Speak EV - Electric Car Forums banner

Voltage sag Going the last few miles

5.2K views 14 replies 7 participants last post by  Automotive Division  
#1 · (Edited)
Hi. New guy here.
I have an electric car with a maximum range of 63 miles. I need some feedback from anyone who has seen the following effects.

I want the maximum possible range out of my EV but when the battery is weak (58 miles or so,) acceleration can be a problem. If I hold down the accelerator pedal too long the battery voltage drops low. The motor pulls a lot of current, and the motor may slow down as the battery rapidly discharges. I have also observed that if the motor stalls, it draws very high current and the battery is soon fully discharged. To get home on a weak battery I have had to accelerate in steps or pulses. I would lift my foot off the pedal for about a second and let the voltage on the battery come back up before pressing the pedal down again to accelerate. There is an alarm that sounds if the voltage on any cell is too low or too high. I have an AC-50 motor and controller with the regenerative feature and LIFEPO4 cells. This may be playing a role in this effect. I would appreciate hearing from anyone who has pondered the effects of voltage sag on batteries and motors. I have asked this same question elsewhere and I got back shrieks of how I am suffering from the effects of an illusion or that I am breaking the laws of physics, motion or thermodynamics. I can't break those laws so I frankly have stopped worrying about it. I am making observations and trying to learn from them to make my electric car better and to make my understanding better.
 
#2 ·
I have experience in this area. What you are seeing is fairly normal.

What car is it?

You say LiFePO4. How many, what brand and what size?

Given you are using an AC50 I guess you are using a curtis 1238?

Sounds to me like you are just running the batteries near flat. The voltage sags significantly more at a lower SOC which has a knockon effect on motor performance.
 
#4 ·
I have experience in this area. What you are seeing is fairly normal.

What car is it?

You say LiFePO4. How many, what brand and what size?


Given you are using an AC50 I guess you are using a curtis 1238?

Sounds to me like you are just running the batteries near flat. The voltage sags significantly more at a lower SOC which has a knockon effect on motor performance.
My car is a 1994 Saturn SC2 converted to electric. I have 2 banks of 36 cells in parallel. One bank is GBS 100Ah and the other bank is CALB 40Ah. Total of 16.8Kwh. Yes, a Curtis 1238 controller. Approx. 2580 Lbs. gross weight. The car performs very well but, I am one to press the limits and squeeze the last mile out of it because: I am working to improve the car. Anything that I can do to extend the range is of interest to me. The classic approach is to add more battery weight. If I can not buy a significantly more powerful battery, then weight reduction is where I intend to gain some advantage. However, back to this knock on effect, could any controller algorithm work to reduce/lift the sag? Could the right ultra capacitor help to get the car rolling and get the motor turning; building up motor reactance and limiting excessive current thus prevent rapid discharge ?
 
#3 ·
Yes it's completely normal if you pull a high C rate from lithium batteries when they are at a low SOC, voltage sag can be dramatic. In OEM EVs they will start to limit current when SOC is very low, and the BMS is smart enough (i.e. It doesn't just look at voltage in isolation) to know that voltage drop will occur in those situations and won't "panic". Lithium and especially lifepo4 will also recover very quickly, I have seen cells under load being 2.5V ping immediately back to 3.1V when the load is removed. That recovery is "fake" in as much as putting that load back will make the voltage drop again.

In terms of a fix for your EV, increasing the size of the pack would be a good idea so you are not routinely running the SOC so low.
 
#5 · (Edited)
Yes it's completely normal if you pull a high C rate from lithium batteries when they are at a low SOC, voltage sag can be dramatic. In OEM EVs they will start to limit current when SOC is very low, and the BMS is smart enough (i.e. It doesn't just look at voltage in isolation) to know that voltage drop will occur in those situations and won't "panic". Lithium and especially lifepo4 will also recover very quickly, I have seen cells under load being 2.5V ping immediately back to 3.1V when the load is removed. That recovery is "fake" in as much as putting that load back will make the voltage drop again.

In terms of a fix for your EV, increasing the size of the pack would be a good idea so you are not routinely running the SOC so low.
Yes it's completely normal if you pull a high C rate from lithium batteries when they are at a low SOC, voltage sag can be dramatic. In OEM EVs they will start to limit current when SOC is very low, and the BMS is smart enough (i.e. It doesn't just look at voltage in isolation) to know that voltage drop will occur in those situations and won't "panic". Lithium and especially lifepo4 will also recover very quickly, I have seen cells under load being 2.5V ping immediately back to 3.1V when the load is removed. That recovery is "fake" in as much as putting that load back will make the voltage drop again.

In terms of a fix for your EV, increasing the size of the pack would be a good idea so you are not routinely running the SOC so low.
Thanks. This is good information. I find that the recovery of both the battery and capacitors that power my car will drop off in a curve that can last a few seconds after I floor it again. Of course the battery can't last forever no matter what I do to tease some energy out of it. As long as I don't exceed 60 miles it is all good. The big issue really is the desired range and what battery I will need and then can I afford that battery. Are better batteries coming? I think that these are the questions that every car maker is asking plus should they use an aluminum chassis to reduce weight? Adding more weight could mean different tires, struts, and brakes. It is always a case of one thing leads to another thing.
 
#6 ·
There is no point in trying to get that last few miles from a Lifepo4 pack, if you are using a BMS its will be doing Top balancing, bleeding down high cells to get the finish voltages similar, so when you take the pack to its limit one or more cells will be going very low voltage, and under load. Thats what causes the cut back. But that point also causes the most damage to the cells. 60 miles is very decent from the pack. Accept that is a very decent range and stop trying to kill the pack. Do you use the BDI feature of the Curtis to give you indication of charge? What % does the cut back occur? Do you use heavy regen?
If you use heavy regen on the 1238, it just measures current, not direction and continues to reduce the BDI % even though you are putting power back into the pack. So doing the opposite to range to what you expect. So in this situation, the controller itself may go to one of its stages of cutback based on the preset BDI %, when you actually have much more than that in the pack. But this cutback is a speed limit rather than the kangaroo effect you are seeing.
 
#8 ·
@satronev you are really pushing your cells to their ragged edge, which is why you're seeing this. If it were lead acid or such then some voltage sag goes with the chemistry, but if you are getting it with LFP then you're going to kill your cells. They should be tough things, LFP cells (I certainly hope so!! [ @Mike Schooling ;)] ) but below 2V per cell you are going to start doing damage.

Buy more cells. Sorry, this is the only answer to more range, though you could also usefully look at aerodynamics and rolling resistance of your tyres. TBH you probably want to increase your capacity by 20% just to do 60 miles.
 
#9 · (Edited)
Yes, I think that you are certainly correct. I have also found that I have one weak cell that is falling below the acceptable minimum voltage before all the others at 63 miles out. If I replace the one bad cell the BMS alarm might not sound as soon. I find that one bad cell can really limit the performance of the car. By performance I mean not only hamster or, gerbil power but, range can fall off sharply. At best my motor has 73 hp. An alternative to adding more cells is to wait about 6 years for lighter cells with 3x the capacity. Some of the car companies are claiming that 300 mile cars and SUV's are in our future along with better batteries. I also have plans to replace the heavy steel hood of the car with lightweight carbon fiber composite. My experience with aircraft has taught me to build strong but light.
In any case, my work is only 30 miles from home. I can charge when I get there and charge at home. That is most of my driving. On weekends I love to attempt the impossible. Anyone planning a Michigan to Florida trip?
 
#11 ·
If you must take the pack very low then it is important that you bottom balance the packs to reduce damage to any one cell, and dont allow the BMS to bleed down at the top, which may mean not charging to the absolute limit. Its always a compromise, but that compromise can be acceptable if you dont go below around 20% and Stop charge at a reasonable level. Ie for thundersky with a bout a 3.33v static cell voltage dont charge much above 3.6v. On a 160ah cell you will get about 3to4 ah between 3.6 and 4.1v but the probability of a pack having cells going highvoltage to get that last couple of ah in is high. Having parallel packs of differening ah is also not an easy approach to adopt as cells will charge in different ways. They really need to be charged separately so that each string is charged to its best condition. If you replace a cell with a pack that uses a top balance BMS then change a cell when the pack is fully charged, and the new cell fully charged. If using bottom balance, then all need to be discharged to the start voltage and static before the pack is charged. When you put in a new cell ensure you follow the makers initial charge routine before usage.
 
#12 ·
Jeremy, as an approximation recognising it is just an approx, on the 160Ah Thunderskys, what open cell voltage would you say 20% charge is which you should avoid going below?

I'm aware that the manual for these cells actually says no harm will come to them so long as they don't drop below 2V, but we don't want them that low of course. I think it was saying to keep them at or above 2.8V? Still very low. I guess that is '0%'. Does that chime with your understanding?
 
#13 ·
Thats really hard to say, as a resting cell with 20% could easily see way over 3v even into the 3.2v they do bounce back when unused. I start a bottom balance at 2.75v and to get a flat cell back into this area probably takes less than 5 to 10 ah. Not a precise science. But in use they drop off even faster under load from 2.75, blink and a 60a load will cause them to go down to 2.5v But even at that they will bounce back to 2.8v . It takes ages to get a pack bottom balanced and stationary. I disagree with the taking down to 2v. Take these down that far under load, and I can almost guarantee they will be bloated. Thats one reason why the Calbs and Thunderskys were designed to be strapped / retained, as once they are bloated the plates get separated, and capacity if badly impacted. If strapped they the gas build up will pop off the safety valve . But unstrapped, I have seen these bloat by 15mm each side. But thats not all as a bloated cell can crush a cell beside it, and at worst can cause an internal short, causing heat damage. I have some that have shorts and melted themselves to the adjacent cells, so one cell going bad can cause multiple cell damage.
Treat all Lithium cells with respect and life is prolonged. Abuse them, and expect to get the max power extracted from them and life is not sweet.
Hence its just not worth trying to take these cells down into the absolute depths of discharge. Unless you wish to keep replacing them.
 
#14 ·
If you must take the pack very low then it is important that you bottom balance the packs to reduce damage to any one cell, and dont allow the BMS to bleed down at the top, which may mean not charging to the absolute limit. Its always a compromise, but that compromise can be acceptable if you dont go below around 20% and Stop charge at a reasonable level. Ie for thundersky with a bout a 3.33v static cell voltage dont charge much above 3.6v. On a 160ah cell you will get about 3to4 ah between 3.6 and 4.1v but the probability of a pack having cells going highvoltage to get that last couple of ah in is high. Having parallel packs of differening ah is also not an easy approach to adopt as cells will charge in different ways. They really need to be charged separately so that each string is charged to its best condition. If you replace a cell with a pack that uses a top balance BMS then change a cell when the pack is fully charged, and the new cell fully charged. If using bottom balance, then all need to be discharged to the start voltage and static before the pack is charged. When you put in a new cell ensure you follow the makers initial charge routine before usage.
All good advice. I bought a charger for single cells so that I could charge each one up to the very top before charging the entire set in series.
 
#15 ·
I realize this is an old conversation, but the subject is still relevant to many. There is a key that has not been said here: Your battery pack usable capacity is not going to be greater than the battery with the least capacity. If one cell is bad it will cause the BMS to shut down or "beep" as you indicated. You need to find that weak link and replace. It may not be the little 40s that are limiting, it could be an old 100ah that is even lower than 40ah. Your batteries are in series and subject to the weakest link.