Polaris Ranger Forum banner

1 - 20 of 158 Posts

·
Registered
Joined
·
202 Posts
Discussion Starter #1 (Edited)
Lithium DIY

All these talks about Lithium conversions made me want to try one. Voltronix kit is nice, but quite a bit spendy, and offers way more range than I need. Why pay
extra for something I won't use ?

I had some conversations here with other forum members to figure out the direction. One thing I wanted to avoid is having to reprogram Sevcon, since
cheap Canbus adapters are nowhere to be found, and "old sevcon software" thread turned into some kind of joke. So I wanted to try to match the Lithium power
pack to the factory Sevcon programming, and figure out the charging problem later.

For the Lithium cells Nissan Leaf modules are quite nice, and I may end up using them, but there one can either do 14S or 16S configurations, which could
be either too little or too much. So instead I decided to do a trial run with a small pack built out of Ford C-Max hybrid cells.

Those are 5.5Ah (or so) cells made by Panasonic. They are rated at 150A continuous discharge (at least that's what the seller claimed), and for both the
discharge current and capacity reasons I decided to go with 4 modules, 15 cells each (switching to 14 later depending on findings).

Below are the pics of the build. Next step is to send the DeltaQ charger for reprogramming to get a correct charging profile going on, and then I can
start doing some range tests. I haven't removed the lead-acid batteries to test for worst case (loaded) scenario. Also note the use of improvised parts -
trying to make it cost efficient :)

















 

·
Registered
Joined
·
255 Posts
On the topic of improvised parts:

- I have used the copper pipes squashed flat, drilled and crimped for cable lugs. My crimper (chinese knockoff) works well with the copper since it is nice and soft. The best (least heat) connection between these improvised lugs, that I have experimentally found, is to stack a flat washer on your stud (or bolt), then stack 2 or more lugs all touching each other, then another flat washer, a lock washer, and finally a nylon lock nut. Tighten the lock nut until the lock washer starts to compress but stop before the washer is completely flat. Not sure why. Just an experimental result.

On the number of cells in your pack:

- If your charger is charging to 55.4V, 14S sounds like a good number of cells for your pack. Wasn't it the low end that dictated 15S? If the SEVCON cuts out at 44V, 14S gives you 3.14V per cell. That's leaving some battery capacity on the table. Your 15S makes the SEVCON drop out (if 44V is right) at 2.93V so it should perform until your BMS decides to shut you down. Without the hesitation that I was seeing.
 

·
Registered
Joined
·
202 Posts
Discussion Starter #3 (Edited)
I think you're right. I did a few runs today with 14s and 15s configurations, and 14s performed quite well. At one point it got charged to 100%, probably caught it before charger went into trickle mode ?
Either way, I did two trips, 1.2-1.3 miles, spending about 40% of the charge each time. Difficult terrain, some uphill. That's while still carrying 700lb of lead-acid batteries. Fun thing about having a small
pack like this is that it charges back up extremely fast :) So next step is charger programming and dropping dead weight. Will post an update.

More thermals right after the last trip:





 

·
Registered
Joined
·
392 Posts
You really cant discharge any 5+ah cell at 150a. Seriously you should aim at around a 3c max discharge. Not 60c, which is what you would have to achieve if your controller was pulling 300a under heavy load. there are no cells made that will deliver 30+c continuously, and to get any usable range you should aim for at least 160ah . If lifepo4 (3.3v nominal)then 16 cells. You must have some form of BMS to monitor the pack and deal with over voltage / charging as they age, and under voltage under load.

There are good reasons why conversions cost as much as they do.

Please note poor crimps are one of the biggest causes of failure and fire. Proper crimps and a crimping tool are not expensive. They are also plated which helps to ensure no corrosion at the joints. Dont solder any of the main power cables, if you get resistance and heat you can get the joints to come apart.

Also dont forget to constrain your cells, They must be stopped from swelling as once they swell they will lose capacity very quickly.

Good luck with the project. But always remember that the vehicle must be safe when left on its own charging, and when being operated by someone who has no idea whats in it and that they have to watch out for low cell voltages.
 

·
Registered
Joined
·
202 Posts
Discussion Starter #7
You really cant discharge any 5+ah cell at 150a. Seriously you should aim at around a 3c max discharge. Not 60c, which is what you would have to achieve if your controller was pulling 300a under heavy load.
I found the government report for that vehicle - https://www.energy.gov/sites/prod/files/2015/02/f19/batteryC-Max3817.pdf

It has a 7.6kW pack made out of 84 cells, with a 310.8v pack voltage (nominal) - it appears it's a 84s configuration, without anything paralleled. The doc the goes to state :

The Medium PHEV battery target performance goals of 37 kW discharge power and
25 kW charge power are shown as a dashed line.
So the pack is designed to discharge at up to 37kW. At nominal voltage that is 119.35A, which is the current on each cell due to pack's series configuration.
In my case there are 4 cells in parallel, meaning the pack is capable of handling at least ~480A of current. 119.35A though is likely not the cell rating,
and I am more inclined to think stated 150A by the seller is actually accurate, though I haven't found the official spec sheet from the Panasonic yet.

So the report above doesn't really stress peak vs continuous discharge, but I found another one :

https://avt.inl.gov/sites/default/files/pdf/phev/fact2013fordc-maxenergi.pdf

This one uses 50.2kW figure for peak draw from the battery. Given the same configuration in 2013 vehicle, that's 161A draw per cell...

Basically, it is a certain different game here when relying on cells made for some electric vehicles with specs significantly exceeding the needs of the Ranger
due to sometimes order of magnitude more powerful drivetrains. One of the things I am trying to do here is to figure out a simple way of achieving the
conversion with reasonable results.

I also have another update on the project: just got the Lithium algorithms programmed into the charger. It now supports the following:

#67: 49.2V
#128: 54.7V
#211: 55.44V
#163: 57V
#53: 58.392V
#123: 59.88V
#164: 60.792V
#135: 62.256V
#177: 65.688V

If anyone is interested in that, shoot me a message and I will share the contact info for the shop that does the programming.
With so many options it is now possible to chose between 12, 14 and 16 series configurations. Last one may require DC-DC converter replacement,
but apparently they can be had for as little as $80.

Good reminder about the crimps and stuff - I took certain calculated shortcuts, but I'm very diligent otherwise.
 

·
Registered
Joined
·
392 Posts
If the pack is 7.6kwh (thats not 7.6kw which is a power figure not a storage capacity) and with a constant voltage of 310v, a 37kw discharge rate would give approximately a 5 c discharge (c being the nominal Ah of the pack). So if the cells are as you stated earlier 5+ah, then thats less than 30a peak.If the pack is 7.6kwh divide that by 310v and you get an ah of the cells of 24.5. So if your cells are 5+ah the the pack must have at least 4 in parallel to get the ah,which means each parallel string of approx 25ah could at 5 c do 125a discharge. Thats achievable. You really cant get normal cells of 5+ah to discharge at 150a.
 

·
Registered
Joined
·
202 Posts
Discussion Starter #9
Correct me if I'm wrong, but I think you're arriving at incorrect results by using C-rate as a way to determine discharge current. Given it is a rate, it is tied to time, and specifically to 1 hour. So yes, drawing 7.6kw for an hour will amount to 1C discharge. Drawing 37kW is not 5C though - it is 5 times 5c, because the pack will last only for 12 minutes at such power draw.

That's where going by basic I = P/V is less confusing, it gives the answer to the instant current value instead of the rate of any kind.

With that said, I think I pulled the specs on the wrong vehicle - cells I have are from HEV, not from PHEV. I will have to follow up on that.
 

·
Registered
Joined
·
255 Posts
Correct me if I'm wrong, but I think you're arriving at incorrect results by using C-rate as a way to determine discharge current.
It is common for a lithium battery to have a max discharge rate based on C. 3C, 10C, 20C are fairly common. 3C would be most common. But it appears that the rate has much to do with how much heat the battery can dissipate. If you discharge at 3C it will only last 20 minutes. 10C, 6 minutes. 20C, 3 minutes.

Your ranger will accept whatever current is required from the battery pack and will put out 650 motor amps ... for maybe 10 seconds to get you up to max speed. After that, current draw drops substantially. I'm told that 100A (what my clamp-on meter shows ... sort of hard to read since it bounces so badly) is high for 'cruising', likely due to dragging brakes, suspension mis-alignment, etc.

Your 4P setup, at 88.9A, would ask for about 22A per pack on 5A cells is 4.4C. That is a bit aggressive, and the batteries may heat up pretty quick. But you only get 13.6 minutes of runtime at 4.4C. And they would cool off while you do whatever you are doing in the yard, after you get to where you are going. Before you put out more energy to get to somewhere else.

But under 15 minutes of runtime spread out over 4 hours should be fine. Given that no one trip is long enough to melt the solder off your cells or put the electrolyte into thermal run-away ;)
 

·
Registered
Joined
·
202 Posts
Discussion Starter #12
Your 4P setup, at 88.9A, would ask for about 22A per pack on 5A cells is 4.4C. That is a bit aggressive, and the batteries may heat up pretty quick. But you only get 13.6 minutes of runtime at 4.4C. And they would cool off while you do whatever you are doing in the yard, after you get to where you are going. Before you put out more energy to get to somewhere else.
The number 88.9A is derived from Ford intending to draw 25kW from a pack that consists of 76 cells and having nominal voltage of 281.2. Considering that voltage, their pack doesn't have
any cells in parallel, meaning the current on each cell is in fact 88.9A, and that's not even the peak according to the docs. So when we 4P such cells, we should be at least at 350A line.

Again, I could be misinterpreting the data, and another pair of eyes on it wouldn't hurt.
 

·
Registered
Joined
·
255 Posts
The number 88.9A is derived from Ford intending to draw 25kW from a pack that consists of 76 cells and having nominal voltage of 281.2. Considering that voltage, their pack doesn't have
any cells in parallel, meaning the current on each cell is in fact 88.9A, and that's not even the peak according to the docs. So when we 4P such cells, we should be at least at 350A line.

Again, I could be misinterpreting the data, and another pair of eyes on it wouldn't hurt.
Sorry.

88.9A rated, should by around 350 for 4P, but much higher for a few seconds of acceleration.

Where did the 5A+ per cell come from ... I missed it somewhere.
 

·
Registered
Joined
·
202 Posts
Discussion Starter #14
Where did the 5A+ per cell come from ... I missed it somewhere.
Oh, the cells are 5Ah (capacity). The concern was that anything above 5C (or even 10C), in other words above 25A, is way too much discharge current for such small cells. I honestly don't know, I am simply
looking at how those cells are used in the vehicles they come from.
 

·
Registered
Joined
·
392 Posts
I used the use of C to illustrate how a 5ah cell cannot reasonably discharge at 150+ amps and live. I also used your 7+kwh pack size and stated voltage to show that they cant just be being used as a single 5a cell, but must be multiple 5ah cells in parall. So using them in a single serial string isnt going to give the desired power that the motors can pull.
In lead acid they used two strings, not so much for the storage (yes its a factor) but to also get the high discharge levels the motors require, at a reasonable voltage. Lead traction batteries do struggle to deliver even 2c and have a decent life. Yes you can pull more but the life is reduced a lot. (High current flows cause the lead plates to warp and lose capacity eventually shorting and losing a cell)

Lithium cells get damaged most when they are heavily discharged under load, or even at moderate loads when they start the rapid voltage decline in the last few ah of life.

So the key to any lithium pack life is to have loads of Ah, never fully charge to absolute max and never discharge anywhere near the drop off point. Thats what OEMS do you dont get anywhere near the max/ min voltages. As the pack ages the spread of voltages at which the cells go high / low voltage will increase. But they mostly stay fairly close voltages through the bulk of the charge discharge cycles. If you force a lithium (any variation)cell low voltage under load, they will swell and once swollen they lose capacity that you cant get back, even by re compressing them (Some cells you cant re compress) Even the most expensive RC Lipoly cells will struggle to do more than 30c for short periods.

Voltronics and myself do conversions with a lowest ah of around 160ah cells in the pack. These will deliver 200amps no problem, they will deliver 300a with a low voltage drop, but we dont let the discharge get too low/ high, its just not worth it when you are selling a product with possibly a decade of useful life if treated well. But above all the cells are not heavily stressed by huge discharge rates at or beyond their design. Even between the Lifepo4 cells there is a lot of difference. Thundersky / sinopoly are virtually the same as CALB, but CALB seem to have a higher quality, and do have a higher discharge rate. And curiously also a higher initial ah, if you test them they are way over the stated ah, which means if you re test them a few years down the line they are probably within the original claimed spec. Ie a new 200ah cell with have probaly 210ah and still have 200ah a few years on. so ten years down the line they will probably still return 175+ amps. So on spec a loss of 25ah is good going. Cheaper cells will usually just make design spec. my tests on new 160ah THundersky cells does give around 170+ ah. A lot of the cars I work on here in the UK have early 160ah thunderskys, a few have failed but most at 8 years old are still returning 140+ah. To prolong their life I change the final charge voltage they were originally given of 4.2v down to 3.65v and they last even longer. When new charging from 3.65 to 4.2v adds only around 2 to 3 ah,just not worth the associated problems of pack balance and cell life.

Good luck with the project, be careful, especially when charging, dont overcharge, if you dont have a proper control system while experimenting, dont turn your back on them during charge. constantly measure each cell, until you know how they work and the cells with outlying voltages, watch those especially closely.
 

·
Registered
Joined
·
202 Posts
Discussion Starter #16
Good luck with the project, be careful, especially when charging, dont overcharge, if you dont have a proper control system while experimenting, dont turn your back on them during charge. constantly measure each cell, until you know how they work and the cells with outlying voltages, watch those especially closely.
One of my initial goals was to figure out the minimum power pack size for the use case. It seems like 1.2kW is too small under heavy load (though would work for light duty), so I am considering going with bigger system. Currently there is a great deal available on LG-based 45Ah modules (2.5kW) arranged as 16S - that is too much with the default Sevcon programming, but I will be taking my Sevcon to the shop to get reprogrammed, and may as well set it up for use with 16S configuration. I will need to change out DC-DC, but that is a minor change. Also my DeltaQ already can correctly charge 14S and 16S without overcharging, so the only remaining concern is cell balancing - I am looking at some solutions for that.
 

·
Registered
Joined
·
255 Posts
I used the use of C to illustrate how a 5ah cell cannot reasonably discharge at 150+ amps and live. I also used your 7+kwh pack size and stated voltage to show that they cant just be being used as a single 5a cell, but must be multiple 5ah cells in parall. So using them in a single serial string isnt going to give the desired power that the motors can pull.
In lead acid they used two strings, not so much for the storage (yes its a factor) but to also get the high discharge levels the motors require, at a reasonable voltage. Lead traction batteries do struggle to deliver even 2c and have a decent life. Yes you can pull more but the life is reduced a lot. (High current flows cause the lead plates to warp and lose capacity eventually shorting and losing a cell)
Thanks - somehow I missed that part.

... be careful, especially when charging, dont overcharge, if you dont have a proper control system while experimenting, dont turn your back on them during charge. constantly measure each cell, until you know how they work and the cells with outlying voltages, watch those especially closely.
+1

Excellent advice on pretty much anything you experiment with! Watch it for a while and make sure it's doing what you expect it to. Watch it, measure it, take video, log data, whatever you need to do.

And for us paranoids ;) .. put in a completely separate system - simpler is better - to shut everything off if something goes wrong in the primary method (BMS) .. like a timer on your charger.

I want to log, in the BMS, when the BMS gets shut down before it is finished. That way I know that something went wrong with the BMS and I need to investigate. But it didn't burn down my shed!
 

·
Registered
Joined
·
392 Posts
Currently there is a great deal available on LG-based 45Ah modules (2.5kW) arranged as 16S
But still the pack really isnt large enough. Why go for such a small pack? Its stressing the cells way too much. Its no benefit to have a decent cell and then load it way way to high. Aim for c or 2c under use. With these cells you are still regularly expecting the cells to continuously discharge at 5 c or greater. Thats not a good recipe for any long life. And it will cause the less than perfect cells to fail prematurely. If there is a good deal on these cells buy enough to make it a worthwhile conversion. ie 4 parallel and whatever pack size you need to go to . You will then end up with a vehicle with a decent range, a long life and a performance thats not just going to let you down when cells start to fail at being asked to constantly heavily discharge.

I constantly battle with people who think a small pack will do just because thats all they can currently afford. But as soon as they build the vehicle and try and use it, the low performance and short cell life come back to haunt them. Long term it then costs a great deal more, as you cant just add new cells to an already compromised pack. If you use a BMS with balancing on the Parallel cells it will take your decent cells down to the level of the poorest, and on charge the pack will stop charging when the poorest starts to go high voltage. You lose all round.
Please think again, you could end up with a really nice conversion. Or something thats going to be a constant headache and after a short while start to eat cells.
 

·
Registered
Joined
·
255 Posts
But still the pack really isnt large enough. Why go for such a small pack? Its stressing the cells way too much. Its no benefit to have a decent cell and then load it way way to high. Aim for c or 2c under use. With these cells you are still regularly expecting the cells to continuously discharge at 5 c or greater. Thats not a good recipe for any long life. And it will cause the less than perfect cells to fail prematurely. If there is a good deal on these cells buy enough to make it a worthwhile conversion. ie 4 parallel and whatever pack size you need to go to . You will then end up with a vehicle with a decent range, a long life and a performance thats not just going to let you down when cells start to fail at being asked to constantly heavily discharge.

I constantly battle with people who think a small pack will do just because thats all they can currently afford. But as soon as they build the vehicle and try and use it, the low performance and short cell life come back to haunt them. Long term it then costs a great deal more, as you cant just add new cells to an already compromised pack. If you use a BMS with balancing on the Parallel cells it will take your decent cells down to the level of the poorest, and on charge the pack will stop charging when the poorest starts to go high voltage. You lose all round.
Please think again, you could end up with a really nice conversion. Or something thats going to be a constant headache and after a short while start to eat cells.
That's quite a battle you have - long term thinking is not rewarded in North America. I'm not sure if it is any better on you side of the pond ...

In my opinion, no one wants 'cheap'. But *EVERYONE* wants 'lowest cost that does the job'. I have noticed that salespeople that supply equipment to our industrial site lead with the high-end systems and features .. then show the scaled-back systems that are just fine, and do the job, later. But no one wants those. And the really cheap stuff is mentioned at the end and is added to the price sheet, since no one wants to buy the cheapest one, and no one wants to buy the most expensive one, the middle of the road one is most often purchased ... unless there is a technical reason to disqualify.

We have some sales personnel that use the same model number for the highest cost, with all of the add-ons and options, with the middle system is the base version ... but you can buy the options later ... sort of making a non-modular system into a modular system of a sort. Another product line entirely is used for the cheap option.

Is there a way to do modular with Lithium packs? Without stressing the crap out of the first cells, and requiring a new battery charger, etc etc?

Like maybe using the smaller pack (say the 16S 2.5 KW pack, arranged as 4S4P ) as the 12V source for the winch and off-road lighting? Maybe run the future 120V inverter from there as well?

I miss the 'reserve' selector on a quad, that gives you enough fuel to get home in the case that you were not paying attention to the fuel gauge. Could the small pack with an inverter be re-purposed to transfer (inefficiently) power to the 48V pack through the charger ... and give you a 'limp home' range?

That seems silly. But I do *REALLY* miss the 'reserve' on a gas quad.
 

·
Registered
Joined
·
392 Posts
Its a bit easier here, as many things are more costly. Therefore throwing money away is a bigger sacrifice. So on a constantly used Ranger where the owners are replacing the lead every two or three years at around £2k a time, if you are in for long term ownership, then a lithium replacement at about twice the cost of a set of batteries is a good starting point. The big plus list for the conversion is the clincher, ie less weight by about 100kg+, no pack watering, better response through the charge (ie power doest start to drop off as the lead loses voltage) and potentially longer range if thats wanted.
You must use lithium carefully, and cold weather makes it even worse. If you need to constantly draw 150 amps then rate the pack accordingly. WHile cells can deliver higher than c (stated ah) and will deliver peak loads, thats not the way to use them. In cold weather high (c or C+) currents cause permanent damage to the cells by moving copper through the plates and blocking the holes in the carbon that the lithium ions move through. Its not reversible. So have a big enough pack to meet normal needs and dont stress it untill its warmed up. At 160ah you have just over an 8kwh pack. Thats really the good starting point and will give comparable performance to a well charged lead pack, and last longer. Add more ah and you will get better range, but little extra performance as the voltage drop change will be very little.

I just dont see the point in creating a vehicle that has a decrease on performance, range and reliability over a lead acid version. Yes you shed weight, and a lot of it, but end up with something thats gives nothing in return.
A bit like taking the fuel tank out of your car, putting in a motor cycle tank, with a tiny fuel pipe and then driving it.
 
1 - 20 of 158 Posts
Top