Battery voltage under load- Mtbr.com
Results 1 to 24 of 24
  1. #1
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548

    Battery voltage under load

    I was poking around with a multimeter on my new light last night and noticed something I didn't expect. I'm sure one of the gurus here will have an explanation.

    It's a 2 x Q5 light, with each led running off it's own DX 3256 driver at 900mA (measured), in parallel with a 4.8v battery pack (4x AA NiMH). With charged batteries, the pack measures ~5.1v with no load. When the light is switched on, the voltage at the battery measures 3.95v. The batteries get slightly warm when the light has been on a while.

    Is this normal? Are my batteries crap?

    Thanks.

    JZ
    It's not about speed, it's about lack of control.

  2. #2
    mtbr member
    Reputation:
    Join Date
    Dec 2008
    Posts
    134
    Thats alot of load off a small battery pack,i think you need more voltage if thats a buck regulator.
    A Buckpuck regulator needs about a volt so you could be losing 2v just on the drivers.
    I recon your dragging the ass out of your batteries.
    Tommo.

  3. #3
    mtbr member
    Reputation:
    Join Date
    Nov 2008
    Posts
    131
    Im not a guru, but... Yes, this is fully normal for AA NiMhs at close to 2A load. Some batteries can handle high current better than others. You would probably get a lower voltage drop if you used eneloop batteries.
    The drivers you use go into direct drive when input is too low. But according to test numbers from the DX forum you still get pretty good output.

  4. #4
    mtbr member
    Reputation: ktronik's Avatar
    Join Date
    Dec 2006
    Posts
    424
    I think the same as the other two guys...

    The voltage difference is what I call 'voltage sag'... its a bit too much load for these batts & getting warm is not good while using...a cheap option would be the next size up, the SubC nimh battery size... these are made to handle that load no probs, less sag...better all round...

    K

  5. #5
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,183
    ..I'm no guru but I agree with the other three...not enough battery. I was going to recommend running them with a 4800mah Li-ion 7.2volt battery but after reading the reviews of the driver on the D/X web site it might not be a good idea to run the driver with a voltage source over 6volts...that is if you believe what one of the reviewers said....

    When I try to drive the same Q5s over 6 volts with these boards (ie. 5 or 6 "D" cells alkaline power) with the same set up and heatsink I am getting 1 of 3 or 1 or 4 board burnout rate! As I said, these boards have performed 100% at 6 volts and under and since they are rated for 9 volts, I should expect the same reliability in a 5 or 6 "D" alkaline maglite mod as a 3 or 4 "D" use. Anyone else have a problem at higher voltages? Can anyone suggest a more robust board?
    ...with that in mind ( unless you have extra drivers ) I think I would just recommend two of the 4X Eneloop AA's sets wired in parallel. You could also try 4x 18650's wired in parallel which should work and run even longer.

  6. #6
    mtbr member
    Reputation: yetibetty's Avatar
    Join Date
    Dec 2007
    Posts
    1,781
    JZ,
    Batteries should never get warm while driving a light,so too much is being drawn from them.
    I would get 18650's and a holder from turboferret (I thnk they are almost ready)

  7. #7
    mtbr member
    Reputation:
    Join Date
    Dec 2008
    Posts
    60
    According to this comprehensive test of NiMH batteries on candlepower forum, if you (we, as I am planning to do the same light) use the eneloop batteries everything should be fine. With two 3256 drivers the average draw should be around 0.75A per LED or 1.5A total. Eneloops have no problem handling that current (they can even handle 10A while still keeping ~1V per cell and 1500Ah). With 1.5A the capacity is above 1.8Ah or over an hour of light.
    I plan to make two lights with 2 LEDs each. One light will be a pure spot the other pure flood. Each LED will have it's own switch and each light it's own integrated 4xAA battery pack. If I will need a longer runtime maybe I'll get a 6V SLA battery.
    As my drivers are still on back order (almost a month now) I am currently direct-driving two leds (in parallel) from 4 old NiMH batteries. No resistor required. I only get about 0.7A per led - yes the batteries are that bad. It all depends on the batteries you are using.

  8. #8
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Thanks guys...I knew the collective knowledge here would come through

    So I'm asking too much of the batteries. I figured I was cutting it close with the 4 x AA setup, but thought I'd give it a try in the name of minimal weight.

    The 18650 option seems unworkable, because in parallel they'd be at 3.7v (not enough to keep the drivers in regulation) and in series 7.4v (8.4v fully charged) which is too much for the driver's 6v limit. But I have some 18650s, so I might do a bench test and see what happens. Maybe since the fully charged batteries are at 4.2v, and a parallel pack will have a high mAh capacity, it will stay in regulation long enough for a good runtime? That will give me an excuse to buy some of Turboferret's cool battery holders.

    The Eneloops seem like an answer. And going with Cat-man's suggestion of running eight of them, 4s2p, keeps the voltage at 4.8v and ups the mAh capacity....never a bad thing for the runtime. What about a 4s2p setup with my "standard" NiMH AAs? Will the extra capacity help me, or am I still going to be pushing the limits of the batteries?

    The interesting thing here is that even with this borderline battery setup, I got almost exactly the predicted runtime (as calculated by the on-line runtime calculator)

    JZ
    It's not about speed, it's about lack of control.

  9. #9
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Do the Eneloop batteries require a special charger, or will the one that came with my EverReady MiMHs be OK?

    dsvilco: one of the reasons I'm trying to sort this out is I want to try your little low-battery warning circuit, but not sure where to set the trip voltage with my current setup. Setting it at 4v would mean it would light up as soon as I turn on my light.

    JZ
    It's not about speed, it's about lack of control.

  10. #10
    mtbr member
    Reputation:
    Join Date
    Nov 2008
    Posts
    131
    Quote Originally Posted by JimZinVT
    What about a 4s2p setup with my "standard" NiMH AAs? Will the extra capacity help me, or am I still going to be pushing the limits of the batteries?
    JZ
    You will gain a lot by using 2 of your ordinary in parallel. You may get more than 0,1V higher voltage/cell at 1A compared to 2A. This gives 0,4V in total, which make a huge diference to the led and driver.
    Take a look at this graph. It is made by SilverFox at CPF and posted in this thread along with eneloops and many other batteries for comparison. Maybe you an find your and see how much you would gain with better batteries.

    EDIT: just noticed there were already a link to the thread and the picture, so I deleted the picture.
    Last edited by gillestugan; 01-17-2009 at 10:55 AM.

  11. #11
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Thanks G. I will have a look at that thread and see if I can decipher it :-)

    Tonight I'll try a bench test of a pair of 18650s in parallel, and 8 x AA NiMH 4s2p, and post my results.

    I'm getting quite an education here!

    JZ
    It's not about speed, it's about lack of control.

  12. #12
    mtbr member
    Reputation:
    Join Date
    Dec 2008
    Posts
    60
    I am certainly not an expert but if you go with a 4s2p setup isn't there a much higher danger of destroying a cell by overdischarging if the two packs are not very well balanced? Eneloop batteries have very little difference in the capacity if discharged at 0.75A or 1.5A so maybe it would be better to drain one pack at the time? Of course, I could be completely mistaken.
    You can charge Eneloop batteries with a standard NiMH charger. I have bought my set with a Sanyo slow charger but there is also a bundle with a fast charger (so, apparently both charger types should work fine).
    As for the low battery monitor, I guess the best thing would be to experiment with your actual batteries. Even better would be to put a variable resistor (something like 5kOhm fixed + 2kOhm variable) so you can later fine-tune the circuit without buying and soldering new resistors.
    How did your experiment go?
    PS: When did you order your 3256 drivers and how long did you wait for them?

  13. #13
    mtbr member
    Reputation: Cat-man-do's Avatar
    Join Date
    May 2004
    Posts
    6,183
    Quote Originally Posted by JimZinVT
    Thanks G. I will have a look at that thread and see if I can decipher it :-)

    Tonight I'll try a bench test of a pair of 18650s in parallel, and 8 x AA NiMH 4s2p, and post my results.

    I'm getting quite an education here!

    JZ
    Yes, I was going to suggest that you do that. Keep in mind a fully charged 18650 peaks out at 4.2volts. That's good enough to power a SSC P-7 LED torch for more than an hour with just one 2500mah cell. Since your LED's are in parallel you have one voltage drop for the LED's and one for the driver's. I don't know what the drop is for the driver's though. Hopefully it will work. If it does it should run a little over an hour ( 1.25hr ) for each ( 18650 ) parallel cell. Personally, my guess is that it will work because those drivers are designed to work with the lower voltages that are usually associated with torches. My fingers are crossed.. ...I hope they work. The Li-ion cells are almost always my first choice when it comes to batteries.

  14. #14
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Well I didn't get around to the 2 x 18650 test last night. Started thinking about how to rig up a temporary 2p cell holder, then I got wrapped up in watching the classic car auctions on the Speed channel, and never left the couch after that But I have high hopes for that setup....the runtime calculator predicts almost 2.75 hours based on a 4.2v voltage and 4800mAh capacity (2x2400). We shall see. I'll try tonight, if I can resist the gravitational pull of the couch.

    dsvilco: I like the variable pot idea for dialing in the batt. monitor circuit. I have a few, but they're 10k, so might be a little imprecise. Should be OK for experimental use though, then measure the pot with a meter and match the result with fixed resistors. This light is too small to squeeze it into, but definitely on the two 35mm triples I'm planning.

    I don't know much about battery balance, but I would guess if I avoid over discharging them, and since they'll be recharged individually (rather than as a full pack, maybe it's OK?

    I ordered my DX3256 drivers on 10/31, and they arrived 5 weeks later

    JZ
    It's not about speed, it's about lack of control.

  15. #15
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Just a quick preliminary note on the 2p 18650 test: I'm not sure it was in regulation for long (if at all?), but it did stay close to full brightness for 1.5 hours or more, and even at 2.75 hours was brighter than the single Q5 running off a buckpuck that I was using for comparison. I'm off to work now, but will post more detail tonight.

    JZ
    It's not about speed, it's about lack of control.

  16. #16
    mtbr member
    Reputation: flyxaos's Avatar
    Join Date
    May 2006
    Posts
    246
    I have built 6 lights with that driver, and Q5's. Every one works great, no board failures, all using 2 of the cheap DX pcb 18650's in parallel at 8.4 to 7.2. Hard to beat that combo.

    I built battery holders out of pvc pipe, and remove the batt's to charge them.

  17. #17
    mtbr member
    Reputation:
    Join Date
    Dec 2008
    Posts
    60
    Quote Originally Posted by flyxaos
    I have built 6 lights with that driver, and Q5's. Every one works great, no board failures, all using 2 of the cheap DX pcb 18650's in parallel at 8.4 to 7.2. Hard to beat that combo.
    Than the batteries are connected serial, not parallel? Do the drivers get hot? What's the average draw from the batteries at that voltage?
    According to DX forum and other independent reviews that board shouldn't be able to reliably handle such a high input voltage.

  18. #18
    I like Monkeys
    Reputation: VaughnA's Avatar
    Join Date
    Jan 2004
    Posts
    886
    I've been using 2 SSC P4's with individual 3256 for about a year without a problem with energizer 4 AA Nimh. I haven't noticed any heat problems, I get about an hour with this setup, I carry a second pack for longer rides. No problems with them at all.
    What do I want to be when I grow up.....Dead!

  19. #19
    mtbr member
    Reputation: flyxaos's Avatar
    Join Date
    May 2006
    Posts
    246
    Quote Originally Posted by dsvilko
    Than the batteries are connected serial, not parallel? Do the drivers get hot? What's the average draw from the batteries at that voltage?
    According to DX forum and other independent reviews that board shouldn't be able to reliably handle such a high input voltage.

    Yes, serial for the batteries. I really don't know voltage or draw, other than whatthe math says they should do. I figured it was cheap enough to try it. It works, so I went with it.

  20. #20
    www.hahntronix.com
    Reputation: mhahn@hvc.rr.com's Avatar
    Join Date
    Mar 2007
    Posts
    252
    VaughnA,

    That seems like kind of a short run time. A decent AA NiMh should be able to provide about 3 WattHours. Each P4 should draw a bit less than 3 watts. So with 12 WattHours of batteries and only a 6 Watt load, I'd expect closer to 2 hours of runtime with your setup.

    Could be your batteries are pretty worn out. What happens if you substitute some fresh Alkaline AAs? If you get over 1 and 1/2 hours I'd look at your batteries.

    Or else the 3256 driver is way inefficient.

    Mark

  21. #21
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    The runtime calculator predicts about 1 hour 40 minutes for this setup (using a led current of 900mA, which is what the 2 boards I used tested out at, and 2 x 3.7vF for the leds). That is about what I saw, running my light on 4 X NiMH AAs. But the batteries were getting pretty warm during operation, which apparently is not a good thing. I think the 2s 18650 setup doesn't give enough voltage overhead to stay in regulation, so I'm thinking of going with 8 AAs in a 4s2p arrangement. Or just enjoying the built-in handwarmer feature and sticking with the 4s pack. I like the small light battery pack.

    JZ

    edit: the efficiency of these drivers is supposed to be ~90% when voltage is 6v or less; it declines slightly above that, but not much.
    It's not about speed, it's about lack of control.

  22. #22
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    Quote Originally Posted by JimZinVT
    Just a quick preliminary note on the 2p 18650 test: I'm not sure it was in regulation for long (if at all?), but it did stay close to full brightness for 1.5 hours or more, and even at 2.75 hours was brighter than the single Q5 running off a buckpuck that I was using for comparison. I'm off to work now, but will post more detail tonight.
    OK, so I tested this light (2 x Q5, 2 x 3256 drivers, parallel) on a pair of 18650 batts., parallel. Batteries were fully charged, reading 4.2v before with the light off, and dropping to 4v when switched on. Unfortunately there was no way to measure driver output current, as this is a completed light and I wasn't going to tear it apart in the name of science. I measured battery voltage and current being drawn from the battery, every 15 minutes for 3 hours. Results are on the graph below. (I couldn't figure out how to change the info on the x-axis, so each segment equals 15 minutes.....1=15min., 2=30, 3=45, you get the picture)



    Honestly, I'm not sure what it all means. I don't see an abrupt change in either of the lines where it fell oout of regulation. Voltage and battery draw declined pretty steadily, which I guess would be expected. It may have been in regulation for close to an hour and a half...that's about when battery voltage was approaching the vF of the leds. (I thought I had read somewhere that these drivers need only 0.2v over vF, but I can't find that reference anymore). That was also about the time that I thought maybe I began to see a decrease in brightness. By 2 hours it had dimmed a bit...but still bright...brighter than the Q5 running off a 1A buckpuck through the same optic that I was switching on occasionally as a reference. At 3 hours it was definitely dimmer (close to the same as the reference light). The battery was down to 3.5v and current draw was under 1A.

    I guess this battery arrangement could be an option. And in practice I'd probably be switching off one driver/led on the long slow climbs so useful runtime could be 3 hours or more. But I did just buy a batch of AA Eneloops (actually white-topped Duracells, which are re-labeled Eneloops according to the folks over at CPF). So I'll be testing the light with them, both in the 4p and the 4p2s configurations. I'll let you know how it goes.

    JZ
    It's not about speed, it's about lack of control.

  23. #23
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548
    I really don't understand how image posting works on this forum....there was a graph in the middle of the post above last night when I posted it. Really...I checked! Now it's gone. Whatever. Here it is:
    Attached Images Attached Images
    It's not about speed, it's about lack of control.

  24. #24
    Lets RIDE!
    Reputation: Jim Z in VT's Avatar
    Join Date
    Sep 2007
    Posts
    1,548

    Eneloop

    Or rather "Duraloop".....

    Just a quick update. (no fancy charts this time ) I did a runtime test with 4 X AA Duraloops, 2000mAh. The runtime calculator predicted ~1.25 hours, and that's exactly what I got. Starting voltage 5.68v, dropped to 4.91v when switched on. For one hour the light stayed evenly bright and the voltage dropped steadily to 4.2v, then started to dim and voltage dropped faster in the next 15 min., so I shut it down. The batteries stayed cool.

    I think the AA Eneloops are the answer. I'm going to rig up an 8 pack, 4s2p. 4000mAh should give me 2 hours. Closer to 3 hours the way I'll use it, switching off one led on the slow climbs.

    Thanks for the help guys!

    JZ
    It's not about speed, it's about lack of control.

Members who have read this thread: 0

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

THE SITE

ABOUT MTBR

VISIT US AT

© Copyright 2020 VerticalScope Inc. All rights reserved.