Battery woes

I know that there have been many posts on the topic of batteries but feel it may be useful to other relatively new teams to share our recent experience. We have 4 pairs which we have run for 2 seasons at a total of 10 events in an F24 & F24+ kit car. We did quite well in our first season but after experimenting with a Nuvinci hub last year we did progressively worse as the season went on with our most disappointing result at the finals. We cooked two electronic speed controllers and reverted to the relay supplied in the kit (This may have been a pivotal moment). Any way, as the new season looms I thought I'd build a BobC discharger - wish I had done this sooner. Not one battery can manage more that 40 minutes and one only managed 11 !!!! So now I have a useful collection of door stops and the prospect of a £500 bill to go racing this year. The moral of this story is......well I'm not sure really. When I used to race a team of karts at least you could see the tyres wearing out ! Hey Ho :(
«1

Comments

  • Next instalment!
    New batteries on the way. Now, shall I "condition" them by cycling a few charges and discharges? If so, at what current? The full 16 amps or shall I re-arrange the resistors to reduce the current to a less aggressive figure like, say, 8 amps. Any advice gratefully received......
  • Yes, cycle them three times. Discharge at 16A if that's the normal current for your test rig, we normally set the minimum voltage at 11V for safety and the absolute minimum is 10.5V

    The really important thing is to make sure the batteries are never left in a fully or partially discharged state for more than a few hours, that is the guaranteed way to ruin them. Get them back on the charger straight after each discharge and ensure they get a full charge.

    Here's a link to one of the posts I wrote last year on the SG1 blog, it might be of some use, note the comment about an initial partial discharge.

    http://greenpowereasybuild.blogspot.co.uk/2016/04/batteries.html

  • The discharge circuit uses a zener to set the cut off voltage and you can't really do much to change that unless you fiddle with the circuit and if you do change it, it makes it difficult to make comparisons with other people who have done the same test.

    http://www.greenpower.co.uk/forum/discussion/comment/25059/#Comment_25059

    We use the standard circuit. See our discharge figures in the link above. I would be interested to see your results Killerwhat from your new batteries for the conditioning. Some people have said it makes a big difference but we have only ever seen a marginal improvement (5 minutes). We also find that the newer design batteries don't seem to have the capacity of older ones, we were getting 2 hours with them on the test rig.

    Also the chargers are quite important, what ones are you using Killerwhat?

    We use CTEK MXS 5.0 chargers, see link for more info
    http://www.greenpower.co.uk/forum/discussion/3462/charging-agm-batteries-nerdy-battery-charger-question/p1
  • Thanks for your replies chaps. We use the CTEK MXS 5.0 chargers too.
    We received our new batteries a couple of days ago and I charged two of them overnight. One has managed 1 hour on the discharger and the other 29 minutes !!
    What's goin' on? I will cycle them to see if this improves but it's not a good start!
  • I would check your tester before you try again.
    Even if you hadn’t charged up the batteries the new ones should give you no less than 70% of the capacity unless they have been sitting around for more than six months, which is usually very unlikely.

    There are two likely causes if it is the tester:
    1. It is drawing the wrong current
    2. It is cutting off at too higher voltage

    1. Check the power resistors are they the correct resistance. If they are ok you are probably fine. You could confirm it by measuring the current. But this is more tricky, most multimeters only go to 10 A and 16 A will blow their internal fuse. A clamp meter is the easiest way of doing this, see if you can borrow one. They are quite inexpensive now and are a useful tool for fault finding a greenpower car. Note though you need one that does dc current, the cheaper ones out there (especially on ebay) only do ac current and will be no use to you. See link,
    http://uk.rs-online.com/web/p/clampmeters/8128910/

    2. Check the voltage cut off by connecting the tester to a power supply. You may have to disconnect the load resistors for this as most bench power supplies won’t be able to supply 16 A. Apply 12 V from the power supply where the battery would be normally connected to the tester, press the start button, the fan and relay should be on. Slowly reduce the voltage until the relay cuts off. This should be around 10.8 V.
  • Was the first discharge a full one or did you do a partial discharge first?
  • Yes, I am beginning to wonder if there is something wrong with the rig too.
    I have a clamp ammeter and it's reading 16 amps so the discharge rate is ok. Maybe the zener is cutting off too soon. Do they vary?
    They were both full discharges.
  • Update: I have checked the current draw and it is indeed 16 amps. I have disconnected the load resistors and checked the cut off voltage and it is 10.8 volts. I ran another discharge on the battery that gave me 1 hour last time and this time I got .......52 minutes !
    What now? I'm at a loss
  • SimSim
    edited March 27
    That does sound like it is the battery that is at fault, everything you have done as proven the tester is fine.

    If the batteries get dropped or knocked in transist that can do it. We killed one falling out a van, it dropped onto tarmac 500mm. Externally it looked fine but the drop caused it cell damage and we had to write it off.
  • Aha!
    I've finally been able to data log the discharge and it turns out its switching off at 12.6v !! The components are all as prescribed so where is the problem likely to be?
  • SimSim
    edited March 29
    All I can think of is; if the cut off voltage at the transistor is 10.8V but the cut off at the battery is 12.6 maybe there is some resistance between the two points. Ensure the current path is nice fat copper all the way from batteries to relay contacts to load resistor.
  • Ok, I'll try that. Thanks for all your patience
  • Hi all,
    We're probably going to build one of these testers too - this thread proves their usefulness!
    However, the idea that one brand new battery had double the capacity of the other
    troubles me. Have you done multiple tests to check that the cutoff voltage is the same each time?
    We're looking to buy four new batteries soon for 24+ so obviously it'd be great to know why the second brand new battery lasted so much less time than the first. We don't really have the money to gamble with batteries at over £60 each!
    Thanks
  • edited March 29
    Here is a pic of the circuit. If you spot anything you think may be wrong please shout !
    https://flic.kr/p/S6KKC9
    https://flic.kr/p/TjB8sA
  • So how do I make the pic visible in the thread then?
  • edited March 29
    Aha, like this !
    photo IMG_0063_zpsh2ndbyr0.jpg
  • I've had an idea. My digital multimeters suggest a voltage drop from one side of the relay to the other. So, if I effectively swap ends and put the zener on the battery side it will respond to battery voltage at the end of the leads rather than voltage on the load side. Does this sound OK ?
  • edited March 30
    Hi Steve,
    got your message, had a little trouble logging in here....
    OK the relay you're using is not the one I used on the original - any volts dropped across this are a worry tbh, also it may need more or different coil current which would also affect the cut off voltage. Would be a shame to butcher your very neat circuit mind you!
    If you decide you're happy with the relay, the next port of call to reduce the cutoff voltage is to get a lower voltage zener. The one in there is 10V and the next "standard" one down would be 8.2V which would probably discharge the batteries more than you'd want to. You should be able to find something in the 9V region in the catalogs (eg RS)
    The way things are going though, I suspect your relay may be degrading and may actually not last too long. Don't move the zener, the circuit won't latch if you hook it to the other side of the relay. Hope that helps
    Bob
  • Looks as if the transistor could have failed. To test it, remove it from the board, then switch your multimeter to "diode". Put one lead on the middle (base) of the transistor. Put the other on the collector and then the emitter. You should see 0.6 V (or thereabouts) in each case. If you don't see 0.6 V then swap the + and - leads and try again. Once you have proven the base-collector and base-emitter junctions are OK, check collector-emitter bothways round and you should get nothing in each case.

    If you find the transistor is OK then get hold of a magnifying glass and very carefully check the board for stray bits of copper or solder shorting out tracks. (It might be easier to do this first!!)
  • Thanks for all your input chaps, much obliged.
    I couldn't get the relay prescribed so I assumed that any 20 amp relay would be Ok.
    I will make more effort and rebuild the rig.
    Thanks again!
    Will keep you posted
    Steve
  • Just one other thought..what is the resistance of the relay coil? It could be that its causing too much current to flow through the transistor and make it fail.
  • BobC can you suggest an alternative relay please, the RS product is discontinued.
  • The relay I'm using:
    https://www.rapidonline.com/panasonic-alf1p12t-20a-pcb-form-a-power-relay-12v-60-5038
    has a coil resistance of 160R, the Omron product has a coil resistance of 155R
  • Your relay should draw around 80mA which the transistor should handle ok. You might find the transistor gets slightly warm, which is also ok. So I would check the transistor (see previous) and look very closely at the tracks.
  • RS Stock No.811-3189 should do it - looks like it's a drop in replacement for the original
  • Thank you both
  • Here's a thought. If we use a Raspi for datalogging why not use it to switch the relay too?
  • Well that's stopped that conversation then !
Sign In or Register to comment.