Electrathon America Forum

Members Login
Username 
 
Password 
    Remember Me  
Post Info TOPIC: ProEV tests an 18650 Lithium pack


Senior Member

Status: Offline
Posts: 168
Date:
ProEV tests an 18650 Lithium pack
Permalink  
 


Hi,

I have written a little about using Rhino Lithium Ion Polymer batteries to run in the Advanced Battery Class. I was out of racing for 3/4 of a year and had left my pack fully charged. This is not the right way to store lithium cells. My understanding is that it is better to store them about 50% charged.

The first two races of this season demonstrated that the pack had suffered during it's racing hiatus. The pack runs out of power after delivering around 800 Whrs. This is down from the 925 Whrs that we got before and way down from the manufacturer claim of 972 Whrs.

USF is doing an excellent job with their battery pack, so I decide to follow their lead (imitation is the sincerest form of flattery smile). The pack is made up of 100 18650 2600 mAh cells. So 2.6*3.7*100 = 962 Whrs. The company seems to be a battery assembler rather than manufacture. They specify the type of cell but not the brand. My guess they build the pack with whatever 18650 2.6 cell they get a good price on.

I order the 72 volt nominal pack which is 5 cells in parallel and 20 in series. They offer the same capacity in different voltages so 24 volt and 48 volt systems are covered.

Here is the link: https://m.aliexpress.com/item/4000165783342.html?pid=808_0000_0101&spm=a2g0n.search-amp.list.4000165783342&aff_trace_key=&aff_platform=msite&m_page_id=5984amp-OnQTdG54bk7KzdX5u6IU5Q

At the time, each pack was $248 with $84 shipping. The site was not very informative but said the pack included a built in BMS and a charger. There was a choice of 1000W 15A BMS or 2000W 30A. I chose the 30A but when I ordered they asked that if I expected to use a higher amp draw, I should indicate this. Since I usually draw 60 amps, that is what I told them. 

I select Anderson connectors. Of course, Anderson connectors come in different colors that do not fit together and the web site does not ask what color you use. They sent black.

The pack has a max instant discharge of 90 amps and a max continuous discharge of 30 amps. The max charge rate is 5 amps.

The pack arrives with a small charger . The only instructions are some warnings on a sticker attached to the pack.

The charger says 84 volts, 1.5 amps. It is meant to charge at 1.5 amps until it reaches 84 volts, then stay at 84 volts while the amps drop to zero. If all cells are at equal voltage, this works out to 4.2 volts per cell.

Ideally, the pack builder assembler the pack with the cells at the same voltage but just in case, the first charge, I disconnect the charger at 82 volts and monitor the voltage. Pack 1 stays fairly steady indicating no cell is at a high enough voltage for the BMS to discharge it. The second pack drops fairly rapidly hinting the BMS is having to work. I let it sit overnight and repeat the process twice more before it holds steady. Then I feel safe charging both packs to full.

They charge to a little over 83 volts.

I discharge each pack at around 1C (13 amps). The sticker says minimum voltage is 60+- 1 volt. I note the time to 64 volts (3.2 per cell) and 62 volts (3.1 per cell). I stop the discharge at 62 volts. The difference in capacity between 64 and 62 volts is about 25 Whrs so 62 volts will be my minimum.

I have found I can get an accurate measure of capacity using a Satiator charger from Grin Technologies. It is easier than recording current and voltage ever five minutes as the battery discharges and the results match when I did it both ways. The Satiator charger gives total Whrs or Whrs that goes back into the pack.

https://www.ebikes.ca/product-info/cycle-satiator.html

I am pleasantly surprised to find one pack tests out at 1019 and the other 1021 Whrs. This is almost 100 Whrs more than I have ever had and 220 more than the damaged pack was supplying.

I talk to a couple of other teams who are running this type of pack. I am warned that the BMS might shut down the pack if I run over the 30 amp continuous current for too long. Also that when the pack sags below the 3 volt per cell limit (60 for my pack), the pack will cut off. Then you must cycle the power to get the pack to reboot.

I will post more as I learn more.



-- Edited by ProEV on Monday 16th of December 2019 07:12:50 PM

__________________

Cliff

www.ProEV.com



Senior Member

Status: Offline
Posts: 168
Date:
Permalink  
 

Time for an update on the new batteries.

First, I have heard from a couple of sources that the batteries might gain capacity during the first 5-10 cycles, so I did another discharge charge cycle on my second pack and got exactly 1,019 Whrs again. So, no sign of increase capacity yet.

To review, I am testing capacity by discharging the batteries at around the 1C rate (1 hour discharge ) and then using a charger that records the energy provided to recharge the pack. This is easier than my more correct method of recording current and voltage ever 5 minutes during the discharge and calculating the energy supplied.

Using the energy put back into the battery would not work for lead acid batteries because they are something like 90% efficient in recharge. This means energy is lost during recharge (as heat), and they require extra energy to fully recharge. Lithium batteries are around 99% efficient in recharge. When I compared discharge calculations to what the charger gave me with the RHINO pack, the results were very close. Knowing this, I felt confident using the charger data to test lithium batteries. Once again, it is never that simple!

I have worked with a number of different types of vehicle Battery Management Systems. Each of these systems balance the cells either by using an isolated charger to boost low cells or resistors to discharge cells that might over charge. None of these systems had any components in series with the pack except a contactor to disconnect the pack.

This BMS seems to be different. My understanding (from conversation and the internet since no manual was included) is that there are MOSFETs connected to each cell. The MOSFETs feed the power from the cell depending on cell voltage, temperature and current. This is very clever since it allows the system to protect weaker cells without shutting down the system. The negative is there is an efficiency hit of going through another component. This combined with higher costs is why I have never run into these systems on the higher amp vehicles I have worked with.

I redid my comparison of discharge vs charge with the 18650 pack and found that a 66 minute discharge from  83.81 volts to 62 volt starting at a discharge rate of 13.6 amps and ending at 10.8 amps gave 953 Whrs but required 1,026 Whrs to recharge. My suspicion is that the BMS is using up about 7% of the energy. 

I ran the pack in a race and ran out of power somewhere between 824 Whrs and 870 Whrs (two different systems for reading Whrs). The pack was hot to the touch after the race. The first lap, the voltage was sagging from 80 volts to a low of 67 volts. After the first 10 laps, the BMS was limiting maxium current. By the end of the race, I was limited to 10 amps. The pack required 1,032 Whrs to recharge. 

My conclusion is that a car running a constant current strategy could get 950 Whrs from the pack and do well. Especially if the team were able to gear their motor to be most efficient at the proper current and resulting RPM.

The ProEV Super Coupe has a high RPM motor that is most efficient at higher amps. Our driving strategy is high amps out of the corner to quickly build speed and then lift and coast to conserve energy. This, unfortunately, puts this BMS out of it's design parameter and makes it very inefficent. (Perhaps as high as 20%). The Rhino pack is probably a better choice for the Coupé.

 



__________________

Cliff

www.ProEV.com



Senior Member

Status: Offline
Posts: 175
Date:
Permalink  
 

Very interesting post. Thanks for the update.

__________________

Build something and race it!

Page 1 of 1  sorted by
 
Quick Reply

Please log in to post quick replies.