New build help. Is this safe?

I bought a 5kw kit off of aliexpress(one of the QS models). It came with the 80 amp controller. Claims can be used with a variety of voltages from 48 to 72. Currently i am powering it with a 20s 3p battery comprised of 21700 40t cells with an 80 amp bms.
So i have some questions

1. Is this battery to small? am I damaging my cells?(I ride a fair amount on it around 30 to 40 mph( hit speed and cruise) it gets me about 10 miles if I'm riding like I'm driving bike weighs around 75 lbs I think I weigh 150.
2. I have a pack that I opened with 100 35e cells and plan to add another 100 for a 20s 10p 72vX35ah to fit into the triangle and move my controlled to the back. would this be a better or worse option?
3. If I took those 100 cells and made a 20s 5p and a 60 amp bms would this be unsafe as i have an 80a controller?
4.Does battery amp discharge stack? (20s 10p at 8a discharge per cell would be a flat 80 amps?)
5.Does charging amperage stack as well? (1 cell in a pack charges at 1a so 10 amps would be max in a 10p)
6. If I use those 100 cells to make to make a 20s 5p and slap a 60 amp bms on it will that be ok and safe?
7. Any suggestions or advice.
The ride is already very comfortable and fast (I've hit 55 on a slight down) but i need to desperately need to increase my range to as close to 50mi. as I can, and move at speed relative to an automobile

Am i allowed to bump my own post

bump

1. If you trust the Chinese seller, then the BMS prevents damaging the cells. You might be wearing them hard by high discharge rate, but batteries are wear items anyway.
2. More capacity is always better.
3. Not unsafe, as the BMS should trip at 60A. You should set the controller max battery current to less than the BMS overcurrent protection to avoid BMS tripping out.
4. & 5. Yes, although the BMS may impose other limits.
6. Only if the cells can support 60A / 5P = 12A discharge each.

Range depends on battery capacity, aero, and speed. Maximize battery capacity, minimize aero, then all you can adjust is speed. With 72V x 35Ah = 2520Wh, you'd have to target 50Wh/mi to reach 50mi. And 50Wh/mi should be around 35mph on an upright mountain bike.

webosplash said:

2. I have a pack that I opened with 100 35e cells and plan to add another 100 for a 20s 10p 72vX35ah to fit into the triangle and move my controlled to the back. would this be a better or worse option?
3. If I took those 100 cells and made a 20s 5p and a 60 amp bms would this be unsafe as i have an 80a controller?
6. If I use those 100 cells to make to make a 20s 5p and slap a 60 amp bms on it will that be ok and safe?

Click to expand...

35E are good for 8A continuous and 13A peak, so it depends on what you mean by safe. 5P would be 40A continuous, when the cells are new, so the 60A BMS won't be protecting the battery from overcurrent; and you'll be killing your cells.
https://www.imrbatteries.com/content/samsung_35E.pdf

E-HP said:

webosplash said:

2. I have a pack that I opened with 100 35e cells and plan to add another 100 for a 20s 10p 72vX35ah to fit into the triangle and move my controlled to the back. would this be a better or worse option?
3. If I took those 100 cells and made a 20s 5p and a 60 amp bms would this be unsafe as i have an 80a controller?
6. If I use those 100 cells to make to make a 20s 5p and slap a 60 amp bms on it will that be ok and safe?

Click to expand...

35E are good for 8A continuous and 13A peak, so it depends on what you mean by safe. 5P would be 40A continuous, when the cells are new, so the 60A BMS won't be protecting the battery from overcurrent; and you'll be killing your cells.
https://www.imrbatteries.com/content/samsung_35E.pdf

Click to expand...

would this also be the case if i always paid very close attention to charging. I thought 60 would be good as the max is 13a i know if i ran them like that they would die pretty fast but i figured if i overshot the thresh hold by accident I wouldn't have to worry about burning the bms. I ordered the extra 100 cells and will do 20s 10p and an 80a bms

webosplash said:

would this also be the case if i always paid very close attention to charging. I thought 60 would be good as the max is 13a i know if i ran them like that they would die pretty fast but i figured if i overshot the thresh hold by accident I wouldn't have to worry about burning the bms. I ordered the extra 100 cells and will do 20s 10p and an 80a bms

Click to expand...

Charging current is always much lower, and while helpful for cell life, is otherwise unrelated to BMS overcurrent protection.

As above, 60A / 5P = 12A/cell, less than the 13A peak discharge, so 60A BMS is okay for peak discharge, but you would want to configure a lower 40A continuous discharge limit in your controller to maintain 8A/cell continuous discharge.

Of course, while better, 10P * 13A/cell = 130A peak discharge, so an 80A BMS would limit what the cells are capable of. You could likewise configure controller with a lower 80A continuous discharge limit to maintain 8A/cell continuous.

webosplash said:

would this also be the case if i always paid very close attention to charging. I thought 60 would be good as the max is 13a i know if i ran them like that they would die pretty fast but i figured if i overshot the thresh hold by accident I wouldn't have to worry about burning the bms. I ordered the extra 100 cells and will do 20s 10p and an 80a bms

Click to expand...

The BMS rating is continuous, so a 60A BMS will do nothing to protect the battery, since it will allow much more that 60A to flow without tripping. I was able to get my 30A BMS to trip on max overcurrent at around 90A, and never tripped under continuous. I've seen 110A flow through my 40A BMS without tripping. The BMS will allow high current for a while before it sees it as exceeding the continuous rating, which in your case is well beyond the continuous rating of your cells. Most reputable battery sellers would advise against using a 60A BMS with 5P of 35E, and in fact wouldn't sell you that combination.

This is the battery idea. Is this going to work( 20s 10p) I know number 18 only has 9 cells. This is my first battery build
(sick of "wasting " money) and I'm trying to make sure everything is overlooked by individuals with more experience.

If any group has 9 cells, then all of them effectively only have 9 cells...because both the current capability and the capacity of the entire pack are determined by the smallest, least-capable parallel group in the series of groups.

So you can save yourself some weight and money and only design *all* groups for 9 cells.


Also, if they are used cells (or old ones, used or not), you should qualify each cell's capacity and capability, internal resistance, etc., to ensure you have well-matched cells so the entire pack behaves the same throughout. If you don't, then different parts of teh pack will have different amounts of capacity, voltage sag, heating, etc., and it won't stay balanced properly either--the worse the mismatch of cells, the worse these problems are. Mismatched cells will also age differently, which increases the disparity of cell behaviors over time.

If you can't avoid mismatches, then try to group mismatches so that they "even out" between all groups, so no group is any worse than any other.

amberwolf said:

If any group has 9 cells, then all of them effectively only have 9 cells...because both the current capability and the capacity of the entire pack are determined by the smallest parallel group.

So you can save yourself some weight and money and only design *all* groups for 9 cells.


Also, if they are used cells (or old ones, used or not), you should qualify each cell's capacity and capability, internal resistance, etc., to ensure you have well-matched cells so the entire pack behaves the same throughout. If you don't, then different parts of teh pack will have different amounts of capacity, voltage sag, heating, etc., and it won't stay balanced properly either--the worse the mismatch of cells, the worse these problems are. Mismatched cells will also age differently, which increases the disparity of cell behaviors over time.

If you can't avoid mismatches, then try to group mismatches so that they "even out" between all groups, so no group is any worse than any other.

Click to expand...

I am going to put a 10th cell there I just want to know if the procedure is correct. I know it may seem silly but i had a thought that current path changing direction lots of times would create heat. Also I do plan to verify each cell. I bought the pack new and it has maybe 20 cycles on it. I am taking every single precaution I can.

E-HP said:

The BMS rating is continuous, so a 60A BMS will do nothing to protect the battery, since it will allow much more that 60A to flow without tripping. I was able to get my 30A BMS to trip on max overcurrent at around 90A, and never tripped under continuous. I've seen 110A flow through my 40A BMS without tripping. The BMS will allow high current for a while before it sees it as exceeding the continuous rating, which in your case is well beyond the continuous rating of your cells. Most reputable battery sellers would advise against using a 60A BMS with 5P of 35E, and in fact wouldn't sell you that combination.

Click to expand...

This is very interesting. I would expect the optimal behavior would be to use the BMS to limit absolute peak current (so as not to artificially limit peak current), and the controller to limit continuous current/power.

It sounds like these BMS are.. well, defective?
Is this expected behavior? Of just Chinese BMS, or higher-quality German or Russian BMS as well?

fatty said:

This is very interesting. I would expect the optimal behavior would be to use the BMS to limit absolute peak current (so as not to artificially limit peak current), and the controller to limit continuous current/power.

Click to expand...

A little analysis may help show why it can be better not to have the BMS limit current (except for VERY high spikes that would indicate a system short that requires fast-acting protection to prevent a fire...and even then a fuse is probably a better idea, and at the least should also be implemented).

One problem is that many (probably most) of the typical ebike/scooter/MC BMS boards don't appear to limit the current quickly enough to prevent damage to themselves, because they don't have sufficient temperature monitoring at the FETs, and end up packaged inside heatshrink or a box with the cells, rather than with a heatsink and fan on those FETs to rapidly remove the heatbuildup. When running at closer to the designed max discharge current, it's closer to the limits of what the FETs can take (because most companies use the bare minimum to get by, rather than using a wide margin past what they will actually apply to the parts), and so it heats up more.... With nowhere to go that heat builds up and heatsoaks the whole board, so when even more current as spikes comes along the heat stays in the FETs more and the FET resistance increases which makes more heat...etc.

Eventually the FETs either overheat and fail, or the solder actually melts on something and flows enough to short across large potentials, and POOF. Sometimes the FETs don't fail dramatically, but just fail stuck on (like a controller often does making the motor hard to turn), and then the BMS no longer is able to turn off the discharge port, so it can not protect against low battery voltage *or* overcurrent....

This is even worse for those that use common charge/discharge port, which puts the two sets of FETs in series, in opposite "polarities", so that if the charge FETs get put into an off state (which shouldn't happen, but does sometimes, based on what I've seen pics of damage for), you're then using the body diode to conduct all the discharge power thru them, and that diode has a lot more voltage drop and thus heating....


Another problem is that even if you have a good BMS and it's cooled sufficiently in whatever way, or simply doesn't heat up, then the only way the BMS can limit current is by completely shutting off the discharge output. It can't limit it like the controller does, because it doesn't have the parts in it to be an SMPS (which is what you need for such limiting, unless you do it like a linear regulator whcih makes MASSIVE heat for the power it can handle, and would be much larger and need more cooling). So every time the current spikes past it's limit it would have to turn off the output and wait for reset.

At least two potential problems with this.

First is general safety while riding; if your battery just shuts off (which you might expect when it's low in voltage / near end of normal range, but won't at other times) then you have no power to anything and depending on the situation you're in, you might have been accelerating hard to merge with traffic or getting out of someone's way, etc., or going uphill or whatever, and suddenly you aren't able to do that and get hit or force someone to swerve, or you practically just stop because of the hill, and if someone's right behind you you get hit, etc. (yeah, these could all happen anyway...but in these situations you might be expecting and anticipating such things with the expectation that the motor will help you get out of them, and if it doesn't you don't have the time to change your action).

Second is if the BMS shuts off, and you're going fast enough, the smoothing effect of the battery on the spikes from the motor, *especially* if you were in regen with high current flow, disappears, and the voltage spikes way up on the controller FETs from the motor current having to go somewhere and being unable to...if this is beyond what the FETs can take, the controller fails and the FETs probably fail shorted, your motor locks up and if you have a front hubmotor you may endo and crash, a rear hub you may just slide out, etc. If it's any kind of freewheeling motor you just have a dead drive system. But any way you slice it you're stuck where you are with a dead vehicle, even if you survive it.



So...it *can* be a problem to have a BMS do any current limiting, depending on the system and the usage to which it is put, especially with typical ebike/scooter/etc parts as most of them are run right up at their parts ratings to begin with, leaving little margin for those current or voltage spikes.


The controller is designed to limit current in a safe, easy to manage way, so that's unlikely to cause any kind of common failure.

amberwolf said:

If any group has 9 cells, then all of them effectively only have 9 cells...because both the current capability and the capacity of the entire pack are determined by the smallest, least-capable parallel group in the series of groups.

So you can save yourself some weight and money and only design *all* groups for 9 cells.


Also, if they are used cells (or old ones, used or not), you should qualify each cell's capacity and capability, internal resistance, etc., to ensure you have well-matched cells so the entire pack behaves the same throughout. If you don't, then different parts of teh pack will have different amounts of capacity, voltage sag, heating, etc., and it won't stay balanced properly either--the worse the mismatch of cells, the worse these problems are. Mismatched cells will also age differently, which increases the disparity of cell behaviors over time.

If you can't avoid mismatches, then try to group mismatches so that they "even out" between all groups, so no group is any worse than any other.

Click to expand...

Given that I add the last cell and all the cells are good, will the pack be safe to build in this fashion?

amberwolf said:

[great analysis]

Click to expand...

Yeah, I tend to believe cheap BMSs cause about as many dangers and problems as they solve. I get the rationale for pack sellers trying to idiot-proof their product, but I'd rather move the balancing hardware off the bike and run the proven fast-blow ANN fuse for overcurrent protection.

fatty said:

Yeah, I tend to believe cheap BMSs cause about as many dangers and problems as they solve. I get the rationale for pack sellers trying to idiot-proof their product, but I'd rather move the balancing hardware off the bike and run the proven fast-blow ANN fuse for overcurrent protection.

Click to expand...

This implies that you are okay going without over- and undervoltage protection, too.

There are circumstances in which I do that-- when using a charger that can't exceed my pack's maximum voltage, and a controller whose LVC will cut off at a healthy voltage for my pack. But these safeguards aren't present or aren't at correct values in most of the e-bikes I work on, so for those bikes I would recommend or install a pack BMS.

The cheap generic BMSes I replace often for pedicabbers are almost always victims of overcurrent, short circuit, or reverse polarity. Packs don't respond well to those things even (or especially) when they lack BMS.

Chalo said:

This implies that you are okay going without over- and undervoltage protection, too...

Click to expand...

Not entirely.
I think overvoltage protection is better managed in a quality charger not subject to mechanical and environmental stress on the bike, and undervoltage protection is better managed in a quality controller.

I can see the appeal of a quality CAN-bus BMS though, so I'm curious to see how Nucular's turns out.

Chalo said:

The cheap generic BMSes I replace often for pedicabbers are almost always victims of overcurrent, short circuit...Packs don't respond well to those things even (or especially) when they lack BMS.

Click to expand...

This speaks to the need for a discrete fast-blow fuse strip, no?

fatty said:

I think overvoltage protection is better managed in a quality charger not subject to mechanical and environmental stress on the bike, and undervoltage protection is better managed in a quality controller.

Click to expand...

Except...you have to wire out each individual cell group to the charger to monitor for overvoltage, and I don't know of any really quality chargers (just RC world stuff) that actually does monitor the cell level stuff--they generally depend on the BMS to do all that. And these externally available balance connections are one more thing to hook up during every charge, and one more thing to have mechanical problems with that may either prevent charging, or cause problems with charging / balancing over time, depending on how the charger is designed to deal with inability to read a particular cell group.

There are no controllers still made that I know of that can monitor for undervoltage at the cell (group) level. (I think Adaptto did...but I don't think you can get them anymore).

So there's no good replacement for a BMS of some type to monitor for cell-level issues while on the bike, other than a human taking their eyes off the road whenever they put the system under load (to watch for voltage dips that indicate a problem on a handlebar/dash mounted display that's got a bunch of battery-voltage-cell-group wires (or some other communication method to an internal module in the battery) running all the way from the battery to there.

If you have a pack good enough to not need cell level monitoring even as it ages, then you have something not typical of ebikes, scooters, etc.


If you know of purchasable commercially available exceptions, or other ways to do these things, please post links to them.

while i appreciate the info i just need someone to answer if my pack design is going to work. Can i make a pack in any shape I can think of as long as it follows the basic rules of flowing energy. Do the twists and turns of my pack effect my power or current in any way.

Yes, it will work fine. Electricity doesn't care about changing direction, it won't cause extra heat. Keep the power connections beefy and make sure the cells don't move in the pack. Use a BMS to keep the battery balanced. Make sure to get one that supports your max power.

Warren

the box im working on for the batteries coming along

www.recumbents.com said:

Yes, it will work fine. Electricity doesn't care about changing direction, it won't cause extra heat. Keep the power connections beefy and make sure the cells don't move in the pack. Use a BMS to keep the battery balanced. Make sure to get one that supports your max power.

Warren

Click to expand...

when you say max do you mean 130 absolute max output, or 80 for my 80 continuous 80a?

webosplash said:

www.recumbents.com said:

Yes, it will work fine. Electricity doesn't care about changing direction, it won't cause extra heat. Keep the power connections beefy and make sure the cells don't move in the pack. Use a BMS to keep the battery balanced. Make sure to get one that supports your max power.

Warren

Click to expand...

when you say max do you mean 130 absolute max output, or 80 for my 80 continuous 80a?

Click to expand...

I am asking, I have 200 35e cells so a 20s 10p would run at 80a as each cell is rated 8a each continuous but the max output on the datasheet says max 13a so do i get an 80 a bms for my cont. rating or my max......................?

webosplash said:

I am asking, I have 200 35e cells so a 20s 10p would run at 80a as each cell is rated 8a each continuous but the max output on the datasheet says max 13a so do i get an 80 a bms for my cont. rating or my max......................?

Click to expand...

Goes back to your assumptions for your used cells, since only you r can test their condition.

EM3EV seems to be a highly recommended and knowledgeable battery builder on this forum. They have the 35E cells among their selection of quality cells. They are conservative with cell ratings, balancing performance and longevity. For 35E, they use 5A continuous and <7A max/burst for their discharge ratings when configuring their packs and BMS. They are using new cells.

https://em3ev.com/product-category/ebike-battery-parts/ebike-battery-in-case/

I'd say drawing 8 amps from 35E cells is possible, but will have some voltage sag, and will heat up the cells quickly. My experience comes from my self made MJ1 packs, which have similar capacity per cell. Drawing 6 amps from each cell will warm up the battery close to their 60c temp limit faster than I can drain the battery. I actually had a cell die in the middle of the pack, which must have been from heat. I replaced it without too much trouble at least, and lowered my BMS temp sensor limit to 52c.