Gpro FBL Unit and Power Requirements

[coolice]

Hey All.

Flight control system power requirements in a model are often over-looked, more so when an already flying model is converted or has it's setup changed to new hardware and so is taken for granted that it will all work as before.

During a recent conversation with a UK masters pilot, it transpired that one of our fellow team members had experienced and inflight problem while testing his newly fitted Gpro into one of his working 700E's to give it a whirl. He was very impressed with it's performance I may add, however on the 3rd flight experienced what I have determined (going by the explanation of what happened and some new data) to be an inflight reset of the Gpro due to a drop in BEC volts.

Now this guy has been using a Kontronik Heli Jive ESC/BEC combination with another setup all ok, the BEC's output of 5amps continuous in my eyes is too low for a 700 class machine, but it was working and so was used as before.
This change to the Gpro has highlighted a potential problem with low rated BEC's again, in similar fashion to the 3GX as Align increased it's performance with new firmware releases and so current draw/demands increased.

Originally the owner was sceptical that the BEC could be at fault and as we all know, after a crash it's very difficult to determine the true cause, but based on the data he then collected after I'd spoken to him about similar issues the 3GX had, I think he's now resigned to the same conclusion.

Two identical models were compared only differing in their FBL units, in this case Align 700N DFC's, with the 1st using a Vbar and the 2nd was flown by Duncan Osbourn with the Gpro. The latter may have been flown harder perhaps, but the great difference in readings was substantial enough in my eyes to prove the theory that it was a power related problem.

Unfortunately I don't have an Eagletree unit to get some nicer readings, so a Watt Meter was employed to record the maximum amps drawn in flight.

The Mini Vbar returned a 3.88 amps max current draw and the Gpro (in Duncs hands) 14.66 amps max. A startling difference.

My purpose of this post is not to put blame on one unit, but to highlight the need to make sure the power supply setup to your servos fitted to your models is up to the job to avoid potential problems.
.

[Angelos]

coolice,
regardless of how gently the FBL may be performing cyclic corrections (gently = less current demand of course), the collective passes through the unit practically unmodified. Thus the maximum current requirement will always be determined by the collective and not the cyclic which makes it largely independent of what unit you use.

However, the average current is very much determined by the FBL hardware and algorithms. Higher average current means your servos will run warmer and potentially not last as long especially if they use brushed motors. The main contributor here is vibrations and other unwanted signals picked up by the gyro sensor together with the good stuff (rotation rates) that we want to measure. If the FBL unit does not filter these properly they will be then taken into account my the stabilisation algorithms. As a result your servos are now constantly moving as they track these unwanted signals, thus consuming more power and getting warmer.

The Vortex I use for R&D has an additional circuit board inside that allows it to measure the current that goes to the servo outputs. Similar thing you tried to do with your Watt's Up however mine has a smaller range of 16A thus making it more accurate to small currents. Every time we test new algorithms we compare the average current between flights and we have made significant reductions on the average current using it as guidance.

[ChrisB]

Can Duncan not fly the Mini V Bar and get a comparison?

This data is hardly going to sell the GPro is it?

A small voice in my head wants to try the Gpro but this tells me Align are still way off the mark compared to the V Bar which i quite happily fly,on Kontronik Becs totally reliably for hundreds of flights,in fact this has now silenced that little voice totally.

[Red3]

I don't think it will stop me trying it I will just take care when choosing how I power it. from what I have seen and herd its worth a go.

thanks for the heads up Ian

[markb75]

im waitting for the 550L dominator then ill have a try

[coolice]

coolice,
regardless of how gently the FBL may be performing cyclic corrections (gently = less current demand of course), the collective passes through the unit practically unmodified. Thus the maximum current requirement will always be determined by the collective and not the cyclic which makes it largely independent of what unit you use.

However, the average current is very much determined by the FBL hardware and algorithms. Higher average current means your servos will run warmer and potentially not last as long especially if they use brushed motors. The main contributor here is vibrations and other unwanted signals picked up by the gyro sensor together with the good stuff (rotation rates) that we want to measure. If the FBL unit does not filter these properly they will be then taken into account my the stabilisation algorithms. As a result your servos are now constantly moving as they track these unwanted signals, thus consuming more power and getting warmer.

The Vortex I use for R&D has an additional circuit board inside that allows it to measure the current that goes to the servo outputs. Similar thing you tried to do with your Watt's Up however mine has a smaller range of 16A thus making it more accurate to small currents. Every time we test new algorithms we compare the average current between flights and we have made significant reductions on the average current using it as guidance.

Hey Angelos.

Thanks for chiming in, with a perspective from a FBL designer and manufacturers point of view.
This may not happen as I'm getting good and bad responses for even posting this.

I see what you are saying about the collective pitch effectively going straight through the FBL unit to the servos, but I the same FBL unit is then using the same servos to control model attitude, so during aggressive collective manoeuvres current draw will increase due to the pitch inputs and FBL corrections to maintain model attitude.
What happens though, in the case of the 3GX for example, that has a "Collective Pitch Acceleration" parameter, which is noted to make sure your BEC (or BED in the software) is up to the job above 50%?
Then we're onto another part of the FBL system artificially accelerating the collective inputs and thus adding to demand.

Vibration is a good one, the tests above were on nitros so can be influenced by this, the brownout was on an electric machine.

As you have openly admitted going through this yourself with the Vortex, which is appreciated and admirable, Align will no doubt do the same along the Gpro's career.
It's not a bad mark against the unit itself, there are quite a few flying now and with very pleased customers, but just a heads up to make sure your other components can cope.

Can Duncan not fly the Mini V Bar and get a comparison?

This data is hardly going to sell the GPro is it?

A small voice in my head wants to try the Gpro but this tells me Align are still way off the mark compared to the V Bar which i quite happily fly,on Kontronik Becs totally reliably for hundreds of flights,in fact this has now silenced that little voice totally.

Hey Buddy.

Good call, it would be nice to have Dunc on the same Vbar machine to have a genuine comparison.

Personally I don't think it's going to effect the Gpro at all, it's not a bad mark against the unit at all, it's just giving people the heads up to make sure their power setup is capable enough.
With Angelos chiming in above saying he had similar troubles in the beginning with the Vortex, this goes to show that no matter what unit you have or comes to the market, all have to start somewhere and get refined along the way. Vortex sales have not been effected by Angelo's and his test pilots constant input to refine a product.
I think this past HeliMasters has shown the Gpro to be a worthy unit, some were being flown I've heard, so it's not a problem as long as it has the power source behind it.
So don't be put off if you want to try it, Duncan likes his, so to does Connor and have not had an issue. Only one powered by a 5amp BEC has potentially had an in flight reset due to a drop in volts.
.

[Angelos]

With Angelos chiming in above saying he had similar troubles in the beginning with the Vortex, this goes to show that no matter what unit you have or comes to the market, all have to start somewhere and get refined along the way. Vortex sales have not been effected by Angelo's and his test pilots constant input to refine a product.

Ian, I think you misunderstood. It is not a problem the Vortex had, it is a design consideration for every PID control loop. If you let noise and unwanted signals reach the PID it will respond with constant unnecessary and unwanted corrections consuming power and wearing out the servos. The Vortex is not the first time I came across it, my very first tail gyro (ds760) had such filtering back in 2004.

What I am trying to point is that the average current (or mAh per flight if you like) is very dependant on the ability of an FBL system to filter out the unwanted stuff from the rotation rates it needs.

-Angelos

[coolice]

Ian, I think you misunderstood. It is not a problem the Vortex had, it is a design consideration for every PID control loop. If you let noise and unwanted signals reach the PID it will respond with constant unnecessary and unwanted corrections consuming power and wearing out the servos. The Vortex is not the first time I came across it, my very first tail gyro (ds760) had such filtering back in 2004.

What I am trying to point is that the average current (or mAh per flight if you like) is very dependant on the ability of an FBL system to filter out the unwanted stuff from the rotation rates it needs.

-Angelos

Evening Angelos.

Sorry, I wasn't saying there was a problem as such, just that through the many hours of testing and further development, you were able to minimise the current draw of the Vortex in this case. Then based on your findings with the much loved DS760 (I loved mine) you drew on that experience to help your future product.
I can see that it is through this constant development and testing that allows a designer and manufacturer, such as yourself, to tune fine tune their product to work at it's best.

I appreciate your input to this thread.
.

[maddog7]

im waitting for the 550L dominator then ill have a try

Me too, it think,

[Alfie]

Surely the reference to how the PID loops control the cyclic servo's is negligible in terms of average current consumption!

I carried out some bench test when the JR DS8915's first hit the market, now i dont have the figures to hand, but im sure they can be found under my RUNRYDER posts.... and off the top of my head those servos pulled just short of 1900mah's stalled! @4.8v's

Now tell me that a high end 700 class will stall all 3 cyclic servos, period.. let alone for lets say 50ms ?

Agreed in a worst case scenario a 5 amp constant BEC should be fine, but it's the peak draw thats the issue, and questionable for how long?

In my experience it's the FBL unit's that are the 'pending' issue, as i believe that the connections to the bus are inadequate, in terms of relying on the age old 3/5 amp servo plugs, in some cases only one connection point.

I always try to connect at minimum of two and preferably three power supplies to the FBL bus, Beastx and 3GX is OK, But Spartan is let down by having two pins not common with the power bus, and non assignable connection for the data pod, meaning splicing a power feed via the pod.! (just to get at least two connectors present!)


These FBL units need to have dedicated power input leads/plugs.

Expecting 3 power hungry servos, plus the tail and controller requirements to all pull of a power bus, which in some cases is fed from one (age old servo plug), is by todays standards, 'NOT GOOD ENOUGH)

col.

[maddog7]

IIRC, when one sets up a Vbar, nowhere does the Vbar software ask the user to input the max frequency of the cyclic servos?
It kinda roughly does for the tail servo but even that isn't exact, like the Beastx or Brain and it's not clear if the choice your making is about the centre pulse width, frequency or both?
This makes me think that the Vbar must only drive the servos at a relatively conservative frequency and this is what keeps the current draw down?

Cool to think it can do what it does without the need to drive the servos to their limit?

This is all just speculation on my part though, I'm certain Angelos will know....

Cheers
D

[maddog7]

I have to say after the failure of an MSH Brain, I too am thinking very carefully now about input power.
My HV servo equipped 700e has 2 dedicated inputs for power directly from an A123 6.6v life pack. According to the Beastx peeps the unit is rated for 60amps across it's power system but I did read somewhere (another forum post, so unsubstantiated) that the new Align HV's can momentarily draw up to 25amps, eek!
STD servo connectors are starting to look a bit weak now huh.....

[coolice]

Surely the reference to how the PID loops control the cyclic servo's is negligible in terms of average current consumption!

I carried out some bench test when the JR DS8915's first hit the market, now i dont have the figures to hand, but im sure they can be found under my RUNRYDER posts.... and off the top of my head those servos pulled just short of 1900mah's stalled! @4.8v's

Now tell me that a high end 700 class will stall all 3 cyclic servos, period.. let alone for lets say 50ms ?

Agreed in a worst case scenario a 5 amp constant BEC should be fine, but it's the peak draw thats the issue, and questionable for how long?

In my experience it's the FBL unit's that are the 'pending' issue, as i believe that the connections to the bus are inadequate, in terms of relying on the age old 3/5 amp servo plugs, in some cases only one connection point.

I always try to connect at minimum of two and preferably three power supplies to the FBL bus, Beastx and 3GX is OK, But Spartan is let down by having two pins not common with the power bus, and non assignable connection for the data pod, meaning splicing a power feed via the pod.! (just to get at least two connectors present!)


These FBL units need to have dedicated power input leads/plugs.

Expecting 3 power hungry servos, plus the tail and controller requirements to all pull of a power bus, which in some cases is fed from one (age old servo plug), is by todays standards, 'NOT GOOD ENOUGH)

col.

Evening mate.

This is one reason why I'm a fan of the regulators that sit between your servos and FBL/receiver, as it puts the power demands on a unit designed to deal with them. I'm using the Gryphon Extreme reg on my 700e.
.