RC equipment reliability

[Frank Skilbeck]

Posted by Martyn K on 20/03/2017 15:27:37:

That a good summary Matty and I agree with most of your points apart from this open ended statement..

"What I am trying to say is those who grew up with analogue RC, low power servos and 4.8V NiCads often still install modern digital gear the way they used to back in the day resulting in in all the expected problems."

What expected problems? Apart from NiCads are replaced with NiMh and for high power requirements LiPo or similar I am puzzled why digital gear should be treated differently.

Martyn

I understand where Matty is coming from, back in the 35 mhz analogue days the 600 mah AA Nicd cells used to hang on to the output voltage pretty well and the analogue receivers would keep working at voltages which had seen the servos go very sluggish. Now with digital receivers any loss in voltage below a certain value causes the receiver to shutdown and reboot when the voltage recovers, add to that people using high capacity (>2,000 mah) Nimh AA cells which drop voltage under any appreciable load and this can lead to problems. I've seen this on large gliders where a 4 cell AA pack has led to brownouts.

Note this isn't just restricted to 2.4, quite a few years back there was similar issues around PCM receivers, if I remember correctly Futaba put in a low voltage warning where the throttle would go to idle (or pulse) when the supply voltage got too low.

[Martin Harris - Moderator]

Posted by MattyB on 20/03/2017 14:38:33:

• Hangover from 1st generation protocols & power sensitive/slow rebooting RXs ...The historic problems certainly hang over one brand to this day though, event though (IMO) there is no hard factual evidence to suggest that they have fundamental issues with their products nowadays
• Manufacturer brain fade - The original Spektrum DX8 botched release. Jeti's broken TX sticks...

I generally agree but it's interesting that you've highlighted how misinformation can be perpetuated by repetition of half truths and rumours...

The Jeti sticks have, to the best of my knowledge, remained the same design and materials since release and hundreds (thousands?) of transmitters appear to survive their flying sessions with their sticks intact. The only broken sticks I've heard of have been subject to physical abuse - including one transmitter that I witnessed being dropped onto a stick - which was replaced for free by the manufacturer with the circumstances being explained! It was discovered that in the two instances abroad that I'm aware of, one had been kicked immediately prior to the failure and the other had been transported roughly in a case with the supporting padding removed...

[MattyB]

Posted by Frank Skilbeck on 20/03/2017 15:55:42:

Posted by Martyn K on 20/03/2017 15:27:37:

That a good summary Matty and I agree with most of your points apart from this open ended statement..

"What I am trying to say is those who grew up with analogue RC, low power servos and 4.8V NiCads often still install modern digital gear the way they used to back in the day resulting in in all the expected problems."

What expected problems? Apart from NiCads are replaced with NiMh and for high power requirements LiPo or similar I am puzzled why digital gear should be treated differently.

I understand where Matty is coming from, back in the 35 mhz analogue days the 600 mah AA Nicd cells used to hang on to the output voltage pretty well and the analogue receivers would keep working at voltages which had seen the servos go very sluggish. Now with digital receivers any loss in voltage below a certain value causes the receiver to shutdown and reboot when the voltage recovers, add to that people using high capacity (>2,000 mah) Nimh AA cells which drop voltage under any appreciable load and this can lead to problems. I've seen this on large gliders where a 4 cell AA pack has led to brownouts.

Yep, that was exactly my point Frank - digital RXs do not "fail gracefully" when starved of power in the same way analogue ones do. On a slope awayday a few years ago I watched a beautiful scale machine go in as crow plus airbrakes were deployed on landing. It had four digital servos on the ailerons and flaps, two on air brakes and two more on elevator and rudder. Picking over the wreckage it turned out a 4 cell 2000mah Eneloop had been used to power the RX, which was flashing to denote a reboot. Case closed.

Edited By MattyB on 20/03/2017 16:38:30

[Frank Skilbeck]

Posted by Martin Harris on 20/03/2017 16:16:28:

Posted by MattyB on 20/03/2017 14:38:33:

• Hangover from 1st generation protocols & power sensitive/slow rebooting RXs ...The historic problems certainly hang over one brand to this day though, event though (IMO) there is no hard factual evidence to suggest that they have fundamental issues with their products nowadays
• Manufacturer brain fade - The original Spektrum DX8 botched release. Jeti's broken TX sticks...

I generally agree but it's interesting that you've highlighted how misinformation can be perpetuated by repetition of half truths and rumours...

The Jeti sticks have, to the best of my knowledge, remained the same design and materials since release and hundreds (thousands?) of transmitters appear to survive their flying sessions with their sticks intact. The only broken sticks I've heard of have been subject to physical abuse - including one transmitter that I witnessed being dropped onto a stick - which was replaced for free by the manufacturer with the circumstances being explained! It was discovered that in the two instances abroad that I'm aware of, one had been kicked immediately prior to the failure and the other had been transported roughly in a case with the supporting padding removed...

It wasn't rumour https://www.rcgroups.com/forums/showpost.php?p=23732797&postcount=876 but it was addressed by Jeti and they strengthened the sticks. But similarly some of the early Multiplex Profi's had a centering arm break (sub-supplier changed the materials apparently) but mines been fine and it isn't treated gently!, also there were some reports of centering spring breakages on some Taranis Txs.

[MattyB]

Posted by Peter Christy on 20/03/2017 15:45:47:

MattyB: An excellent summary! However, as one of those who came to the hobby before proportional control was actually available - analogue or otherwise - I feel duty bound to point out that we *had* to install our equipment carefully!

Reed and relay receivers (some had as many as 12 relays on board!) had to be mounted "just so", and with as many as seven wires going to each servo, heaven help you if your wiring was not tidy!

Yep, I have seen some of my Dad's old 7 wire servos in a dusty "archive" box at the back of one of his cupboards, and by all accounts they sound like temperamental beasts! I am sure that in those days installs were done with extreme care in order to give the best chance of success. Those installs were certainly not the ones that led to complacency in the 2.4 era!

[MattyB]

Posted by Frank Skilbeck on 20/03/2017 16:41:09:

Posted by Martin Harris on 20/03/2017 16:16:28:

I generally agree but it's interesting that you've highlighted how misinformation can be perpetuated by repetition of half truths and rumours...

The Jeti sticks have, to the best of my knowledge, remained the same design and materials since release and hundreds (thousands?) of transmitters appear to survive their flying sessions with their sticks intact. The only broken sticks I've heard of have been subject to physical abuse - including one transmitter that I witnessed being dropped onto a stick - which was replaced for free by the manufacturer with the circumstances being explained! It was discovered that in the two instances abroad that I'm aware of, one had been kicked immediately prior to the failure and the other had been transported roughly in a case with the supporting padding removed...

It wasn't rumour but it was addressed by Jeti and they strengthened the sticks. But similarly some of the early Multiplex Profi's had a centering arm break (sub-supplier changed the materials apparently) but mines been fine and it isn't treated gently!, also there were some reports of centering spring breakages on some Taranis Txs.

Yep, as Frank says above Jeti have stated they strengthened the sticks after reports of breakages in the early days for the DC/DS-16. They claim they have been tested to 40kg of load which may be the case, but whether they failed through mis-handling or a manufacturing defect nobody really knows. In my book I would still say letting them out of the door like is an error on Jeti's part, as something like a stick on such a premium product should be thoroughly over-engineered to avoid a failure - doing so has essentially zero consequence to weight or price.

Edited By MattyB on 20/03/2017 17:12:50

[Nigel R]

"digital RXs do not "fail gracefully" when starved of power in the same way analogue ones do."

And a problem for all kinds of kit where microcontrollers have replaced analogue circuitry, not just RC gear.

Often compounded by software being written by folk who may have no electronic hardware background or appreciation.

Typically, SW would need specific bits added to deal properly with brown out conditions. If you don't know you need that, you won't put that in. You might also be working with a hardware setup that doesn't tell you when brown out is happening...

[MattyB]

Nigel, I have no idea how they have addressed it in the end - you clearly have more knowledge than me. What I can tell you is that everyone (including Spektrum) now seem to have now got their RXs to the point that they only reboot at voltages below those that most servos cease to respond.

I am guessing, but given that's the case I doubt they have integrated anything in software to deal wit the brown out conditions - they have just concentrated on minimising the chances of it happening by reducing the minimum voltage of operation, and rebooting and reconnecting faster if a reboot does occur. Even so it remains critical to think carefully about how you power your RX and servos nowadays, especially if using big digitals that are current hungry.

[Nigel Dell]

sorry senior moment posted in wrong thread!

Edited By Nigel Dell on 20/03/2017 17:48:31

[Martin Harris - Moderator]

Posted by MattyB on 20/03/2017 16:52:58:

Yep, as Frank says above Jeti have stated they strengthened the sticks after reports of breakages in the early days for the DC/DS-16. They claim they have been tested to 40kg of load which may be the case, but whether they failed through mis-handling or a manufacturing defect nobody really knows. In my book I would still say letting them out of the door like is an error on Jeti's part, as something like a stick on such a premium product should be thoroughly over-engineered to avoid a failure - doing so has essentially zero consequence to weight or price.

The strengthening was done, I understand, as a confidence boost after an unproved instance of "spontaneous" failure. There was certainly no recall of early units and no further reports that I am aware of. As I inferred, an example of very little evidence perpetuating a probable myth.

[gangster]

Guys. I am not sure I understand the comments about software needing to be written to deal with brownout conditions. I may have missed the point here. As stated in an earlier post and confirmed by my own experiments with 3 different dsmx receivers and an orange dsm2 the receivers keep on working until the servos go slow. They also recover instantly I have not experimented with spekky dsm2 but it is well documented how long of them take to recover. Not sure what is needed in software. What point have I missed as far as I see nothing will work if the voltage drops too low I have also notice the engine stops if you don't give it any fuel

Edited By gangster on 20/03/2017 19:05:45

[Don Fry]

Very well written Matty.

And the edit is nowt to do with me, but my inability to keep up, and in particular get to live with predictive spell checkers. 

Edited By Donald Fry on 20/03/2017 19:20:17

[Former Member]

[This posting has been removed]

[MattyB]

Posted by Steve J on 20/03/2017 19:27:20:

@MattyB

While I agree with most of what you say, I must pick you up on a couple points.

Only the earliest Spektrum receivers had a slow signal reacquisition time. The problem was fixed by the Quick Connect firmware which was released early in 2008. The fix was simply for the receiver to record the frequencies that the transmitter was using and check them first on boot.

The lower voltage limit for most Spektrum receivers has been 3.5V (i.e. the input voltage for a 3.3V low dropout regulator) since the first DSM2 receivers (AR7000 3.5-9.6V, AR8010T 3.5-9V).

Yes, I am well aware the slow reacquisition was eliminated long ago, hence why it titled that section "Hangover from 1st generation protocols and power sensitive / slow rebooting RXs". Re: operating voltages for Spektrum, 3.5V may be the stated minimum on their TXs but many do better than that - this 6100E gets down to 3V before it fails. It was not always so, but despite the fact they don't quite match FrSky (who go down to ~2.8V) it's still good enough IMO to dispel the persistent "Spektrum Brownout" excuse from the early days when their reboot voltage was 4.1-4.2V.

Edited By MattyB on 20/03/2017 20:22:30

[Nigel R]

gangster, mattyb, the effects of brownout are something you can mitigate in SW, obviously you can't fix the issue because it's an external problem (as far as the RX's software is concerned).

For our situation here; it is electrical loads causing volt drop, you would probably have several lists of loads to remove when the voltage dropped to different levels. As a first line of defence you'd stop driving (or return to a failsafe position) non-essentials; flaps/gear/aux stuff, then as a second line (if voltage doesn't rise, or drops further to a lower level) you might return to a standard full failsafe position for all controls. Subsequently if voltage drops so low the CPU can't operate you'd want the 'warm reboot' operation to be performed much, much quicker than the cold start (first switch on). Fixing warm reboot might involve radio protocol changes (e.g. to allow the RX to continue working without a full negotiation with the TX) as well as SW fixes (store your last operating state to allow instant return to last know positions and so forth).

This kind of strategy is what goes into full size power systems (my day job), the big boys get highly concerned if your electrical system goes down unnecessarily and/or stays down in an emergency.

From the sound of things, Spektrum have concentrated on sorting the warm reboot time, which at least gets you out of a hole when the battery goes south.

[Denis Watkins]

Well done Matty for this comprehensive summary,

And should be compulsory reading for installers new and old

[Nigel R]

hear hear, Denis

[Don Fry]

Posted by Nigel R on 21/03/2017 08:34:48:

gangster, mattyb, the effects of brownout are something you can mitigate in SW, obviously you can't fix the issue because it's an external problem (as far as the RX's software is concerned).

For our situation here; it is electrical loads causing volt drop, you would probably have several lists of loads to remove when the voltage dropped to different levels. As a first line of defence you'd stop driving (or return to a failsafe position) non-essentials; flaps/gear/aux stuff, then as a second line (if voltage doesn't rise, or drops further to a lower level) you might return to a standard full failsafe position for all controls. Subsequently if voltage drops so low the CPU can't operate you'd want the 'warm reboot' operation to be performed much, much quicker than the cold start (first switch on). Fixing warm reboot might involve radio protocol changes (e.g. to allow the RX to continue working without a full negotiation with the TX) as well as SW fixes (store your last operating state to allow instant return to last know positions and so forth).

This kind of strategy is what goes into full size power systems (my day job), the big boys get highly concerned if your electrical system goes down unnecessarily and/or stays down in an emergency.

From the sound of things, Spektrum have concentrated on sorting the warm reboot time, which at least gets you out of a hole when the battery goes south.

But is not Matty saying that the RX works to voltages where the the servos are dying from lack of volts. So how does the battery find the power to move 4 or 5 servos. Keep it simple, I just shove this stuff into airframes, and remember schoolboy physics. I can follow the arguments of the above couple of pages.

[Nigel R]

"But is not Matty saying that the RX works to voltages where the the servos are dying from lack of volts. "

 

But not so with the original DSM receivers; therein lay the problem we were commenting on. They would brownout at a relatively high supply voltage, so I understand.

If your worst case load drags the supply voltage down to such a level that the RX continues to function, then all is well with the RX, and there is nothing that needs to be done other than keep on keeping on.

Spektrums solution, noted above, was they tackled two problems; (1) they made the RX restart quickly (by remembering what the TX was doing) and (2) improved the ability to deal with a low volt power supply. Or in other words, the brown outs don't happen as much, and when they do they're not as bad.

The other stuff in the middle is probably beyond what most of us care about or need, in all honesty. Maybe the guys flying big & complicated stuff might like it.

At the end of the day, if we can stick in a big enough battery then none of this happens anyway. It's all only relevant when the battery is a bit marginal. Or fails.

I thought the glider example was an excellent illustration of how it goes wrong; low altitude, low speed, activate landing controls, lose radio link, splat.

Edited By Nigel R on 21/03/2017 10:46:03

[MattyB]

Posted by Nigel R on 21/03/2017 10:44:59:

"But is not Matty saying that the RX works to voltages where the the servos are dying from lack of volts. "

But not so with the original DSM receivers; therein lay the problem we were commenting on. They would brownout at a relatively high supply voltage, so I understand.

Correct - the DSM and some first generation DSM2 RXs had brownout voltages in the 4.1-4.2V range. That proved well within the window of vulnerability if using an aging 4.8V pack as the power supply for high power/large numbers of servos. Coupled with slow re-acquisition after reboot it did cause losses, but these issues have long since been addressed by Spektrum in their newer products (any RX purchased from 2009 onwards should be fine).

[Former Member]

[This posting has been removed]

[MattyB]

Bruce from RCModelReviews (tests on his site and Youtube channel), and a personal test I did of a friends DSM2 RX about 6 or 7 years ago after he experienced a brownout - sorry, I can't remember the model number. As I stated above not all Spektrum DSM2 RXs exhibited high reboot voltages, only a subset of the early ones.

Edited By MattyB on 21/03/2017 13:58:14

[Former Member]

[This posting has been removed]

[Peter Christy]

Posted by MattyB on 21/03/2017 12:37:03:

Correct - the DSM and some first generation DSM2 RXs had brownout voltages in the 4.1-4.2V range. That proved well within the window of vulnerability if using an aging 4.8V pack as the power supply for high power/large numbers of servos.

And that hits the nail squarely on the head!

It never ceases to amaze me that modellers will spend months building a model, spend hundreds of pounds on RC gear, and then power it using a battery and switch that they pulled from the last wreck!

I used to have a friend who ran a model shop, sadly no longer with us. He used to sigh with relief when a heli flyer came in! "Why?", I asked him. "Because they think nothing of spending £100 on a servo, if that's what's needed! Fixed wing flyers argue about the cost of the cheapest servo / batteries / switches I stock, despite the effort they've put into building, and the amount they've spent on receivers and engines!"

I have to say that looking around some of the local clubs, I can see where he was coming from......!

--

Pete

[john stones 1 - Moderator]

Nope, fixed wing flyer here, my gears fit for purpose, also flown helis as well and trust me i've met some iffy modellers there.

John