ESC cutting out on one type of battery?

[Denis Watkins]

Here is an electric control line flight timer from Dens models, that may help the lads decide on the issue

**LINK**

Edited By Denis Watkins on 27/08/2020 07:59:00

[Dickw]

Posted by Ian Whittaker on 25/08/2020 11:06:14:

Hi,

I was at the field yesterday watching some chap with electrical snags. His model would run for approx 4 minutes on a 2100mA, 3 cell pack until the low voltage cut on the ESC kicked in. With a 2900 pack (also 3 cell) it wouldn’t run beyond about 30 seconds before stopping. The battery showed as 80% after. We tried a ground run with a multimeter monitoring the battery voltage and it seemed stable and well above the cut value. This was the case with 2 separate 2900 packs, albeit both from the same manufacturer.
The set up was in an electric C/L model incidentally. There was a device acting in place of the receiver so as to allow him a 30 second delay to energise the model then get to the other end of the lines for flying, and using the low voltage cut of the esc to stop the motor at the end of the flight.

So, back to the original problem!

There are two possibilities that have not been clearly checked out yet as far as I can see.

1/ Dodgy (high resistance) main connectors on the 2 x 2900 packs resulting in excessive volt drop as the heat up under load. It could be the connectors or the soldering. The volt drop across the connectors under load needs to be checked - monitoring battery volts alone won't help.

2/ The bigger cells have a lower internal resistance and are providing more volts and current under load than the 2100 packs, and the ESC doesn't like it. Need to compare the current/voltage when running the 2100 and 2900 packs

Right out of ideas after that.

Dick

[Engine Doctor]

It no good relying on the 80% capacity left according to a checker. Test the the individual cells by running them with the checker connected to the balance lead and see in real time what's happening to each cell. It could be just one cell causing the problem. Just because they are new doesn't mean anything, they could still be faulty and it will only show under load.

Similar situation occur with all types of cells/ batteries. Car batteries can show a healthy voltage when charged but as soon as the starter is pressed they collapse only to rebound and show excellent voltage .The true test is when the battery is working.

[Chris Walby]

I had 3 new lipos with a model and after checking them before initial charging found their charge levels all over the place. Spoke to the supplier and returned them for new ones, they were better, but did not last that long.

My point is that some suppliers might not sell that many lipos of a certain size for a very long time, by the time they do they were really next to useless compared with "brand new ones"

The supplier would keep exchanging them, I just lost the will to drive 35 mile round trip to swap them multiple times.

[Bob Cotsford]

If it's draining a 2200 pack in 3 or 4 minutes then it's drawing an average 40 or so amps, more when it increases demand to accelerate or climb steeply. It sounds like a wattmeter needs to be used to check the max current and to be sure the esc isn't being cooked and on the edge of thermal shutdown. What is the esc, and is it configured for LiPos?

[Ian Whittaker]

The esc is a Robotbirds one. It's rated at 100A so should have plenty of headroom I'd have thought? As far as it being set up for a 3 cell Lipo, surely the fact that the 2200mA pack works fine would indicate it's ok?

Im still confused by it - and having spoken to Lawrence this morning we are no wiser.

Thanks for the help so far people, I'm sure we'll get to the bottom of this...

[Martin Harris - Moderator]

Is there sufficient airflow around the ESC - not forgetting an adequate exit path. Even if it's rated for a true 100A it would assume reasonable clearance and airflow.

Let's hope their electronics are better than my experience with a model and customer service were!

Edited By Martin Harris on 27/08/2020 13:58:47

[PatMc]

Posted by Ian Whittaker on 25/08/2020 23:21:48:

 

for what it's worth his 2100 pack showed as 3 milliohms per cell, his 2900 battery 6 per cell. As a reference my 5200 and 5000 packs are about 2 per cell.

The 2900 should be lower IR than the 2100 (2200?)cells.
As I said earlier the IR should be inversely proportional to the cell capacity, if they have the same C rating.

Overlander only list one 3s 2900, it's 35C, but they list 4 different 2100 ranging from 25C to 80C.
Which 2200 battery is being used ?

This raises another point - 3s 2900 isn't a popular size of battery whereas 3s 2200 is very popular.
The 2900's may be old stock that have deteriorated in storage.
Or they may be being used outside their realistic (true) C rating.

It would be interesting to know what current is being pulled at the operating throttle setting. A 100A ESC seems OTT for 2200 or 2900 batteries.

 

Edited By PatMc on 27/08/2020 14:14:33

[Ian Whittaker]

A couple of interesting points raised..

the esc is in the direct airflow of the prop - it's a profile type model and the speed control is simply attached to the 'fuselage'.

The pack that works well is a 2200, but not Overlander. The two that are causing issues are Overlander - both new, bought for this model at the same time. They are 35C I think (from memory, but certainly no less than 30C).

The esc is 100A purely to provide a bit of headroom and also to future proof himself should he choose a larger model next.

Edited By Ian Whittaker on 27/08/2020 15:03:41

[Lawrence Drennan]

Thanks for everyone input this thread.

The 2900ma Lipos are 35C (77amp), 65 (154 amp) burst, Giant Power- I test ran them today on my e-soarer and they performed OK.

The 2200ma are 35C, 70C burst.

Its not a big deal to move on and use the 2900's for other models, get some more matching 2200ma packs but it would be good to understand whats going on here.

[MattyB]

Posted by PatMc on 27/08/2020 14:11:02:

Posted by Ian Whittaker on 25/08/2020 23:21:48:

for what it's worth his 2100 pack showed as 3 milliohms per cell, his 2900 battery 6 per cell. As a reference my 5200 and 5000 packs are about 2 per cell.

The 2900 should be lower IR than the 2100 (2200?)cells.
As I said earlier the IR should be inversely proportional to the cell capacity, if they have the same C rating.

Overlander only list one 3s 2900, it's 35C, but they list 4 different 2100 ranging from 25C to 80C.
Which 2200 battery is being used ?

This raises another point - 3s 2900 isn't a popular size of battery whereas 3s 2200 is very popular.
The 2900's may be old stock that have deteriorated in storage.
Or they may be being used outside their realistic (true) C rating.

It would be interesting to know what current is being pulled at the operating throttle setting. A 100A ESC seems OTT for 2200 or 2900 batteries.

I agree - everything points to dud packs for me, new or not. That is a very high IR for such a pack - plugging it into the JJ604 Objective Lipoly calculator give and effective max real world draw of 59A (assuming 6 milliohms per cell); that's only ~19C! By contrast the 2100 has a real world max draw of 65A or ~31C. Yet more evidence that the C-ratings on packs are effectively just a marketing tool; they give little indication of what you will really get!

Posted by Lawrence Drennan on 27/08/2020 19:31:01:

The 2900ma Lipos are 35C (77amp), 65 (154 amp) burst, Giant Power- I test ran them today on my e-soarer and they performed OK.

The 2200ma are 35C, 70C burst.

Its not a big deal to move on and use the 2900's for other models, get some more matching 2200ma packs but it would be good to understand what's going on here.

Put them on the wattmeter for a final check under load, but I think the 2900s need to be returned as faulty. At best they appear to have aged and lost a lot of performance on the shelf, at worst (and more likely given what you are seeing) one or more cells is collapsing under load. What current does the e-soarer pull on the meter compared to the CL model?

Edited By MattyB on 27/08/2020 21:51:07

[Lawrence Drennan]

Put them on the wattmeter for a final check under load, but I think the 2900s need to be returned as faulty. At best they appear to have aged and lost a lot of performance on the shelf, at worst (and more likely given what you are seeing) one or more cells is collapsing under load.

Thanks Matty, I agree this seems most likely is down to the battery packs- I'm sending them back to the supplier.

[Peter Christy]

I've got 4 Overlander 2900 3S packs. They are used in pairs to provide 6S for my "Lama" helicopter. Not had any issues with them in nearly two years of operation:

I'm using XT60 connectors.

I still think the most likely explanation is a poor connection somewhere at the battery end - maybe the connector itself, or possibly internally in the pack.

The ONLY way to settle this once and for all is to measure the voltage arriving at the ESC UNDER LOAD. Simultaneously checking the voltage on the balance lead would give an indication as to whether it is the connector or the pack itself at fault.

Until this check is carried out, discussion is just speculation.

--

Pete

[MattyB]

Posted by Peter Christy on 28/08/2020 08:20:50:

...I still think the most likely explanation is a poor connection somewhere at the battery end - maybe the connector itself, or possibly internally in the pack.

The ONLY way to settle this once and for all is to measure the voltage arriving at the ESC UNDER LOAD. Simultaneously checking the voltage on the balance lead would give an indication as to whether it is the connector or the pack itself at fault.

Until this check is carried out, discussion is just speculation.

Yes and no. To prove for absolute certain the 2900 batteries are the issue an on load test would be helpful, but based on the IR readings alone it's clear both packs are faulty/aged and should be returned anyway. I always test the IR on new packs as soon as they arrive to validate they are in the right ballpark, and return any that are clearly faulty before use (has only happened on a couple of occasions, and in both cases the retailer swapped them without question).

[Peter Christy]

Matty: My point is that if you carry out the test as I describe, it will quickly determine whether the problem is in the connector or in the pack itself.

If it is in the connector, it should be a quick and easy job to fix. If it is in the pack, it doesn't matter if its IR or a poor connection between cells, it needs returning.

Without checking both the voltage at the ESC simultaneously with the cell readings at the balancer while under load, it is impossible to guess where the fault lies.

--

Pete

[MattyB]

Posted by Peter Christy on 28/08/2020 13:43:36:

Matty: My point is that if you carry out the test as I describe, it will quickly determine whether the problem is in the connector or in the pack itself.

If it is in the connector, it should be a quick and easy job to fix. If it is in the pack, it doesn't matter if its IR or a poor connection between cells, it needs returning.

Without checking both the voltage at the ESC simultaneously with the cell readings at the balancer while under load, it is impossible to guess where the fault lies.

Yes it should, but my worry would be that even if the test pointed at the connector you could swap that out and still have a battery with high IR, except now you can't return it as you modified it! Batteries are the one thing in modelling where I just return them if they show any indication of being fault; once used or modded they are yours forever whether they work or not...

[Keith Berriman]

A little confused here with the stated milliohms expected on lipo batteries. Just spent last few days setting up a Wot4e (blue and white) model with a battery may be 4 years old that has given reliably service. Coming to final motor check on Wattmeter I could only get 250 watts and dropping quick ( motor rated well above 400 t ) so up the prop an inch and try again. This time even less 200 watt whats going on thinking ESC etc so have a coffee and start again.

So charged up a newish battery from last years spend and away we go , watt meter on and whow just short 500 watts so drop the prop and inch and 400 watts just what was needed for this model.

So reading this thread in depth I picked up on internal resistance etc something I have not gone into so far set the charger up and found the following results 25--65--27 milliohms on duff battery all other batteries same reading all cell varying between 16 to 20 milliohms

So my question am I reading a wrong scale when someone early said this reading should be a lot lower and do I take readings on fully charged cells or storage mode

Many thanks in advance for any responses

[Frank Skilbeck]

The readings on your "Duff" battery are suggesting the middle cell is faulty, try and pull too many amps and the volts collapse and the ESC reduces the power to protect the battery which is why you see it drop off really quickly. Looking at your faulty battery it has a total resistance of around 117 milliohms, so assuming the volts at the ESC are 10v the 200 watts is 20 amps, the voltage drop across the battery at 20 amps is 20 x 0.117 = 2.34v

Also the chargers only measure the resting cell resistance and with a load on it could be much higher, What volts was you watt meter reading under load if it is less than 3.3v per cell (i.e. 9.9v on a 3s) your LVC is kicking in, Another check is to plug in a battery checker into the balance lead and check the individual cell voltages under load.

Edited By Frank Skilbeck on 30/08/2020 16:22:22

[MattyB]

Posted by Keith Berriman on 30/08/2020 14:59:32:

...So my question am I reading a wrong scale when someone early said this reading should be a lot lower and do I take readings on fully charged cells or storage mode

Many thanks in advance for any responses

This article should give you the basics needed to measure and utilise IR for tracking the health of your packs. The most important thing to remember is that you need to measure IR at a consistent temperature and state of charge in order to be able to compare measurements meaningfully.

[Allan Bennett]

If I remember correctly, I read somewhere a long while back that with a 'good' battery you should expect an IR that's 12,000 divided by cell capacity. So, if that's the case, the 2900mAh packs mentioned in this thread should ideally have an IR of about 4 miliohms.

[Keith Berriman]

Many Thanks for the responses and links I can only assume my charger is possibly missing the decimal point in the readings. If I add in the decimal point to all my readings then I am in the ball park expexted readings of others

[MattyB]

Posted by Keith Berriman on 31/08/2020 11:43:52:

Many Thanks for the responses and links I can only assume my charger is possibly missing the decimal point in the readings. If I add in the decimal point to all my readings then I am in the ball park expexted readings of others

What charger is it? Ichargers definitely show the decimal point, but I don’t know about other makes. It should say in the manual though I’d have thought...

[Keith Berriman]

My charger is SKYRC e660 and read all the pages just out of interest I have another charger that can measure internal resistances ( GT Power bt Robotbirds ) it shows the same figures as the Sky unit. ?..?

I thank you all again for your responses and i will just carry on flying , charging , discharge as I have done many times and occasionally check internal resistances if a battery comes in a little warmer than usual