Spektrum RX repair and tech info.

[PLMS]

Since there isn't much on RCGroups about Spektrum receiver repair I'd like to jot down some things I've learned recently about them. May as well be here where it can be built on by the forum and found by others.

First thing I've found that probably most Spektrum users know is that the AR7000 receivers won't bind or run without the remote receiver plugged in, but will run if one is removed while the system is already running. A friend was trying to use his main AR7000 RX without the remote, easy 'fix' that one. That info was very useful with this next fault though...

Second thing. Same friend had a faulty AR6000 receiver. It died after a crash. The fault was that it would not light the 'lock' led or bind anymore.
I had a look at it for him and after trying other things I probed the two RX chip outputs on the top PCB with a occiliscope. One chip had an output, one didn't...
The output from one chip looks like about a 100 odd bit long binary 'word' or frame. Moving the TX sticks would result the binary data changing on the scope, so I knew one chip was seeing the TX even if the main micro-controller chip was not happy enough to 'Lock'.
So the second bit of useful repair knowledge gained there, is that on a non-remote receiver like the AR6000, both RF chips must be working for the RX to lock or bind. Just like in the full range receivers...

Third thing. Now knowing that one chip worked, one didn't, and that two are needed for the RX to bind or run, I took a stab in the dark and linked the output of the working one over to the same pin on the one that didn't. Just a jumper, nothing more. In the photo that was data1 to data2.
Bingo the RX would now bind and also run.
This means that the micro-controller on the bottom PCB does need to see two valid inputs before it will allow binding or running mode, but also that it doesn't know that it's reading the same RF chip twice ! That's very handy to diagnose a fault...

OK, now knowing which RF chip was not outputting a data frame and proving the system was otherwise working I resoldered the pins on the RF chip (under a magnifier) and also the associated crystal.
The faulty RF chip sprang to life, the RX now works fully...
Happy friend.

A photo of the AR6000 and where the data pins are on the top PCB is attached. That was taken 'after' my resoldering job on chip1 and the crystal BTW.

It seems likely that you could get rid of the remote receiver on the AR7000 (to save weight maybe) and link the data output of the main receiver chip to the input side of the remotes data pin... Might be useful to someone somewhere.

Feel free to add info here too if you know something more about them technically speaking.

[Rodney]

Very interesting PLMS, thanks for the info.

[richg99]

Thanks, I have a defunct ar6000. I'll see what I can see with this input. Rich

[PLMS]

Since my last post I've seen an AR7000 with a probable cracked solder joint. Vibration makes the RX do a restart. This time it seems to be on the servo headers...

[markw365]

It's common for solder joints to crack on stuff like header pins. Saw this all the time repairing various consumer electronics (monitors, printers), automotive (honda main power relays). Something about a pin or something hanging off the board with lots of leverage on the solder joint. :)

Mark

[PLMS]

It's common for solder joints to crack on stuff like header pins. Saw this all the time repairing various consumer electronics (monitors, printers), automotive (honda main power relays). Something about a pin or something hanging off the board with lots of leverage on the solder joint. :)

Mark

Hi Mark,
Yes, and it also seems to me to have become more common since 'lead free' solder was made mandatory. I see it regularly in late automotive electronics myself (lots of vibration). Designers need to add mechanical strength that doesn't rely on just the solder pads in the increasingly thin circuit boards.
Martin

[dmccormick001]

Very interesting. Our club has seen a lot of unexplained Spektrum crashes, but every time we've asked the "experts" at the factory, or quizzed the company reps at some fun-fly or swap meet or something, the answer has always been that "the technology is sound, so it must be something you're doing wrong." So they've suggested that we didn't charge the batteries fully, or we didn't have something connected securely, or whatever, and that's been the way of it. No matter that many of the pilots involved have been doing this for a long time and are quite experienced. Now I have to wonder if some of the problems haven't simply been due to poor manufacturing and/or poor design of some of the core components, instead of problems with the 2.4GHz Spektrum per se.

And with regard to the vibration and it's effect on solder joints, if I'm not mistaken the Spektrum folks specifically recommend that, to prevent overheating of the Rx, you should NOT wrap a Spektrum Rx in foam as has been the standard for 72MHz receivers for years. Maybe that recommendation should be re-thought, huh?

[PLMS]

Very interesting. Our club has seen a lot of unexplained Spektrum crashes, but every time we've asked the "experts" at the factory, or quizzed the company reps at some fun-fly or swap meet or something, the answer has always been that "the technology is sound, so it must be something you're doing wrong."

I don't want to open a can-of-worms in this thread, but yes I agree, the quality and robustness of the construction of the receivers has dropped since 2.4Ghz came in (at about the same time as lead free solder).
It's just my opinion, and I've kept no statistics, but I 'feel' that I see more crashes from momentary or permanent control loss now with 2.4Ghz than I did with 36Mhz FM.
Sure some were from bad installations, but quite a few have had no explanation other than intermittent circuit faults, particularly evident when they repeat (if the plane survived to be flown again).

[dmccormick001]

I've always felt like the 2.4 equipment was rushed to market before it had been tested sufficiently to work out these kinds of bugs.

[AndyKunz]

And with regard to the vibration and it's effect on solder joints, if I'm not mistaken the Spektrum folks specifically recommend that, to prevent overheating of the Rx, you should NOT wrap a Spektrum Rx in foam as has been the standard for 72MHz receivers for years. Maybe that recommendation should be re-thought, huh?

Futaba is the one with the overheating problem.

Proper installation is the key. Make sure your transmitter and receiver can see each other (antennas, that is) w/o something conductive in the way and you'll be fine.

The only crashes I've had with my Spektrum have either been caused by poor installation (I knew it when I did it, just was too lazy to fix) or pilot error.

None of my Spektrum receivers have foam anywhere near them. Just Velcro holding them to the fuselage.

Andy

[dmccormick001]

Well, I'm kinda like PLMS, I'm not trying to get a brand war started, but in my club, the F.A.S.S.T. (Futaba) systems have had very, very few failures. Of course, they were also quite a bit later in getting into the market in the first place. The Spektrums seem to be the ones that suffered most from the unexplained re-booting problems, which they tried to solve by adding a capacitor to the power rail. Even after that, we still had a lot of crashes in which the Rx seemed to just lose the bind with the Rx. 9 times out of 10, the plane would go into an aileron roll to the left (I think it was left) and crash. And those are the instances I wonder about, if they weren't do to much simpler problems like the ones suggested by PLMS, with the construction of the hardware, rather than defects in the technology, as many assumed beause it's such a new and different concept.

Many of the crashes we've had have involved planes in which the radio system was installed perfectly, by guys who new exactly what they were doing. As I recall, there were even several such unexplained crashes happening as recently as this summer at the SEFF event, to pilots who were very experienced in electric flying/building.

[PLMS]

Well, I'm kinda like PLMS, I'm not trying to get a brand war started.

And those are the instances I wonder about, if they weren't do to much simpler problems like the ones suggested by PLMS, with the construction of the hardware, rather than defects in the technology, as many assumed beause it's such a new and different concept.

Yep, lets avoid a brand war in a technical discussion thread about repairs :)

Yes, my opinion too. Spektrum has VERY solid over-the-air linking, and very good range (about 2km in my ground tests).

The things to look out for after a lock-out crash (with any brand) are:

Will the receiver still link (not lost it's binding) with TX.
Do the servos work fine while the RX is tapped lightly with a stick (non metal).
Do the servos work fine while the RX casing is lightly torqued (twisted).
Do the servos all work fine if the servo plugs are gently moved about on the RX header pins.

These tests work best if a buddy runs the TX, continuously moving most servos smoothly.

If you fail any of these then there could well be an issue with soldering on the RX PCB. That's how most have been detected.

Martin

[dmccormick001]

Those are great suggestions for post-crash testing. Most of us probably just hook everything up, wiggle the sticks a few times to see if they work at all, and go fly.

[Jimmi]

Very good info guys, thanks and keep it comming :)

[pilotpete2]

Futaba is the one with the overheating problem.
Andy

Was, as in past tense ;)
And foam has nothing to do with overheating, since no RC receiver will "overheat" from the small amount of heat it dissipates, only exposure to high ambient temperatures and direct sun will create a problem.
Independent testing has shown current production FASST receivers to be good to 110C :cool: , but then what do I know about hot weather flying :rolleyes: :D
Cheers,
Pete

[dmccormick001]

I don't remember which 2.4GHz manufacturer it was, but I'm almost sure one of them specifically recommended that you not use foam to wrap the Rx. But I might have been wrong about the reasoning behind it, it may not have had anything to do with heat.

It just seems to make sense to me to wrap the most expensive and critical part of the plane's radio system in foam or something to protect it. I was taught to wrap the Rx and battery both in foam to protect them from crash damage and from the effects of the high-frequency vibration caused by a 2-stroke engine completely "wound out". I've never lost a plane to a radio system malfunction, and I still fly on channel 46.

[Gerry B]

Even after that, we still had a lot of crashes in which the Rx seemed to just lose the bind with the Rx. 9 times out of 10, the plane would go into an aileron roll to the left (I think it was left) and crash. And those are the instances I wonder about, if they weren't do to much simpler problems like the ones suggested by PLMS, with the construction of the hardware, rather than defects in the technology, as many assumed beause it's such a new and different concept.


I thought I would share my near crash and what the problem was.

I was new to Spektrum. I followed the manual exactly for binding using the Rx battery/switch harness. I couldn't get a flashing light just one flash when powering up the Rx. Having had no experience with Spektrum I thought maybe the manual is wrong and just carried on with binding and everything seemed to work. I did a range test and tested the surfaces and all seemed fine. I did my flight and on the first circuit it rolled over to the left knife edge, I thought it was gone but managed to get it back and land right away.

What I learned was that the switch harness I had used was a two wire not a three wire so the bind plug wasn't doing anything. I plugged the bind plug directly into the battery/data port and all was fine. I haven't had any trouble since.

[dmccormick001]

I can certainly see how that would cause a problem. I'm sure the manufacterers get blamed for a lot of things that just aren't their fault. Doesn't Spektrum caution you against that somewhere in the manual? Looks like they'd have a big "ATTENTION..." in the section on binding the Rx.

That may expalin some of the crashes, but I personally know of three or four involving very, very expensive 35% or bigger planes that were built by experienced guys who used only the finest equipment money could buy. Two of them I witnessed, and on the maiden flight, the plane lifted off, flew about two seconds, went into a left-hand roll and crashed, never responding to controls again.

[Gerry B]

I know...it was a stupid mistake on my part. I should have looked into it further. I am sure there are a lot of other problems for the crashes too. I still don't trust it yet for my more expensive planes, I have seen many 2.4 crashes where the cause wasn't found.

[Jimmi]

The reason being not to wrap the rx was futaba had a problem with overheating. I'm sure thats been taken care of, but I don't wrap any of my 2.4 rx.

[bnrusso]

I think that finding the broken joint and repairing the solder connections was a smart, good fix. Good Job!! I too worked in microcirtuits and did some chip soldering repair under a microscope.
I'm concerned that most should not attempt this as it requires a very small pointed tip, a steady hand and good magnification. It's super easy to short adjacent contacts and very hard to remove shorts without damaging the circuit or the board.

If you have a broken RX and wish to fix it and fly it (only in an inexpensive foamy), be aware of the consequences.
1. Spektrum uses 2 frequency outputs to double the chances that it won't be interfered with another single frequency. Having only 1 output puts this spectrum fix in the same class as the cheaper Chinese 2.4 radios (2nd or 3rd class systems).
2. By jumpering the output data signal, although it would bind there is still only 1 output channel. If the 2nd chip began to operate I don't know what would happen to the plane.
3. Not using the second RX on the 7000 limits the 360 degree coverage that Spektrum thought was so important for a non parkflyer (long range RX). That's why they added the 2nd RX. This is an insurance policy and I would not remove the 2nd RX for the sake of weight or some other convenience.

[sonicj]

i have 3 spektrum surface rx's. all 3 have had their antenna's fail due to the brittle low strand count wire and the harsh nature of their environment. (seriously, what are they thinking sometimes!?!?!) i've already replaced the broken wires with some lengths of wire more suitable for the application. i would prefer to use a remote antenna though! possibly a diy with a pattern tuned appropriate for the application or maybe something harvested from a old wifi card, router or 2.4ghz cordless phone. this would allow optimal rx placement for routing wires / configuring component layout with compromising antenna placement.

my main concern is with the sheild of the coax... ie: what do i do with it since the stock antenna is a single wire? anyone here have experience with such a mod to their spektrum equipment? any suggestions for antenna designs or signal patterns that might be suited for a 1/18th scale rc car? cheers!
-sj

[David T]

SJ, I agree the Spektrum aerials are a weak link. Newer receivers have been improving this. If the frayed strands touch nearby components it would cause all kinds of intermittant problems.

I would not have thought you would have a range problem with surface vehicles. However, signal transfer is maximised when the Tx and Rx aerials are parallel. I'd have thought this was easiest to achieve with both vertical.

Router coax will be a good option for a remote aerial. Spektrum have various aerial designs. So to recommend how to replace them with this it would help to have some good quality photos of the top and bottom of your receiver. I need to see the components and traces on the board.
David.

[sonicj]

I would not have thought you would have a range problem with surface vehicles. However, signal transfer is maximised when the Tx and Rx aerials are parallel. I'd have thought this was easiest to achieve with both vertical.

Router coax will be a good option for a remote aerial. Spektrum have various aerial designs. So to recommend how to replace them with this it would help to have some good quality photos of the top and bottom of your receiver. I need to see the components and traces on the board.
David.
david, thanx for the info! i only experience range issues when the antenna snaps off :D then again, i have the rx mounted at the highest point for optimal reception(top of subchassis). whats optimal for reception isn't always the most practical or preferred placement for the rx, especially when space is at a premium! (see photo)

the surface rx's only use one antenna.

google turned up this mod to a sr300 link (http://www.rcuniverse.com/forum/fb.asp?m=8657072) thats pretty much what i had in mind. i have the sr300 and two sr3300t. there is little if any difference in the antenna section of the board as i recall, but i'll check to confirm. the last photo in that post is what im after.

[David T]

Coax has an inner core which needs to solder onto the existing solder point. The wire in the core would normally go through the board and be soldered to both sides to reduce the risk of ripping the pads off the board.

The outer braid screens the core and provides structural integrity. To be effective it needs to be attached well to the ground plane, ie the large area under the receiver. One way is to add the UFL connector as shown in the link. Another is simply to unwrap a few millimeters of braid and twirl that into pigtails to solder direct onto the board. Consider making a pigtail on both sides of the cable to give two ground solder points to spread the physical stresses.

You then need to route the coax out of the Rx in a way that does not put undue stress on the solder points. Consider gluing or stitching the coax to the board so that external bending stresses are supported well.

If you use a router aerial you can remove the 'rubber duck' part if you wish or cut the wire shorter. Inside it will simply have about 30mm of the inner core exposed. It probably has a 'cone' of similar length soldered to the braid which is not essential. The end result you are seeking is to have all but the last 30mm of the inner core sheilded.

Digikey sell short lengths of coax if you want to buy some.
dt.

[PLMS]

my main concern is with the sheild of the coax... ie: what do i do with it since the stock antenna is a single wire? anyone here have experience with such a mod to their spektrum equipment? any suggestions for antenna designs or signal patterns that might be suited for a 1/18th scale rc car? cheers!
-sj

I've done two antenna wire to coax transplants on AR6100 receivers. Both successful in flight testing in park foamies.

I had the thin coax available from other devices. The inner conductor just solders to where the original active element goes, the braid is twisted into a larger strand and is soldered under the board on to a grounded area. The other original antenna wire is a ground plane and I just removed it completely.

While coupling into the new coax antenna is not going to be perfect it does seem to work well in models where the short open wire was blocked by batteries and servos.

I'd not use a WiFi antenna 'unless' it's in a model car. Any antenna with gain will have a flattened radiation pattern, not nearly omnidirectional. That's not good in aircraft that can roll to any angle in flight.

Martin