VESC voltage cut-off values considering voltage sag

I rode my dual-VESC board today which has a 10S3P battery made out of Samsung 25R cells. The cut-off values are set to start at 35V and end at 33V - I chose those values to keep my cells healthy and happy (I also charge to just 4.1V per cell instead of 4.2V). However I noticed some serious performance loss climbing a hill because of the voltage sag driving my cells almost down to 3.3V. What kind of values are you guys using so that the performance stays usable while not hurting the cells too much? @whitepony @Ackmaniac maybe you can help out here?

When charging the battery received only ~4500mAh until being fully charged back to 41V. So there must have been quite some juice left.

1 Like

Your cells are rated to 2.5V as cutoff voltage. But from 2.8V to 2.5V they go down rapidly. So i would recommend to set the cutoff start to 30V (3.0V a cell) and the cutoff end to 28V (2.8V) a cell. And just handle them normally and have fun. It think it is not worth to save some cell life to get some more charge cycles. If you think you will do more than 10000 km with this battery then it is OK. Otherwise i wouldn’t care. People are always scared about their batterys instead of heaving fun with them. And in your graph i can see that you only use like 18 amps in maximum on the battery for each VESC. So that is 36A in total. Your pack would be good for 60A. So you are not even pushing the limits. I only would keep the voltage at 3.85V a cell when i want to store them for a long time. Otherwise you should enjoy them instead of being worried all the time.


here’s the curve for 15A.

So yeah, at 33v youre using almost %50 capacity. I would set it to 3.1v and 2.8v. At 3.1V you still have 10% left. And if you stop them at 2.8v you will keep them plenty happy since they go down to 2.5v

E: 15A/cell. Translates to 45A at the battery. So this is a little conservative but there’s no graph between 10 and 15 online.

Graph for 10A/cell

1 Like

there is something else going on, this is not normal sag! using a loop key? spin up the motor without load and check live monitoring. i had a very similar problem with my melting loop key on the trampa the moment strong currents were flowing. i bet something in your setup is getting really warm :grinning:

I am using a Vedder Antispark Switch with a 40A fuse. Battery current is set to 20A in each VESC and Motor current set to the recommended value by Jacob (40A per Motor).

this is not voltage sag, reduce your setup to an always on direct connection to vesc and try it again!

that aside, my settings are soft off 33V, hard off 30V! I have next to no sag with lg hg2!

How can I try this again when this was while going uphill? And what should I look for in the no load scenario?

just try revving your motors on your desk while vesc is hooked to your comp - i could see that melting loop key voltage sag the moment a stronger current started flowing (in the vesc voltage live monitor)! a normal setup wouldnt show any sag in that scenario. if you cant reproduce that issue on your desk, rule out uneccessary components like the vedder switch -> always on direct connection vesc to battery.

Ok I opened it up with the limited tools I have right now (I am at the parents in law over christmas and new years). I cannot connect the VESC directly to the battery because they use different connectors (VESC XT60, Battery 4mm female bullet) and I don’t have the equipment needed - not even a Multimeter :scream: I noticed however that the ground bullet connector was not inserted all the way - seems like the vibrations might have loosened it. This is before opening it - the slow increase in voltage worried me a lot.

After pushing the bullet connector in all the way again:

Would love to test it now but it is already dark and freezing cold :cry:

that was the problem in another melting loop key thread!

LG HG2 are significantly better at handling voltage sag than 25Rs. Here’s the comparison.



So although the cause may have been a bad connector acting as a fuse. The voltage sag on his setup is real.

1 Like

And again, this is why I’ve gone back to lipos. Li-ion cells sag a lot. You need a massive pack to get rid of sag. As a result of sag, you end up getting less range, along with stressing your cells more. An HG2 at 20 amp continuous and going up to 100% and close to 0% will only get 400 charge cycles. Obviously, your not pulling that all of the time, but it will have an impact on the cells life. Instead of the 1500 cycles, I have had about 150 charge cycles and have experienced quite a bit of decreased capacity.

Until a better battery technology emerges, we are stuck with either less safe batteries (lipos) or lots of sag (li-ion).

There will always be out liars that don’t have problems. Likely due to the fact they are light, and ride up little to no hills. You can’t deny though that wen I climb a 20% grade hill, I pull 80 amps+ on a 12s4p. I sag a lot, and I could have predicted this by just looking at the 20 amp discharge curve of the cell.

you end up getting less range -> you need to get the cutoff right. Watt hours are watt hours. Doesn’t matter if Li-Ion or LiPo. As far as i know you did set the cutoff to 3.3V for each cell which is too high and a result of my first post of this thread. stressing your cells more -> You don’t stress the cells more because that is what they are build for. will only get 400 charge cycles -> that is roughly 10000km for a 10S4P

The only difference will be your max power output compared to a Li-Ion and the max speed because of the motors kv which depend on the voltage…

So at 80 amps under load for a 10S4P this would be 20A for each cell.

Fully charged Li-Ion can handle 3,7V at 20A load: 3,7V (cell under load) * 10S * 80A = 2960 Watts

Fully charged Li-Po can handle 3,9V at 20A load: 3,9V (cell under load) * 10S * 80A = 3120 Watts (160 watts more)

Close to empty Li-Ion, so that we reach the voltage cutoff for the VESC at 2,8V: 2,8V (cell under load) * 10S * 80A = 2240 Watts

Close to empty Li-Po, so that we reach the voltage cutoff for the VESC at 3,3V: 3,3V (cell under load) * 10S * 80A = 2640 Watts (400 watts more)

So the the LiPo Batterys can deliver more power at different discharge levels. But that’s it. So if you only want to blast up the steepest hills then it makes sense. But i think many people don’t even get close to those watts. And if the cutoff end is set wrong for Li-Ion cells (like 3.3V) then you reduce the battery to half of what it would be capable. So 2.8V would be right.

your post looks like an answer to @evoheyax post - not mine. I was the one using 3.3V as cut-off though - not him.

We had this discussion before and found out that he also set this value as the cutoff. Sorry to use this thread here for that but i want to avoid that the myth gets created that the Li-Ion cells are not good.

Most people who say there Li-Ion cells perform badly just simply had the wrong settings. You just simply need to get the settings right with the cutoff start and end and then the Li-Ion cells work great.

Lipos can do a bit better but they are more dangerous and you need more knowledge about them. If you know what you do then it is fine. But if you don’t stay away from them.


It’s cool - just thought you confused him with me.

How are LiFePO cells with voltage sag? I just ordered a space cell.

Space cells are not lifepo but liion afaik. They even use the exact same cell as I am (samsung 25r). I think it depends on the cell just like with liion. A123 cells for example are good for 70A - so will also have low sag.

Sorry, I don’t mean to hijack your thread.

Just to clear things up, I used to have no low power cutoff. I ran them till I sagged and my lcd Volt meters turned off under sag (they bounce back up to 3.2 or so after the load is removed). I still had shitty range (and less than my smaller lipos). My point is with liion, you can’t get the cells down to 2.8 when there’s no load. When your at 3.3 or 3.2, under heavy load, they sag down to 2.8 instantly. My point exactly is that they can sag so badly, that you can’t even use half of the battery. Yes, it’s normal for li-ion to sag like this, the charts tell it all. But this amount of sag is not ideal for heavier guys who live in a hilly area. It’s so unfavorable that I call it a problem, because I can’t sell a board to the community with a battery that performs like this.

What other settings would I have wrong? I could lower my max amps from the battery more, but then I’ll stall on steeper hills (or go up at 10 mph).

Wait, in the enertion website, on the SPACE cell 3 page, it says Battery Chemistry: LiNiCoMnP