Homepage
Forum
Drivers
HMM
MTM
Feed
I’ve decided to expand battery life of my old T21. Battery holds ~40-50min . I need a lot lot more.
In Theory, I would like it to hold 4-6 hours during lectures @ the collage w/o plugging in.
New battery (even extended 4.4ah instead of 3.6 ah) would give me 2, maybe 2.5 hours top.
I bout new battery last year and it’s already half-dead. So I thought, for the same price (roughly) I would buy a good Li-poly battery with higher capacity and hook it up in parallel to the existing one – to greatly improve battery performance.
So I bought off Ebay this pack:
http://cgi.ebay.com/ws/eBayISAPI.dll?Vi ... 0243459258
It consists of 3 Li-poly cells 5Ah each, and very high discharge current. I know that Laptop does not use very high current, but this is still very good. This is because most batteries that don’t hold up charge too good due to high internal resistance and not capacity loss. (I’ve tested it many times) – I read it also @ battery university. For example my T21 battery after charging has (4.18v*3)=12.54v. After unplugging power it immediately drop to 11.62v (which is 3.87v per cell).
This would not happen with high-discharge li-poly cells. Those cells are rated 12C which is 5*12=60Amps. T21 can consume maximum 2-2.5Amps, 1-1.5 on average. So 2 amp load on battery that is designed to hold 60 amps – it is piece of cake. Voltage would drop very insignificantly.
I have connected this new pack and battery in parallel. First, I discovered that T21’s battery was highly out of balance (there is 3 sets in series of 2 parallel cells each , i.e 3p2s) Some had significantly higher charge level than others. So I had to balance them first (by charging manually all of them to 4.2v) This would give me in theory (just in theory 4.4Ah+5Ah =9.4Ah) Now the old pack is nowhere near it’s rated 4.4Ah, but by connecting them In parallel, the old pack has to supply much less current, and thus it can deliver much more capacity, than before.
This is how it goes: pictures have numbers and explain everything by themselves: I will add here some comments.
Pic6….. there are 4 wires coming from the battery so that each of 3 series sets can be monitored separately. I connected all those 4 wires to my cells.
Pic8:….. pay attention how this pack is out of balance. I wrote voltages on cells with the marker.
Pic9…… global + and – have to be very thick to support large currents and not drop any voltage. Rest 2 wires are not that important.
Pic15…. You can notice extra thickness that they add to the laptop. But this is not that much, believe me. Weight is almost the same. Those cells are very light. I decided to position them below dvd-drive because this place isn’t heated compared to beneath sd-ram or hdd. I believe I could put them there, but I just wanted them to feel more comfortable. In theory, I could easily add another 3 such cells and wire them all in parallel. But I hope this will be enough.
---=== Test results, problems and questions====---
First I charged all cells manually by myself (each series pack separately), disabled low and critical battery warnings – I got around 5.5 hours of maximum load (max brightness, H264 movie playing in loop). I did not test the battery under low load since I have no time. Maybe I will test later.
Than the laptop would not charge the pack fully. It would charge it very fast (40min) and would say 100%. After that pack only worked 1-1.5 hours max. I measured the voltage when it showed 100% and it was 11.3v (11.52/3=3.84v per cell) instead of 4.2v. Clearly it did not recognize the new capacity. The only way to charge it – was to charge it manually again.
After that I googled around to find out how to reset fuel gauge, and didn’t find anything for my old t21. The only thing I found – was updated power management and battery maximizer for WinXP. At least it charges now fully (12.51v)
And works for many hours, But the problem is that it still can’t recalibrate the gauge. At the very beginning of discharge it shows 1hour 15 min remaining. After couple of hours it reaches 1-3% (3 min left) and stays there for 2-3-4 hours before shutdown.
So the disadvantage is that I have to guess and approximate how much energy left. And if I forget to send it to stand-by in the right time – I can really loose my work.
Do you guys have any idea how can I solve this? – I would like to have some indication at least 10-20 mins before battery _really_ runs out . If I don’t find any solution – I will try to make some simple circuit with led(s) that will light up when voltage is below some threshold.
Thanks for reading.
If you have any suggestions – you are welcome
mod edit: added pic warning to subject and shortened long link.
before it would not power on on battery, and would die instantly if i unplugged AC.