DIY Electric Car Forums banner

201 - 220 of 305 Posts
J

·
Guest
Joined
·
0 Posts
Then you only have about seventy things confused. The biggest one is that there is no NEED to avoid going below 2.5v during discharge. You have violently misread the manufacturers data.

You do not want to DISCHARGE the batteries to the point where the STATIC voltage - that is the voltage of the cell at rest, is below 2.5v. They will routinely sag in voltage under load and going below 2.5v is not a problem at all. But when you cease loading, the battery should rather quickly recover to a voltage that is ABOVE 2.5v.

That you cannot determine state of charge from voltage is comical. The curve IS relatively flat, if you are accustomed to monitoring lead acid cells. But this is a GOOD thing. It makes it very easy to determine SOC with quite a bit of accuracy. But you have to be able to monitor relatively small changes in voltage accurately. Buy a decent multimeter and you're pretty much there.

It is true that you can't measure it very well with a piece of wire and a five gallon bucket.

I have no concept of what your grousing about cable drops. They vary with current, and cable size and length. What's not to understand? I have a six year old grandson who understands it well enough.
 

·
Registered
Joined
·
8,109 Posts
You do not want to DISCHARGE the batteries to the point where the STATIC voltage - that is the voltage of the cell at rest, is below 2.5v. They will routinely sag in voltage under load and going below 2.5v is not a problem at all. But when you cease loading, the battery should rather quickly recover to a voltage that is ABOVE 2.5v.
Have you tested how low you can momentarily go before the cell won't recover to above 2.5 or so?
 

·
Registered
Joined
·
4,958 Posts
Actually I THINK the maximum C rating should not draw the cell below the low voltage cut off. This would naturally also mean that the closer to full charge the battery is, the more of that C rating you have access to before the voltage will draw down below the 2.3-2.5 cut off zone.

I performed what I believe was a 3C test on a battery that was rated at 5C and voltage dropped to about 2.62V per cell while under load with the battery starting out on a full charge. After the test, the voltage popped back up to about 3.3V per cell.

This is only my limited experience, and the test was on a prismatic LiFePO4 battery made by sieden instead of SE, so its not a perfect comparison, but from what I observed the battery should be able to deliver the rated peak current without dropping below the low voltage cut off by any significant amount.
 

·
Registered
Joined
·
44 Posts
Dear jrickard,

Lighten up a bit would ya. :)

My reference to a possible failure to understand voltage drop in cables was not aimed at you. I don't know anything about your cell testing work. Have you posted any?

This was just my guess as to the possible source of some problems in results posted earlier in this thread or in threads that lead to this thread. 12 gauge wire was mentioned. But don't get me wrong there either. Anyone who uses their own time and resources to do tests and post the results for the benefit of others deserves praise.

When I said voltage vs SoC was "too flat", I hope you understood I meant "too nearly horizontal", not "too linear". Yes, linear is good. Yes, you can determine LiFePO4 SoC from a rested cell at a known internal temperature. But that's not terribly useful in a vehicle being driven. There no way to know the average internal temperature of the cells at any instant (all you can measure is case or terminal temperature and there is a big thermal time constant) and there's no way to rest them for long enough. Integrating amp-hours with peukert compensation is more useful there.

Yes, taking a cell below 2.5 V (e.g. down to say 2.0 V) briefly under heavy load is nowhere near as bad as taking the rested voltage down there. But the idea that it does no damage at all is debatable. In any case, the BMS we plan to use does not allow the low voltage alarm point to be a function of current hence our desire to know more about the max discharge current vs temperature relationship at 2.5 V.

Now, my friend ... What are those other 68 things I am confused about. ;)
 

·
Registered
Joined
·
44 Posts
Have you tested how low you can momentarily go before the cell won't recover to above 2.5 or so?
Not tested, but I expect you could go to zero volts, a dead short, pulling maybe 25C from a Thunder Sky depending on temperature, provided it didn't last for more than about 1 second, and it would still bounce back above 2.5 V. Don't try this at home kiddies. Hot burny. :eek:

The question is more, "How much capacity do you permanently lose, and how much does internal resistance permanently increase, every time you do this, or any other momentary excursion below 2.5 V?"

There's nothing magic about 2.5 V either. Even when you stay above 2.5 V, the lower you take the rested voltage, the more capacity you lose per cycle. The same goes for high voltages on charge. It's all tradeoffs. They only have a limited calendar life anyway. For lead-acid the economic optimum was typically around 50% depth of discharge. At least with LiFePO4 it seems likely to be more like 80% DoD.

You can think of a cell in transmission-line terms. Some parts are electrically closer to the terminals than other parts. You can model it as a number of sub-cells in a ladder with resistances between. When you momentarily pull the terminals down to a very low voltage, the sub-cells electrically far from the terminals are still at the rested voltage while those close to the terminals may be fully discharged and undergoing destructive secondary reactions. Soon after you remove the load, the sub-cells near the terminals will be quickly recharged by those far from the terminals, but may have suffered permanent loss of capacity.
 
J

·
Guest
Joined
·
0 Posts
I've done quite a bit of testing of these batteries, and have about 5000 miles on the car using the THunderskys. I tested several types of batteries on GEMS before settling on Thunderskys for the car.

I do have a test rig that can do up to about 100 amps on a single cell, and about 45 amps on a 4 cell batt. It will maintain constant current and graph the voltage, or cycle the battery on and off in a kind of driving simulation.

The batteries do sag under load and it is rather variable. You can see this best not at the cell level, but across the pack simply because the voltage swings are larger and easier to measure. But they also vary from the fully charged condition and increase as you approach the lower end.

The temperature actually doesn't matter. Yes, I'm familiar with what it does, but in the end, you'll come to the same conclusion. Track it and compare it all you want, in the end the curve is the curve.

In practice, you don't have to predict the static voltage from the sag voltage. It's the sag voltage that is a moving target. You can easily see the static at the stop light. I don't wait all day to see what it EVENTUALLY recovers to. If you remove the load at the stop sign, and give it a couple of seconds, that's static enough. If you are below 3.0 volts actually, you don't have very long until it is 2.5 volts. You're already on the losing side of the curve - basically once the 3 disappears, it is time to be pretty close to the house. Their just aren't many amp hours between 2.99 and 2.5.

Damage accrual? Well you've got a couple of things going on here. The life cycles are different at 70% and 80% discharge, so of course the less you discharge them, the better. Under that philosophy, which is more or less inarguable, pick any number you like.

My point was do not be alarmed at 2.0 volts per cell when accelerating. But the 2.5v, while also not a religion, is pretty important at rest. Below that, we are advised that copper shunting can begin to form, and some lithium plating that will decrease the life of the battery. I think, while this is similar, it is also somewhat DIFFERENT from the normal life cycle curve from DOD. At 2.5v and below, the effect changes from gradual wear to more of a damage function.

And I was immediately confused when starting out about WHICH 2.5 volts we were talking about. The factory, and the distributor, both assured me separately, if not quite in the Queen's English, that it was the static voltage they referred to. The sag voltage of course has a lot of variables, current level, duration, age of cells, and current DOD of cells. So a measurement of cell sag voltage to static voltage is of limited utility.

The classic test is to draw a specific and constant amount of current from the cells and observe the cummulative amp hours provided during the test, and the changes in voltage across time. But I will routinely terminate the test at 2.0 volts expecting a recovery to ABOVE 2.5 in essentially every case.

So basically, among the 68 items, I'm seeing a LOT of focus on temperature, and I've personally not found it very useful. I guess it possibly would be in extremis, but 20-35C it just isn't very.

But I am developing a strong preference for Chinese vendors. I've tested maybe 20 TS 90Ah and 160ah batteries in 4 cell or single cell groups. I haven't found a 90Ah cell or battery putting out less than 94 Ah, and most are up around 100 Ah. I pulled 104Ah from one battery last week.

I've also developed some maturing views on BMS systems. First I couldn't find much in the way of BMS. Then I tried to do my own. Then I went back to shopping. Soon I decided they weren't very important.

More recently I'm coming to view them as actually dangerous. These batteries provide a LOT of power. And almost anything you do to it in a car, as opposed to on the bench, is a bit of a problem. You either have lots of light weight wires running everywhere (even at the battery level, much less this nonsense about monitoring individual cells). Shunt regulators are probably an outright fire threat much more dangerous than having a cell out of balance. And the computer systems are just laughable.

FIrst, you don't need all that information while driving. Second, there's nothing you can do with it while driving if you had it. And worse, it's not something you can explain easily to your daughter if she wants to drive the car.

But finally, they don't work. None of them. They just don't. I worked for DOD contractors for a dozen years, and the guys who do EMI shielding live in a world of their own, because they are all driven squirrely by the problem within just a few years. I actually laughed out loud when I saw the THundersky BMS. It was a touch screen, a computer, and about 80 lightweight wires to wire to each cell.

Gentlemen, in the rear of your car is a 50-150kW AM radio station, whether you are tuned in or not, sporting 80 light wires of varying lengths AROUND it is just comical. If you don't blow the computer entirely out of the water, the information it displays will be utter nonsense. Works pretty good on the bench, as long as there are no electric motors in the vicinity.

We use an EVISION with a very simple balance circuit that splits the pack in half and measures one side against the other. It displays a very simple LED bargraph. When it is centered, your pack and ALL of its cells will be balanced closely enough. It will deflect several bars over very minute variations in voltage across the two pack halves, indicating an imbalance. It takes a little experimentation to determine just what that is for any given pack voltage. But you get the idea pretty quickly.

If it isn't IN balance then it must be out of balance. We do have some terminal strips and wires bringing all the 4-cell batt voltages up for easy access with a voltmeter. If there IS an imbalance, we go measure all the batt voltages, which takes about 180 seconds.

IF a battery is out of line with the others, say more than a .3 or .4 volts, we pull the battery and take it to the bench.

And there we determine if any cells are damaged, out of balance, etc. Often we can bleed some or charge some and get it pretty close. Then discharge the battery pretty low. Then charge it up to the full 4.25 per cell. Then discharge it again. If it seems to be staying inline, we'll consider putting it back in a car.

But that does bring up another subject. Batteries are changing pretty fast. Don't buy a PACK for a car, without buying SEVERAL extra "batteries" of 4 cells each. It is kind of a bit of a problem after a pack has cycled a few dozen times, to marry an individual battery back into the pack. You have to kind of carefully balance it with the others. I use a little hand held 250 watt 5 ohm to bleed small amounts, and have a larger resistor array for more serious stuff. And I use a regular little 12 volt smart charger to "add a little" to a battery.

But over time, the battery business is changing. We're already on order for some Blue Sky 100 AH that have a different SE chemistry and entirely different voltages. I have a LOT of TS laying about fortunately. But it occurs to me that if I was three years down the road, with a scant pack in my car working pretty well, and I did have a cell go bad (approximately inevitable), where would I THEN get a matching battery - even in chemistry and capacity, never mind same batch from the factory?? Putting a BlueSky SE in a pack with TS batteries is out of the question. And that is the transition we've made in a year.

So I would urge you to provision SEVERAL extra 4 cell batteries if you are doing one pack in a car. These batteries have the potential to last 7-10 years at some level of capacity. But inevitably you will have problems with some cells before others. And over that period of time, the current chemistry and manufacturer not only may not be available then, but almost assuredly won't be available then. To have to pull a perfectly good pack and replace ALL the batteries because you cannot find replacements for a single cell seems absurd.

Jack Rickard
 

·
Registered
Joined
·
4,958 Posts
I have been looking at all the ideas about PCM integrated BMS and was wondering if it was really worth all the fuss. Since these batteries don't explode like LiCO2 batteries and are much more forgiving at the top and bottom voltage limits, I think its a little extreme to have an active management for each individual cell that is always working and micromanaging the battery pack.

The only thing my setup has is a low energy cell balancer built into the charger that comes on near the end of the recharge cycle, so there is one wire going to every connection, but after charging, the battery is on its own.

Pulling higher AH ratings out of the cells was something I observed as well in my limited testing. In my case I got about 5% extra, but from cycle life data that SE sent me (which I feel is plausible), the output can peak at 110% by the 50 cycle mark, then drop back down to about 100% by the 500 cycle mark, 90% by 1000-1500, and so on. Not bad IMO, not bad at all.

So you figure the shelf life is about 7-10 years? I think thats reasonable, since the older LiPos can sometimes last 10 years and LiFePO4 should be more stable.
 

·
Registered
Joined
·
1,148 Posts
So I wonder what will happen with significant regen if the only balancing is in the (mains) charger. The battery will get most of its charge from the mains, but probably most of the extreme charge current and voltage from regeneration.
 

·
Registered
Joined
·
44 Posts
I've done quite a bit of testing of these batteries, and have about 5000 miles on the car using the THunderskys. ... <much wisdom> ...
Thank you so much for that post, Jack. A distillation of real experience like that is worth its weight in gold. :)
 

·
Registered
Joined
·
401 Posts
I have been looking at all the ideas about PCM integrated BMS and was wondering if it was really worth all the fuss. Since these batteries don't explode like LiCO2 batteries and are much more forgiving at the top and bottom voltage limits, I think its a little extreme to have an active management for each individual cell that is always working and micromanaging the battery pack.
There is at least some risk. See http://www.batteryvehiclesociety.org.uk/forums/viewtopic.php?t=1825
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #212 (Edited)
I've seen them short circuit a single cell of Lifepo for over 10 min (the TS brand) and it not do anything but smoke.

Likely the fire originated from the heat itself due to something else.

However active management is not necessary to prevent it if it was a short... relays can be employed, fuses, etc.

I'd suggest that was one in a million shot, or the manufacturer lied about the chemistry ... and probably improper battery placement.
 
J

·
Guest
Joined
·
0 Posts
>There is at least some risk. See http://www.batteryvehiclesociety.org...pic.php?t=1825<

This is an excellent idea. Read it carefully and read between the lines. This IS one of the cases I was referring to about burning a car to the ground. The guy had a "BMS" all right. Shunt balancers mounted on each cell.

I've played with these a lot. And if you are set on them, I'll tell you what to use. On eBay there is a whole bunch of RC helicopter guys using balancers from Hong Kong. Typically, they will "balance" six to eight cells. They do so at the 250 milliamp range. And they are sufficiently well designed that I don't THINK they will go thermal - largely because they are so small. Over the course of a week or two, they will help balance your cells, very slowly. They don't do any particular good while charging, but note that your car speeds a lot of time just sitting around.

I played around with a variety of shunts, most of which I destroyed. I designed and built several prototypes myself of a pretty nifty circuit that could actually shunt 7 or 8 amps and had a very precisely controllable turn on voltage. They could be produced for about $15 per cell. And they would allow you to totally balance your pack during charging.

The problem is the power. Semiconductors tend to change characteristics when faced with rising temperatures. They have a very bad habit of rating these MOSFETs and so forth at very high power levels that are almost entirely based on your ability to heat sink them "properly" which means massively. But there is a limit to how much heat sink you can put on top of a cell, and of course, the heat sink gets HOT. I could get these circuits to work MARVELOUSLY when operating under the expected conditions. And they probably would work.

But if I cranked things up a bit, admittedly to somewhat unrealistic levels, they went into thermal runaway and the current took off to the sky - eventually burning the MOSFET to cinders and really pretty spectacularly.

My assumption is that if I can take a perfectly good one, and easily get it to burst into flames with a little overvoltage, what would happen if I had one that failed or wasn't good from the beginning. The failure mode is catastrophic.

I can current limit them using somewhat expensive high power resistors. But I've moved the problem. The heat isn't in the heat sink and semiconductor, now it is dissipated in the resistor. But it still gets hot.

Not unmanageably hot on the bench. But quite warm. The problem is, mounting them in a car, first I have 72 cells, that means 72 of them. Second, a car isn't a test bench. It rolls, true enough. But it also vibrates, shakes, rattles, groans, leaps tall speed bumps in a single bound, smashes into pot holes and curbs. Etc. Finding room and securelymounting such a circuit on 72 individual cells is a problem.

Ok, well, let's just do it at the battery level. Uh-oh. Now I'm doing 4 or 5 amps at 14 volts - we're up to 50 watts when it's working well.

The basic problem is that you have a LOT of cells, and we're dealing with a LOT of power, and in shunting it resistively, that is cummulatively a LOT of heat. Of course, we're charging at night, ideally while sleeping or otherwise distracted by the pleasures of the flesh. Errr... while our car is unsupervised and quietly burning to the ground. Ultimately, the batteries are more reliable than the circuits, however simple, that we design to protect them. And if a $2 MOSFET turns cherry red and gives up the ghost, we have a great possibility of burning a $50K car to the ground. All because someone told us we HAD to have a "BMS" when using LiFePo4.

Regeneration. Currently don't do it as I'm running a DC series motor. We are in the throes of the Mini Cooper Clubman which will do it. I see a LOT Of angst about regeneration overcooking the batteries. I wish. I have some ideas about curing the problem, but unfortunately it is not a problem.

First, the kinetic energy that keeps your car rolling down the road is at its peak efficiency when used as kinetic energy to keep your car rolling down the road. In other words, don't key your regen to your accelerator. Normally, if you take your foot off the accelerator, you will be more efficient letting the car glide than you can ever achieve by putting it back into the battteries.

We will kick in regen as kind of a first detent on the brake pedal. About enough to make the lights come on but before we are actually applying pad to disk. This will let you kick in regen with a slight touch on the pedal, with further pressure required to actually brake. And if you adjust the regen to a level where this results in a pleasant and useful deceleration, it just doesn't make that much current or voltage. You can't take energy out of a pack to accelerate a car, and then convert forward motion BACK to electricity to put in the battery, without a lot of loss. And we tend to brake, using regen, much more gently than we tend to accelerate. The result is that overcharging batteries with regen I personally believe to be not only a myth, but more akin to a fantasy. You can recover a bit of energy with regen. But real world range extension I would look for 7-8%. It's just not going to be a problem. You would have to come out of the garage, and immediately down a 4 mile long 10% grade to be able to get into a position where overcharge would be a problem. If you drive a block and a half before the 4 mile downgrade, it won't be a problem.

Every way I turn on this, it winds up being the CHARGER that is the issue. We're now using a Brusa NLG-11 and it allows us to program as many as 7 stages of charge alternating between constant current and constant voltage. We basically add energy quickly for most of the charge, and then finish with a kind of gentle alternating charge where we bump the voltage a couple of volts, hold it there until the current declines to nothing, and then shut down for 10 minutes or so. Then repeat it. And after doing this a couple of times, we'll simply hold it precisely at our target voltage for an hour at the end.

And the MOST key element, is that we do NOT charge these batteries to the specified 4.25 volts. Indeed, the manufacturer notes that you can extend the life of the cells quite beyond the 3000 cycle limit simply by charging to 4.10 instead.

Real world? We're charging to 3.75 volts. Up to that point, the cells are really remarkably well balanced. After that point, they go completely squirelly with different voltages all over the place, with some zooming to 4.5 while others don't reach 3.95.

So we just don't go there. We charge hard to about 3.6. And then we charge very gently and in a kind of alternating charge cycle up to 3.75 volts. From 3.75 to 4.25 on my pack, I figure I'm giving up 3 amp hours, or about a mile and a half range, to avoid all the problems.

I discuss all this in a video at http://www.youtube.com/watch?v=TUzPzX9A4lw.

IN the car DRIVING, I'm watching a couple of items. First is my pack voltage. Small variations at the cell level show up as readily detectable changes when you multiply it by 36. So I start at 129 volts or so, and QUICKLY drop to 3.3 x 36 or about 118 (when stopped). Then it will very gradually and predictably decrease to 108 volts (3.0 x 36), before dropping off a cliff from 108 to 90. Somewhere in that 108 to 90 range, I need to make the garage or source of AC.

Second element is the little balance bargraph. As long as it is within a dot or two of center, my cells are balanced. If it starts to creep away from that, I need to stop and put a voltmeter to the batts to see what's going on. This is very infrequent. But it has happened. Particularly when in the 100 volt range.

I DO have the ability to monitor amp hours into and out of the pack both when driving and charging and similarly kWhrs. That is mostly secondary information. For example, I have 180 Ah of capacity that is REALLY up around 200 ah. So if I have taken out 150 Ah and I'm down around 108 volts that all makes sense.

But if I've taken out 90 Ah, and I'm down around 108 volts, that really doesn't jive. I need to take a closer look.

Metric Minds EVISION is really a good instrumentation system. It's a bit of an install project. And it does cost about $750. Despite my earlier railing about computers, it basically IS one. But it isn't running data lines all over the car. It has a CAN bus from the board to the display. All other connections are analog. And indeed it can get twitchy at times. But it mostly works and at full stop, it reads pretty well. I would have the display brighter. In full sunlight it is not readable.

Jack Rickard

http://evtv.me
 
J

·
Guest
Joined
·
0 Posts
As long as I'm raving on about the problems with these batteries, for those new to the topic, I'm a HUGE fan of them. And again, the reasons are NOT what I thought they would be going in.

Yes, they are lighter. I've got 550 lbs in a car that if I tried to just put that many Trojan 12vs in would come to about 1500 lbs. And they have MORE capacity for that weight. I rather think they do NOT put out as much peak current. That may be the way I have it configured, but that's my sense.

The real benefits are all about life time and charging. We subliminally know a lot about batteries that is just not so with these batteries. Lead acid, Nicad, and Nimh all have some issues we are almost TOO familiar with. Things about completely charging them, and completely discharging them. Even basic lead acid batteries you do not want to charge them halfway and then go drive. Lead sulfation is very real and I've played a lot with the desulfators (they do work by the way).

LiFePo4 are a bit different. You NEVER have to fully discharge them. And really, you never have to fully RECHARGE them. Not ever.

What this means is that I go to the store, and come home. I plug in the charger, even if I'm going to be there 20 minutes. It's just not that hard to do. Then I go to a friends house, stop by the car wash, and then out to the airport. I plug it in there and charge it as well. Then home, back on the charger.

These batteries like that long flat area in the middle of the curve. They DON'T like the overcharged end, and they don't like the over discharged end. But you don't EVER need to go there, to condition them, to equalize them, to balance them, to avoid sulfation, to recondition them to eliminate memory effect or ANY of that. They just like it in the middle.

Life of 7 to 10 years? In reality of course I won't know for 7 or 10 years. They do have some calendar deterioration. But all the life cycle data are about how much capacity is lost over the life of the battery. The batteries will still WORK, but you won't have as much range. You are going to lose some by cycling. And you're going to lose some because of the trip of the planet around the sun. But I have more range than I need. By charging early and often, I think that is what we'll get.

Another little misunderstanding is that if you drive the car, and then charge it, you have "used up" "a cycle". No no no no.

In TESTING the batteries, by charging it completely, and discharging it completely, and doing that 1000 times, they can measure the decrease in capacity and show it to you on a graph.

You do NOT lose nearly as much, by orders of magnitude, if you do not fully discharge the batteries. So if going from 100% to 20% charge is a cycle. Then 100% to 40% twice is not a cycle. Going from 100% to 40% about 10 times, MIGHT be equivalent to a cycle. And going from 100% to 60% about 100 times might be equivalent to "a cycle." Avoiding the two ends, overcharged, and overdischarged, is how you extend life.

And so the most important things to me about these batteries is that they are basically "life of the car", part of the expense of the vehicle, not the operating costs.

And second, they are actually much easier to live with. IF you can get a good charger ( a bit expensive I'm afraid) that you can plug in and walk away from knowing it will CHARGE the batteries without OVERCHARGING the batteries (or burning the car to the rims). If you do so, you just plug it in a lot. No water. No conditioning. No topping. No equalization. No fooling with it.

You're basically down to not running them down too far. It is my belief that this is what the automobile manufacturers are struggling with, and the preoccupation with hybrids and range extension.

1. Nobody is going to buy a factory made electric car without a manufacturers warranty on the battery.

2. All batteries can be destroyed by overdischarging them, but particularly the LiFePo4 batteries, which are at this point expensive.

3. 10,000 motorists run out of gasoline per day NOW with ICE cars. The general public are idiots, and can't be trusted to not run out of gas. We already know this.

4. Everyone interested in an electric car is concerned about the range.

Simple. We put a gasoline engine in the car with a generator. If the battery voltage decreases to 3.0 volts per cell, we automatically start the gas engine and start charging the batteries. The driver can then get home without us having to buy them a battery pack. And we'll feed on their range fear by making this a FEATURE - more range.

I view hybrids as inheriting the drawbacks and complicating disadvantages of BOTH technologies in the same vehicle. But I would be very surprised to see much in the way of a plug-in electric production car that doesn't sport this feature, and for those reasons.

The alternative is simply to build a cutout in the controller, that disables the car when the static voltage reaches a certain minimum level. This is actually a WEE bit complicated, but can easily be done and is the proper solution. The problem is that if I miscalculate then, I can't just pour in a gallon. I have to have the car put on a flatbed towtruck and taken back to the house. Which would be annoying if I were two blocks away.

Jack Rickard
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #215 (Edited)
Simple. We put a gasoline engine in the car with a generator. If the battery voltage decreases to 3.0 volts per cell, we automatically start the gas engine and start charging the batteries. The driver can then get home without us having to buy them a battery pack. And we'll feed on their range fear by making this a FEATURE - more range.

I think the answer is just a less boxy, non-steel body... that'd fix the range problem... all for a cheap cheap price (battery costs of like $5000 tops for a 200 mile range if done right)

A solid BMS/charger system is way way cheaper than basically everything else.. at least if you're looking at a maximum 110v wall outlet and want to use all of it.

Talking like $300-400 tops for both ($100 BMS $200-300 charger)
 

·
Registered
Joined
·
56 Posts
I keep hearing about cheap, relaible BMS. I'll definitely be needing such, as my LiFePO4s are on their way. Some of the links go to systems that aren't made to handle my 30 cells, and the maker won't answer my emails.

So where can I really go for a BMS that will do the job with lithiums and not break the bank?
 
J

·
Guest
Joined
·
0 Posts
I agree, in looking somewhat out to the future, you are quite correct.

We are being sold a bill of goods on the cost of the electric car from the manufacturers.

With just ordinary economies of scale, the price of the car should actually be lower. No exhaust. No radiator/cooling. No smog controls. No "fuel system" (a touchy point with me. Suburbans and Escalades have a fuel pump IN the gas tank. They ALL fail. And it is $1500 to have one replaced.).

Etc.

But you are also correct in that the core proficiency of our current automotive manufacturers is the pressed steel body. If we want to be light, strong, and efficient, the current technology is of course carbon fiber.

I like the motor in wheel concept. Eliminate the entire drivetrain. PML Flightlink.

I have been personally STUNNED by the current levels required above 65 mph. I had NO IDEA. So aerodynamiics is important. I don't want an electric car to look like an Aptera frankly. But aerodynamics is important.

But I probably am on a different page with regards to chargers. THE most expensive item in my car after the batteries, and so of course the most expensive single item, is the charger. They are not $100. And as I've described, other than some instrumentation, the charger IS the "BMS" in my system, if you insist on my having one. It's a Brusa NLG511. We are using a NLG513 in the Mini.

They cost $3600 currently. The motor and controller COMBINED were about that. But that's where you need the technology. I could PROBABLY get a Manzanita PFC-30 to work at $2700, but I like the programmed charge cycles of the Brusa.

How it charges, how difficult it is to charge, how safe your car is while it is charging, etc. all center around the charger, not the rest of the components.
And of course, it is the limitation on charge time as well.

Most people are unaware of it, though it is spelled out on the data sheet, kinda/sorta in English, is that the batteries are NOT the holdup on the amount of time it takes to recharge. I can easily recharge my entire pack in 40 minutes with no harm, and if I'm willing to accept a little shorter battery cycle life, 20 minutes.

You can recharge these LiFePo4s at up to 3C quite safely. For a 90Ah battery, that's 270 amps. I have them in parallel, so all I need is 540 amps at about 130v and I'm good to go in 20 minutes.

Now I laughingly noted in one video that there isn't any 540 amps anywhere. But it occurs to me that there is. I use 540 amps in the car at about 80 mph. Where does it come from? The batteries.

So if I had batteries, and a controller, and some 2/0 cable, I ought to be able to do this.

So picture a SECOND bank of batteries, a little larger than the one in your car. And you charge THAT slowly with a little charger in the garage. You could pull in, hook up, and in 20 minutes you're near enough fully charged.

This then could be the model for charging stations. Not massive chargers. massive battery banks, with admittedly also large chargers. But the car to charging station interface is about connecting batteries to batteries. Not swapping batteries. And not "charging" for 10 hours. This could be done NOW with current chemistry. No breakthroughs.

In operation for me, I just don't need the range or the quick recharge. But you could do it now, and really rather easily. Separate high current connector on the car. Old abandoned Lead acids in the garage. And any old charger.

Hook it up. Use a PWM controller to turn it up. Check your voltage. Get a cup of coffee and take a pee. And turn it down. Disconnect and roll.

It will work NOW with the batteries you are using. Anybody could build one. Charge them with solar panels if you must.

Connectors? I would use two L6-30s and wire all four poles to the cable. Have one connector for positive, and one for negative. And hope it doesn't melt. Otherwise, just a couple of posts with wing nuts and standard lugs on the cables would work better, it would just be more to hook up.

Jack Rickard
http://evtv.me
 

·
Registered
Joined
·
2,123 Posts
Jack, with all due respect to you ( I read lots of your posts at the TS Yahoo list ) spending $3600 on a charger is INSANE, I don't care how good it is :eek:

My $1000 Zivan NG3 does almost everything ( OK, it has 3 phases, not 7) just as well.

Also, properly designed shunt regulators with feedback bus to the charger and the controller for HVC and LVC do not pose any fire risk. You just need to calculate Wattage properly and use it with correct charger, capable of low current at the end of the charge. Yes, regulators get warm, very warm, but not even close to a fire hazard.

You are telling a good story, but you seem to contradict yourself sometimes...

LiFePo4 pack can be designed properly with matching charger and BMS and it doesn't have to cost same or more than the pack itself.

I am very satisfied with my 40 cell pack, have been running it very hard with overnight charges, all cells are balanced and behave very well. I use Paktrakr, which is awesome, yes, it glitches sometimes due to EMI, but its mostly reliable, and I use my own revision of Brian's Volt Blochers with HVC and LVC feedback bus, feeding back to controller and Zivan charger.

When it comes to battery management, some people tend to go too complex and some too simple, and the truth as usual is somewhere in between, IMHO.

Still, I truly appreciate the time you take to share your info, its very helpful if one can read between the lines :)
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #219
Jack, with all due respect to you ( I read lots of your posts at the TS Yahoo list ) spending $3600 on a charger is INSANE, I don't care how good it is :eek:
spending $1000 on a charger is insane... even if it monitored 75 cells actively too (which his $3600 charger almost certainly doesn't do)

I personally am a HUGE fan of a simple analog BMS for each cell like the goodrum/fetcher for $100 (for 24 cells), because you don't need jack over that utility... it even has a cut off at "empty"

Not to mention the perfect charger for that BMS is just a straight power supply... which as I'm sure everyone realizes 1600w power supplies cost like $40 in parts.
 
J

·
Guest
Joined
·
0 Posts
Well, that's what makes the world go round. Different takes on the same data.

I would urge you to keep an eye on those shunt regulators. I have tested the Volt Blochers specifically, and indeed have drained a cell dead to zero volts with one of them. I'm not certain why. They are very similar to my LM431 design and ought to work.

As long as the shunts work, they work fine. But if one of them fails, I can assure you they DO pose a fire hazard, whatever your thoughts on the subject, and I've already put out several fires. Small fires on the bench fortunately. That's the point. You're trusting your car to a few semiconductor components which give off enough heat for you to note it, when working properly. Given the environment, I don't buy it.

As to the chargers, yes, the Brusa is quite expensive. And the Manzanita is quite expensive. And yes, I guess I could rig up a charger with some 50 amp full bridges I have sitting here, that indeed I ordered and have been eyeing for some days.

But note that you can get anything to work while your standing there watching it. The problem is that I need to be able to plug it in and walk away, or it doesn't work for me.

In truth, the only Zivan I have any experience with is the NG1. It was only "programmable" by sending it to the factory. I ordered a new one for a GEM specifying the parameters I wanted. After six weeks, I got an e-mail basically telling me I didn't know what I was talking about and that that would never work for SLAs or GELS. They hadn't even STARTED on it. They specifically noted they don't do LiFePo4s (if that is what I was thinking) So I cancelled the order.

If the NG3 can indeed be programmed for voltage and current, switches between constant voltage and constant current, and you get 3 stages, and terminates the process reliably, it would indeed be overkill. And I have had anomalies with the Brusa - it will magically decide the input voltage is about 350 vac and terminate and I've never quite tracked that down. Fortunately it is quite rare.

So I don't agree that $3600 for a charger is insane. But on the other hand, I don't have to carry a torch for them either. it's well made and has features I like, but I'd love a $600 version. If you can take that $40 worth of parts and make it do all that reliably, sign me up at $600. That ought to make you a pretty handsome profit.

What I don't understand about the Brusa is how it measures the battery voltages. And I would note that it doesn't do it precisely accurately. It typically reads a volt beyond what I've set before it switches to constant current. But the problem is that it is a little difficult to measure the voltage of a battery when you are applying a charge voltage to it.

I can only guess they have some way of dropping the voltage momentarily until current flow drops to zero, and measuring it at that point, bringing it back up to continue charging. But if I WAS to put some bridges and caps together into a power supply, I could probably make it programmable with an Arduino or something similar as a microcontroller, but how do I measure voltage of the BATTERY accurately WHILE I'm charging it??? It's kind of difficult to ramp the voltage output down until the current zeros out, take the value, and then ramp it back up to the voltage we're supposed to be charging at, once per second or so. And the batteries move around, slowly, when you do so.

And this IS the time your cable drops, mentioned a page or two back, start to get into the game to mess you up.

So I don't have a good charger design in mind. And that leaves me pretty much with a Brusa or Manzanita, and currrently I favor the Brusa.

Sorry, Zivan isn't happening for me. I have taken an interest in DeltaQs, but so far haven't seen anything in useful voltages - say 130-200 volts.

So I'm still looking for solutions and receptive to any information you guys have. But the propensity to repeat myths and admonitions endlessly starts to look like the usual attempt to "type yourself smart" by repeating each others advice. And that doesn't work for me. If I can go into the back of the garage with a piece of wire and a five gallon bucket and repeat your experiment myself, much better. The way this started was NOT to criticize the guy for doing it, it was to try to understand what useful information was being gained. The process was admirable. I didn't understand the test objective.

And I think he was actively learning some things I've kind of previously discarded as unimportant ( a view ALWAYS subject to revision in the face of new evidence I would note). I not only reserve the right to my opinion, but to change it regularly, frenetically, and in the face of all reason.;)

I've got a great design for a flying capacitor "charge shuttle" that will balance these cells quite well. But with the PCBoard, the clock, the drivers, the mosfets, and the caps, it's about $400 in components to do four cells. Oh, and sufficiently abused, it indeed bursts into flames as well.

Jack Rickard
 
201 - 220 of 305 Posts
Top