New GEM Owner - Converting To Lithium Batteries

GEM Forum
1

Hi all. I’m a recent 2002 GEM short bed utility owner, and loving it. I love it so much I recently sold my other cars. :cool:

The current batteries are 3 year old, but still give a decent range of about 15 miles. I’m dealing with it right now, but it requires spreading out trips to multiple days.

In January, I’m ordering 25 prismatic 180Ah lithium iron phosphate (LiFePO4) cells to replace the aging lead acid pack. Instead of having about 6Kw of usable energy in the lad pack, I’ll have over 15Kw of energy. The empty weight of the GEM will also drop by 20%. Both of these factors should give me about 3 1/2 times the range of the lead acid batteries. Finally making range worries go away completely. Four of these cells fits perfectly in the space that one of the lead acid batteries goes now, so the lithium is basically a drop-in replacement. The on-board charger will work with lithium cells after being sent back to the factory for $100 reprogramming.

The best part is the rated cycle life on the lithium pack is 3000 cycles to 80% depth of discharge vs about 350 to the same DoD with the lead pack. The Lithium cells are rated at 5000 cycles at 70% DoD, and lithium does not have a “calender life” like lead acid batteries do, and have no self-discharge. So I expect these cells to last essentially the life of the vehicle, and maybe more. By that time, they will be obsolete, and better cheaper batteries will probably be available.

The only down side is the upfront price. The lithium pack will cost about 3 1/2 times a lead acid pack, so I better plan on keeping the GEM for awhile. But the batteries can out-last 10 or more lead acid packs and can pay for themselves if owning a GEM (even a series of GEMs) for 10 years or so.

1 Like
2

What vendor are you using to buy the LiFe?

3

[QUOTE=doogie;9802]What vendor are you using to buy the LiFe?[/QUOTE]I ordered through Current EV Tech.

4

[QUOTE=bgeery;9804]I ordered through Current EV Tech.[/QUOTE]

Are you able to detail the items you are purchasing as an aid in trying to determine how to retrofit my 2005 eS gel batteries? For example, besides the batteries what other components will you need to convert to the LI system? Since you probably have a Zivian charger vs. my DeltaQ, I’m not sure if mine can be reprogrammed.

Additionally, please keep me in touch with any nuances you may experience in dealing with a vendor so far away in western Australia. Thanks!

5

[QUOTE=doogie;9817]Are you able to detail the items you are purchasing as an aid in trying to determine how to retrofit my 2005 eS gel batteries? For example, besides the batteries what other components will you need to convert to the LI system? Since you probably have a Zivian charger vs. my DeltaQ, I’m not sure if mine can be reprogrammed.

Additionally, please keep me in touch with any nuances you may experience in dealing with a vendor so far away in western Australia. Thanks![/QUOTE]

Yeah, from what I read, Delta-q will not reprogram a GEM charger. They will reprogram this model, just not the GEM version of it. So in your case, you are looking at buying a new charger as well. You might be able to buy a used Zivan NG1 and have it reprogrammed from ZivanUSA in California.

Current EV Tech is located in Washington, not AU. I picked them because I didn’t want to wait 1-2 months for a shipment directly from Thundersky in China. These are 160 AH Thundersky cells. The 180 AH China Aviation Lithium Battery Company (CALB) cells are the exact same size, so would also fit fine. They have a US warehouse and I’ve been told they can be talked down to $1.12/AH if you negotiate. The TS cells have a 5000 cycle at 70% depth of discharge, the CALB cells 3000 cycles at 70% DoD. But, the CALB cells have a slightly higher continuous current rating.

I’m using 25 cells. That’s limited by the high voltage cutoff of the motor controller. 4 cells fit in the place of one lead battery, with one string having 5 cells (I’m placing that string where the front most forward battery sits now.) I’ll use the same battery tie-down “rods” that are used with the lead batteries, but I have to drill new mounting holes so I can rotate their direction by 180 degrees (cross the top of the cells from front to back instead of side to side like on the lead acid cells.) Extra rods can be purchased at auto parts shops.

All I ordered was the 25 cells. They will be strapped together as I requested. When I receive them, I’ll use a 0.05 ohm 300 Watt power resistor to individually bleed the voltage of each cell down to exactly 2.500 volts. This is “bottom balancing”, and is done so that if I even drive the pack to empty, no one cell is destroyed by cell reversal. I’ll then stick them in the GEM and charge them all as a pack with the reprogrammed Zivan NG1 charger. The charging profile will be set to slightly undercharge the pack, thus protecting the weakest cells from overcharge and maintaining the bottom balancing.

That’s about all that’s involved in converting a GEM. Oh, I also plan on adding an Xantrex Link-10 meter to monitor the Amp-hours remaining in the pack, giving an accurate fuel gauge. I just happen to have one on hand. It needs an isolated power supply and 100V voltage scaler, both purchased from Belktronicx.

If I missed something, just let me know.

6

Thanks so much for taking the time to list all the detail! Since I do not feel comfortable with having to outsource all of the items, I’ll continue to wait until a total retrofit package is offered for the GEM.

In the meantime, best of luck with your upgrade.

7

Greetings, I’m a newbie to the forum but very interested in the LiFePO4 conversion idea. I have a California Roadster converted to 72V with a delta Q charger so that part should be pretty easy. I’m wondering how you’re arriving at your KWH ratings. I’m taking a 12v X 130ah x 6 cells and arriving at 9360 AH for LA Batteries and using 3.2v x 160ah x 25 cells and arriving at 12800 AH for the LiFePO4 cells. Please correct me if I’ve done this wrong? My particular setup uses 9ea T875 8v batteries rated at 170ah
so it works out to 12240AH. I was hoping for an energy gain and to get rid of the LA batteries but don’t see much gain unless I’m not doing something right. I guess the gain would be in size and weight. Please let me know if these number look right. Thanks

8

[QUOTE=islandman;10241]Greetings, I’m a newbie to the forum but very interested in the LiFePO4 conversion idea. I have a California Roadster converted to 72V with a delta Q charger so that part should be pretty easy. I’m wondering how you’re arriving at your KWH ratings. I’m taking a 12v X 130ah x 6 cells and arriving at 9360 AH for LA Batteries and using 3.2v x 160ah x 25 cells and arriving at 12800 AH for the LiFePO4 cells. Please correct me if I’ve done this wrong? My particular setup uses 9ea T875 8v batteries rated at 170ah
so it works out to 12240AH. I was hoping for an energy gain and to get rid of the LA batteries but don’t see much gain unless I’m not doing something right. I guess the gain would be in size and weight. Please let me know if these number look right. Thanks[/QUOTE]

We have a Peukert Effect with lead acid batteries. Lead acid Amp-hours are rated at a twenty hour discharge rate, and we are discharging them in 1/10 that time so lose much of the effective capacity. LiFePO4 cells are rated at the one hour rate, and also effectively have no Peukert Effect anyway. Bottom line is that in EV use you can cut that 170Ah capacity of lead acid in half to get the real world energy available for driving our GEMs. Also, in my original message I was planning on using 180Ah cells, but ended up using 160Ah cells due to cost and availability. There are 180Ah and 200Ah cells that are basically the same size and weight (but more money.)

160Ah cells seem to be a perfect match for the GEM. I’m seeing 160Amps max on acceleration and cruising seems to hover around 50mps +/- 10Amps (it’s hilly around here). This all bodes well for cell life with such low current requirements in an EV application. Many full size EVs seem to hover around 1C (in this case 160A) on cruise and up to 3C in acceleration. Within the specs, but still a little rough if you want to maximize cell life.

I took a 23 mile trip the other day around town (I’m marking on a map all the dirt roads to avoid) and ended up at 142 watts per mile. That give an expected max range of 74 miles to 80% discharged. If I had gone with 200Ah cells a 93 mile range would be within reach. My average trip is under 10 miles so I saved the money instead.

Regarding using the GEM’s DeltaQ charger for LiFePo4-- I don’t think it will work long term. DeltaQ will not reprogram the GEM version of this charger so you are stuck with the Lead Acid charge profile already in there. The voltage setpoint will certainly be wrong for 25 cells. You could try with just 24 cells instead (but buy 25 anyway for a spare and if you change chargers). Try it, but assume you will have to swap out for a Zivan NG1 or NG3 at some point.

I’d also like clarify cell placement issues. I converted a e825 Shortbed utility model that places four battery trays under the front (only) driver/passenger bench seat and two under the front hood. There was enough room to fit the LiFePO4 cells in the existing battery locations. It turns out that the standard two seaters instead of placing the two batteries under the hood instead place them over/around the rear axle. These battery trays apparently would not be big enough for the LiFePO4 cells to fit and a battery box would have to be made for under the hood like the utility models. Some angle aluminum bolted together and to the frame would do the job and then rerouting the battery cable from the back to the front.

9

[B]Here are my most relevant posts from the Yahoo NEVs Group for those that aren’t Yahoo members.:[/B]

I noticed that one of the group’s polls is asking about interest in a Lithium
Iron Phosphate (LiFePO4) upgrade kit. 99% voted they were interested. Even though there is currently no kit available, I can at least provide you a complete parts and source list to make the process just about as easy, and probably cheaper than a kit would be.

  1. 25 LiFePo4 Cells. Either the 160 AH Thunder Sky cells with a 3000 cycle life to 70% discharge from Thundersky or the 180 AH CALB cells with a lower 2000 cycle life to 70% discharge, but higher capacity range in the same size cell from Products (I hear they can be talked down to $1.12/AH). Request the cells be strapped as 1 5-cell battery and 4 4-cell batteries. These will fit right where the current Lead-Acid batteries fit (at least on the utility models).

2a. You will have to drill new holes for the battery tie-down bolts, 90
degrees from the current holes. Get 6 pairs of 12 inch tie-downs for only $10.23 from eBay My World - mo-stuff For the hold down “bars”, buy a composite decking board and have it cut down to correct lengths, then drill holes for hold down bolts to go through.

2b. If you have batteries located between the rear tires (non-flat-bed models, and depending on the size of your lithium cells, you may have relocate them by building a battery box under the front hood. This can be done in a few hours using aluminum angle, hack saw, hand drill, and nuts and bolts (or $20 hand riveting gun).

  1. Get your Zivan charger reprogrammed by Elcon to charge your new batteries. If you don’t have a Zivan, you will have to buy a new charger, as DeltaQ will not reprogram their chargers for use with lithium cells (and the GEM version of the charger is not reprogrammable anyway.) You can buy an Zivan NG1, NG2 or on of the Elcon chargers.

  2. Cut off your Zivan NG1 charger temp probe and hook up a temperature-stable 1.4K to 1.75K Ohm resistor like the Vishay/Dale Metal Film Resistor, Through Hole, 1/4watt, 1.43Kohms, .1%, 10ppm.

  3. If you plan on bottom balancing the cells before you install them into your GEM, you will want an accurate voltmeter Tenma Compact True RMS Clamp Meter | 72-7226 (727226) | Tenma and a .05 Ohm 300 watt wire-wound power resistor Wire Wound Resistors - 0 Ohm to 0.499 Ohm and a 3.3 volt power supply 3.3V Power Supply - 6.0A Single Output They also sell a larger 30 Amp version for a few bucks more.

  4. By opening up your charger and adjusting the top potentiometer, you can increase the output of your charger to as high as about 13 Amps. That will give about 30% faster charging. I’ve been running like this over a year without a problem.

Well, my GEM e825 short bed is upgraded to 25 160Ah Thundersky LiFePO4 cells. I followed my parts list, and almost everything went smoothly.

Getting the Zivan NG1 reprogrammed turned out to be the hardest part, and I still don’t have it back yet, after over two weeks. The guy that does the programming is a bit of an jerk, but there’s no choice but to work with him. I should have the charger back via UPS by Jan 3rd, and I’ll get it installed and stat charging the pack for the first time. Total cost was $143, but that also includes about $85 for several mechanical updates that had not yet been preformed on my old original NG1.

The under-seat batteries were a bit of a bear to get placed as they are a bit taller than the stock lead acid batteries, but I did manage to get them in with a little thought. Remove the main disconnect switch to give room to install Batt 4. Then disconnect the parking break cable to make room to install Batt 3 then Batt 2 via the driver’s side. Then slip in Batt 1 and try not to short it on the frame! Then reconnect the parking break cable being careful not to short the tops of either Batt 2 or 3.

Batt 1 to 4 and Batt 6 are four cells each. Batt 5 is five cells. Batt 1 to 4
are under the seat, Batt 5 and 6 are under the front hood. I had to remove the plastic battery tray under the hood that’s for batt 5 and 6, and cut off the front half (the tray for Batt 5) to allow the five cells of Batt 5 to fit in that space. There is plenty of aluminum angle there still supporting the battery. The back half of the tray needs to remain, as there is nothing else supporting Batt 6.

The cable going from Batt 4 to Batt 5 turned out not to be long enough, so I figured out a workaround. I drilled a hole in the fiberglass under the seat to mount the main disconnect between Batt 3 and 4 so I can reach it from the cabin now. I ran the cable that supposed to go from Batt 4 to Batt 5 to the main disconnect, and then ran one of the other cables that was long enough from the main disconnect to Batt 4. All the other long cables that could not be replaced were all (just) long enough. Most of shorter battery-to-battery jumper cables were not long enough and I already had some longer cables to replace them. I could have just bought starter cables from a auto parts place as well.

Also, the under-seat batteries were a little difficult to hold down. With the
batteries in place, there is no room to slid the wood over the tie-down (J)
bolts. I ended up instead just cutting a slot so I could slid in the bolt
through the slot and then secure with the included rubber washer and wing nut. It seems to work fine and the slots don’t seem to have weakened to wood too much. I also had to round the front edge of the wood for Batt 2 and Batt 3 so it would clear the fiberglass without hitting.

Anyway, not a difficult project. It took me about 8 hours to remove the old
batteries, clean things up, measure and drill the holes for the new J bolt
tie-downs, remove and cut and reinstall the plastic battery tray, install the
new batteries, and wire the batteries up. It probably took about 3 hours to cut and drill all the wood for the tie-downs. It’ll take about an hour or so to install the Zivan NG1 again.

I just took my first couple of mile test run of my '02 GEM e825 short bed
utility with about 80% charged LiFePO4 batteries.

The acceleration is incredible. I used to be able to just floor the throttle
and I could handle it, but with the lithium cells I have to ease into the
throttle or else it just seems too much. Part of the extra acceleration is from the 96 pounds shaved off the battery pack weight, but most if it is just from the higher voltage that is maintained by the lithium cells.

It also gets right up to speed fast. With stock rims and tires and motor, I
used to get to 25 MPH max if I’m on level ground. Now it quickly gets up to 28MPH and the speed limiter kicks in, so no more accelerator to the floor driving style. I’ll have to get the controller reprogrammed to disable the limiter.

The lighter weight also seems to make the suspension a little harsher, but not too bad. I know from hauling stuff on the short-bed, the suspension seems to smooth out with a second person or about 175 lbs of cargo. I guess that will change to about 270 lbs of cargo.

Anyway, I’m impressed with the increase in performance.

I spent the whole day wiring up the Link 10 amp-hour meter so I can get an idea how much energy I’m consuming. Took my first real trip of about 10 miles. The Link-10 shows the trip used 147 watts per mile so my range to 80% discharged is 72 miles with the 160 Ah pack. The 200 Ah cells are the same weight and about the same size, and would give a 90 mile range.

Made the perfect instrument housing for the Link 10 amp-hour meter from two 2"
ABS plumbing caps and a short ~1 inch section of 2" ABS pipe. The link 10 fits
like a glove. I didn’t want to buy a ready made instrument pod because they are
really too big.

The short section of pipe is cut to fit down snug in the front cap, but not be
long enough to really lock down into the bottom cap, making disassembly easy.
To hold the instrument firmly in the housing, I used hot glue along the outer
lip of the Link 10 and along the inner edge of the mounting hole in the cap.
It’s not going anywhere, but I can break it free cleanly if I need to.

To mount the housing to the instrument cluster, I thinking of just using a hose
clap bolted through to hold it. Still haven’t decided yet. Right now the hose
clamp is just taped on for experimentation.

YOU CAN ALSO FIND THESE IMAGES IN HIGHER RES AT http://imgur.com/a/wDjvk#0

1 Like
10

Looks like a nice and major upgrade to a basic LSV. I have said for some time there should be a new category and I believe I read about it in Nevada. Called MSV - or Medium Speed Vehicle, I think the spec was to allow speeds to 40 Mph, or maybe 45 Mph - I don’t remember.

It is obvious that with the 160 Ah TS Cells you have the juice to go the speed, now what about crash protection at the higher speeds?

11

It would be nice to be able to travel on even 40MPH roads. I’m in a rural town, and all the major roads are 40+ MPH. Plus many dirt roads that I avoid. I had to to take a map and mark all the 40+MPH roads and all the dirt roads and figure routes around all that. Even have to use parking lots to cross from one street to the next in a few spots.

Even with all the restrictions it’s still worth it to me. I don’t even own another vehicle anymore. Owning a car made no economic sense. Doing the math, with just the money I save in insurance (only paying $78/yr for the GEM) and savings in car payments, I could rent a car 26 days out of the month! I’d pay over the weekend (3 days) for $32 total or for the same price for one weekday. And this is at the daily rates, not weekly. I don’t even have to worry about the overpriced collision insurance on the rental car because my credit card benefits cover that for me for free. The rental company even picks me up and drops me off after the rental. So far I’ve only needed to rent a car once in the last three months, so I’m way ahead of the game financially. The GEM is also simple enough for me to work on myself so more saving there. I would never try to work on a modern car myself.

Even though my pack are way over-sized for the size of my driving area (60AH cells would have been adequate), I still enjoy the feeling of not ever worrying at all about range, and choose to charge up whenever it happens to be convenient. Sometimes once or twice in a month.

The lithium powered GEM really has been one of the best and smartest decisions I’ve ever made in my life. :thumb:

1 Like
12

Greetings,
Glad to here your LiFe conversion is working out so well, I’m still looking at converting my California Roadster. I live on a small island in the Caribbean so I don’t need much range but we have some short but VERY steep hills here so the standard 72V system struggles, I think the Lithium batteries would allow a much higher discharge rate and give me the amperage necessary to make it work. I have one question about your system that I didn’t see addressed, what are you using for a BMS ? The consensus of opinion is that if you’re investing in LiFe Batteries that you should have one or your asking for troubles down the road. Just wondering what you’re doing about that. Thanks Mark

13

[QUOTE=islandman;10551]Greetings,
Glad to here your LiFe conversion is working out so well, I’m still looking at converting my California Roadster. I live on a small island in the Caribbean so I don’t need much range but we have some short but VERY steep hills here so the standard 72V system struggles, I think the Lithium batteries would allow a much higher discharge rate and give me the amperage necessary to make it work. I have one question about your system that I didn’t see addressed, what are you using for a BMS ? The consensus of opinion is that if you’re investing in LiFe Batteries that you should have one or your asking for troubles down the road. Just wondering what you’re doing about that. Thanks Mark[/QUOTE]

Yes LiFePO4 will voltage sag mush less than Lead-Acid putting more watts into the motor and giving you extra power on your climb, but you will still be limited by the current limits of the controller (300 Amps IIRC). First step might be to attach some instrumentation to let you see what kind of voltage sag and current you are presently drawing climbing that hill. A Cycle Analyst is a good device giving a ton of data and is under $170 with the shunt.

Regarding BMS systems. I don’t bring up BMS systems because it leads to useless debate. :deadhorse: I don’t feel the need for a BMS and believe they are more trouble than they are worth. If you feel the need for a BMS there are plenty of them out there to choose from.

PS: A great video podcast on DIY LiFePo4 electric conversions can be found at www.EVTV.me

14

Where did you buy the batteries from? do you have a link?
Thanks

15

This is for those converting an older GEM, like mine, with the Zivan NG1 charger with the temperature compensation probe. The temp comp keeps throwing off the adjusted voltage, and it’s not needed for lithium charging.

The problem is you can’t just unplug the probe plug from the charger, cut the probe off and leave it open circuit, or close the circuit by tying the wires together. In all these cases the charger will sense the probe resistance is out of range and refuse to charge the pack.

The simplest workable solution seems to involve cutting off the probe and replacing it with an appropriate value temperature-stable resistor. The charger will still sense resistance in the correct range, but it will not change value much like the thermistor in the original probe does.

After I cut off the probe, I measure the probe at 38.4K Ohms at about room temp. I read about 38.2K Ohms around 69 degrees F, and about 38.8K Ohms around 85-90 degrees.

Looking for resistors around this value, I can’t find anything good with the super-stable .01 ppm resistors. The best I found (and what I just ordered) was a Vashay/Dale PTF5638K600BYEK, 0.1% ⅛ watt 38.6K Ohm resistor with a +/-10ppm temperature coefficient. Mouser has them here for under a buck each.

If I calculated correctly, 10ppm would keep the impedance to within +/-11.6 Ohms with a +/-30 degree C ambient temperature swing from between 120 F in the summer and 32F in the winter. I’d don’t see that level of fluctuation being a problem. I’m going from ~55.0 Ohms per degree C of temperature change with the original probe, down to 0.39 Ohms per degree C with the new resistor. Good enough for me!

I’ll know for sure that the new resistor will work once I get it in a few days, but I have very little concern it won’t work as expected.

16

[QUOTE=bgeery;13178]After I cut off the probe, I measure the probe at 38.4K Ohms at about room temp. I read about 38.2K Ohms around 69 degrees F, and about 38.8K Ohms around 85-90 degrees.[/QUOTE]Update: Despite the measurements showing I needed ~38K Ohm, when I received the resistor and installed it, the charger world not function.

I found further info at this page that indicated that 1.4K to 1.75K Ohm works. I ordered another temp-stable resistor from mouser: Vishay/Dale Metal Film Resistor, Through Hole, 1/4watt, 1.43Kohms, .1%, 10ppm.

Hooked it up, and the charger works great! No more temp compensation messing up my voltage adjustment. 10PPM stability will keep the resistance to within +/- 2.1 Ohms over a +/- 30 C temperature swing. Very stable.

17

Many Thanks for this valuable info detailing how to convert a lead-acid GEM 825e to LifePo4!

I’m in the process of converting my 825e to a (25) CALB 100Fi setup, but I need some help with selecting the most appropriate charging algorithm to have ElCons install in my Zivan NG1 charger.

Clearly the goal is to maximize cell life by avoiding over-charging and while
Electric Conversions has presented me with a long list of wildly different charge algorithm-options, they have been has been most UN-helpful otherwise…:rolleyes:

Jack Rickards at EVTV appears to be a very knowledgable source and provides a recommended charge profile for these 100 Ahr cells (available on their store site @ blog. evtv. com /store… ) – which summarizes to:
V4 = 3.6V
i2 max = Ah/1
i3 = Ah/15

This profile avoids over-charging by cutting off the charger at i3 = 100Ahr/15 = 6.67 amps

Unfortunately none of the profiles offered by ElCons match Rickards’ recommendation exactly. :confused:

The closest seems to be ElCons’ “500” algorithm which provides:
V4 = 3.6V
i2 max = Ah/1
i3 = Ah/60

However this profile doesn’t cutoff until current falls to i3 = 1.67 amps, so I’m concerned that this profile will over-charge my (expensive) LifePo4’s leading to premature death. :eek:

I’ve repeatedly asked Mark at ElCons if it’s possible to amend the 500 algorithm to install an i3 =Ah/15 but he has yet to reply to any of my emails and I can’t ever get him on the phone.:shocked:

Does anyone know if ElCons is able to install the custom charge algorithm I want in the NG1 or am I limited to just the options ElCons has offered?

Also, is the NG1 actually capable of delivering an i2 max = Ah/1 (=100Ah for these CALB 100Fi cells) as ElCons 500 algorithm would imply…or is the NG1 limited to an i2 max = 16Ah, as suggested in one of the posts in this thread?

Perhaps I’ve overlooked something…?

Thanks for any insight you can offer.

18

Nice. You have the latest CALB cells there. How do they fit? My single 5 cell string only fits under the hood, in the front-most battery position, after I cut away the battery tray plastic and just set it directly on the angle for support. I imagine with 100 Ah cells, you might actually get then all to fit under the seat alone.

Welcome to the wonderful world of Elcon. Here’s the thing: Elcon offers no lithium profiles for the Zivan chargers. They will make one for you-- for $600 IIRC. We have to get by using one of the stock Zivan profiles that they already have (mostly LA, but some nicad, ect.) You tell them what Elcon profile you want, and they find the closest Zivan profile they have. The reality is, you are going to get whatever they choose to give you anyway, so don’t fret about it, and just work with what you will be given. In the pictures in this thread is attached the charge profile (supposedly) that was used by Elcon on my Zivan NG1.

The nice thing is, we can go into the Zivan and tweak the voltage and current pots to whatever we like. The Elcon chargers are locked in stone.

Just let them know what you want, and accept however they choose to program it best. Once you get it, you will have to monitor the first few charge cycles to make sure it’s not overcharging the cells. You will probably have to go into the Zivan and adjust the voltage and/or current pot to tweak things.

You may or may not have to replace the temperature probe with a temp. stable resistor, as I had to. I asked that the temp probe be disabled, but found out it still was messing up my voltage adjustments.

Lithium batteries are easier to take care of and charge than Lead, they are just much less forgiving of abuse.

From your message, I can’t tell if you’re going to be using a BMS or not, nor if you are top or bottom balancing.

If you are not using a BMS, and are going to bottom-balance the cells: Upon charging the pack, find out which cell is the smallest (one that rises to the highest voltage first). Monitor this cell, trimming the voltage pot in the Zivan to insure that cell doesn’t over-charge. The other cells will be somewhat undercharged, but all will have stored the same amp-hours of energy. It will take a few charge cycles to find the right voltage adjustment to get the pack as full as possible, without overcharging the weakest cell. I find the charge profile does not act like what is shown on paper. You may or may not notice that as well. In any case, you will be able to tweak things and make it work for you.

The ultimate solution would be installing this replacement chip in your Zivan, making it fully user programmable. With it, we could program exactly the profile wanted. I don’t know if he is still selling them. Also, these only work as-is on the NG-3 right now, but it’s believed they should work on all the Zivan models, once the programming is tweaked for them. I have one of these chips, but have not yet played around to get it working on an NG-1.

Jack’s target of 3.6V works better on chargers that have more ideal charge profiles than what Elcon will put on your NG1. I find 3.5 Volts per cell works better for me on my 160Ah cells. Try it low (like 3.4 V per cell), and increase it little by little each charge cycle, until your lowest capacity cell (highest voltage cell, if bottom balanced) settles to about 3.38 Volts max, 24-hours after being charged (and main disconnect turned off, so no parasitic loads on the pack.) Plus, don’t forget that the controller in the GEM is limited to 86 volts, or an average of 3.44 Volts per cell with 25 cells.

Also, If you don’t need all the range your cells have to offer, undercharging will extend their life further. No need to charge the pack to 100% if you don’t need to use all that range for your trips. Undercharging extends pack life, unlike with Lead where it leads to shorter pack life.

This reply is getting long, so let me know if I forgot to address a question you had.

19

Thanks so much for your detailed reply, (which I just found as I went through my email–it was a busy week…). It’s a pleasure–and very reassuring–to have the benefit of your experience and insights!

I’m excited about the new CALBs too…just got them and they’re still sitting in boxes in my garage, so not sure about the fit yet. I hope to find time to pull out the NG1 & the old LAs this week so I’ll keep you posted.

I do plan to bottom-balance (no BMS), but I’m still exploring exactly how best to do it (in the car, on the workbench…?). It sounds like the process is pretty manual so I’m looking at getting a PowerLab 6 or 8 ( revolectrix . com ) on Jack’s recommendation, as a way to get it done efficiently and this device looks useful for monitoring/documenting/analyzing charge etc as well.

I particularly appreciate your insights on charge profiles & process with the NG1 & on working with Elcon. The charger algorithm profiles Elcon sent me to choose among look very different than the graphic of what they installed on your NG1. Per your suggestion I will look for an algorithm that tops at 3.5V and then tweak. (btw, I had no idea I could tweak the NG1 post programming. Is there a Zivan manual out there that details this process? The manual I found on the Zivan site doesn’t address this feature.) Once I get my NG1 reprogrammed, I’ll need to assess what they’ve actually installed and locate the current/voltage potentiometers. I’ll look forward to comparing notes with you on this.

I am still wondering about max charging current on the NG1: is it actually 14 amps (as your charging profile graphic indicates) or can it go higher? It seems desirable to maximize charging amps in order to minimize charging time, and Jack believes this won’t have a negative impact on the cells.

Btw, How long does it take to fully charge your 160Ah cells from 80% discharge (I expect my 100Ah charging time will likely be a bit less…)? Also, what sort of range are you getting with the 160s? I’m excited to see what I can do with my new CALBs here on the hills of San Francisco…

And thx for the heads-up on the replacement chip. Sounds ideal, if it’ll work. By the time I’m ready to try it, perhaps you’ll have gotten yours to work. (fyi their site says they now only sell in lots of 10 minimum.)

Once I get the system up & fine tuned, I’m hoping to include some 12v power for a sound system, seat heat and some LED lighting. Still researching whether to go with a DC/DC step down transformer, or to install a completely separate 12V system, (to prevent parasitic loads from unevenly discharging my main power pack). Any thoughts would be appreciated!

20

[QUOTE=timo;15481]Thanks so much for your detailed reply, (which I just found as I went through my email–it was a busy week…). It’s a pleasure–and very reassuring–to have the benefit of your experience and insights!

I’m excited about the new CALBs too…just got them and they’re still sitting in boxes in my garage, so not sure about the fit yet. I hope to find time to pull out the NG1 & the old LAs this week so I’ll keep you posted.[/QUOTE]No problem. I’m excited to have another lithium powered GEM running around somewhere (yours will make a total of three LiFePo4 powerd GEMs that I’m aware of, not including Jack R’s which he only put together temporarily as a test.)

It may take some time to get the NG1 reprogrammed. It took them over 2 weeks to get mine back to me. For a month or so, I left the LA batteries installed and charged them individually with a standard 40 Amp car battery charger.

Just jumper together the two wires that have spade connectors that lead to the NG1 (these wires just hook into the AUX relay on the NG1, and inhibit the GEM from operating while charging). The GEM will then function without the charger being installed.

I do plan to bottom-balance (no BMS), but I’m still exploring exactly how best to do it (in the car, on the workbench…?). It sounds like the process is pretty manual so I’m looking at getting a PowerLab 6 or 8 ( revolectrix . com ) on Jack’s recommendation, as a way to get it done efficiently and this device looks useful for monitoring/documenting/analyzing charge etc as well.
If you are willing to spend the money, why not get a PowerLab! I’m cheap, so I just used big fat 300 Watt 0.05 Ohm resistor to do the bulk discharging manually on each cell. I watched the first one really closely, but after that I knew about how long I had until I needed to disconnect it for the other cells. Good thing I never forgot. :slight_smile: IIRC, took me about a day to do each cell. Then went back over all of them again in the final day. Worked well enough for me, but a PowerLab would be neat to own, and certainly makes the job easier.

I particularly appreciate your insights on charge profiles & process with the NG1 & on working with Elcon. The charger algorithm profiles Elcon sent me to choose among look very different than the graphic of what they installed on your NG1. Per your suggestion I will look for an algorithm that tops at 3.5V and then tweak. (btw, I had no idea I could tweak the NG1 post programming.
Elcon originally sent me very different graphics as well. I think what you are probably seeing is Encon profiles. They will then select something from their stock Zivan profiles that is close to what you choose.

The Elcon chargers have a lot more flexibility on what settings they can adjust. With the Zivan profiles, they can only choose the value for I1 and U1. The rest of the values are locked to a ratio of those values, depending on the profile. For example, in the profile in my NG1, I2 is always 32% of whatever I1 is adjusted to, and U2 is always 92.9% of what U1 is set at. Why not just let use choose from all the stock Zivian profiles? I don’t know why, but they don’t like disclosing them. They are much more open to sharing the elcon profiles.

Elcon will give you something close enough to work with, but it will need tweaking to be right, especially if you are bottom balancing (I wouldn’t let Elcon know that, btw.)

Is there a Zivan manual out there that details this process? The manual I found on the Zivan site doesn’t address this feature.) Once I get my NG1 reprogrammed, I’ll need to assess what they’ve actually installed and locate the current/voltage potentiometers. I’ll look forward to comparing notes with you on this.

I am still wondering about max charging current on the NG1: is it actually 14 amps (as your charging profile graphic indicates) or can it go higher? It seems desirable to maximize charging amps in order to minimize charging time, and Jack believes this won’t have a negative impact on the cells.

Btw, How long does it take to fully charge your 160Ah cells from 80% discharge (I expect my 100Ah charging time will likely be a bit less…)? Also, what sort of range are you getting with the 160s? I’m excited to see what I can do with my new CALBs here on the hills of San Francisco…
It’s not in the manual, because you’re not supposed to be in there touching anything. :slight_smile:

Yeah, I get 14 Amps out of the NG1 with voltage trimmed to 85 volts (1200 watts-- 20% higher than stock.) I drilled a hole in the case so I could adjust the trim pots without having to remove the charger from the car (but I still have to take off the top half of the dash.) I’ve been running at 14 Amps for a year without a problem, and it gets 100-110 degF out here in the summer, so the charger seems to be fine with it. I have blown the internal fuse inside a Kill-a-watt meter (after using it without issue for some months), so it must be drawing a peak power very close to what a 15 amp 120VAC socket can provide (1800 watts).

While charging, I adjusted the current trim pot until current would not further increase. Don’t keep turning the trim pot past the point where the current stops rising, as the charger seems to act strangely in that case. Also, watch out when you have the NG1 open-- the metal parts are live with 120 VAC! That’s anther good reason to drill the hole for access.

Don’t worry about charging your cells too fast with your NG1. Your cells are rated at 1C quick charge, and C/4 for standard charging. Your puny NG1 is practically a trickle charge! The 35-40 Amps available from the NG3 would be a better charger for someone using most of the pack capacity each day.

Regarding your question on how long it takes to charge: I generally don’t take my batteries down that far, but I do from time to time. From 80% DoD it takes probably around 12 hours. Longer than I would like, but that’s a lot of driving to get the pack that low.

And about my range? IIRC, I’m averaging about 145 watt-hours/mile. So with 11.5KW-h available in my pack down to 90% depth of discharge, I have a range of about 80 miles.

Do you currently drive your GEM on those insane SF hills? Do you have the stock 5HP motor on your GEM? No problem with overheating, or too slow going uphill, or too fast going down? With those hills, you want to max out the pack voltage by using all 25 cells. A 9HP motor upgrade could make sense (unlike most that get them just for extra speed.) Reprogramming the controller might also help with performance and tweaking the regen settings might be a good idea while you’re at it. I can’t think of a more challenging city for any EV.

And thx for the heads-up on the replacement chip. Sounds ideal, if it’ll work. By the time I’m ready to try it, perhaps you’ll have gotten yours to work. (fyi their site says they now only sell in lots of 10 minimum.)
Interesting. Maybe ask them to place you on a waiting list until they get another nine people interested?

Once I get the system up & fine tuned, I’m hoping to include some 12v power for a sound system, seat heat and some LED lighting. Still researching whether to go with a DC/DC step down transformer, or to install a completely separate 12V system, (to prevent parasitic loads from unevenly discharging my main power pack). Any thoughts would be appreciated!
The current 12V DC/DC converter used for the lighting is already sized to handle a stereo. And everything you listed, except perhaps the seat heating.

For the seat heating, I’d probably use the full pack voltage along with
Flexwatt heat tape under the seat/backrest fabric. You could probably use a standard light dimmer switch to control the temp. Might need one of the newer electronic versions, that a compatible with LED and CFL bulbs. Then you need a solid state relay or contactor to turn it on/off.

Here are some other links I have regarding heating (talking about heating battery boxes, but same basic idea.)

Underfloor Heating | Radiant Ceiling Heat | Radiant Heating
Buy Flex Watt Heat Tape for Less
Battery Box Design - DIY Electric Car Forums

Oh yeah, have you thought about what you are going to do about a fuel gauge? The built-in SoC meter will always read full charge with the lithium pack. If you need a suggestion, I’d point you to the high current version of the Cycle Analyst with the 0.5 Ohm shunt. For under $150, you can’t beat the features. I think it could be mounted right on the top of the instrument pod, maybe using the two screws that hold the left/right halves together. Or as you plan on adding a stereo, right above the stereo would be another good location.

Good luck, and we want pictures, or it didn’t happen!