To UBEC or Not to UBEC? That is the question.

[Tim Mackey]

So, I am replacing my PC9, and would like to canvass opinion on the most suitable power supply. The powertrain is a given - it will utilise the motor, ESC, prop ( 14 x 10 ) and battery ( 6s ) from the original, as it flew exceptionally well on it. The fatal crash was of course, caused by failure of a 4mm gold bullet connector /soldered cable joint on the main power leads, and I shall be using 6mm non caged gold plated connectors throughout on the new model - with double checking of all soldered joints for integrity .

I am using the original 2 cell A123 pack for the retracts

So...the question is - radio power supply? I based my servo selection on 6v operation, and the 6V 5A UBEC from the original seems fine.

TBH, I still feel inclined to use the same system - UBEC hard wired into ESC input leads - but I can just imagine the comments, if I should be so unlucky as to suffer the same fate with this one .

What I would like is 6V UBEC as main supply, and a separate back up battery which would kick in ONLY if the UBEC supply failed. Powerbox systems have confirmed that they do not produce such an item, there redundancy systems being based on using two identical batteries - and both powering at the same time.

[Tim Mackey]

To Quote from just one post above,

"I think if I was convinced to use a seperate battery, I would set the BEC to 6V, and have a 6V battery wired in parallel with blocking diodes, with two diodes on the battery. That way I could use a small AAA receiver battery, and fit it to the model I was flying at the time.

That way the battery would not normally be supplying any power, and would solely be a back up."

 

This would seem to fit my criteria, but what do others think?

How exactly will two diodes on the battery allow isolation, yet immediate back up if required, without interfering witrh normal 6V UBEC operation etc.

If it was this simple and cheap, why are we spending many pounds on complex electronic systems such as the powerbox units ?

[Steve W-O]

Each diode drops .6V, so from the BEC set to 6V - .6V from one diode, you would normally run on 5.4V. The battery with two diodes would have a voltage of 6V - 1.2V which would be your 4.8V. Because this is lower than your BEC output through the diode, little current would be taken from the batteries, unless the BEC failed.

 

There are some drawbacks, but it should be functionable, easily tested to make sure.

 

 

Maybe I haven't seen a major problem, lets see what people say

[James M]

Hi Tim

 

Would this not work with a U-Bec in one port and a battery in the other ??

 

Powerbox 12

 

James

[Phil 9]

I am not sure diodes will work either? do you have a circuit diagram

 

for redundance you need some kind of switching circuit to islolate the battery untill required I can not see how you can achive this with diodes.

 

the simple answer is as described with two batteries both powering at the same time. Or for this model just one 6v battery pack for rx and servo power this seems to work fine for ic models.

 

Timbo why not try the separate battery method this time it is a simple alternative to the UBEC

[Tim Mackey]

Posted by James M on 30/03/2011 16:31:02:

Hi Tim

 

Would this not work with a U-Bec in one port and a battery in the other ??

 

Powerbox 12

 

James

 

Nope...Powerbox have confirmed to me by email that they produce no systems that will accommodate a UBEC and a battery. I am assuming that this is because they so called redundancy systems are probably just two identical PSUs wired in parallel, and draining simultaneously. I do find it hard to decipher exactly how these so called back up systems actually operate - the PDFs of the manuals do really explain the MO well at all

[James M]

Oh well that would have been a simple solution

 

Can't the make one ??

 

James

[James M]

Hi Tim

 

How about this

 

AR9100

 

James

[Tim Mackey]

Posted by James M on 30/03/2011 16:59:52:

Oh well that would have been a simple solution

 

Can't the make one ??

 

James

They haven't offered to do so

[Tim Mackey]

Posted by James M on 30/03/2011 17:04:40:

Hi Tim

 

How about this

 

AR9100

 

James

Its pretty pricey, and also, I'm again not sure that it really offers more than high current capacity ( which I dont need - and I suspect nor do many purchasers of this ) and two battery inputs which are simply parallel connected internally. No real redundancy as such.

[Dusty]

Tim have you looked at the Jeti Max Bec ?

 

I use these in bigger more valuable models, just another 2p!

[Tim Mackey]

Hi Dusty - thats basically just a linear regulator, and not even a great one either, as its only capable of maximum 16V input ( Im using 6s = 25V ). No redundancy either. Ok so it has soft switch, but a digiswitch from powerbox does too, and its less dosh!

Im more than happy with the UBEC I have as its HV capable , switching ( more efficient ), and 5A rated.

[Peter Beeney]

Tim,

I would consider that it’s all a matter of personal choice, which ever way you go. I’m happy to connect any of these items up to any other, the system will work ok. It’s not until something goes wrong does it become apparent where the weaknesses are. Which are invariably the situations you didn’t foresee to start with but are blindingly clear (sorry!) once it’s happened.

I certainly don’t wish in any way to attempt to contradict Steve W-O in his post about about the diodes but I would tend to call them forward biased rather than blocking. I’m not sure that I would want to actually block any circuits, that might be a bit negative too. (Sorry again).

As Steve has invited comments perhaps we could look a bit closer at the voltages. The UBEC is a steady 6V output. The voltage of a fully charged 5 cell pack might be 6.5V. Less 1.2V, the diode bias, it’s 5.3V. So the UBEC would remain in charge, but if it failed the pack would seamlessly take over. However, when the battery voltage fell to 6V, the receiver voltage would only be 4.8V. So, from then on down, and the pack voltage will fall, you are on your own; in DB’s (Dave Bishop) idiom, “You can get lower than that!” However, in practise, this might not be a big problem anyway.

The drawbacks are that you still have to check, and charge, the battery charge state very regularly and you have to test each system regularly, it might be possible that you wouldn’t know if one side had failed. But again, I’m sure most unlikely.

I have actually done this in a model, after some serious bench testing that is, but in slightly different circumstances, a small battery backing a faulty battery that I was testing. It’s definitely one option.

With regard to the ‘unforeseen situations’, I did mention one-such in a post in this thread, 25/03/2011 14:12:46, the model that crashed. This was definitely a real ‘back-to-front’ occasion; and the pilot did not anticipate it at all!

As far as the expensive backup systems are concerned, I’ve never really seen the point of these anyway. And, as it so happens, a while back a friend who flys large scale planes had a problem. He’d done quite a lot of flying at a Scale meeting and was then asked for one more for the TV cameras. It crashed! He had two batteries, 5 cell, connected via a backup box. He thought it was the packs, but they tested ok, so we checked some voltage drops across the whole system, using his packs after discharging them down by about two thirds. What we found was enough to convince him to throw out the box, just use one battery and change it for the other one half way through the flying session.
I think these things are required in some large models but I wonder just how much testing has actually been done under working conditions? I have a feeling that a genuine backup system designed for equipment that mustn’t fail might not work exactly like these. I’m sure the kosher item will perform in all fault conditions.

It probably all depends on how great or otherwise you think the chances of a failure are, as to how important some consideration is given to a separate rx supply.

PB

 

[Tim Mackey]

Thanks Peter, your kinda confirming what I suspected, and that is that there aint really a commercial "proper" failsafe redundancy system available to us. Most f these so called systems are really nothing more than a fancy way of connecting two batteries together, both feeding the system by a lesser or greater degree. Your point about checking and monitoring both constantly, and charging the battery, as well as the danger of that battery with diode actually operating in flight without even knowing it - possibly to low a voltage etc etc, makes me think the following

Whatever system is used, its introducing further potential points of failure. In most cases simple ( but good ) is best. I have used the UBEC / ESC-BEC method on almost every one of my models over the years, and have yet to prove conclusively that the method has ever actually caused a model to crash. The PC9 crash was not actually due to a failure of the "method" used, but a poor solder joint - which could occur at any point in the chain. Even a simple battery supply could be considered potentially more complex than a UBEC ( a cell could fail, a link could fail, a wire could fail, the necessary switch could fail, I could easily forget to recharge it, etc).

Unless a convincing argument sways me, I am sticking with my original preferred methods, in the same order of preference as I originally stated in the threads linked to earlier.

And - apart from using 6mm connectors, and triple checking my soldering - the new PC9 will have the exact same radio power system.

[Tom Wright 2]

I recon that will represent a system with due diligence well done. Mind you a hv 7a ubec is so cheap to buy these days the added margins could increase reliability just a little bit more, best of luck with the new build Tim ,and there is one consolation as its always easier second time around.

 

TW2.

[Steve W-O]

Posted by Peter Beeney on 30/03/2011 19:34:50:



I certainly don’t wish in any way to attempt to contradict Steve W-O in his post about about the diodes but I would tend to call them forward biased rather than blocking. I’m not sure that I would want to actually block any circuits, that might be a bit negative too. (Sorry again).

.

PB

 

 

 

 

Yes, it relies on the forward bias to differentiate between the two sources, but they block the dead circuit in a failure, and reverse polarity, which in this use is unlikely to be a problem.I don't think the voltage would be any more of a problem than a normal 4.8V pack, as a 6V pack with two diodes is exactly the same, 4.8V and dropping on discharge. (I have intentionally used the nominal voltages, the fully charged voltage is only relevant here when considering the maximum permitted voltage)

 

 

It would be very simple to make a microprocessor designed circuit, to switch between the two inputs, but how would you power that? Another small battery? Or diodes between the two sources again?

 

As for forgetting to charge batteries, that is always a possibility, I have seen people take of with an IC powered plane, lose control, then remember they haven't charged the battery. Have never understood how people can not check all their batteries the night before, and an LED indicator is only about £3  (they are quick enough to respond to a low pack when you waggle the sticks to apply some servo load)

 

But forgetting to charge batteries is just pilot error, you can't eliminate pilot error entirely unless you automate it, but any measures to avoid "aircraft failure" must reduce the overall possibility of an accident

 

No matter how many precautions you take, there will always be the possibility of an unpredicted area of failure, just need to look at all the effort put into full size, or the space programme to see that, the only 100% way to ensure no crash is not to fly it

Edited By Steve W-O on 31/03/2011 01:26:33

[Steve W-O]

Here is a suggested circuit to consider.

 

The parts to the right of the letter "A" on the supply rail may be optional, but i would fit them, though it will work without. Total cost without the optional parts about 75p.

 

 

I think I would be quite blunt and say if you don't know what it is about, don't do it.

Obviously the well being of your model depends on what you do with it and how you do it.

 

 

 

 

 

I can't see the picture, so I hope it looks OK. (all pictures are filtered out by our system)

[Chris Bott - Moderator]

Steve that's a pretty simple system, I've often wondered about doing just that but with shottky diodes. and a 4.8V pack.

 

What do you think about adding a trickle charge circuit from the regulator to the battery? i.e. a properly sized resistor that wouldn't allow much drain from the battery if the regulator happened to go S/C

 

The one worry I'd have about this system would be an unnoticed regulator fail. Would I keep flying all day blissfully unaware that my battery was being used/

[Steve W-O]

I( personally wouldn't worry about a trickle charge from the main battery, the backup battery would only be used if the main supply failed, and using low self discharge NiMh recharging before each session, it will give hours of flying time.

 

Like I mentioned before, these battery level indicators are only a few pounds, put one across the output of the UBEC before the diode, and you know it is working. A rougher indication could be a single LED in the same place, as usually the UBEC will work OK or not work at all. It would be simple to make a warning system, but as the backup batteries would give a full days flying, hardly worth it.

 

I would hope that a person would check the UBEC before the days session, by leaving the backup battery out until the main supply was confirmed.

 

 

 

Edit. Sorry didn't comment on the idea of using Schottky diodes. The faster switching time may or may not help, but the lower forward drop would actually decrease the effectiveness of the circuit, as the voltage drop across one diode and two diodes ensured that the backup battery only supplies current if the main source fails. I think the narrower range with Schottky dioes would result in the backup battery being used when there was a high drain on the main supply lowering its voltage slightly.

Edited By Steve W-O on 31/03/2011 11:17:57

[Peter Beeney]

Steve,

There might be an even simpler system, just plug the rx pack, minus the UBEC and via a switch, of course, straight into the rx! That works too! Unless an emergency back up system can be seen to work properly, under any fault conditions, I’m not sure how worthwhile it is.
As Tim said, - Even a simple battery supply could be considered potentially more complex than a UBEC - but I would have to consider the the UBEC always needs a battery supply of some sort, (I’ve always wondered about the odd name, BEC?) so maybe you might consider it’s just another simple battery supply with the added complication of the UBEC? That being the case. then you might also consider that this same circuit could be open to the same problems, - ( a cell could fail, a link could fail, a wire could fail, the necessary switch could fail, I could easily forget to recharge it, etc). -
In other words, I don’t think I can consider the UBEC in isolation, it’s always going to be part of a complete circuit, the failure of one segment will result in a complete breakdown.

To try and figure out what might happen here, I’m actually just doing a little experiment with some battery voltages. One problem is the discrepancy between quoting nominal voltages and seeing the actual on line voltages. So far it’s not looking very good.
One point, is the UBEC limited to being adjusted to 6V, or is that variable? From a quick Google, I suspect it’s only 5 or 6 volts. So I will assume a 6V UBEC output.

Hopefully back later with some working figures. Might take a while, slow discharge.

PB

 

 

Edited By Peter Beeney on 31/03/2011 12:39:09

[Steve W-O]

Sorry Peter, I'm a bit lost, I think those points have already been commented on.

 

I'll just answer the question about UBEC voltages, the ones I have are switchable between the two voltages, not variable

 

Lokk forward to hearing the results of your tests (and of course about the tests themselves)

[Tim Mackey]

I have been out flying most of the day, so will have a look through the latest posts soon.

[Peter Beeney]

Just my very few facts and figures concerning UBEC’s, Diodes and Batteries. Rightly or wrongly. All I’ve done is charged and discharged a pack, noting the levels at which various voltage points are reached.

Firstly, the lump. - A home-made pack, in June 2005, 5 cell, Uniross 1000mAh AAA size Ni-MH. It’s capacity at the moment, circa 900mAh. Used regularly in a sport model, 6lb weight, Irvine 53 power. I gave it a fixed top-up charge, at 400mA.

When I checked the fully charged open circuit voltage it was 7.385V. I started a discharge at 100mA to simulate a slow, in the air, type of battery usage. The battery voltage reached exactly 6.600V after 129mAh had been discharged.

Turning now to the nicely drawn little circuit diagram; my guess is that at the junction of the UBEC output and the rx battery, with the main pack connected but without the rx battery in circuit the voltage will be 6 - 0.6 = 5.4V. If the rx pack is now connected, and if the voltage is higher than 5.4V, it will supply the current. The UBEC will remain shutdown. Thus the fully charged battery will feed the rx demand until the voltage drop is equal to the UBEC output of 5.4V. With 2 diodes connected in the output this will be 5.4 + 1.2 = 6.6V. This is reached after 129mAh has been discharged. So there is still plenty left in reserve, certainly better than I first thought it would be.
However, I can say I have experimented with this stuff in the past, batteries and regulators connected together, and with an ammeter in each circuit you can clearly see this happening, the regulator gradually taking the load as the battery voltage falls. But due to the slight voltage swings on the regulator as the load changes the battery continues to discharge slightly on sudden demand, so it may go down a bit further, say to perhaps 6.5 or 6.4V, or even lower. I think it would really require a full-blown test to correctly establish the exact figures.

Regarding the points about the various techniques here, perhaps I might put my views a touch more clearly. This is totally without prejudice, and is purely my opinion, based on my observations and personal experience.
I would much prefer not to use the UBEC driven from the main power supply system, one reason is because I would consider it to be the most unreliable, perhaps by some measure. I’m sure UBECs themselves are inherently very dependable, but that’s totally irrelevant if any other part of the system fails. The backup battery may be one answer, but perhaps it’s also not without it’s problems, and looks to be maintenance intensive. I’m convinced that a separate rx battery has a much better track record, albeit there are caveats attached. Nevertheless, again these would apply to any modellers battery systems, not least the one that says that modellers are very reluctant to generally apply any TLC to their battery packs. In my experience anyway. How does that square with all the little intricacies of the backup pack? Or perhaps it’s just that I can’t see the justification for trying to fiddle about with what I think may be a bit of a tricky area when I have an already proven alternative to use.

But, I guess, as always this is an individual choice, it seems many pilots simply use a UBEC. As I’ve experienced two main battery failures, but with the luxury of a simple separate rx battery to rescue the situation I’m just biased anyway.

Just my usual upside-down way of looking at it, utterly insignificant!

PB

[Chris Bott - Moderator]

Peter I agree it's good to come at it from your other angle.

I wonder what it would be like with a 4.8V battery, a 6V UBEC and shottky diodes (0.2V drop).

 

I can't help still preferring a UBEC to a battery though. That way I have a brand new fully charged supply to the radio every flight.

I've yet to have any power battery failures in flight and I can't see flight batteries being any less reliable than a separate Rx battery. We do look after them well, making sure every cell is charged properly unlike Rx batteries.

I use A123 flight batteries almost exclusively and would say that they are probably the most reliable of all.

Even when a flight battery goes low, the load is taken off it and it then bounces back with plenty for the radio for a deadstick landing.

[Peter Beeney]

Chris,

Using your figures, I’d guesstimate that the UBEC voltage would be 5.8V, and the fully charged 4 cell would be 5.908 - 0.4 = 5.508V. So it would remain fully charged. A bit theoretical, but I can’t think that’s far adrift.

If you prefer the UBEC, I’m sure that’s absolutely fine! As I said, in terms of reliability, probably unbreakable. I just consider it a part of the whole system though, and that’s maybe not quite so good. There have been four complete motor power failures just in this thread, how many more lurking about out there unsung? I have seen a number of ESC failures, which includes the BEC, one was fairly visible, as I related earlier.
I do agree with you, I’m sure the A123s are brilliant as a power source, you are most unlikely to get any trouble from them.

In my experience, larger size emergency and back-up systems are rarely called on to work in anger; occasionally never; but when they are it always seems to be when you least expect it. Or it was when I was doing it for real. I’m sure most electric model pilots will never have any cause for concern, like yourself, they have a formula and routine that works. And stays within it’s working budget limitations, it doesn’t need to get seriously overloaded, one possible cause of some of the breakdowns!

My underlying concern always leans toward the Safety angle, though, and now after more than thirty years of poggering over it, I’ve only manage to conclude that it’s nowhere as simple a subject as you would think it would be! It’s certainly not always that straight forward. But again, that’s another story.

Good Luck!

PB

Edited By Peter Beeney on 31/03/2011 22:09:21