Is this accepable..

[pgkevet]

.as an explanation for my crash?

I hate not having an answer to why stuff happens so bear with this longish post.

The crash was a simple flip to inverted and then heli plummetted to the ground on it's head. I can find nothing on the carcase to explain that and the only two things that make sense to me are that:
a) I gave the wrong input despite remaining convinced that I didn't..
b) I gave the correct input but didn't get a response. Usually with glitches and brownouts and all the usual excuses one would expect some sort of erratic behaviour or a failsafe response.

Lets go back a year (said it was longish)..
I had a predator gasser that decided to go walkies on it's own. I was happily flying a simple circuit and at the end of a long run it refused to turn on rudder and kept straight. I banked and it kept straight banked. Ineplicably it decided to turn itself 180 befre going out of sight and came back. I toggled throttle cut and my remte earthing killswitch with no response until heli has flown past me 100yds when it finally decided to cut it;s engine and piro down an auto. At the time i flet that it had to be an onboard power issue on such a heavy bird with it's servos etc and even though i found no sign of power problems I ended up modding it to totally dual systems etc.

It used to have an align reg before which despite my extensive bench testing never showed a problem.

Of course after a year in my parts boxes THAT is the regulator I shoved on the Blitz since the onboard glow is handy.

[B]SO [/B]If it was capable of a transitory power issue that allowed heli to fly straight before could it have been responsible for another freak situation where a power drop held cyclic but collapsed collective immediately after the power usage of a flip??? Power drop would also bock failsafe throttle cut..and even if it didn't there is a 3 sec delay..it hit the ground before 3 secs..

Either way it's going in the bin unless someone wants a free align reg?

(and I fully accept that one would normally expect power failures to give some erratic effect but this is on futaba 2.4)

[stumpe]

Might be daft, but was you in idle up?

Lee

[pchristy]

Putting my engineer's hat on, I have to say that I never use any form of regulation - nor will I!

In recent years, every complaint I've dealt with of mysterious radio issues has been traced to regulators, dual batteries, or some other form of complicating the power supply!

Modern digital servos - especially when you have 5 of them - can draw an awful lot of surge current, and whilst the regulator may well be able to cope with a nice steady drain, it may not be so happy with spikes in the current demand!

Basically there are two types of regulator - linear and switch-mode. Linear regulators dissipate the "unwanted" energy as heat, and are very inefficient in our applications. They work best when used for a specific - and constant - design load, and not when the load is varying wildly.

Switch mode regulators are MUCH more efficient, but generate copious amounts of RF energy. Unless VERY carefully shielded, this is bound to have an impact on signal reception.

Even dual battery systems can cause weird and unexpected issues. Some years back, I supplied a 459 MHz system to a guy for a large scale model with twin spark ignition engines. Because I knew where it was destined, I checked that radio over EXTREMELY carefully. You can imagine my surprise when he turned up a couple of days later complaining of radio issues!

It turned out he had a "failsafe" twin battery system! Eliminating this, and reverting to a single nicad pack resolved all the issues instantly!

More recently a fellow club member, and competent 3D pilot, started complaining of glitching. Again, eliminating the lipo/regulator set up and reverting to nicads completely cured the problems!

Now I know there are people out there who use regulators without problems. If you do, good luck to you! But if you have any odd radio problems, the FIRST thing to do is eliminate the regulator, and go back to a simple 4 cell nicad system. This will usually provide a complete and instant cure!

If you think you need more torque or speed than 4-cells can provide, then buy better servos!

This problem will not go away until the manufacturers recognise that they need to standardise on higher voltages for the servos (receivers are rarely the problem). Once we can get everything running on a 2-cell lipo setup WITHOUT the need for a regulator, the problem will go away!

At the moment we are in a transitional phase, and if you have any strange and inexplicable radio issues, DITCH THE REGULATOR!

[pgkevet]

(Definately in idle up and wasn't first flip of the flight just first held inverted)

[quote=pchristy;539076]Putting my engineer's hat on, I have to say that I never use any form of regulation - nor will I!.......
.....
At the moment we are in a transitional phase, and if you have any strange and inexplicable radio issues, DITCH THE REGULATOR![/quote]

Can I pick your brains then? On my current gasser I did a lot of playing about with power options. I discovered that a single subC Nihm pack would still throw a ripple of lights on a cheapo battery indicator when i stirred the sticks on the bench. That heli has an onboard generator as well and I found that my best combination was dual AA nihm into 2 different rx sockets and the generator into a third. I also had to eliminate all junctions and Y leads before I could get the battery display to hold to only 1 led drop on stick stirring.

On a gasser weight isn't such an issue (it weighs so much anyway) but I'd like to keep weight more reasonable here and a decent number of flights on a charge.

So 4 or 5 cells (with stepdown to tail)? Nihms or Nicads? Usable lifespan of packs? Capacity for 4xbls541's 1xbls451? Brand or type of cell?

[Andy from Sandy]

The capacity of the pack will be goverened by the amount of weight required to get the CoG correct.

Sub C cells probably offer the best in low internal resistance and power output (think C rating).

Sanyo cells are very good.

[Nick B]

My 450 lost power while inverted a few times last month,it only did it while inverted and found the cause to be the deans conectors.While at the field i waited for the heli to arm and the gyro to settle and then pulled the deans apart by 1mm,didnt have a prob after that and has been ok since:)

Dont know it helps at all in your scenario though:(

[pchristy]

These are my personal views, based upon what I do with my models....!

Best batteries JR or Sanyo 1900 mAH NiCADS, going through a JR Heavy Duty switch harness. (Thats the one with 3 wires going to all the plugs/sockets, and the charge lead seperate, ie not combined with the switch).

Next best choice: Sanyo Eneloop NiMHs.

Do NOT connect NiCADS or NiMHs in parallel! (ie: put two batteries into the receiver on seperate connectors). If you do this, the bad battery will pull down the good one!

Reasoning: The JR and/or Sanyo 1900 cells are physically robust and have a low internal resistance. They are quite capable of delivering the current required by 5 digital servos provided the connectors and switch are in good condition. The JR heavy duty switch has a very good reputation in this application.

Higher capacity "Sub-C" cells tend to have the plates at a much closer spacing, leading to the risk of internal shorts through mis-handling (vibration, impact, etc).

NiMHs generally are less suitable, having a higher internal resistance (unable to deliver as much current without the voltage dropping), a higher tendency to self-discharge, as well as being physically less robust. However, the latest Eneloops seem to go a long way to addressing these concerns. Sanyo cells in general have an excellent reputation!

Whilst it is acceptable to connect LiPOs in parallel, this should never be done with Nicads or Nimhs. Twin battery setups usually have diodes between the packs to prevent one pack from trying to "charge" the other, to the detriment of both. The problem is that these diodes can also limit the current and cause unexpected voltage drops under sudden heavy loads.

I've no experience with on-board generator set-ups, but I would expect these to use the battery pack for surge currents, and to provide a "trickle" charge (though it might be quite a high trickle!) in between.

Nicads seem to thrive on being abused, within reasonable limits! Its too early to tell yet with Nimhs, but I would expect similar general rules to apply.

However, Nimhs do not like being left on continual trickle charge. Whereas you could leave a nicad on a very low current charge almost indefinitely without harm, you mustn't do this with nimhs!

My advice would be to use a single 1900 Nicad (JR or Sanyo) or 2000 Eneloops. I would advise sticking to 4-cells unless there are compelling reasons for using 5. If you use 5, you will still need a regulator somwhere if you are using a high performance tail-rotor servo, and this could still cause problems (though less than using a reg for everything!).

Rather than run a generator, I would simply top the batteries up with a field charger every other flight. This will "excercise" the batteries in a way that they appear to like, without the risk of overcharging or not loading them enough to maintain performance.

Cheers,

[Ade_Law]

i would say that in this case the issue is the regulator. However my experience is the opposite of pete's.

I once spent a long time trying to trace an intermittent momentary lockout which after changing bearings, RX, switches it turned out to be a tab on the nicad that had become unwelded! I have also seen so many nicads and nimhs go "lazy" especially at low temperatures or been left on the shelf for a month or 2.

another point is that in my experience nicads cant supply the peak amps we are seeing. I had one of those inflight battery checkers that showed the minimum voltage. on my dyna-x with a good quality 4cell nicad the in flight i saw sub 4v at the rx. when i switched to the old style fromeco linear regs the in flight minimum rx voltage went up to 5v when set to 5.8v this jumped up again when i made a loom that supplied the cyclic servos directly rather than it going through the rx.

based on this, I dont think its fair to say that regulators cant supply the transient currents.

however i agree that it is about time that everything started running on 2s with a solid state fail "on" switch but i also feel that its about time that they updated the connectors we use too. Some of these servos can pull several amps, I have seen these servo connectors get warm when fast charging rx packs at 2amps. a single servo will push a single connector, using them as battery connectors is just foolish.

Ade

[pchristy]

[quote=Ade_Law;539105]I once spent a long time trying to trace an intermittent momentary lockout which after changing bearings, RX, switches it turned out to be a tab on the nicad that had become unwelded! I have also seen so many nicads and nimhs go "lazy" especially at low temperatures or been left on the shelf for a month or 2.[/quote]

I would agree that physical damage can be a problem - but a regulator has far more connections to break than a nicad pack! And I have seen some truly appalling soldering in mass produced items in my time!

[quote=Ade_Law;539105]another point is that in my experience nicads cant supply the peak amps we are seeing. I had one of those inflight battery checkers that showed the minimum voltage. on my dyna-x with a good quality 4cell nicad the in flight i saw sub 4v at the rx. when i switched to the old style fromeco linear regs the in flight minimum rx voltage went up to 5v when set to 5.8v this jumped up again when i made a loom that supplied the cyclic servos directly rather than it going through the rx.[/quote]

I've never witnessed drops that low myself. I believe a far bigger culprit is the style of connector we use. As you correctly point out, these are marginal even for a single servo, and the battery connector is supplying them all! Spektrum seem to be moving in the right direction with some of their high-voltage receivers, using decent spade terminals for the battery supply!



[quote=Ade_Law;539105]based on this, I dont think its fair to say that regulators cant supply the transient currents. [/quote]

Be fair! I didn't say they couldn't! I just pointed out that linear regs are very inefficient, and switch mode ones are electrically noisy!

Not long ago I was asked to test a cheap Chinese switch mode regulator for an importer. On a bench test, the "Tx aerial removed" range dropped from 8 feet (4-cell nicad) to 1/4" (regulator) (Yes, thats a quarter of an inch!!!)!

I just don't like putting anything between the battery and servos when the load is changing so hugely and so quickly. Its asking for trouble!

[quote=Ade_Law;539105]however i agree that it is about time that everything started running on 2s with a solid state fail "on" switch but i also feel that its about time that they updated the connectors we use too. Some of these servos can pull several amps, I have seen these servo connectors get warm when fast charging rx packs at 2amps. a single servo will push a single connector, using them as battery connectors is just foolish. [/quote]

Agreed! See my earlier comments!

:cheers2:

[pgkevet]

In the case of my gasser with the generator.the generator is fully capable of running the whole system so long as the engine is fired up.. so the batteries act partly as a throttle control for startup and also as a backup for the generator. The reason for two was the voltage drop on cyclic servos on the bench and they get topped up when the generator is active.

I also have a habit of checking each pack independantly (dual switches) to the voltage display before first flight on the gasser ..but really don't want another heavy and fiddly system here with all the extra wiring

A solution here based on guesstimates of past experience is that these servos are going to pull about 350ma per tank but a 1900AA pack (or at least those NihM's I've used before) drop off within 2 tanks which worries me if there was some problem/excessive draw on a flight.

And even JR switches have been known to fail. My testing on futaba and HK heavy duty swtches certainly showed a voltage drop across those switches using NihM's.

I accept the comment re parallel packs pulling each down..but surely if there was a cell down in the single pack it would be just as much a problem??

Dual 4/5 cell 1900 AA's through a single switch (but charged independantly) with a simple stepdown to the tail seems elegant and lighter than 3400 subC's with dual displays on their charge leads (if they are two lead packs)

Does that seem a rational compromise??

[Ade_Law]

nice reply pete, sorry i miss read a bit of your post about the linear reg transients.

quite some time ago i conducted a very simple bench test. With results that scared me. Take a traditional nicad, HD switch, rx, servo system,

Stall a servo, measure the voltage at the battery connector, and at the servo connector. Dont stall the servo for too long you dont want to burn it out! It is scary the voltages you can see with gear that has had just one seasons use.

The trouble is that if you then start adding flybarless systems or cyclocks etc theres an extra length of cable and connectors between the battery and your nice high power servos so the voltage drops even more which is why I made the loom in the attached picture. The servos plug into the 3 connectors on the left, then into rx/whatever so the power goes direct to the major power consumers directly.

Im not sure that the efficiency of the linear regs is an issue. yes they waste some power as heat but its more than made up for from the capacity gain from a lipo weighing the same. Your experience will be different tho as you use smaller nicad than I used to. I used to do 4 flights on a sub C nicad this doubled with the same weight fromeco+lipo and was fast to charge and easier to monitor the safe fly voltage.

Ade

[pchristy]

[quote=pgkevet;539127]In the case of my gasser with the generator.the generator is fully capable of running the whole system so long as the engine is fired up.. so the batteries act partly as a throttle control for startup and also as a backup for the generator. The reason for two was the voltage drop on cyclic servos on the bench and they get topped up when the generator is active.[/quote]

OK, but bear in mind what I said about nicads being happiest when used heavily. Slight use and constant re-charging is a recipe for short life!

[quote=pgkevet;539127]A solution here based on guesstimates of past experience is that these servos are going to pull about 350ma per tank but a 1900AA pack (or at least those NihM's I've used before) drop off within 2 tanks which worries me if there was some problem/excessive draw on a flight.[/quote]

OK, so 2 flights is going to pull 700 mAH out of 1900. Plenty of safety margin there, I would suggest!

And by "topping up" after 2 flights, and measuring how much the pack takes, you are regularly checking for excessive consumption, which should enable you to spot trouble before it happens. Your in flight charging system will give you no such warning!

[quote=pgkevet;539127]And even JR switches have been known to fail. My testing on futaba and HK heavy duty swtches certainly showed a voltage drop across those switches using NihM's.[/quote]

Anything can fail! But my experience has been that the switch JR use in this particular harness is about the best currently available.

I still have a small stock of Japanese "Noble" switches, which were the "best of the best", but sadly, no longer available!

[quote=pgkevet;539127]I accept the comment re parallel packs pulling each down..but surely if there was a cell down in the single pack it would be just as much a problem??[/quote]

Its not quite that simple.

For a start - statistically - with twice as many cells, you are twice as likely to have a failure!

More importantly, the individual cell voltages in Nimhs and nicads is nowhere near as precise as it is in Lipos. With Nicads and nimhs it is quite possible to have differences of several hundred mV between otherwise identical packs with the same levels of charge.

Lipos are self regulating in this respect - Ni*** cells aren't! Trust me - do NOT connect nicads or nimhs in parallel - its a recipe for disaster!

If you must do it, you should put a diode in series with each pack before they are paralleled, so that the lower voltage pack doesn't pull down the higher. But as I said previously, this introduces problems of its own....!

[quote=pgkevet;539127]Dual 4/5 cell 1900 AA's through a single switch (but charged independantly) with a simple stepdown to the tail seems elegant and lighter than 3400 subC's with dual displays on their charge leads (if they are two lead packs)

Does that seem a rational compromise??[/quote]

I prefer Sub-C nicads as they are more robust. The ONLY "AA" cells I have trusted in a helicopter are Eneloops - and then only because they don't do them in sub-c sizes!

[pchristy]

Ade: No problems! This is a complex subject, and not everyone shares my opinions! If you are managing to use regulators without problems, then good luck to you.

BUT - if ever you start having unexplained "funnies" with the radio, this is the first place to look!

I completely agree with the comments you make about the wiring looms and connectors we use. A 3D flybarless heli being flown aggressively is pushing the power supplies we use - be they regulators or batteries - to the very limit! And sometimes beyond!

You may wonder why the manufacturers aren't moving more rapidly to higher voltage systems, which would eliminate many of the problems. Its probably because hard-core 3D useage probably only accounts for a tiny fraction of their overall sales. The vast majority of their customers will be quite happy with 4 "AA" 1100 mAH cells!

And in any case, whatever advances they make in power supplies, it will never be enough!

A few years back, when the FAI raised the contest engine size limit from 10cc to 15cc (60s to 90s), it was done because "then no-one will need to run 30% nitro - they'll have enough power on straight fuel"!

:rofl:

I can see something similar happening here!

[pgkevet]

I'll digest all that and make some sort of decision... Thanks

As a final thought perhaps we should consider Li-io cells? 5-cell Ni** with a nominal voltage of 6 really seem to charge up way higher. How would power/weight reliabiltiy go with 2-cell LiFe.. what do they peak up to and should a servo rated as 5-cell Ni** handle them?

[pchristy]

My only experience with LiFe cells is from some years ago when they were used as an alternative to Lead/Acid as the starter batteries for a standby generator!

The main problem is the peak voltage when the things are fresh off charge. a 4-cell Nicad or Nimh is nominally 4.8v, but its not unusual to see nearly 6 when they are fresh off charge.

And this is why many people feel that performance is dropping off quickly. It isn't! Its just settling to its "nominal" voltage, where it will remain for some time. Nicads and nimhs have a very flat discharge curve!

As I said earlier - we are currently in a state of transition. I would be loath to spend a lot of money on battery technology right at the moment, as I get the feeling that the manufacturers are gearing up for a big switch over to something more suitable!

I have no "inside knowledge" here! Its just a gut instinct!