Redundancy in larger models?

[Tim Kearsley]

Just looking for some opinions chaps.

I'm assembling a Hangar 9 Spitfire, which, at around 81 inch wingspan is the largest model I've yet owned. Adding everything up, there's quite an investment up in the air every time this beauty takes off - do you think I should be thinking about things like redundant power for the Rx etc? I'm taking the electric power route I should add.

Any thoughts from you lads who build large would be very welcome.

Thanks,

Tim.

[Jon H]

Personally i dont bother.

I have 4 models of that type/size and each runs a very simple setup with a 3000-3600Mah 6v sub C battery pack running through a futaba heavy duty switch direct into the rx. I use a variety of servos ranging from 3.5-10kg with some digital and some standard and get somewhere around 15 flights from the battery.

Obviously it is your choice, but i have not had any issues with my futaba gear and i regularly monitor the health of my batteries to minimise the chances of an issue creeping in.

I dont like adding more and more stuff to a model needlessly as the more stuff you add, the more things there are to go wrong.

Even when i was assisting with an over 20kg spitfire we didnt have redundant batteries for the rx. We did have two rx's, with controls shared between the two as this was a better guard against a failure.

[Peter Christy]

Posted by Jon Harper - Laser Engines on 09/06/2017 08:50:12:

I dont like adding more and more stuff to a model needlessly as the more stuff you add, the more things there are to go wrong.

Exactly!

--

Pete

[Rich too]

Mmmmm, I agree to an extent, but switches are (apparently) a common point of failure, therefore two switches/batteries appears to be good insurance.

My personal preference for petrol models is two switches, two batteries, plugged into rx separately, and a Tech Aero ibec for ignition powered by the receiver rather than a separate ignition battery.

I am in favour of simplicity and do not use power boxes/expanders etc..

However, I do not generally fly electric so cannot comment on redundancy for these models!

[flight1]

The KISS principle is always best, 0ne good quality battery and switch (or 2 switches) and with regular checks and maintenance is all that is needed for safe flying, you only have one battery in the tx and that is never questioned as unsafe or unreliable.

Edited By flight1 on 09/06/2017 09:22:42

[Engine Doctor]

Having lost one expensive model a few years ago to battery (one cell in the pack failed ) failure I now use a Two battery and switch set up for added security . Not a lot of weight in a big model and doesn't affect flight performance I understand other post who choose to use a simple set up and use the KISS principle but for me any larger model with considerable investment in time and money I use and recommend a redundant RX power supply. 

Edited By Engine Doctor on 09/06/2017 09:27:19

[Tim Flyer]

On my "big" Wots Wot. I have a single large LIFE battery. It has a thick high current charging lead with Tamiya connector and a Separate Futaba output lead. What I did was use both leads into the RX via two different heavy duty switches. My battery has a massive 4500ma capacity ( also use it as it adds needed weight to the nose ). I got the battery connectors and switch gear mail order at componentshop.com in Wales. I have experienced switch/connector failure in the past so having two switches and connectors to the big battery seems to work well. This model is "thin skinned" with plenty of free area around the RX aerials so 1 receiver is fine.

[Andy Meade]

It depends what you mean by "large" I think. My 110" A10 uses a dual battery system with one receiver. My 133" C17 uses 4 batteries - 2 per receiver, and a receiver redundancy bus system. Everything I fly now uses telemetry to protect myself and others from dead cells or signal quality / interference problems.

I think we have a duty of care with big models to ensure that everyone and everything we fly near is safe. IMHO a 20kg Spit with just one battery should not be flown, and these days falls under the remit of the CAA / LMA.

[Martin Harris - Moderator]

I'd very strongly agree with the simple approach - but this, to my mind, includes a 2 battery / 2 switch set up. If you really don't want a second battery then a dual switch / single battery would add a great deal of protection against a very common point of failure for a very small weight/cost penalty.

[Rich too]

Posted by flight1 on 09/06/2017 09:21:14:

The KISS principle is always best, 0ne good quality battery and switch (or 2 switches) and with regular checks and maintenance is all that is needed for safe flying, you only have one battery in the tx and that is never questioned as unsafe or unreliable.

Edited By flight1 on 09/06/2017 09:22:42

If I could use two in the tx I would! Although the tx environment is a lot kinder on batteries/switches/electrics than a petrol powered airframe.

Edited By Rich2 on 09/06/2017 10:16:00

[Matt Jones]

Posted by Andy Meade on 09/06/2017 09:46:18

I think we have a duty of care with big models to ensure that everyone and everything we fly near is safe. IMHO a 20kg Spit with just one battery should not be flown, and these days falls under the remit of the CAA / LMA.

I agree with Andy and take a similar approach of multiple batteries and smart switches to minimise risk, I have a power distribution board in my biggest. For the scratchbuilders among us we invest too much work in big models to lose them due to an avoidable failure.

[Martin Harris - Moderator]

Posted by Engine Doctor on 09/06/2017 09:26:09:

Having lost one expensive model a few years ago to battery (one cell in the pack failed ) failure I now use a Two battery and switch set up for added security . Not a lot of weight in a big model and doesn't affect flight performance I understand other post who choose to use a simple set up and use the KISS principle but for me any larger model with considerable investment in time and money I use and recommend a redundant RX power supply.

Telemetry saved my Chipmunk in similar circumstances - a cell failed in flight and a low battery voltage warning prompted me to make an immediate landing. There are many who try to tell me that telemetry is just a gimmick...

[Peter Christy]

Whilst I can see the logic of dual switches, experience has taught me that dual batteries and any kind of power box are likely to cause more problems than they solve.

Case 1: A large scale twin, fitted with a radio that was thoroughly tested in a "hack" model. Short range, and constant servo twitching when in range. Replace two battery and switch box arrangement with a single high quality battery and switch, and all the problems disappeared!

Case 2: 3D helicopter which had been flying well all season. Owner replaced 4-cell NiXX pack with lipo and regulator. Result - glitches every flight! Revert to 4-cell NiXX, and previous reliability restored.

Batteries do not just fail. Unless they are cheap packs, they tend to fail slowly - something that should be picked up in routine maintenance. And unless some kind of switching arrangement is in place, there is always the danger of the "bad" pack pulling down the "good" one. Any switching arrangement will increase the apparent internal resistance of the power supply, something best avoided when using "digital" servos.

Dual receivers: How do you decide which one is right? Full size passenger aircraft use a minimum of triple redundancy and a polling system - 2 out of 3 wins. With just two, there is no logic system that can decide which receiver is giving the correct output, unless one is giving no output at all. Again, in this day and age, receiver failure is rare, and is often preceded by strange behaviour that should be picked up in pre-flight checks.

KISS rules, OK?

--

Pete

[Jon H]

Posted by Andy Meade on 09/06/2017 09:46:18:

I think we have a duty of care with big models to ensure that everyone and everything we fly near is safe. IMHO a 20kg Spit with just one battery should not be flown, and these days falls under the remit of the CAA / LMA.

I agree Andy, but the model in question had two rx's with independent power sources and was fully inspected and passed by the LMA.

Also to clarify my earlier comment, really big models do need a somewhat higher degree of redundancy, but an 80 inch spit is not that big in the grand scheme.

Batteries tend to be reliable for the most part, and i personally have never had a switch let me down but i do avoid the cheap ones. Remember too that public enemy no1 for a switch is vibration and compared to a big petrol engine the electric will be in a different league.

As for telemetry, i dont really see the point in anything beyond the rx voltage being reported.

[Jon H]

Posted by Peter Christy on 09/06/2017 11:03:04:

Whilst I can see the logic of dual switches, experience has taught me that dual batteries and any kind of power box are likely to cause more problems than they solve.

Case 1: A large scale twin, fitted with a radio that was thoroughly tested in a "hack" model. Short range, and constant servo twitching when in range. Replace two battery and switch box arrangement with a single high quality battery and switch, and all the problems disappeared!

Case 2: 3D helicopter which had been flying well all season. Owner replaced 4-cell NiXX pack with lipo and regulator. Result - glitches every flight! Revert to 4-cell NiXX, and previous reliability restored.

Batteries do not just fail. Unless they are cheap packs, they tend to fail slowly - something that should be picked up in routine maintenance. And unless some kind of switching arrangement is in place, there is always the danger of the "bad" pack pulling down the "good" one. Any switching arrangement will increase the apparent internal resistance of the power supply, something best avoided when using "digital" servos.

Dual receivers: How do you decide which one is right? Full size passenger aircraft use a minimum of triple redundancy and a polling system - 2 out of 3 wins. With just two, there is no logic system that can decide which receiver is giving the correct output, unless one is giving no output at all. Again, in this day and age, receiver failure is rare, and is often preceded by strange behaviour that should be picked up in pre-flight checks.

KISS rules, OK?

--

Pete

 

i pretty much agree. On occasion things will just fail (like ED's battery) but in general its unusual these days.

The dual rx setup i used on the spit, and will use again in future was that one rx has left aileron and right elevator, and the other rx has the opposite. The other channels were evenly spread out as well so if you had an rx go down you could have flaps and throttle, or engine kill switch and undercarriage. it seemed logical to me and worked well. Not that we ever had to test it in flight.

Edited By Jon Harper - Laser Engines on 09/06/2017 11:11:28

[Nigel R]

" Full size passenger aircraft use a minimum of triple redundancy"

They don't always. Sometimes dual. Sometimes quad. Sometimes no redundancy at all. Other times the redundancy allows a graceful failure with lower levels of control as bits fail.

It depends on analysis of rate of failure, severity of failure, then down to what logic is needed to make a safe fallback.

If the receiver is designed to allow it, and is linked and communicates with its twin, then it could negotiate and vote between the various signal sources. Receivers already do this to some extent - any receiver with a satellite will choose which signal source to use, dual aerials go so way toward this goal. Joining two receivers would then give you more sources to pick from. However it would take some fairly special design to be able to redundantly control single servos from two receivers.

OTOH at a more fundamental level, you could arrange something like split ailerons or split elevators as in Jon's case, whereby one receiver gives a measure of control in the event the other ceases. A bit like split brake circuits in cars. Probably the simplest and best option at our scales.

[Martin Harris - Moderator]

I mostly agree Pete, and have a strong aversion to the use of relatively complex voltage regulators but I would take issue with the absolute reliability of a battery. In most cases, you are reliant on the manufacturer's connections hidden under the shrink wrap and these have been known to fail, even on "good named" packs. Plug/socket connectors are vulnerable and only as good as the crimping tool used to terminate them.

My own experience (on a decent quality NiCd pack) was that the failed cell was in a reasonable aged pack that charged/discharged as expected, gave normal voltages on and off load and didn't exhibit excessive self discharge in storage.

I'm not convinced that the "pulling down" scenario is likely to be terminal - with 2 reasonably charged 4S NiXX packs, a failed cell is unlikely to drop the combined output voltage of the pair sufficiently to lose control and I always perform separate voltage checks pre-flight in the manner of a mag check - again, a great benefit of telemetry...

I use diodes for LiFe pack systems which give as good protection as anything against a massive failure or short circuit although primarily to limit the pack voltage.

[Bill_B]

I would suggest using a Jeti dual redundant magnetic switch and LiPo/LiFe batteries rather than archaic heavy duty mechanical switches and NiMHi cells.

Edited By Bill_B on 09/06/2017 12:09:04

[John Lee]

I fit an Optipower ultra-guard as a backup in my larger models. It can be retro fitted, it keeps things very simple & will only kick in should there be a Rx power failure. I've never had to use it for real but check its operation prior to each flying session & it has been 100% over 3 years.

[ChrisB]

On all my large models from around 80" and upwards I use a single receiver with 3 satellites (spektrum) , 2 hd switches and 2 subc 3300 5 cell nimhs. I don't use any regulator they just plug into 2 separate rx channels and that is that. The powerbox type systems are ok but because they go through a common route,if that fails then everything fails. As others have said, more to go wrong.

[Jon H]

Posted by Bill_B on 09/06/2017 12:08:40:

I would suggest using a Jeti dual redundant magnetic switch and LiPo/LiFe batteries rather than archaic heavy duty mechanical switches and NiMHi cells.

Edited By Bill_B on 09/06/2017 12:09:04

With more than more than double the failure points of my setup i will stick to my battery thanks. Im also not sure the weight of an rx pack is going to make much difference to a 20lb warbird that already has a full pound of lead under the cowling.

[Bob Cotsford]

Since the advent of crimped connectors on rc gear to replace the dreadful soldered joints of the 70's I can only think of two models lost to 'radio failure' - one a definite NiMh cell loss and the other a suspected switch. Now I prefer two batteries and two switches even on 120 size models.

While this may have more failure points, the 'point' is that the redundant paths mean that a failure in either path is not critical, and the failure can be found and fixed without the need for a spade as would be the case with a single power feed.

[Andy Meade]

Exactly Bob - the complexity goes up, but the SAFETY of the system also does. "...double the failure points.." - quite right!  ALL of those have to fail i norder to bring the aircraft down, not just one.

Simplicity would be great, but doesn't protect the airframe and other people from the number one cause of crashes (after dumb thumbs) - Rx power loss.

And the "my transmitter has a single battery so what's the point" is a poor argument.  The Tx isn't subjected to engine vibration / bumpy landings / oil and moisture ingress / sat in a cold shed all winter.

Edited By Andy Meade on 09/06/2017 13:45:03

[Rich too]

Posted by Bob Cotsford on 09/06/2017 12:58:31:

Since the advent of crimped connectors on rc gear to replace the dreadful soldered joints of the 70's I can only think of two models lost to 'radio failure' - one a definite NiMh cell loss and the other a suspected switch. Now I prefer two batteries and two switches even on 120 size models.

While this may have more failure points, the 'point' is that the redundant paths mean that a failure in either path is not critical, and the failure can be found and fixed without the need for a spade as would be the case with a single power feed.

I agreed Bob. This appears to have opened a can of worms!

I am not sure about electric as the vibration killing the electrics is greatly reduced, but I frequent the FG website and redundancy of any type is pretty much universal.

A lot of guys use three batteries - one for ign - and I cannot understand that, as it means an extra battery and no redundancy on the ign!

I guess, as with many things like this, until you experience failure (which fingers crossed, I never have), there is a reluctance to change.

ps I am also going to stop using the old type switches and will move to the pin flag switches in future.

Edited By Rich2 on 09/06/2017 13:35:27

[The Wright Stuff]

I think there is an argument that says that redundancy is not just down to hardware, it's also a state of mind. Duplication through increased complexity acts to improve safety ONLY if:

a) the user fully understands the technical implications of adopting a more complicated approach, and

b) the user does not become complacent as a result of the added hardware.

Otherwise, yes, there are situations where the redundancy is well-intended, but overall risk goes up! Can one honestly say that they will check each battery just as thoroughly as if there was only one? Appreciate this may sound cynical, but I have overheard at the field, a conversation along the lines of:

"Are you going to check the 'X' again?".

"Nah, it's alright, there's a backup anyway"