case the problem was corrosion on the male pins of the connector which was not visible by eye causing my problem so for added protection I would highly recommend any one working on this to give a shot of deoxit or any type of contact cleaner.
I also want to change the connector with the one that locks when we know the correct part # .

 

[Wibble]

Thanks for that story. It annoys me so much that they spoil an awesome machine with silly things like this!

What people need to remember is that the poor Phantom starts life in a very harsh environment - a few months at sea in a shipping container! Salt laden air will corrode bare copper pretty quick!

 

[vk3bq]

they should be using somehitng like this

http://www.molex.com/molex/products/fam ... connectors

but molex connectors are expensive.. and the unit is built to a cost.

but it will rely on the NAZA side of the system to change the plug, rather than the socket on the board (i gather the GPS is not customer DJI designed * don't have a phantom yet to check myself) i assume the gps daughter board is provided form outside DJI?

my question is, has anyone who has performed the magic plug fix had a fly away or other issues? post this "fix"?

the next question is, looking at pics and info I can find (*again don't have a unit to check directly) i would gather the flight controller that is taking the GPS feed from the GPS unit wouldn't be so poorly written as to LOOSE ITSELF if the NMEA GPS string stopped refreshing? typically 1 refresh a second (can be faster, but not that much faster. the software should remember its LAST known location, it would have a period of waiting for the next string before it GAVE up and changed modes? the fact its 4 bits of wire on a header suggests 4800bps or 9600bps serial NMEA string data. the sampling will be "generally slow" and the logic of the programming should have a hold/maintain/wait as such, your in car GPS doesn't suddenly JUMP 15000KM in location when it looses sats, it might after a while complain that sats are lost, but most modern system will continue to display last location and some even try and predict the path based on last received samples, especially when you enter a road tunnel.. but who knows, i don't have a quad to test and explore.

so dji, play with the GPS programming and add some smoothing and averaging and a hold/recall function if the nmea feed is critical? if this is the issue.

and hello forum, visiting to get some background before I invest in a unit!. most likely the phantom2 with H3-3d (or) h4-3d gimbal.. if I buy, I'm sure you will see more of me!

Andrew
VK3BQ.COM <-- ham radio operator / electronic engineer.

 

[RichWest]

Welcome the forum! Your input is greatly appreciated.

The molex connector series you suggested seems correct for the application, vibration. Actually better than the material I was trying to source and I found the cost to be about the same. This would be a better original solution, but sure cost could be a factor for DJI or that it's made in Japan. Who knows. The only drawback within the present needs is the pitch for the pins, 2mm versus 2.54mm on the existing connectors. At this point, at least for me, my ordered material will be my solution.

To answer your other question, "anyone who has performed the magic plug fix had a fly away or other issues? post this "fix"?
". Well at the moment I am having full success both in regards to GPS (12) and full FPV resolution. No fly away issues. Actually my confidence level is increasing with every flight, never imagined a simple fix could resolve multiple issues. At least for me...I hope I didn't jinx my success.

copyhere's suggestion to clean the contacts is another great idea. Added contact cleaner to my order....

 

[mad in nc]

when i had the V3 open, I tried the "wedge" fix and even outside I'm still not able to get above 6 sats... Even when I flew through three batteries 7 before opening the cover was the best I could see then

Have ordered the Deoxit Gold and will try that before pulling the pins. Its (GPS connector) been on and off a total of less than 20 times and hoping the cleaner will solve the issue. Per the GNSS iPhone programs I should be seeing ~ 9 sats but when in my office I tap on the shell, I get between 3-5 sats.... so I'm hoping its the connector. Goal is to get the V3 all tuned up, new motor wires, GPS connector situated before the weather turns good here...

 

[vk3bq]

Guys (and girls). Understand that this PLUG can never change the "amount" of GPS sats the quad sees. The antenna and "gps" smarts are done in the GPS pcb.



The plug feeds a "I assume" TTL/serial data feed to the controller system.

It's how this feed is fed into the controller. And what happens when this feed is lost. (On the controller).

We need someone with an electronics background to measure the GPS on a CRO. And. See if it's TTL serial. And then maybe connect it to a serial port and "read the GPS output.

Wiggling this connector might improve the "data feed quality" but won't make magic sat improvements.

 

[robinb]

However if the connection is poor electrically the data can get scrambled and all sorts of nasty things could happen.

Depends on how much error checking if any is done on the data stream.

 

[N017RW]

Guys (and girls). Understand that this PLUG can never change the "amount" of GPS sats the quad sees. The antenna and "gps" smarts are done in the GPS pcb.



The plug feeds a "I assume" TTL/serial data feed to the controller system.

It's how this feed is fed into the controller. And what happens when this feed is lost. (On the controller).

We need someone with an electronics background to measure the GPS on a CRO. And. See if it's TTL serial. And then maybe connect it to a serial port and "read the GPS output.

Wiggling this connector might improve the "data feed quality" but won't make magic sat improvements.

Here you go:
http://www.rcgroups.com/forums/showthread.php?t=2071772

 

[mad in nc]

Guys (and girls). Understand that this PLUG can never change the "amount" of GPS sats the quad sees. The antenna and "gps" smarts are done in the GPS pcb.



The plug feeds a "I assume" TTL/serial data feed to the controller system.

It's how this feed is fed into the controller. And what happens when this feed is lost. (On the controller).

We need someone with an electronics background to measure the GPS on a CRO. And. See if it's TTL serial. And then maybe connect it to a serial port and "read the GPS output.

Wiggling this connector might improve the "data feed quality" but won't make magic sat improvements.

I hear you but in my case by even "tapping the quad GPS sats come and go... thus the connector integrity is first on the list to resolve,.............

 

[RichWest]

Quick update on the connectors. I did receive my first set(s) in today and expect another shipment on Thursday from another source. I won't be able to work on the material until this weekend. Once I've tested, I'll post part numbers and additional options for those interested in going a step further for the DIY fix.

I can confirm that the DJI connector pins are not plated, just copper on my example. Which is not good if you have any concerns about corrosion. If your going to clean with a Contact Cleaner, make sure you clean both the pins on the board and terminated wire pins (female). Possibly use a pipe cleaner with CC to get to the contact fingers. No need to get crazy, you're just trying to work the surfaces with the cleaner.

I hope this effort will resolve everyone's lack of GPS count. I would be even better if this reduces or eliminates the "fly away" issues many have had. I doubt that will be the case, but it's one less thing eliminated for a possible cause.

 

[flyNfrank]

Guys (and girls). Understand that this PLUG can never change the "amount" of GPS sats the quad sees. The antenna and "gps" smarts are done in the GPS pcb.



The plug feeds a "I assume" TTL/serial data feed to the controller system.

It's how this feed is fed into the controller. And what happens when this feed is lost. (On the controller).

We need someone with an electronics background to measure the GPS on a CRO. And. See if it's TTL serial. And then maybe connect it to a serial port and "read the GPS output.

Wiggling this connector might improve the "data feed quality" but won't make magic sat improvements.

I agree....to disagree with you totally on that one. Yes I know about right now you are feeling a fire burn wondering if you are going to get more mad as you read, and I'm afraid you will. Technically speaking the plug connects creating a link. Anytime the performance of that plug fails it will effect the link between the GPS antenna and communicating component. When it fails, it either breaks connection or it changes the amount of resistance going to the component. Basically the sensor in which reads the resistance makes it changes based on the amount of that resistance. Regardless which of the two events take place, there will be a bad repercussion from it. Connection break will result in the quad dropping. A resistance issue will result in a fly-away. There is a good chance you can save the quad if it is in the process of dropping or flying away. To save it,.......Hang on a second and let me say that just because I'm going to say these details doesn't mean you should slack off on doing the mod. Ok, to save it, grab the S1 switch and switch out of GPS Mode and back into GPS Mode. This worked for me after experiencing a total connection loss and dropping from the sky.

RichWest, how many plugs do they include for what you paid them?

 

[vk3bq]

The force is strong in this one? Resistance is futile?


I'll take my engineering hat off then

So. Does this connector feed the device its power?

Does it have a serial / bus feed ?

What is the resistance you talk about. Is it a rebel base on the planet Hoth?

Have a read of the link to the "reverse engineering the NAZA GPS" posted a couple of lines up and see how this works.

The GPS smarts are on the pcb. This is just a feed of the output of that.

Check it with science. Get a CRO. Look at what's on those 4 boys of wire.

And may the force be with you.

 

[flyNfrank]

Better yet explain how you want all to believe you are in a freshly created account and pretty much each of your post are argumentative in one way or another. Smell that rat? Why not use the account with your usual screen name?

Btw, it doesn't matter which account you decide to post from, not one person in this thread has to prove anything to you. If you can't except the fact it is what it is, then fill free to bounce on. The info in this thread has saved a great deal of grief. And just know it all took place prior to you showing up and putting your dress on and bumping your gums.

 

[vk3bq]

You didn't even know a round cyclinder thing was a capacitor. Yet have an insight into this magic plug issue. With engineering to back it up.

With resistance. I'm just trying to understand what's going on.

Trying to use an electric engineering, scientific method. My background. I asked some questions.

However you seem to know the answer. I'll go back to wriggling my plugs.


Thanks
Andrew. (I used my name my first reply)

Who frankly does not give a dam about your theories. Until we methodically work through them. And from what i can see and read. Are just that. Theories and your opinion. So. Scientific method. Please. Otherwise it's just flyin in the wind......

 

[mad in nc]

vk3bq - thanks for your input. Based on your site and hobbies I respect the work you have done with antennas, baluns etc. in the RF field of Ham Radios.... A great community.

The four wire GPS connector I believe has the TTL logic bus and power feeds to the GPS device (from memory where I saw it before). With what Richwest stated the male/ female plugs being copper from DJI shows the "cheapness" of the connector used and I would never use that plating on this critical telemetry component.

I'll be getting some contact cleaner shortly and try to clean the connector first, check sats, then upgrade the antenna to the 4mm 5db device and I can only hope for short term improvements. If RichWest comes through with the "right" connector and pins and when that is changed in my P2+ that all I believe I can clean up without a major change or overhaul...

I look forward to your contributions. Enjoy the Aussie summer!