WOT4 Becomes Racing Car

[Geoff S]

I think I had a similar problem on my Riot now I come to think of it a year or so ago. I checked the charge on the 3S2200 LiPo before I fitted it and it showed fully charged (ie 95%+ I suppose) but it seemed very sluggish on my attempted take-off. When I re-checked the battery it was still showing fully charged but a fully charged 2S rather than a 3S.

I just forgot about it until your post reminded me, BEB. You live and learn

Geoff

[The Wright Stuff]

Posted by Biggles' Elder Brother - Moderator on 06/03/2018 15:12:42:

But I put it on the battery checker,....98% - as I thought. Hang on! 13.6v? That's very low for a fully charged 4s - should be more like 16v+. So I check the cell voltages - two at 4.16V (fine), two at 3.0V! What! Looks like two cells have gone down on the battery.

Which battery checker was it, BEB? Measured using the balance lead?

I'm a little confused as to how it could read 98% if it 'knew' there were 4 cells and not 3?

[Chris Walby]

The feature (not a problem as its outside the design of the ESC) is that the resting voltage when the ESC is powered up thinks its a 4S, but as the voltage collapses under load so the ESC backs off thinking the lipo is discharge. It all becomes self limiting (been there had that, but mine had just taken off!).

The twist is that if you stick the lipo on a resistive load you can get full current all be it at a reduced voltage the alternative is the cell goes open circuit. The difference will be the RX will loose power and you won't have any control (but then as the voltage recovers the failsafe activates) so it just dead sticks in.

There is no real way of testing lipos without a big (as in high power) resistor and logging individual cell voltages and even then there is a view that stressing lipos does them not long term good!

[Ron Gray]

But isn't the correct time to check a battery's state just before you fly, both by % but also cell by cell?

[Andy48]

Posted by Biggles' Elder Brother - Moderator on 06/03/2018 15:12:42:

So what can we learn from this? When I was working with the UAVs we logged every battery into and out of the drone recording individual cell voltages. I'm not so diligent as a retired hobby flyer! But it does show that just taking a few seconds to flip through the cell voltages at the end of charging, even if you don't record them, is a very good idea!

The answer is to plug in the lipo voltage sensor to each model when you fly it, and set up a warning in OpenTX when the lowest cell drops below 3.4v (and critical battery level set at 3.2v). That way you would have instantly got a low battery warning when you opened the throttle. You were perhaps very lucky that the 2 failed cells were as bad as they were. Had they had a little more charge in them, you might have been able to take off then lost power whilst still low down.

I have every model with a SPort lead permanently plugged into its receiver, and then swap voltage and current sensors from model to model as I fly them.

[Andy48]

Posted by Chris Walby on 06/03/2018 16:28:19:

There is no real way of testing lipos without a big (as in high power) resistor and logging individual cell voltages and even then there is a view that stressing lipos does them not long term good!

Yes there is. BEB uses a Taranis, and the current and voltage sensors available for OpenTX will give you precisely these readings.

[Biggles' Elder Brother - Moderator]

OK to try to answer TWS's question - ie how can the checker show 98% with two duff cells - here's how:

First, here's te battery, a 4000mAh 4s Zippy;

Next up, here's the checker I used - a pretty standard item I think:

Finally, here's the battery's balance lead connected to the checker:

Note, 5 wires on the balance lead - as you would expect of course - and the checker tells us the battery is now 95% charged with a voltage of 12.43v. It also tells us (in very small letters at the top centre) that the battery is a 3s!! So, despite a 4s being connected, the checker says 3s - and of course 12.43v is a healthy charged voltage for a 3s - so its 95%.

OK, I should of course have seen the "3s", and the 12.43v should have rang alarm bells (as it did this morning). But I'll be honest - sometimes, you're on auto, you plug the battery in, it says 95%, you think good-oh, and off you jolly well go. I shall be less trusting in future! This is feature of battery checkers we need to be aware of.

BEB

[Biggles' Elder Brother - Moderator]

Andy, yes of course I could use a voltage sensor etc. with my Taranis and I do in my more complex models. But this is winter hack. I fly it on grey cold days and really I just want it as simple as possible.

I have more than enough complex models. Sometimes you just want to plug in and fly! I think I just need to remember to battery check cell voltages pre-flight - not just the percentage charged - to protect myself. It's true that might not give me a very accurate reading, but it will be good enough to spot any major problems - and that's all I need to know.

One or two of you might have noticedI that have a tendancy to over-engineer and over think solutions most of the time! Noooo, surely not, perish the thought I hear you say! But just now and then I can manage technologically minimalist as well!

BEB

[Devcon1]

Its good to know the cause in the end.

I have a FrSky LiPo sensor installed in all my models, for the relative low cost and only requiring a second connection to be made whilst installing the battery it works for me. I notice FrSky have released a new micro LiPo sensor which means it will tuck away very neatly.

What was interesting BEB was looking back at your description of the event didn't make low power an obvious culprit but hard evidence proved otherwise.

Just shows we need to keep an open mind when fault finding..

(Like the time I convinced myself a new FPV install which lacked picture on my goggles was down to to a failed Video Transmitter when, after swapping in another one I realised I'd not taken the lens cap off the camera... Doohhh)

[Biggles' Elder Brother - Moderator]

No Devcon, good point. We can often form false impressions in the heat of the moment. The model felt OK on the ground - taxi and initail acceleration seems nomal - but obviously died off dramatically, by which time I suppose my attentions were er,.. "elsewhere" as in "Why won't thisdamn thing rotate???!!!"

BEB

[kc]

BEB should take heart that none of the other forum experts predicted the cause either!

What did Sherlock Holmes say about when you have eliminated everything else .......

[The Wright Stuff]

Posted by Biggles' Elder Brother - Moderator on 06/03/2018 17:28:43:

Note, 5 wires on the balance lead - as you would expect of course - and the checker tells us the battery is now 95% charged with a voltage of 12.43v. It also tells us (in very small letters at the top centre) that the battery is a 3s!! So, despite a 4s being connected, the checker says 3s - and of course 12.43v is a healthy charged voltage for a 3s - so its 95%.

OK, I should of course have seen the "3s", and the 12.43v should have rang alarm bells (as it did this morning). But I'll be honest - sometimes, you're on auto, you plug the battery in, it says 95%, you think good-oh, and off you jolly well go. I shall be less trusting in future! This is feature of battery checkers we need to be aware of.

Thanks for that BEB, I think it is a point well worth making, and the photos add emphasis that is only a good thing.

At the risk of going off topic, I am still a little perplexed to why the checker (and I have the same model) is so easily confused. After all, the total voltage is measured between the 1st and 5th pins on the JST connector, not between pins 1 and 4. It would be more understandable if the end cell was completely open circuit, because then the checker wouldn't register any connection at the 5th pin, but then it would be summing only the first 3 cell voltages, and not 4.

Perhaps I am overthinking this, but why go to the effort of providing the multiple JST connections in hardware, but then not recognising the significance in firmware? Am I missing something?

[J Moyler]

Well done for finding the cause. There is nothing worse that not finding the answer.

You did the troubleshooting well, I always like reading your posts because I often learn things.

[J D 8]

One small snag of electric power I suppose, An electric motor at 70% sounds much the same as one going flat out, where as with an IC engine ones ears can let you something is not right.