Ebike scooter project QS 4kw V3 72v, Kelly KEB72601

[smeagol222]

Ah yes internal resistance testing. How do you 'class' yours? What numbers should I be looking out for?

Another issue I've run into the the new 1/0 cable is fitting into the SB120A terminal. I guess I could prune a few of the strands to fit inside then hammer crimp.
I'm wondering if I should just leave it at hammer crimps and not bother with soldering at all- or if its possible to do both, solder then crimp? I would think crimping after solder would break the bond I'd just made with the solder?

[marcexec]

I test at 1A to 2.5V.
Anything under 120mOhm is used for "power" use cases ( e-kick scooter for daughter, NiCd replacement for power tools).
Above 200 is a dud, but then the capacity is in the 4-500mAh range anyway.
Best one I found yet was 82mOhm (HP laptop printer with it's own battery).

Edit: just crimp properly and you won't look back.

[amberwolf]

I'm wondering if I should just leave it at hammer crimps and not bother with soldering at all- or if its possible to do both, solder then crimp? I would think crimping after solder would break the bond I'd just made with the solder?

A proper crimp won't need solder. I'm not sure what "hammer crimp" would be though.

If you do solder and then crimp, then yes, the solder will fracture.

If you were to crimp and then solder, then the solder could still form a "seal" against potential corrosion if that's what you're after--but the crimp will be a much better electrical connection, even by itself.



Regarding these:

they don't all seem to be all that accurate; there's a thread about them here:
https://endless-sphere.com/forums/viewt ... 14&t=84244&

[smeagol222]

Thanks for the info. I'll just make sure I do a good crimp job. A crap those loggers are shit huh? I will have to go through that tread to see what the quality alternative is. Don't really wanna be paying $25usd each tho. Also anything I seem to order from hobby king has a crazy shipping price added to it. Will have to try and find a good source in canada

EDIT:
I could only find these 3, they're pretty expensive for shipping from USA warehouse (without import duties). Cellmeter might be best value?

price for 4pcs incl. shipping to me $65.65USD
https://hobbyking.com/en_us/hobbykingtm ... 2s-6s.html


price for 4pcs incl. shipping to me $39.33USD
https://hobbyking.com/en_us/hobbykingtm ... alarm.html


price for 4pcs incl. shipping to me $30.01USD
https://hobbyking.com/en_us/battery-monitor-2-6s.html

[mistercrash]

Celllog 8M has been trusted to work well and be accurate by many in here for years. A Google search will show you where to get them.

[smeagol222]

Cool thanks Mistercrash.

Since I've got the new 1/0 cable I'm going to use the old 6 AWG to make my 4 packs of 5S into one battery using XT90. the (S) signifies the XT90 series adaptor plugs.

I came up with an idea for a gauge increaser using 2 ring terminals bolted together (with heatshrink of course). That should be okay right- at the point of all 4 cells becoming series (current & voltage increase), the wire turns into 1/0 awg which makes sense right?

[smeagol222]

Celllog 8S has been trusted to work well and be accurate by many in here for years. A Google search will show you where to get them.

Is the 8M just as good as the 8S if you don't need the PC data logging feature? The 8S is nearly double the price- if you have the 8S what are the advantages of collecting data for PC analysis- catch your leaves fading early on?
http://www.progressiverc.com/celllog-8m.html
http://www.progressiverc.com/celllog-8s.html

Also while I'm at it can I ask what voltage differences you're looking for? If by any chance you had these when you had your 18650 packs (or another 3.7v cell) what differences were you looking for to warrant a balance charge? (1/10th volt difference or more?)

[mistercrash]

I corrected my post above. If you want CellLogs get the 8M, it has the same accuracy as the 8S but does not have data logging. Data logging is cool to see what each of your cells are doing while under large loads. I bought a single Celllog 8S to log data for my 5S groups in my battery, one at a time doing the same ride with my scooter. I thought it was a good way to see if I had a cell that would show bigger voltage drops than the others under load.

About voltage difference, I make very short rides to work so anything lower than 0.1V is not a big concern on a 60ah battery. If I was regurlarly riding until I heard my LVA go on then I would monitor more closely to have all my cell as close to each other as possible. There's a lot of info about that in the forum but there are also a lot of different opinions so it gets confusing.

[mistercrash]

Cool thanks Mistercrash.

Since I've got the new 1/0 cable I'm going to use the old 6 AWG to make my 4 packs of 5S into one battery using XT90. the (S) signifies the XT90 series adaptor plugs.

I came up with an idea for a gauge increaser using 2 ring terminals bolted together (with heatshrink of course). That should be okay right- at the point of all 4 cells becoming series (current & voltage increase), the wire turns into 1/0 awg which makes sense right?

A gauge increaser? This is how I understand it, a 6AWG cable connected to a 1/0AWG cable is still a 6AWG cable. But it doesn't really matter since the ring terminals or the XT90 connectors may be more restrictive than the 6AWG cable. Something in there is going to get hot and may melt. But I may be wrong.

[amberwolf]

If you have a series of different gauges of wire, the smallest gauge is the limiting factor. If you had two 6 gauge connections in parallel that directly go into a 1/0, it's less of a problem than if the 6 gauge go into another 6 that then goes into a 1/0....

The connectors are usually higher resistance than the wire, so smaller connectors also tend to have higher resistance than larger ones, but you cant' directly compare different types of connections, without knowing or measuring their actual resistance.

[smeagol222]

Bear with my guys I'm learning as I go, thanks for the continued advice really appreciate it

I know I should just use the 1/0 awg everywhere but since I want to do this modular battery this is my guage dilemma with wiring up.

I want them to be easily separate-able by unplugging XT90 connectors (which apparently are rated for 90amp cont. which is the peak in my system) for when I need to balance charge, and for normal use.

I would think that 6awg on the first level connecting two of the 5s10p packs together creating 37v would be okay? Perhaps I should just remove the last series adaptor and 6awg and create my own 1/0 series harness (connecting the two 37v packs together). I could then still keep the extra series connector for bulk charging purposes?


Series
I have them plugged together in series when inside ebike using these adaptors and 6awg. Each group of 5s10p would be 18.5v. LG MJ1 cells are 10a cont each so x10 = 100a?


Parallel
I have them plugged together in parallel these adaptors and 6awg. for balance charging (bottom set of adaptors would need some wire to connect them since its a tight fit). Since the 1/0 wont fit into XT90 I could change guages using these terminals


And then..
Then use these big Anderson SB120 plugs for continuing 1/0 to the main contactor, then controller etc..

[Doctorbass]

Both Anderson and Xt connectors are excellent connectors when original brand and not chineese copy.

Personally i have tested the Xt150 and the 10mm bullet connetors and these , when properly soldered and have MAXIMUM copper strands in contact with the solder cup of the connectors) can be un order of magnetude of 100 microohms or less witch is equal to 1 inch of gauge 10 wire.

I am using these 10mm bullet connectors also availlables here: http://www.progressiverc.com/10mm-bulle ... pairs.html
for my drag racing Zero withc take about 350A each for 10 second and i use 2 in parallel. these are connected to 2AWG and fit perfectly.

6Awg is usually good for about 150A battery current continuous.

when i want to merge two or more smaller gauge wire to one larger, i use copper tube that i sand inside and then crimp and solder.

Doc

[smeagol222]

Thanks Doc. I will re-evaluate, might grab some of those bullets since I'll need the celllog 8m's from there anyways
XT90's are rated at 90a continuous, but at what voltage? What would be the peak amps?

[smeagol222]

After 9 weeks finally got my 24S 72v Lifepo4 20ah battery today
http://sun-ebike.com/72v20ah-lifepo4-ba ... p0022.html
Specs are:
Rated discharge current: 65A
Peak Discharge current: 85A


Came with pretty high gauge wires (I guess its the minimum for this power level) If I chop off and replace with 6 Gauge and my big Anderson SB120 plug it would be better ya?


Also and this little USB converter for charging cell phone etc.



Seller told me has these cells inside



I plugged it in and it charged for only a few mins before the green "fully charged" light came on. Measured the voltage and it came to 81.7v
As I understand Lifepo4 cells are 3.2v x 24 = 76.8v so I guess fully charged is around 3.4v per cell?


Charger seems pretty hefty too

[Animalector]

Some numbers not adding up there... Lifepo4 is 3.6V per cell full charge, your charger has it right.. If it only measured 81V after full it has undercharged... Perhaps it needs to balance.

[smeagol222]

So Lifepo4 is 3.2v and 3.3v and Li-ion is 3.7v?
Fully charged 24 x 3.6 = 86.4v
http://www.cobox-ebikes.com/296/basic-u ... fepo4.html

Anyways I've left it on the charger and it seems to be charging on and off. Perhaps since its a high amp charger it has a timer when voltage is approaching cutoff (88v). I will leave it and see. It has a BMS inside (at least I friggen hope so).

On the product page it says "Charging cut-off voltage: 88V (for a new battery after several cycles)" I wonder after its been fully discharged and re-charged back up the BMS will kick in and balance out the cells a bit more

[Animalector]

On and off is it doing its balancing. Leave it like this for a while ( hours) . They have used 3.64V as the charging voltage which is fine.

Enjoy, looking forward to seeing it all running

[smeagol222]

oh good thats a relief - yeah tell me about its its been ages. I can't wait! Another problem I thought of is the kelly controller programming. My other battery is going to be more powerful. I wonder if kelly have a bluetooth dongle to program it on the fly. Or even better the one set of programming will be good for both batteries

[mexicanhanu]

awesome work so far ! subbed

[smeagol222]

I heard back from the seller they said to fully discharge and fully charge again. Just gotta build the ebike first then hehe or hook up a DC converter and that 70watt light bar for like a day or so

Did a quick measure of the battery, one side was only off by 0.5cm from the online product description. Fits perfectly though- going to cut out that bar and move it forward so I have room for my modular 18650 batterybloc battery behind it


Did a quick re-render, going to be a tight fit which will be good- dont really want stuff knocking around


I got my JST-XH parallel plug and the little extensions for balance charging. Going to have them running off each of the 10p5S stacks


Hopefully the CellLog 8M will accept my JST-XH 5s cables, the description says 2-8 so surely right?
http://www.progressiverc.com/celllog-8m.html


Got my battery blocs- these things are awesome. Also got the 12v battery one 4s4p took about 10min to assemble super useful! I have this 12v 60watt pressure washer gonna use it for that



Gonna have some fun changing over this tubeless tire from this craptacular 500w wheel to the shiny QS one . Anyone got tips for that? The local ebike dealers said they'd do it for $40CAD ($30USD) maybe its worth it- I'm just scared they gonna mess up my nice QS wheel

[smeagol222]

On a side note I got my little JST 4S connector to balance my 12v battery blocs battery as practice for the 5S packs I'll parallel balance.

Works like a charm!

[smeagol222]

I got my 4 little Celllog 8S today. These things are super tiny. Came with a tiny CD with user manual and stuff and a little alarm cable I'm guessing I could hook up to 4 dashboard LED's?

I tested by plugging into my 4S4P 12v 18650 pack which has been freshly charged/balanced. It seemed to detect correct li-ion type "Lilio" and then the alarm started going off saying it was "over" voltage? All the cell groups were around 4.18v - 4.16v. I applied the load and they dropped down to 3.8v - 3.9v difference showing 113mV
Is the default settings for cell group voltage differences fine? Or do I need to customize this? I've heard the rule of thumb is to balance the battery if it becomes a tenth of a volt out- which is 100mV right?

18650 cells are 3.7v nominal full charged are 4.2v, does this mean every time I fully charge the alarm is going to be going off until I apply the load (not a big issue since I plan to wire the cell logs with separate switches).



UPDATE:
One important thing I completely forgot to do was balance out the capacity of the groups of cells -there was a mAh diff of like 300 surely not good. So after taking it apart (easy) I regrouped the 4P cell groups and used a spreadsheet to figure out the difference (got it down to 28mAh)

[mistercrash]

Your Celllogs are set to LiLo. Set them to Lipo and the alarm will trigger at 4.2V, not 4.1. The tiny CD is the instructions.

[smeagol222]

Oh I see. I thought lilio sounded closer to li-ion than lipo

[macribs]

Are metal sheets used on the battery blocks really steel? I mean wouldn't copper or nickel be more suited for carrying large amount of current?
Adding all those steel plates will it make an effect on the end power you think? Reason I ask is because I remember one of the DIY spot weld threads he got nickel strips that was actually steel strips nickel coated and ended up ordering new ones to avoid the extra resistance.