NEV battery pack... replace 72v 200amp with...?

R

redorblack

100 W
Joined
Sep 26, 2009
Messages
142
Picking up a 2007 Revolution R1 electric car on Wednesday. It has the original 72v lead acid battery pack (6ea. 12v 200ah batteries... I think they are advertised as 200ah... it still gets around 30 miles range with hills). From switching my bikes from lead to LiFePo4 Headways, I know there is performance/range to be gained from less weight and more true capacity of lithium... but EV chargers, BMS, and batteries cost a freaking fortune still.

Is there a charger/pack solution that won't break the bank or burn me to the ground if I don't watch the charge/discharge cycles like a hawk? I'm tempted to build a pack out of LiPo RC packs and use a server power supply (Liveforphysics! Help!).

If I went with Manzanita Micro, I'd be looking at about $12K to put together a 72V 180Ah CALB pack with boxes/regulators, and a 30A charger (just based off the website pricing). All real nice... but the car is costing me under $3K and is a nice driver for what it is. That doesn't strike me as bang for the buck...

Any better ideas? And anyone familiar with the Revolution R1, maybe under one of it's other names... it's the 72v AC version listed here on the Chinese site. http://hyev.cn/content/en/product/ev/HY-B22.html I wonder if it's as easy to hotrod as my Bafang controller was by upping the voltage for my bike.
 
RC packs would be nice but i personally would not trust a pack made of 160 of them... lol..

Good news with lithium is that you get about 100% of it's capacity, unlike SLA where you get in reality about 50%..
So a 100-150ah pack would do the trick just fine.

Thundersky makes 160ah bricks for this application. I think those would work well.
Headways would work, but are you willing to make your own packs or pay someone to do it? they don't come in large AH groups from the factory like thundersky does, so i dunno if they would cost more..
 
Parallel 16 AH headways would work.
6 in parallel would be good for a single pack of 3.4 V 96 aH or 12 for 192AH then series the subpacks
get 22 of such at $24 a cell = about $3200, Get a couple goodrum BMS and you could charge with an SLA charger through the BMS.
So for $3500 ish you get a 96 AH pack

The size of said battery is your real concern.
Once assembled and cables run, encase the entire thing in high temp WAX as a temperature mitigation.

Just rig the BMS board with an OFF switch on the negative cable

3G for 8 years of gas not bad.
 
I'd be inclined to go with the thundersky stuff if you want 200 ah, just for the simplicity. I would think the same charger would work, provided you matched the voltage close enough to the original pack, IE, 24s lifepo4. Then have 24 low voltage warning lights or some kind of at least sub pack voltage monitoring like 6 12v low voltage monitors.
 
I have a GEM e4 that has six 12V gel batteries in series. Instead of replacing the gels, what I'm going to do is add a 24s PSI "booster" pack in parallel, with a specialized version of our BMS. The special part is that it has a LVC Active Cutoff circuit that sort of works in reverse. What it does is it keeps the booster pack disconnected (using six IRFB4110 FETs in parallel...) until the gel total voltage dips below 71.5V, which only happens when there's a max load, like going up a hill. I've done quite a few tests and the controller has a limit of 160A, but since the gels can't keep up with this, the voltage sags, going down to about 68-69V, even with a full charge on the gels.

As soon as the PSI pack is switched in, the gel voltage will rise above the cutoff, so there's an adjustable delay of about 10-20 seconds before the PSIs are cutoff again. The rest of the BMS is pretty "standard", with a "regular" LVC function that will also cutoff the booster pack, and a similar HVC cutoff function for charging. For the latter, there are six more 4110s, basically in a back-to-back configuration with the active cutoff FETs. The HVC FETS are controlled by any of the HVC circuits tripping, which will cut charge current power after a 1-2 second delay. The GEM's built-in DeltaQ 15A charger is set to charge the gels to about 86.3-86.4V, which conveniently works out to 3.60V per PSI cell, which is what the shunt circuits are set to. The DeltaQ already has a low current, auto-shutoff function, which will activate when the current drops down below about 150-200mA. If any of the PSIs are out-of-balance, some of the shunts will be on, so the current will be above the cutoff. The HVC protects against really out-of-balance conditions, and won't let any one cell get above about 3.70V. Anyway, when the PSIs are balanced, the shunts will be off, and the current can drop below the DeltaQ's cutoff point.

With the 100Ah-rated gels alone, my e4 gets about 18-20 miles. I live in a fairly hilly area though, which is causing the controller to max out quite frequently, so the gels are getting worked pretty hard. The speed drops quite a bit, as the voltage sags. This booster pack will keep the speed up, as the PSIs will shoulder the 160A max loads, but it will also mean that I should get more usable capacity from the gels. I expect I'll see the range go up to around 25 miles.

Initially, I'm going to try this with a "single-p" setup of just 24 PSI 12Ah cells in series. I'm sure they can probably handle the 160A without sagging too much, but we'll see. Eventually, once I prove the concept, I plan to go to a 24s2p/24Ah configuration. If nothing else, this should increase the longevity of the PSI cells. Shown below is what the boards look like. They are designed to go on the front of the PSI's "Lego" assembly blocks. The there are two 12-channel shunt circuit boards, and then the control section, all stacked together. There will also be a "cover" board that will have a big VAL-U-LOK connector, so that I can plug in another special board I have that has three 8-channel CellLog units. This is mainly for test purposes, so I can see what is happening when in use.

View attachment 2x12 PSI LiFePO4 BMS v4.3.3i.png

As you can see, the only external connections needed are two battery cables that are used to connect to the main gel + and - connections. Operation is completely automatic. The booster pack will mount on a tray, right above the gels, under the rear seat.

-- Gary
 
GGoodrum said:
I have a GEM e4 that has six 12V gel batteries in series. Instead of replacing the gels, what I'm going to do is add a 24s PSI "booster" pack in parallel, with a specialized version of our BMS. The special part is that it has a LVC Active Cutoff circuit that sort of works in reverse. What it does is it keeps the booster pack disconnected (using six IRFB4110 FETs in parallel...) until the gel total voltage dips below 71.5V, which only happens when there's a max load, like going up a hill.


As you can see, the only external connections needed are two battery cables that are used to connect to the main gel + and - connections. Operation is completely automatic. The booster pack will mount on a tray, right above the gels, under the rear seat.

-- Gary
Click to expand...

Cool, Which PSI cells are you planning on running? What do you think this pack will cost all together? I haven't caught up on the BMS thread in a few weeks... is it ready for prime time?
 
redorblack said:
Cool, Which PSI cells are you planning on running? What do you think this pack will cost all together? I haven't caught up on the BMS thread in a few weeks... is it ready for prime time?
Click to expand...

I plan on using the 12Ah PSI cells, 24 initially, and then probably 48, in a 24s2p configuration. The "regular" BMS is still being tested. We are still trying to get all the gremlins out. Andy found a problem with how the HVC signal was interacting with the charge controller and although we think we have that sorted, we still had one issue where the relay was getting latched the wrong way. Because of this, we decided to bite the bullet, and redo the charge controller without the relay. Richard and I are testing this new FET-based scheme this week. The booster pack BMS uses a modified version of this new scheme.

-- Gary
 
Ok, lipo concept time... consider these 10s nano packs.
http://www.hobbyking.com/hobbyking/store/uh_viewItem.asp?idProduct=14609

37v, 5Ah... so lets say I'm contemplating building a 2s10p setup for 72v and 50Ah. That is only 3.6Kwh capacity, but it's also about 60 lbs vs somewhere around 600-720lbs for the lead that is in there... so maybe shaving over a 1/3rd of the weight of the vehicle off. In the gas world of drag racing, every 10 pounds dropped equals adding about 1 horsepower and this would be like adding 60+hp. My suspicion is that I'd get dramatically better performance at lower draw, but with no voltage sag/lower weight it would also climb much better. Any way to compare/calculate expected range between lead and lipo at these weights to see where the happy place would be in price/weight/range/performance?

The gentleman I got the vehicle from said he's been commuting 20 miles a day, then recharging, but has gone 30 without charging. The advertised range was 60 miles... assuming new batteries (they are 4 years old), I suspect they took the 6kw rating on the motor and divided that into the theoretical 14.4KW of lead (if drained over 20 hours... oops) to come up with 2.4 hours run time. Take that and multiply by 25mph it's supposed to be limited to and you have... Tada! 60 miles even. Going with the classic "You get 50% of the rated Ah out of lead if your lucky" and maybe I have 4 year old lead batteries in pretty damn good shape. I've been commuting 6 miles round trip with the car sitting outside in sub freezing temps and no idea how far/hard I can safely push it. The horrible translation from chinese manual says to not discharge past 50v... I'm assuming under load... and I got down to 51v pulling a small grade trying to maintain 30mph in 19 degree weather at the end of a commute with everything on. I assume the pack was warmer due to mass and me running the heater for about a 1/2 hour before heading home (off the pack though..)

Anyways...

Under $3k for charging/balancing/battery packs assuming 3.6Kw capacity sounds about right. I think I'd get a bit less range than now, but could add pairs of packs in parallel as needed. Trying to wrap my head around the amp readings at voltages when the lead batteries are sagging vs. what I could expect from lipo. The other night it could only hit 130 amps or so at 51v... let's call it 6.5KW. If a warm lipo pack stays at 70v or so under load at 200amps we are talking 14.4 KW available... I think the controller is limited to 200 amps... but am I correct that means I could see double the power available at 500lbs less weight?
 
I think the numbers might be even better than you think, for a 20s10p 50Ah LiPo setup. First of all, no way that under a 200A load, the voltage will drop much below about 78V, even if they are cold. So, using your metrics, for a 200A load you should hit peaks closer to 15.6kW.

How this translates into range is hard to figure. I know the 50% rule sorta works with my setup as well. The stock motor in my GEM is rated for 7hp, or about 5.2kW. The six 12V gels are rated for 100Ah, so nominally, 7.2kWh. Dividing that by 5.2kW gives a run time of 1.38 hours, times 25mph yields a range of 34 miles. The GEM rates the range at "up to 30 miles", so that is close, but mine has never seen anywhere close to this. The original gels were four years old when I first got this, about four months ago, and I've only been able to get a max of about 18-20 miles. Part of the reason is that I live in a hilly area, and the gels are getting worked pretty hard. Last month a couple of the gels started dying, and my range was suddenly almost in half, down to 10 miles. I then bit the bullet, and bought six new gels. The new batteries made a pretty big difference in performance. I got a lot less voltage sag, going up hills, and my range went up to about 23-24 miles. Since then, I've made a few more changes. I swapped the stock motor for one of the "true" 7.5hp motors from ride-4-fun.com, and I changed to some larger, 14 inch wheels/tires (also from ride-4-fun.com...). The motor has a modified speed sensor, which "tricks" the computer into thinking it is going slower than it is. That, coupled with the taller tires, means I know have a 40mph+ golf cart. :roll: :lol: Pretty fun, though. :) Anyway, my range is down a bit, back to about 20 miles, but that's better than I thought it would be. I think it is because the motor isn't working as hard, going up the hills.

My GEM controller has a current limit of 160A. Going up the worst of the hills around here, with the gels at about the 80% level, the total voltage drop is down to about 68-69V. I haven't tried measuring this when the gels are closer to empty, but I'm guessing it will be down closer to 60V. I can't say what it would be in "cold" weather, as today it is about as cold as it ever gets here and it 54F degrees. I don't own a heavy coat, but I am wearing long sleeves today and I might break out a windbreaker. :roll: Anyway, what I'm trying to figure out is how much of an effect not having any voltage drop might have on my range. I won't be maxing out the controller, going up all these hills, so it should have a positive effect, but how much? I'm not really counting on the "extra" capacity from the booster pack, but it can provide a bit of reserve, if the gels are drained down empty, their resting voltage will be low enough to kick in the booster pack so it could provide a bit more range.

I'm also considering doing a version using some of these nano-tech Lipos for the booster pack. These are so strong that I could use a "single-p"/5Ah configuration, but double that would probably be better.

-- Gary
 
Those 10s Turnigy Nanotech LiPos at $108 each is just too good a deal to pass up, so I just ordered 8 of them. Actually, I had to break it into two orders of 4 each, in order to be under the shipping limit. Anyway, I'm going to give these a shot, along with my 24s PSI setup.

-- Gary
 
Well, I just pulled the seat belt latches and the battery cover in order to see and service the batteries. Turns out the batteries are marked as 140ah (c20... maybe they mean 20hr you think). They weigh about 78 lbs each. No room in there for anything else without removing the lead... so revised estimates are less weight savings, but not 'needing' as big of a pack in order to match or exceed the leads mileage performance.
The battery
https://picasaweb.google.com/lh/photo/Ucd5NGGR9EP2Yk2ZqUb0Xw?feat=directlink
The pictures I've taken so far of the car
https://picasaweb.google.com/Scott.McElhiney/RevolutionR1#
 
Yes, c20 means capacity when discharged over 20 hours. My gels have a c20 rating of 108Ah and a c1 rating of 98.5Ah. I looked into getting flooded batteries, as replacements, because in the same form factor, the capacity is greater (150Ah...), but I was told that out here, where it can get over 105F in the summer at times, the flooded lead acids wouldn't last long, maybe two years. The gels last about four years.

-- Gary
 
You must log in or register to reply here.
Back
Top