Kalkoff Battery Problem?

[oriteroom]

Hi all

I was out cycling yesterday and when I started I had 4 lights up on the Panasonic battery and still two on the handlebar control. With quite limited use (my wife and I only really use our power for bad hills or strong headwinds) the handlebar control suddenly went from two lights to a flashing single light without any intermediate stage of a constant single light. (Of course this was so sudden I'd not checked the number of battery lights up - only normally ever do this when we get home and then it's probably only dropped one light). This 'loss of available power' was after about 15 miles of riding with very little use of the power and I feel this was quite abnormal behaviour (we normally get about 180 miles from a fully charged battery given the way we cycle).

I swapped my battery with my wife's bike and it behaved normally, and my battery on her bike acted as it had on mine (suggesting whatever it is it's related to the battery). She's a lot lighter then me so doesn't need as much power to get home!

I noted that the previous (last) time I was also towards the bottom of the charge range it did the same of not going thru' a static single handlebar light phase. When I recharged it last time from a flashing light to full charge it took 8.4 Amp Hr charge which seemed about right from all my earlier readings. I did the press and hold test on the battery and it showed all 5 cells OK. I've noted the number of lights remaining and charge accepted for both mine and my wife's batteries since they were new. My battery is about 21 months old and had only 20 charges (covered 3,300 miles)

This time however I monitored the recharge carefully and each light went out in an orderly way, but finally the battery only took a charge of 7 Amp hr. That's the sort of 'charge acceptance' I've got in the past when the batteries (mine and my wife's) have still had 2 lights remaining on the battery - not a flashing single one. From my battery recharge data from a single flashing light I'd have expected 8.4 to 8.8 Amp hr. I've read all Flecc's great information about battery deterioration, but there seems to have been a sudden and step change in the performance of my battery. The 'press and hold' test' alledges that all the cells are OK!

My concerns are:

Has the battery significantly and suddenly deteriorated indicating an internal fault? and/or,
Shouldn't I still get an orderly indication from the handlebar indicator rather than a sudden jump from 2 lights to a flashing single light?

Any similar experiences or insights would be a help.

Thanks.

Mike

 

[JamesC]

Hi Mike

I had a similar experience with one of my two batteries on the ProConnect.

It tended to occur when riding on rougher surfaces - cobble stones or cinder tracks - and I suspected that it was a momentary break in the circuit between battery and motor control.

The battery support allowed a little bit of looseness between the battery and the bike, which is probably most evident with the solid forks of the standard ProConnect.

This led me to trying a pad underneath the battery, made from one or two thicknesses of inner tube rubber.

The problem has not reoccurred since damping out the battery vibration.

Flecc's writeup on the Panasonic unit on his website describes the mod.

James

 

[Fecn]

It is most definitely the cold. I have been keeping a record of all my battery charges for the last year, and can report that there is apparently a drop of around 30% capacity when the battery gets below about 5C.

I will post some graphs of this later on. I am still in my evidence-collecting phase at the moment, waiting for the weather to warm up again.

The behaviour I have noticed is that when you first take the bike out and everything is warm, it's great.. good as new. After you've been out and about for an hour or more, suddenly and without much warning, you're down from 3 lights to a flashing one (That's what happened to me). Measuring my recharges, I'm down from 2.21Ah-per-led in August, down to 1.54Ah-per-led in the sub-zero conditions.

This behaviour I'm getting now the weather is cold is the exact same behaviour I had last year before I got the battery swapped by 50cycles. As I still have another 6 months on my warranty, I thought I'd wait until the weather warmed up again and see if the battery recovers before seeking a replacement.

The most energy I've managed to cram back into my one-year-old battery since the start of November has been 7.69Ah - By contrast my wife's battery which is now two years old (but not heavily used) still takes 9.53Ah on a full charge.

By all reports, when the weather warms up again, you'll regain your missing battery capacity.

 

[WALKERMAN]

My ProConnect battery was replaced as it was part of the earlier 'faulty' batch. It is now 14 months old and is down to 4 lights.
About a month ago I left it out in a freezing cold wind all day and when I set-off for home it dropped to a flashing light on the controller long before it usually did. The very cold weather certainly did affect it.

 

[Mussels]

By all reports, when the weather warms up again, you'll regain your missing battery capacity.

That's what I have found myself. Last winter the capacity went down and recovered in the spring, this winter it has dipped again and I expect it to recover. Fortunately I have a bigger battery so it's not so noticable.

 

[JamesC]

The behaviour I have noticed is that when you first take the bike out and everything is warm, it's great.. good as new. After you've been out and about for an hour or more, suddenly and without much warning, you're down from 3 lights to a flashing one (That's what happened to me). Measuring my recharges, I'm down from 2.21Ah-per-led in August, down to 1.54Ah-per-led in the sub-zero conditions.

Am I correct in thinking that a sudden reduction from 3 lights to a flashing one translates as the battery voltage sinking suddenly ?

It seems very possible that the battery voltage could sink suddenly, either through low battery temperature or an intermittent poor connection.

When it is a an intermittent bad connection, the battery indicator lights recover sometime later.

I am sure that you are right that the main culprit will be temperature at this time of year. My experience of subzero cycling is limited !

James

 

[oriteroom]

Thanks for your replies so far.

James - the single light is also flashing on the battery itself, not just the handlebar control, so more likely internal battery 'effect' than poor external connection? My wife's battery behaved perfectly on my bike and my battery was still 'troublesome' on hers for the trip home.

Fecn - look forward to seeing your data. I've also been collecting recharge data for both my and my wife's batteries (sadly not right from the beginning but for at least 12 months out of the 20 months I've had them). 8.8 A/hr the most I've packed into them (but from 2 lights on battery not exhaustion). Typically I've been getting recharges at about 2.7 A/h per 'light' but the last two on my battery have dropped to 1.7 A/h/light and yesterday's was down to 1.4 A/h/light. My wife's battery is showing no such effects and the batteries share the same conditions (we always go out on them together, the only difference being that she's 4.5 stone lighter!!). Altho' 'chilly' yesterday it was not particularly cold around the hills of Dorset, and not sub-zero.

Fecn - why was the first battery replaced? because of this effect. Like you I've only a few months left on warranty.

Regards.

Mike

 

[tillson]

I too am experiencing exactly the same battery characteristics as I did last winter. Nearly all of my cycling involves travelling along exactly the same route as I commute to and from work, so I have a constant against which to measure the battery performance.

When the bike was new in June 2007, I could travel to and from work and arrive, just as the second light went out on the handlebars. Occasionally, I would even make it home with just one light extinguished.

As time went on, the first light would go out at a progressively earlier point on my constant unvarying route, and the same would happen with the second light. Eventually, during late November 2008, I had reached the point whereby the last light would be flashing long before I reached home. This was in stark contrast to arriving home with just one light extinguished when the bike was new.

I carried out the 5 second button push capacity check and found that the battery capacity was indicating less than 8 Ah (4 lights). I reported this to 50 Cycles in November 2008 and they exchanged the battery for a new one.

Initially, the new battery performed well. Just like when the bike was new in June. However, by January 2009, I was arriving home again having cycled my "constant" route with just the final light flashing on both battery and handle bars. The 5 Second button push capacity check had also fallen to 4 lights, or less than 8 Ah.

Again I reported this to 50 Cycles in January 2009, and they again replaced my battery. This third battery initially performed like the new one did back in June 2008 and I was initially arriving home with just one handlebar light out or the second just having gone out.

By the end of February 2009, the same had happened. Arriving home with the final light flashing and 4 lights on the capacity check. Following advice on here, I ignored it for a while and just carried on using the bike as normal and it never let me down and the battery never became exhausted.

I continued doing this until about April 2009, well into when the weather had turned warmer. The battery showed no sign of improvement as the weather improved.

I then decided to see just how far the battery would take me. The final light on both battery and handlebar started to flash after about 10 miles. I new it would continue to at least 20 miles because that is how far my commute is. It actually went 36 miles before the battery cut out, so i had covered an additional 26 miles on the final flashing light.

After the battery had cut out and I recharged it and the capacity indication returned to 5 lights. The battery performed like new in that I would arrive home from my commute with just one light extinguished. (This characteristic had been discussed on Pedelecs and was not my discovery). The battery continued to perform flawlessly all through the summer of 2009 with no need to fully discharge.

Since about October, the battery has followed exactly the same pattern as it did last year. The final light starts to flash after just 9 miles and the capacity check is showing just 3 lights.

When the weather turns warmer, I will do a full discharge run again and I expect the battery to return to similar discharge characteristics as it displayed in the summer of 2009.

 

[Fecn]

oriteroom, My battery was replaced for exactly the same reasons as tilson. I was finding that it went down to one flashing light after at little as 10 miles (one time at 6 miles, but that was through some thick thick mud). The capacity indicator went down to 4 out of 5, so I got 50cycles to swap it for a new one (everyone else on the forum seemed to be doing it at the time).

I got my new battery on 1st Feb 2009, and have kept track of every charge it's had since then using a mains watt/hour plug-in meter. My wife's Agattu was second-hand to us, but was originally purchased in May 2008 and given very little use before we had it.

In the plot below, my battery is the round orange blobs, and my wife's is the square black blips. My battery has had a total of 65 charges since 1st Feb 09. My wife's battery has had a mere 17 charges in the same time.

The lowest set of orange dots are all of the most recent winter set of charges. The x axis is the number of LEDs lit on the panasonic battery before it went onto the charger. 0.5 is 'Flashing', 0 is cutout-reached. The y-axis is the calculated Ah absorbed by the battery assuming an 85% charger efficiency (based on former tests). Because the Watt-hour meter I'm using is of limited resolution, absorbed charge readings are all in steps of 0.3Ah (e.g. it's either 9.84Ah, or 10.14Ah... I can't do any reading in-between)



If anyone can think of a better way to show this data, let me know and I'll chart it. It's a shame that this chart doesn't show the outdoor temperature, but I think that the line of low-reading results is quite pronounced anyway.

FWIW, my theory is that as the prismatic cells in the panasonic pack cool from the outside-inwards, the electrolyte material in the centre of the cells becomes discharged more rapidly than the material at the outer edges. This cooling effect changes the cell discharge curve from it's normal shape to one where the voltage drops far more rapidly at the start, but holds a lower voltage for a longer time at the end. The poor battery meter on the unit doesn't really stand a chance of reading accurately is the battery chemistry changes it's behaviour.

 

[Old Timer]

Well.I did some maintenance on my wifes bike on Saturday and did a short ride on a full battery. I didn`t bother to take the battery off the bike and it was in the cold shed for two nights. Just half a mile from home on the next ride I lost one light and nearing the top of a hill close by the second light was flickering. the bike still ran OK and I managed a run and back home. I`m not surprised because I always keep my Li-ion camera batteries in an inside pocket in the winter till needed and the performance is double that of a cold battery.

I always bring our bike batteries indoors, I let them warm before charging and leave them indoors till needed.

I`ve never seen any battery I`ve used that didn`t drop off performance in the cold.

Don`t tell me I`ll have to convert a hot water bottle to wrap around the battery in the cold

 

[tillson]

Fecn's cell cooling theory fits perfectly with the battery cold weather characteristics that I have experienced.

 

[Tim]

It's around this time of year thoughts turn to a range of knitted battery cosies in a variety of colours and patterns.

 

[JamesC]

The lowest set of orange dots are all of the most recent winter set of charges. The x axis is the number of LEDs lit on the panasonic battery before it went onto the charger. 0.5 is 'Flashing', 0 is cutout-reached. The y-axis is the calculated Ah absorbed by the battery assuming an 85% charger efficiency (based on former tests). Because the Watt-hour meter I'm using is of limited resolution, absorbed charge readings are all in steps of 0.3Ah (e.g. it's either 9.84Ah, or 10.14Ah... I can't do any reading in-between)

View attachment 1227

Your chart shows the cold weather performance very well.

Looking at the spread of orange dots on 0.5 (battery used down to 1 flashing light), and then the tight group on 0 (battery used completely to cut-out), we can see clearly that there is considerable capacity remaining in the battery during cold weather when the final light begins to flash.

Popular advice to users is NOT to exhaust the battery down to cutout very frequently; better to treat the range as defined by when the last light starts to flash.

So in Winter, less of the capacity is available and we experience a significant loss of range between full and the flashing light.

In Summer temperatures, the distance available whilst the light is flashing is, say, 4 miles whereas in Winter it might have increased to 10 miles.

If the final light starts to flash early due to low temperature, is it reasonable to use the bike for say a further 5 miles at low battery voltage, or does this place damaging stress on the battery ?

James

 

[Fecn]

My guess is that it is probably quite bad for the battery trying to drain it right the way down when it's cold.... However... there's a cutoff anyway, so it's most likely not a problem.

The picture I have in my mind of what's going on at a chemical level within the battery is something like a half-frozen ice-cube (where the centre remains liquid, but the outer is solid). The liquid bit in the middle of the ice cube can be drained/replaced without affecting the ice around the edges. It's not the best analogy, but it hopefully explains what I think is happening.

We know that LiIon cells can be damaged if the cell voltage falls too low, but I suspect that in cold conditions, it may be possible to over-discharge the ice-cube-liquid-centre, whilst the outer frozen edges and not taking part in the charge/discharge process. If the theory holds out, then it could be possible to damage the inner-liquid bit whilst leaving the frozen outer bits unaffected. This would show up when the weather warmed up as a loss of Amp-hours of capacity.

Of course.. this is only a theory...

In reality... the low-voltage cut-out on the bike should prevent the cells reaching too-low a voltage no matter how many amp-hours of capacity are diminished by the cold.

 

[flecc]

I agree, I'm sure battery won't be damaged by fully discharging to cut-out when capacity is reduced by cold conditions. The safety margin is essentially the missing capacity, in effect the chemistry looking after itself, this being the way virtually all chemical reactions respond with temperature change.

When the temperature rises again, the chemistry responds by returning the additional safe usable capacity.
.

 

[NRG]

There's some good data on cell performance vs temperature on this site including how Ri is affected...

Battery Performance Characteristics - How to specify and test a battery

 

[JamesC]

Thinking back to Mike's (Oriteroom) situation, I guess that his position is actually worsened in Winter by the fact that he is so sparing with the use of the battery !

By the time he needs a bit of assistance, the battery has become very chilly and is less willing to help.

Someone like me who uses assistance throughout a journey does at least get the benefit of a warm spell early on.

In fact, I think that inadvertently I have quite a good solution because I carry a second battery wrapped up warm in the saddle bag.

When the light starts to flash on Battery 1, I switch batteries.

Quite likely, Battery 1 then recovers in the saddle bag and would do another stint - I must check that.

James

 

[oriteroom]

Rain Stopping play but...

Hi all

Been doing some more measurements and testing. Will report yesterday's events but today is on hold as it's raining in Dorset and whilst I'm allowed to get soaked the Pro Connect's not allowed out to play in the rain. Note that ambient temperature was about 4 to 5 degC, not sub-zero.

Had a chat with Tim at 50cycles (Lloyd's gone) and he said to wait and see what the battery was like for the next power cycle and if it behaved normally or went straight from 2 to flashing 1 red light on handlebar control. Bearing in mind we don't usually use a lot of power this might take a month (our typical drain rate is 40mA/hr per mile cycled or 300mA/hr per hour cycled), I decided to put the battery through a test ride, using 1:1 power assist all the time with some fairly hilly bits (especially at the start of the ride). I decided to monitor my fully charged battery to exhaustion, measuring the range and subsequent charge uptake. I plan to do the same with my wife's battery in my bike over the same course to get a comparison, but this is the 'experiment' that's on hold due to rain.


Detailed battery performance on constant power assist. Power usage 2900mA/hr per hour cycled, and 360 mA/hr per mile cycled. (Sorry it's not a pretty table!)
Handle bar light................Battery Lights...........Miles...........Average speed
3 change to 2--------------- 4----------------- 7.1------------- 12.7
2--------------------------- 3---------------- 10.8------------- 12.5
2 (after hr cycling)---------- 3---------------- 12.4------------- 12.4
2 change to 1--------------- 2---------------- 15.3------------- 11.8
Flashing 1 light-------------- 1---------------- 18.1------------- 11.9
Fast flash/cut out----------------------------- 28.7------------- 11.6

At 18 miles I was starting to get disappointed! I thought if there's only <10% left then the battery will die at about 21 miles. I was VERY surprised (but not after reading tilson's post) that I 'squeezed a further 10.5 miles of power assist AFTER the handlebar light started to flash.

Recharge data was also interesting:

Starting charge rate was at 55Watts, peaked at 61Watts when around 3 lights were back, and after 5 lights there is a power down which lasts about an hour from 60 Watts slowly down to a residual level of 3.5W. I've noted that the charger uses 3.5W when no battery is connected so I've always adjusted my recharge figure by 3.5mW/h charge.

The recharge took 5.5 hours to put all the light out (and residual charger current down to 3.5W). I thought the good news was that the battery took a total charge of 270W/hr or a total recharge of 10.4A/hr. As part of my ongoing 'checks' I intend to repeat the test ride again to see if the discharge characteristics are the same (particularly the amount of power left when the handlebar light starts to flash.

As I said I also intend to do a comparison on my wife's batery on my bike.

Fecn - I'm going to see how all my recharge data fits on your graph (will put data in a separate note). We seem to be living in similar 'time dimensions'!. You and your wife both have Kalkoff's, your battery's been recharged more than hers, you measure recharge characteristics as well!!! As an aside, was it you who also bought a Thule towbar bike carrier designed for pedelecs - if so how's it going, I've still the earlier EuroClassic.

Tilson - liked the summary - informative. You must have covered quite some assisted distance on your comutes.

Everyone - What's the assembled wisdom on what mileage a 'decent' Kalkoff battery should assist for on 1:1 assistance. Is my battery looking about right?

As reports used to finish when I wasn't semi-retired - 'Further research in the areas identified in this report is required!' TO BE CONTINUED

MIke

 

[JamesC]

Further to my last post, is there any evidence that the battery condition lights can reverse WITHOUT visiting the charger ?

I don't think that I have ever seen the final light become solid due to the battery temperature increasing at the end of a ride except by commencing to charge.

I suspect that the software might prevent a regression in the sequence.

James

 

[NRG]

Batteries are self heating in use, to an extent... so if use is very low then the internal temp will drop quickly reducing performance, used more the internal temp will rise and help performance. There's a thread on ES about cold battery performance and some ideas on keeping them warm....I've no idea if they work...

Endless-sphere.com &bull; View topic - Lifepo4 batteries and cold temperatures