liveforphysics
100 TW
Do you ride a 250w bike or something? lol
Homemade Battery Packs
- Thread starter DrkAngel
- Start date
The original source of many of the cells in various laptops, toolpacks, etc., is quite likely China, regardless of what company "makes" or sells the packs the cells themselves are in.miniebiker said:
So even making your own battery pack out of cells from other packs doesn't guarantee you're not using those "cheap made in China batteries", unless you specifically use only those cells that you can absolutely trace the origin of back to their original factory, to ensure it is from the place you're wanting them to be from, and the level of quality you want them to be.
miniebiker
100 mW
- Joined
- Aug 5, 2011
- Messages
- 35
amberwolf said:The original source of many of the cells in various laptops, toolpacks, etc., is quite likely China, regardless of what company "makes" or sells the packs the cells themselves are in.miniebiker said:
So even making your own battery pack out of cells from other packs doesn't guarantee you're not using those "cheap made in China batteries", unless you specifically use only those cells that you can absolutely trace the origin of back to their original factory, to ensure it is from the place you're wanting them to be from, and the level of quality you want them to be.
So far all the packs I have seen said country of origin and where they were put together.
The packs all say assembled in China.
Cell origin says either Japan, Korea or China cells.
The ones I have that tested best said made in Korea, are pink and say Samsung on them.
I found a data sheet for them too.
The cheap ones I was talking about are the no name ones LP was suggesting DA get his name on ?
miniebiker
100 mW
- Joined
- Aug 5, 2011
- Messages
- 35
liveforphysics said:
It is important to know how many watts you need when you build a homemade battery pack.
It is also important to read the data sheets from the factory on the batteries being used.
It is also important to test the batteries so they can be matched.
Know what your batteries can do so you can build a pack that will go the distance and not get hot or run out.
miniebiker
100 mW
- Joined
- Aug 5, 2011
- Messages
- 35
DrkAngel said:
My testing so far discharging at 1 amp shows very good performance with little temp rise.
DrkAngel
1 GW
450w EZip - cruising at 18-20mph.liveforphysics said:
25.9v, 31.2ah Li-ion (homemade) pack = 50+ miles per charge, cruising about 3 hours @ 18+ mph, motor only - no pedal assist.
miniebiker
100 mW
- Joined
- Aug 5, 2011
- Messages
- 35
DrkAngel said:
With a 9 lb battery and a 5.5 lb motor. And a bullet proof brushed motor too ! That's what I'm talkin about !
A setup that total would add 25lbs or less to the weight of a bicycle and go 50+ miles on a charge.
http://tncscooters.com/product.php?sku=106118
DrkAngel
1 GW
Homemade packs are working nicely.
I've gone about 1000 miles, using just my 25.9v recycled Lipo, in EZip, pack.
I've been using my Li-ion builds, for the past 3 years, thought I should concentrate on testing the Lipo.
My first Lipo builds were built from the "defective" packs, the ones that would not properly charge in the 4 bay laptop, bulk, charger.
25.9v.
The good news?
I have 200+ "good" packs, that fully charge, and more importantly, maintain 80% charge after sitting for 3 months!
(4 of 5 capacity meter leds)
4-5 kWhs of SONY cells (only tested at 11.1v 21.6ah, with 110v AC inverter, then replacing 2 - 6v cells in 12V florescent lantern.)
7-8 kWhs of unlabeled cells
No where near as cheap as Li-ion, but typically in much better condition, (Higher percentage of 100% usable)
For my ebike packs, I leave the individual cell fuses.
For my 18v B&D (Black & Decker) Power tool pack upgrades, I will bypass the built in fuses.
Of course I will closely monitor, as I apply maximum demand, during durability testing.
I'm guessing the 8" chainsaw should demonstrate the highest demand.
Testing will use an amp meter between the tool and an external pack, (temperature monitored , of course!)
Still shopping for eMotorcycle ... and eCar candidates-components ...
I've gone about 1000 miles, using just my 25.9v recycled Lipo, in EZip, pack.
I've been using my Li-ion builds, for the past 3 years, thought I should concentrate on testing the Lipo.
My first Lipo builds were built from the "defective" packs, the ones that would not properly charge in the 4 bay laptop, bulk, charger.
25.9v.
The good news?
I have 200+ "good" packs, that fully charge, and more importantly, maintain 80% charge after sitting for 3 months!
(4 of 5 capacity meter leds)
4-5 kWhs of SONY cells (only tested at 11.1v 21.6ah, with 110v AC inverter, then replacing 2 - 6v cells in 12V florescent lantern.)
7-8 kWhs of unlabeled cells
No where near as cheap as Li-ion, but typically in much better condition, (Higher percentage of 100% usable)
For my ebike packs, I leave the individual cell fuses.
For my 18v B&D (Black & Decker) Power tool pack upgrades, I will bypass the built in fuses.
Of course I will closely monitor, as I apply maximum demand, during durability testing.
I'm guessing the 8" chainsaw should demonstrate the highest demand.
Testing will use an amp meter between the tool and an external pack, (temperature monitored , of course!)
Still shopping for eMotorcycle ... and eCar candidates-components ...
DrkAngel
1 GW
Most all digital meters will read 100th volt.SamTexas said:
Only way to verify accuracy is comparison against "known" good, or "Standards" certification.
Never had any use for 5.23mV accuracy. (.00532 V) 100,000th V
SamTexas
1 MW
Which one is that? How accurate is it?DrkAngel said:
I'm looking for one that can also accurately measure mV with 2 decimal digits, e.g. 5.23mV.
DrkAngel
1 GW
Cell capacity-condition easy "quick check" ...
After evaluating hundreds & hundreds of cells, I noticed that the best cells were the ones with the "best" bleed down profiles.
Typically I charged all cells to a 4.2v resting voltage.
With a quick charger, the finish voltage must be slightly higher to attain this "resting" voltage.
My standard procedure was to confirm 4.2v, after 1 hour removal from charger.
Then ...
Voltage, of each cell is tested, and recorded at 24 hour intervals.
1 week of testing gives a very good cell evaluation, I use this method to compare cells of same manufacture and capacity.
Initial voltage drop seems to give a good indicator of capacity-condition.
This method seems fairly reliable for Li-ion, as well as for Lipo.
I have used this "yardstick" method to match cells for my larger pack builds, (84 - 120 cell packs), with good success.
Taking this method, to the next level, I intend to do my future cell matching, using a preliminary cell voltage of 4.25v.
While towards the high end of acceptable charge, daily "bleed down", should earmark the very best cells, (The ones closest to brand new condition), indicated by the 1st 24 hour voltage drop.
Typically I try to parallel together a dozen cells, or more, then "gang charge" to specified voltage, before separation and testing.
Oh! It is important to have a digital voltage meter ... capable of measuring 1/100ths volt!
Stay tuned for my 44 Magnum!
The Lipo packs I recycle, are nicely built in 3S, 11.1v, "bricks".
It has been bothersome adding 1S, to build 7S 25.9V and 10S 37V packs.
Fortunately, a 12S 44.4V pack charges nicely to 50.4V.
The 48V controllers, I've been eying, have low voltage cutoffs of 41V & 42V ...
41V / 12S =3.4V
42V / 12S = 3.5V, which is my personally recommended, optimal DOD (Depth Of Discharge)!
Update: Optimal DOD revised to 3.7+V (Li-ion - LiCo 18650 cells)
3.5V per cell, low voltage cut-off, should work nicely for 3.7+V minimum, actual, discharge. ("Voltage sag" under throttle.)
Test pack will use 2160mah cells, in a 12S 10P configuration. (.96kWh)
44.4V - 21.6ah
Which can be easily paired with Twin Sister Pack for 1.92kWh pack.
44.4V - 43.2ah
Yeah ... I stocked up on cells!
Initial charging will be done with multiple iMax B6 as 22.2V sub-packs,
or ... slower ... 22.2V dual charging from single charger.
After evaluating hundreds & hundreds of cells, I noticed that the best cells were the ones with the "best" bleed down profiles.
Typically I charged all cells to a 4.2v resting voltage.
With a quick charger, the finish voltage must be slightly higher to attain this "resting" voltage.
My standard procedure was to confirm 4.2v, after 1 hour removal from charger.
Then ...
Voltage, of each cell is tested, and recorded at 24 hour intervals.
1 week of testing gives a very good cell evaluation, I use this method to compare cells of same manufacture and capacity.
Initial voltage drop seems to give a good indicator of capacity-condition.
This method seems fairly reliable for Li-ion, as well as for Lipo.
I have used this "yardstick" method to match cells for my larger pack builds, (84 - 120 cell packs), with good success.
Taking this method, to the next level, I intend to do my future cell matching, using a preliminary cell voltage of 4.25v.
While towards the high end of acceptable charge, daily "bleed down", should earmark the very best cells, (The ones closest to brand new condition), indicated by the 1st 24 hour voltage drop.
Typically I try to parallel together a dozen cells, or more, then "gang charge" to specified voltage, before separation and testing.
Oh! It is important to have a digital voltage meter ... capable of measuring 1/100ths volt!
Stay tuned for my 44 Magnum!
The Lipo packs I recycle, are nicely built in 3S, 11.1v, "bricks".
It has been bothersome adding 1S, to build 7S 25.9V and 10S 37V packs.
Fortunately, a 12S 44.4V pack charges nicely to 50.4V.
The 48V controllers, I've been eying, have low voltage cutoffs of 41V & 42V ...
41V / 12S =3.4V
42V / 12S = 3.5V, which is my personally recommended, optimal DOD (Depth Of Discharge)!
Update: Optimal DOD revised to 3.7+V (Li-ion - LiCo 18650 cells)
3.5V per cell, low voltage cut-off, should work nicely for 3.7+V minimum, actual, discharge. ("Voltage sag" under throttle.)
Test pack will use 2160mah cells, in a 12S 10P configuration. (.96kWh)
44.4V - 21.6ah
Which can be easily paired with Twin Sister Pack for 1.92kWh pack.
44.4V - 43.2ah
Yeah ... I stocked up on cells!
Initial charging will be done with multiple iMax B6 as 22.2V sub-packs,
or ... slower ... 22.2V dual charging from single charger.
DrkAngel
1 GW
DrkAngel said:
My initial use will be on my 24V EZips ...
Running a 24V motor at 44V might seem like a good way to burn up it up?
There are many stories of these motors being burnt up, running at 36V.
Some have gone to extravagant measures of, cooling, ventilation etc.
Well ... I do not propose this option as an uninformed or lightly considered manner.
I have run multiple bikes at 37V, for thousands of miles, with no problems.
My key to motor 'survival" is not additional cooling, it is prevention of heat production!
Full throttle draw at a dead stop is 30+ Amp, full throttle at top speed is 10- Amps.
Full throttle at 5mph might produce 80% waste heat and 20% usable power.
DC motors seem to have an efficiency range, this range moves - dependent on throttle position.
Throttle, "effectively", regulates the voltage to the motor, the "efficiency range" of the motor seems to be in the 50-80+% of top speed, nearing 80%+ efficiency & 20%- wasted heat.
With the EZip 24V:
Throttle % : Equiv. V : Max mph : Best Pwr
....25...........6V.......... 5....... 3-4 mph
....50...........12V.........10.......5-8 mph
....75...........18V.........15...... 7-12 mph
...100...........24V.........20......10-16 mph
With the EZip 44.4V:
Throttle % : Equiv. V : Max mph : Best Pwr
....14...........6V..........5.........3-4 mph
....27..........12V.........10.........5-8 mph
....40......... 18V..........15........7-12 mph
....45......... 24V..........20 .......10-16 mph
....67..........30V......... 25........13-19 mph
....81......... 36V..........30........15-24 mph
....90 .........40V..........35........18-30 mph
...100......... 44.4V.......39 **......20-35 mph
As long as the throttle is regulated and the motor kept in the efficiency groove, heat should be manageable.
24V full throttle at 5mph (80% wasted heat)
Should prove much more damaging than -
44.4V full throttle at 30 mph (20% wasted heat)
44.4V full throttle, @ 30mph, produces less than 1/2 the heat as 24V full throttle @ 5mph!
Moderate initial throttle, combined with pedal-assist, keeps my motors in the pink, (moderate temperature).
Personally, I have difficulty understanding eBikers that pump more and more amps into their motors.
2x the amps might produce 50% more low speed torque, but it also produces 300% of the damaging heat.
With only a slight, if any, increase in top speed.
But that might just be my frugal spirit being offended by the speed, sorry, power demons!
** "Max Mph" does not reflect wind resistance! Still ... 44.4V should produce 30++ mph!
37V provides 27mph, with pedal assist, past 30mph.
alsmith
100 kW
It's good to have a plan- I hope it works for you. Why not fit a temperature probe in the motor by the windings so you can watch what is happening- it seems sensible given the feedback on overvolting these motors, and the probes aren't too expensive. I got one to fit into my motor so I can see what happens for peace of mind. It's an overcautious step for me but better to be safe.....
DrkAngel said:
DrkAngel
1 GW
44.4V Magnumalsmith said:
I've put 3000 miles, @37V, on one motor ... will try to "burn-it-up", as my test bed.
I have 3 new, spares.
No ... I won't really try to kill it ... but not going to treat it any nicer than "my" usual.
My EZip goal is, melding smoothly with local 30mph traffic.
No need to go any faster than that.
Beyond that will be my eMotorcycle project ...
DrkAngel
1 GW
44 Magnum - TURBO
Possibly, just so that I could add the Turbo moniker, I'm resurrecting one of my older ideas.
The implementation of an additional throttle return spring.
Preliminary idea, was a method of regulating top speed, as could be used for seniors, on a motorized tricycle designed as a mobility device.
A simple set screw on the throttle could suffice.
Next stage was the possibility that instead of a "stop" screw, a heavy spring could engage at a specific throttle position.
Logically, I could set this point where "legal" 20mph cruising is provided, (about 24V equivalent).
Beyond that point, additional required effort would serve as a firm reminder that throttle "Turbo" should only be used in emergencies ... or when cops aren't looking.
Alternately, I could "switch" a proper value resister into the HALL throttle circuit.
A simple switch, or even an appropriately labeled, TURBO momentary contact button, could bypass the resister and provide the same function!
Possibly, just so that I could add the Turbo moniker, I'm resurrecting one of my older ideas.
The implementation of an additional throttle return spring.
Preliminary idea, was a method of regulating top speed, as could be used for seniors, on a motorized tricycle designed as a mobility device.
A simple set screw on the throttle could suffice.
Next stage was the possibility that instead of a "stop" screw, a heavy spring could engage at a specific throttle position.
Logically, I could set this point where "legal" 20mph cruising is provided, (about 24V equivalent).
Beyond that point, additional required effort would serve as a firm reminder that throttle "Turbo" should only be used in emergencies ... or when cops aren't looking.
Alternately, I could "switch" a proper value resister into the HALL throttle circuit.
A simple switch, or even an appropriately labeled, TURBO momentary contact button, could bypass the resister and provide the same function!
jamo96
100 W
Here are a couple of battery packs I've made recently. They are 14s 5000mah Turnigy 20c lipo contained in a 150mm x 150mm x 75mm enclosure. They fit very well. I have soldered a harness so that 2 x 7s jst-xh plugs are available from the outside for balancing on my hyperion charger. So gives me a total of ~55V 10Ah.
Finished pack. Contains two 3s and two 4s packs.
Unfinished pack. Contains four 3s packs and is missing a 2s pack (its in the post)
With the Cell Logs connected.
Finished pack. Contains two 3s and two 4s packs.
Unfinished pack. Contains four 3s packs and is missing a 2s pack (its in the post)
With the Cell Logs connected.
SamTexas
1 MW
Nice box. Where did you buy that box? Is it an electrical junction box?
jamo96
100 W
Yes I very happy with the way they turned out. The batteries fit very nicely inside the boxes. I had 15s in there initially but went back to 14s for my charger. They are just general electrical termination boxes from B & R:
http://www.brenclosures.com.au/pj.htm
They were about $15 each from a local electrical supplier. Can probably find them cheaper online somewhere.
Probably should have just made two 7s 10Ah packs but they were not made at the same time. But overall very happy!
Improvements: Need to think of someway to mount the cell logs better, possibly add an LVC alarm output and maybe a temp sensor.
http://www.brenclosures.com.au/pj.htm
They were about $15 each from a local electrical supplier. Can probably find them cheaper online somewhere.
Probably should have just made two 7s 10Ah packs but they were not made at the same time. But overall very happy!
Improvements: Need to think of someway to mount the cell logs better, possibly add an LVC alarm output and maybe a temp sensor.
DrkAngel
1 GW
25.9V Li-ion - Lipo chargers - available again!
Free shipping on $75 order w/code: "THANKS"
Tenergy Intelligent 2A Charger for 25.9V - $23.63 Works fine with, all I've tried, 24V EZip-IZips, probably all 24V controllers?
I also got 1 - 18.5V 2.5A Lipo charger, for charging 18V B&D power tool pack, lipo conversions, about same price.
Tenergy Intelligent 2.5A Charger for 18.5V - $22.81
Free shipping on $75 order w/code: "THANKS"
Tenergy Intelligent 2A Charger for 25.9V - $23.63 Works fine with, all I've tried, 24V EZip-IZips, probably all 24V controllers?
I also got 1 - 18.5V 2.5A Lipo charger, for charging 18V B&D power tool pack, lipo conversions, about same price.
Tenergy Intelligent 2.5A Charger for 18.5V - $22.81
pavlik1
10 W
Here is one more, not quite finished yet.
the box
the bat
looking for BMS
16s app 20Ah
Weight 9.8kg
the box
the bat
looking for BMS
16s app 20Ah
Weight 9.8kg
DrkAngel
1 GW
My 25.9V 26Ah recycled Lipo EZip build has been doing great for the last 1000+ miles!
Finally, it developed a voltage fluctuation on 1 bank.
So I desoldered one leg of the parallel, and put a one amp draw on the individual pairs.
It was easy to find the weak cells, right on top of my "A" stack.
Replacement would be very simple, but I wanted to try one of my experiments.
Since the voltage variance was fairly minor I figured one additional cell on the 12P bank, should equalize things nicely.
Well, everything is packed pretty snug, no room for another lipo cell.
So ... I wired a single 18650 Li-ion cell into the weak bank, and placed it on the lower side-6pack.
Result:
After 2 charge-discharge cycles pack is back in balance. within 1/100th at the high and low voltages.
Of course I will, eventually replace the weak cells, but 1st, I have to give the Lipopackpatchjob a proper trial.
Which might help me seriously consider the option of building larger packs with spare cells built-in.
The spare cells would be part of the pack, but not wired in-circuit.
If one bank suffers, an additional cell could be easily tacked in.
Or, bad cells clipped out and the replacements clipped in.
A quick fix for longer trips, or when shop not available.
Finally, it developed a voltage fluctuation on 1 bank.
So I desoldered one leg of the parallel, and put a one amp draw on the individual pairs.
It was easy to find the weak cells, right on top of my "A" stack.
Replacement would be very simple, but I wanted to try one of my experiments.
Since the voltage variance was fairly minor I figured one additional cell on the 12P bank, should equalize things nicely.
Well, everything is packed pretty snug, no room for another lipo cell.
So ... I wired a single 18650 Li-ion cell into the weak bank, and placed it on the lower side-6pack.
Result:
After 2 charge-discharge cycles pack is back in balance. within 1/100th at the high and low voltages.
Of course I will, eventually replace the weak cells, but 1st, I have to give the Lipopackpatchjob a proper trial.
Which might help me seriously consider the option of building larger packs with spare cells built-in.
The spare cells would be part of the pack, but not wired in-circuit.
If one bank suffers, an additional cell could be easily tacked in.
Or, bad cells clipped out and the replacements clipped in.
A quick fix for longer trips, or when shop not available.
DrkAngel
1 GW
Notes:DrkAngel said:Sorry, this is the raw data of my monitored discharge test.
Will graph later, but wanted to post the results.
Quick stats on mAh per 100th V.
Li-ion (LiCo) 2600mAh used cells.
Charged to 4.25V
.2C discharge rate
..V... mAh
4.23V .. 4
4.22V .. 5
4.22V .. 6
4.20V .. 9
4.19V .. 18
4.18V .. 37
4.17V .. 38
4.16V .. 38
4.15V .. 38
4.14V .. 38
4.13V .. 38
4.12V .. 38
4.11V .. 40
4.10V .. 40
4.09V .. 42
4.08V .. 43
4.07V .. 44
4.06V .. 44
4.05V .. 44
4.04V .. 42
4.03V .. 43
4.02V .. 42
4.01V .. 43
4.00V .. 44
3.99V .. 44
3.98V .. 43
3.97V .. 45
3.96V .. 46
3.95V .. 45
3.94V .. 45
3.93V .. 47
3.92V .. 48
3.91V .. 47
3.90V .. 45
3.89V .. 43
3.88V .. 42
3.87V .. 44
3.86V .. 45
3.85V .. 57
3.84V .. 57
3.83V .. 78
3.82V .. 82
3.81V .. 86
3.80V .. 88
3.79V .. 89
3.78V .. 86
3.77V .. 94
3.76V .. 98
3.75V .. 103
3.74V .. 102
3.73V .. 105
3.72V .. 120
3.71V .. 125
3.70V .. 100
3.69V .. 100
3.68V .. 100
3.67V .. 78
3.66V .. 74
3.65V .. 72
3.64V .. 74
3.63V .. 72
3.62V .. 71
3.61V .. 70
3.60V .. 45
3.59V .. 43
3.58V .. 43
3.57V .. 24
3.56V .. 20
3.55V .. 20
3.54V .. 20
3.53V .. 17
3.52V .. 14
3.51V .. 12
3.50V .. 10
Output is reasonably steady from about 4.18V till 4.1V.
Slowly increases from 4.1V till 3.85V
Then increases sharply from 3.85V till it peaks around 3.7V.
Below 3.7V output decreases rapidly till it reaches minimum usable at 3.6V.
Based on these figures:
For Li-ion (LiCo),
Charge-discharge voltages should be,
Maximum 4.18V
Minimum 3.6V
3.85V to 3.65V is the muscle bulge of the discharge curve.
To take full advantage of this "bulge" I would recommend "bottom balancing", at 3.65V.
Charging above 4.2V is worthless, and unnecessarily damaging.
Same for discharging below 3.6V.
Monitored discharge was done with an iMax B6.
iMax B6 voltage readings were not used, voltage was determined by meter directly on battery.
The iMax displayed a noticeable voltage sag, that varied at different points of the discharge.
The meter flip-flopped between voltages as the voltage dropped, I marked the point when lower voltage was constant for 10 continuous seconds.
Be aware! These results are for a specific cell type-formulation and might have no relation to other brands-types-formulations!
Recycled Lipo will be tested next.
New RC Lipo after that.
Maybe SLA ... for a laugh!
Smoothed the graph, to moderate ... human error ... ?
DrkAngel
1 GW
mAh per 100th Volt graph, prompted revision to 3.6V maximum DOD! (18650 LiCo cells)DrkAngel said:
Kin
10 kW
Just wondering if you got the PM I sent asking about tips for sourcing used LiPo. Was looking over at RCgroups, thinking they might have some stuff (but a lot of it is excessively high C rating or excessively low Mah).
