Headway 38120 P cells as motorcycle SLI battery replacement

marcexec

1 kW
Joined
Mar 10, 2009
Messages
433
Location
Dublin, Ireland
I found data under
http://www.powerscaner.com/product_1.php?id=157
(not my source)
weight is 338g on my scale though


The intention is using them on my (ICE) bike: http://www.spritmonitor.de/en/detail/378249.html, http://en.wikipedia.org/wiki/Suzuki_RF400R

The current SLA battery makes her a plug-in hybrid already - needs to charge overnight to be able to start in the morning. :roll:

some pictures:




I used the voltmeter to monitor the actual voltage while driving, tops out at 14.41V +.07 drop through cabling, switch and fuse(measured) gives me 14.48V, or 3.62V per cell. It's a little on the high side but within spec.

Cells came in today. Many thanks at this point to "Jack Bauer".
Cell voltages are 3.29, 3.31, 3.33 and 3.34.

I think I will have to balance them first as they are essentially bulk charged. Maybe I will build an adjustable power supply and do a few cycles before putting them in.
Any thoughts on a CellLog from Hobbyking?
 
I will start monitoring some cycles, first some balancing

cycle 0 . 1a . 1b . 2 . 3 . 4a 5*
Batt
1 3.34 3.60 3.40 3.48 2.31 3.23 2.97
2 3.32 3.60 3.40 3.49 2.64 3.23 0.00!
3 3.30 3.51 3.41 3.57 2.13 3.22 3.33
4 3.29 3.49 3.42 3.52 1.86 3.22 3.33

a = hot off "charger"
b = settled

Edit:
My circuit with LM338K got too hot to touch, will have to mount it on a heatsink.
Was charging fine @3.65V, tapered off from .85A to .5A.
Will have to re-design and use a power supply with less than 12V.

Now using PC power supply for cycle 2.
cycle 3 - after a week in the bike, curiously drained over 2 idle days
cycle 4 - balance charging
 
The batteries are shipped from the factory at about 50% charged. Quite a bit of voltage variation is to be expected. Also the cell voltages will be a bit unstable until they have been cycled several times.
 
Got around to buy some copper pipe (3/4 inch, sold by the foot - guess the metric system has not gotten through to the Irish countryside yet) and M6 bolts for the bars.
Cut the pipe into 60mm lengths, then cut it longitudinally. Carefully flattened and folded at 20mm, then folded once more i got three nice bars:



Put the full pack into the bike for testing.


Not surprisingly at ~13.5V battery the bike idled much better right from the start. I saw peak voltages (with revving) of 14.58V, but that was without the headlight on.

I will now add balance leads and wrap the pack somehow to protect it from the elements.
 
Seems like a cell died last week.

I had some gremlins with the bikes electric system (alternator, reg/rec) as well as the stupid alarm system that was in drains batteries.
The alarm is out now, I'll see what I can do with the cells.
Measured cells on cycle 5 above.
 
I am currently rebuilding and equalizing the pack.

After charging all the cells individually they rest around 3.33V, ΔV is 18mV.
What should be the maximum difference on LiFePo4 cells like the Headways in a pack?

Or, when do you guys equalize?
 
Nice work! I've been building a few of these packs myself. IMHO look at each cell as "how full is each one with water" (analogy only). I believe that
each cell needs to be full prior to being put in series (charges and final balance is faster with my charger this way). I use a Venom Pro Charger Plus I
purchased from Robot Market Place. I charge each cell individually - Build the pack - Re-charge - Balance Charge. Some people will charge and discharge each
cell (cycle them) a few times first. Also, I use a DB-8 from Chargery to monitor the cells, it will also use the cells own power to level each cell with each other while in stand-alone mode.
LiFePO4 cells apparently is like taking care of a Lead Acid Battery. But look at the LifePO4 Cell as an individual SLA. Never let one get to full or to empty.
How often to BALANCE? A fun pun..... like a women can change her mind.........each cell will have it's own character and by watching each cell you will get to know each cell individually and can treat it accordingly. I believe that there is "No" one rule, just general rules. Apply general rules as it applies. :D

Quote.... Everyday is a learning experience for me! Peace man!

Tommy L sends.....
 
marcexec said:
Put together a charging circuit based on a LM338K.
It's crude, but works nicely. I am still supplying 12V so I cool it actively.



Pic with pack:

I got a PM on giving some details about the circuit.
It's really simple:
Starting point is http://www.eleccircuit.com/lm338k-adjustable-power-supply/
If you remove the transformer, bridge rectifier and the switch and throw in random capacitors that are lying around instead of the mentioned ones.
I just used a trimpot that I have lying around in masses and selected the other resistor accordingly to get to 3.65V.
at 5A the dropout voltage is 3V, even at 1A it's still ~2V (http://www.national.com/ds/LM/LM138.pdf, page 3) so 5V doesn't cut it.
I just use a re-purposed ATX PC power supply for 12V so I have (12V-3.65V=8.35V) * ~4A (peak) = 33.4W heat.
Between +5V and +12V you'd have 7V you might use as well, depending on your ATX PS.

Let me know if you want detailed pictures or the resistor values.
 
Here is a well deserved, real-world update:
I have been using this battery for 2 weeks now in my scooter which I use for commuting.

I just put a CellLog to it:

resting voltage (installed) 13.45V
delta: 9mV

What do you think?
 
600km on the pack in the scooter now,
just held the CellOg to it:
13.48V
delta: 38 mV
I was wondering about that and switched the ignition on for 60 seconds - drained the surface charge right away, came back to 9mV afterwards.
 
marcexec said:
Here is a well deserved, real-world update:
I have been using this battery for 2 weeks now in my scooter which I use for commuting.

I just put a CellLog to it:

resting voltage (installed) 13.45V
delta: 9mV

What do you think?

After more resting, I find that my headways 38120P sit around 13.34. Delta: .009v as well. My balancer won't even kick in. :)
Fresh back from a ride I'm up around 14.4V (3.6v per cell) :)
Great battery! I have one in my VTX1800 Honda, My friend has one in his lawn tractor! They're awesome! \\m//
 
small update:
Pack is still going strong, moved from the bike in the first post to a 125cc scooter (1600km), another RF400 (4500km) and now a RF600 (http://www.spritmonitor.de/en/detail/561736.html) where the SLA was not strong enough to start the bike in the recent cold snap in Ireland.

Only change is a short cable with a ring terminal so I could keep the connectors in the bike original.
Makes for a better drop-in replacement.
 
My pack is 4s2p Headway 38120p cells for a total of 16 a/h.

Running great still as well. I did notice that one parallel group was starting to head to 4v.
I installed a Bleed Circuit that automatically bleeds off voltage above 3.65 or 3.7v.

I sell them if you need them.

http://rawvelocity.com/batterieschargers.html

Scroll down to the "Bleed Resistor Circuit".

This has kept my pack from cells going over 3.7v.

Tommy L sends.... \\m//
 
Just for people wondering if this hold up, yes it does, no problem - now for 5000 km daily driving the RF600.
I'm pretty sure I would have needed another SLA for this winter by now.
 
Subject: Headway 38120 P cells as motorcycle SLI battery replacement

marcexec said:
Just for people wondering if this hold up, yes it does, no problem - now for 5000 km daily driving the RF600.
I'm pretty sure I would have needed another SLA for this winter by now.


I've been running my 2P4S headway as a start bat for 3 years. Love it!
But watch for a cell to run. It took about 1 or 2 seasons then one parallel cell group started to hit 4.1v.
I put a BLEED resistor network on each cell group. Bleeds off 3.7v and higher :)

Tommy L sends....

4402268.jpg
 
Back
Top