What is voltage sag?

It's the particular chemistry of battery you are using trying to impregnate you through the wrong orifice. Look up Peukert effect http://en.wikipedia.org/wiki/Peukert's_law
 
It also stumbles over the Low-Voltage-Cutoff (LVC). If you drain your battery too low, the extra-low state of charge can damage the battery, so...most controllers have an LVC setting to cut the power when your battery is low.

If you have a large battery, then, when you accelerate, the voltage may dip a tiny amount because you are drawing so many amps from it. Usually not a problem. If your battery is too small for the application, hard acceleration will cause the voltage to dip so low, that the controller thinks the battery is low and then it cuts the power (not realizing its just a temporary dip).

Sealed Lead Acid (SLA) is the worst chemistry, regardless of the size of the battery. Even if your commute is only a couple miles, you may need to have a 20-mile battery to avoid heavy voltage sag tripping your LVC when you accelerate from a stop on a steep uphill and your pack is at a half-charge.

LiPo is the least affected by voltage sag. They have a high C-rating (Current), so you only have to buy a pack with the minimum miles you need. Of course a bigger battery means you can charge it up less often, and also a bigger battery causes the loads to be more gentle on the pack leading to the maximum possible cycle life.
 
Voltage sag is basically a lowering of voltage in response to electrical resistance.

All batteries have differing levels of resistance inside them, noted as 'internal resistance'. You will get some sort of voltage drop when you apply a load ( IE hitting the throttle on your bike ! ), and that amount varies on the IR of the battery you have, the higher the amount of amps required, the worse it is.

That is why you see "C" ratings on batteries. It is, loosely, an indication of their internal resistance.

The more energy your controller/motor demands, past the C rating of the battery, the drastically worse the voltage drop will get. Lead Acid batteries have such bad voltage sag, that they invented a term for it, lol.. Peukert's effect.

One example would be if you ran a set of 5ah 1C cheapo cells at a 2C draw ( 10amps ).. instead of giving you 36 volts, they'd give you more like 24-32v, limiting your top speed ( and beating up on the poor batteries too )

Make sense?
 
Thankyou, thankyou and thankyou for clearing that up. You always get more than you bargain for on ES :D

and Yes, it makes sense!
 
Sorry to resurrect this old thread but is the way of designing out voltage sag to make sure the battery pack can deliver enough current?
 
You either have to upgrade to a cell that provides more current, or use the same cell and make the pack much larger with more cells in each parallel string.

Also use wire that is large enough to avoid the wire being a bottle-neck.
 
Builditgood said:
Sorry to resurrect this old thread but is the way of designing out voltage sag to make sure the battery pack can deliver enough current?

Voltage sag is always going to exist, but to minimize it you oversize your pack, or at least oversize the discharge rating. At a bare minimum you should be selecting cells that are rated at continiously delivering at least twice the current draw that you will draw from the pack. For example if your controller can draw up to 25A, your battery configuration should be able to do 50A+.

Also higher voltage lower current motors will have less voltage sag across the conductors. E.g. A 100v 10A motor would be generally much more efficient overall than a 10v 100A motor.
 
Thanks. I've got a controller that requires 50a continuous and 80a peak at 60v so have paired it with 16s1p 16ah multistar lipo pack that I'll charge to 60v (rather than its max of 70v) to hopefully give some extended life. The cells are rated at 10c continuous and 20c peak so my maths tells me the ratios are 50/160a or 0.3 of continuous discharge and 80/320a or 0.25 of peak discharge. I know lipos suffer less from voltage sag than lion and lifepo4 but would designing a pack with these safety margins minimise voltage sag for all chemistries?
 
Is it not a good idea to only charge them to 3.75v per cell
 
Builditgood said:
Is it not a good idea to only charge them to 3.75v per cell

3.75v/cell on hobby packs means they are mostly empty (say 10-20% SOC).
 
Thanks liveforphysics. I read somewhere that it is a good idea to put an extra cell in so the pack only ever gets charged to 90%. For example if you need a 14s build a 15s. The problem with this is it seems is that since the cells start off effectively partially discharged they are already someway along their discharge curve.
 
Builditgood said:
Thanks liveforphysics. I read somewhere that it is a good idea to put an extra cell in so the pack only ever gets charged to 90%. For example if you need a 14s build a 15s. The problem with this is it seems is that since the cells start off effectively partially discharged they are already someway along their discharge curve.


I think you may have your voltage mixed up with your amp hour/capacity.

Multistar cells should be charged to 4.10 or 4.15v . When they get down to 3.6 when left to rest for a minute or two they are pretty much empty.

As stated above in another comment, add another cell in each of your parrallel groups for more capacity, rather than adding more voltage. This will make your battery sag less as there is less stress being put on it when pulling amps out which causes the sag. Trying to pull too many amps from a cell cause sag.

By the way, multistars are good for a CONTINUOUS discharge of about 3C. Short period peaks will be higher, maybe 5C. But my experience with them shows that i get quite a bit of voltage sag when doing a 4C load..... 5 or 6 volts on a 20s 30ah pack made from 4s 10ah multistar bricks. When the pack is on 76v (3.8 per cell) the sag brings the cells down to 70v(3.5 per cell)which is about as low as i like to personally see. My lvc is 68v.
 
Back to the original question,, I like to think of voltage sag under load as heating your cells. In general this is true, if you pack sags a lot under load, the cells are warming up.

Fix this by adding more cells in paralell is one way. Lower the c rate by adding cells or paralleling a second battery to the first, and they tolerate discharging better. Without losing so much to heating the cell due to its internal resistance. Pull hard on a cell with high resistance, even more becomes heat than if you pull easy on it.

Or, get better cells. Or lower the draw/ go slower.

Re undercharging,, I like to undercharge just slightly on my lipo packs,, so a 4.2v cell gets 4.15v. If the ride will be short, sometimes I only charge to 4v. No less than that.

On that pack, I do like it to be 14s vs the current standard for 48v of 13 cells in series. No change in sag,, but the extra cell does bring the lowest voltage you get at the end of a ride up by 4v which is nice at that point. I don't like the 13s pack so much, once it's voltage is drained to the last 10-20%. Below 46v is pretty sluggish.
 
If your pack has a BMS, then...it is my understanding that the cells should be charged to 4.10V per cell at a minimum for the BMS to accomplish any cell-balancing. Due to common manufacturing variances, the internal resistances of the individual cells will vary slightly.

I'll make a gross exaggeration as an example. If a charger bulk-charges a common 13S / 48V pack to 4.2V per cell, then the whole package is charged to 4.2V X 13S = 54.6V. However...one string of cells might be charged to 4.1V per cell, and another string of cells might be charged to 4.3V per cell. The combined average is 4.2V, but some of the cells are actually going to be damaged by the high voltage.

With some BMS's it's hard to see if it's "doing its thing", and...you may even need to charge to 4.15V per cell to get adequate balancing accomplished.

Another thing, The published "continuous" rating of each cell is what I consider it's "peak" current output, under real world conditions.

20A 25R
15A 30Q
10A PF
etc
 
many thanks all. it sounds like the way to design a lipo pack (so as to not over charge it and give it a better life expectancy) is to have the average cell charged to 4.15v off the charger.

e.g 60v /4.15v = 14.46 cells. round it down to 14 cells so max pack voltage = 14*4.15v = 58.1v. putting another cell in series to make it 15s would reduce the cell voltage to 60v/15 = 4v which is getting close to where the cells start to get sluggish. so it is better to keep the cells at a higher voltage if possible and then add a second pack in parallel if more current is required to reduce sag.

i'm hoping to bulk charge the pack using this charger from em3ev. http://em3ev.com/store/index.php?route=product/product&path=35_37&product_id=120
the problem is that it only says it's for NCM/Li Ion not lipo. i guess i should use battery medics and low cell voltage alarms as well
 
First of all the usual warnings.

Don't store, or charge RC lipo packs where you would not build a fire. You don't go falling asleep while that stuff charges, store it in a basement under your bedroom, etc. My own setup is most of it stores in an old refrigerator in my yard. What I'm actively using stays in a fireplace, where I charge it. This is no joke,, houses have burned from lipos that caught fire. Don't let them get dinged, or puffed much and keep using them. They will slightly puff as they age, but one cell really swells, it gonna burn.

All that,,, might be more for the other guys reading,, you may have lipo experience.

Charging to 4v,, that's just leaving too much capacity on the table for me. I already leave a lot of it on the table at the bottom most of the time. I don't want to carry 10 ah, only to have 8 ah,, even if most of my rides only use 6. Sometimes I want 9 ah distance.

I have the same charger, more or less. I set the highest voltage to 58v. for 14s pack. This gives me a bit more space at the top. I can have a cell .12v out of balance, and it still remains below 4.3v when I charge it. I tend to keep my packs in storage at around 3.7v per cell after a use. Pack at maybe 50v or so. Night before a ride, I charge to 55v, then finish in the am with a 58v charge.

No bms,, this is NAKED charging and discharging. So I must check individual cell levels at top of charge fairly often. If I do discharge 100%,,then I know the pack will need a balancing. When I do a deep discharge, I put a voltage alarm on the weakest pack once the whole pack is below 50v. If I stop at 49v,, no need for the alarm because I know if the pack is balanced or not.

Any pack that will stubbornly go out of balance constantly gets taken out of the 14s assembly. I won't tolerate ANY weak cells in my lipo packs. TOO dangerous to do that.
 
Thanks Dogman Dan. I hear what you're saying and it is all invaluable advice that I've taken on board but please forgive me for pushing on this one point. You said "Charging to 4v,, that's just leaving too much capacity on the table for me. I already leave a lot of it on the table at the bottom most of the time. I don't want to carry 10 ah, only to have 8 ah..." which is all understandable but what if you had 4no 4s lipo packs? Would you put one of them through the trauma of splitting it in half to get 14s or just charge the 16s to 60v? Thanks again
 
I spoke to a mate who is into RC cars and he said there shouldn't be a problem using an under charged pack. There was no science to his answer though :?
 
Builditgood said:
Thanks Dogman Dan. I hear what you're saying and it is all invaluable advice that I've taken on board but please forgive me for pushing on this one point. You said "Charging to 4v,, that's just leaving too much capacity on the table for me. I already leave a lot of it on the table at the bottom most of the time. I don't want to carry 10 ah, only to have 8 ah..." which is all understandable but what if you had 4no 4s lipo packs? Would you put one of them through the trauma of splitting it in half to get 14s or just charge the 16s to 60v? Thanks again

I've been there with the 14S conundrum and 4S packs. Gotta use 6S with 2qty 4S -or- 2S and 3qty 4S. It sucks but I can practically guarantee 3.75/cell top SOC will suck even more. Do not top charge 16S of RC Lipo to 60V and expect to ride much more than around the block.

Can you guess which of these 8 RC Lipo cells will run out of charge and quickly dive off the cliff before the others?

View attachment 1
1-8DischargeZend.jpg
If you picked purple, you win.

Plus, a "resting" 3.75V/cell is much much different SOC than a "loaded" 3.75V/cell as shown in the graph above.

Don't, please don't...
 
Thanks ykick. Charge them up to the top it is then - minus maybe 50 millivolts for a factor of safety
 
Just want to add to the previous information with a real world example to minimize voltage sag of otherwise saggy cells.
Below is the battery in my cargobike, and the cells are crap for high discharge applications.
But put enough cells in there and each cell will have an easier job for a specific power output, and it becomes a low discharge application.

In this case, the battery only need to put out 0,2-0,5C for enough performance.
This also means I typically ride 5-6hours on a single charge. The drawback is ofcourse the weight.
Which is less of a drawback on a cargobike. The main reason for this compromise, is to be able to use cheap cells and still get decent energy density + life expectancy.
This does in turn give the required distance between charges and removes range anxiety.

The cell being used is the LG D1. 3000mAh 4,35V capable cell.
I charge the cells to between 4,15V and 4,2V depending on occasion and expect to have around 2500mAh if fully discharging (down to 3,0V).
This still equals 4kWh+ of energy.

Today there are more cheap cell options than 3years ago!

VLJ46_zpslqntvrjk.jpg
 
Nice battery Wheazel. So the trick is to increase the number parallel of packs to get the max discharge of each cell down so that sag is negated. For lipos this appears the max discharge should be < 1/3 their c rating. I don't know if that's the same for 18650's etc.
 
That is generally correct, but as others have said earlier, the voltage sag will not disappear, it will just become smaller.
Just like when you are hunting efficiency in any system, you will never reach 100%.
If you parallel enough cells, something else becomes the bottle neck. (maybe the main cabling, or connections inside the battery for example)

I see 18650 cells as better manufactured, better qualitycontrolled lithium cells, more skewed towards the energy density direction compared to rc lipo.
With todays lithium technonlogy, there is always a compromise between energy density and charge/discharge capabilities (very much simplified).
It is no coincidence that high drain 18650s generally have lower capacities. The same applies to rc lipo. Look at the multistar cells compared to 50C rated cells.

Therefore it is just as important to choose the right cell characteristics depending on application.
 
Back
Top