Charging setups for high voltage packs

Batteries, Chargers, and Battery Management Systems.
Post Reply
Danzzigger   1 µW

1 µW
Posts: 3
Joined: Mar 30 2018 9:53am

Charging setups for high voltage packs

Post by Danzzigger » Apr 14 2018 2:13am

Hi all, first post. I have used the search function for the last few weeks to gain as much info as possible about my questions. That being said, if you have links to other posts in regards to my questions, please do so.

Bit about myself, I've come from the RC scene, started with nitro, got annoyed... paid the big bucks to get into electric and have never looked back. I have some basic knowledge of electronics so I understand most things until they get technical... then I use google :)

My intentions are to build a E bike with a 2004 DH comp frame. Been basing a lot of my stuff on Doctorbass's, Obiwan007, Ohbse, GCinDC and I feel like there's a few more I missed out.

Anyway I'll stay on track to my question in regards to chargers.
How are people charging their big packs? From what I'm gathering, a lot of people are using a 20+ cell BMS to charge the pack. Bypassing the discharge on the BMS as most of the cheaper ones don't have the amp rating capacity required to do so.
Then finding a charger to charge at the required voltage is the next hard thing. So I saw that a lot of people are using meanwell chargers and tuning them to get the voltage they require. From what I'm understanding, all it is, is a Constant voltage and Constant Current Powesupply. So technically it isn't a charger, but just a power supply, the BMS is doing the charging function? I don't understand huge amounts about charging cycles and how different chemistries work, I know that nmih is trickle charged at the end of its cycle. Is there not more to it than pushing a whole lot of 'power' (term used loosely) into a battery and then balancing (discharging) the cells to the same capacity as each other at the end of the cycle?

So I bought one of these ages back, was going to use it to power a stereo to take onto the job in a systainer. Got it cheap on ebay. These are like $25aud or something.
https://kamboti.unpatti.ac.id/2015/11/1 ... er-supply/

They're pretty light too, weighed it at 1.26kg 13.5v x 69a = 931.5watts for $25 :)
So I'm aware of the adaptto controller and their bms system which apparently takes any 12+ voltage power source and bumps it up to the required voltage using what it seems is a step up toroid or something? Is it possible to do this without going down the adaptto route. I'm doing all of this on a really tight budget. So I'd rather get a Kelly controller or something similar. Trying to keep everything high voltage so as to reduce loss in efficiency. Should I just split the pack so that its only like 4 lots of 5s for charging purposes? BMS for each?
So to give you a run down on what I'm specifically trying to do;

2004 DH comp bike
qs205 1000watt hub motor
A speed controller to match hub motor - thinking about Kelly, unsure of model
120 Samsung cells - fits in the triangular frame, kinda want it to remain stealthish / slim. 20s 6p 74v-84v 9ah

Image

I have a mate whose sourced 246 used samsung INR18650 - 15M SDI 121 (1500mah), 16 samsung INR18650 - 20R M SDIEM 125 (2000mah) and 8 (2500mah) batteries - haven't opened the pack yet. All the battery packs I got were 'broken', ie BMS stuffed, or maybe single cells gone. I'm in the process of charging, discharging, matching etc. Which leads me to another question, how do people most accurately test their cells? should I youtube it? I've just been charging at 0.5c, then discharging at 1amp, writting down the results.

Lastly, what have people been doing to clean up the original spot welds? I tried this, and it works pretty effectively.

Image

Image

Is there a better method?

Thanks guys, I've really been learning a great deal going through this forum!
Last edited by Danzzigger on Apr 15 2018 6:06am, edited 1 time in total.

User avatar
amberwolf   100 GW

100 GW
Posts: 25457
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Charging setups for high voltage packs

Post by amberwolf » Apr 14 2018 3:51am

Regarding efficiency losses, its probably a wash as to whether HV or HC is better. HV means generally more expensive controllers, BMS, etc., and switches/fuses/breakers have to be rated to break at least the hot-off-charger voltage (or they might continue arcing and cause a fire. HC means thicker wires, better cells or more parallel cells to handle the current draw, etc.


A watt is a watt, so whether it comes from HV or HC doesnt in and of itself matter, as long as it gets to the wheel.

Regarding the MW PSUs, they are the chargers; all a BMS does is detect when cells are full and shut off charging (whether using a charger or a PSU) while it drains down the highest cells so that the lowest ones can then be charged some more.

If you use cells that are high quality and dont use them at the edge of their performance ratings, theyll probably stay balanced without a BMS (like the EIG NMC cells I use on SB Cruiser and CrazyBike2; theres plenty of good ones out there as long as you arent pushing them really hard).


I dont know any cheap safe way to turn the 12v PSU into an HV charger. There are RC chargers that can turn its output into higher voltages, but to do it for a 20s pack youd probably have to charge only half the pack at a time, assuming you use a 10s RC charger. You couldnt use two 10s in series if theyre fed off teh same 12v PSU, becuase the chargers are almost certainly not isolated from input to output, so itd short across half your pack and cause a fire at worst, and burn up wires / blow fuses at best.

Its probably cheaper / simpler to buy a couple HLG series Meanwells that will charge the pack at the rate you want, and wire them in series (and make sure you build the pack (or buy one) so it can handle taht charge rate without being near its performance specs. (so it lasts longer))

Regarding buildng a pack from used 18650s, if its going to be a performance bike you will need to be sure all the cells are specd for teh current draw you will need to pull from them to get teh peformance youre after. And then keep in mind used cells may last a long time or may die immediately when put to use, or anything in between, depending on what usage they saw before and how old they are, etc.


Even new cells will need to be selected on performance, so youre using htem well within their ratings. The harder you push htem, the more voltage sag, and the less performance you get out of them.


Regarding cleaning up the old welds---if you leave the old tabs on, you can solder or weld to *those* without having ot clean up the cells (and risk damaging them).

Danzzigger   1 µW

1 µW
Posts: 3
Joined: Mar 30 2018 9:53am

Re: Charging setups for high voltage packs

Post by Danzzigger » Apr 14 2018 9:25am

Thanks for the reply. I was actually thinking of reusing the old solder tabs, as that means I don't need to invest in making a spot welder as I can solder 'off' the cell, not getting it too hot.

I've charged 3 cells so far, and I've got a bit of variation, but I wasn't sure whether it was cause of my charging process. my results were 1785, 1884 and 1940. All from my 20R cells (2000mah).

The bike itself is not going to be overly powerfull if I end up going for the qs205 I have a odd 11kw of potential discharge at my disposal, if it only needs say 2000watts at peak (burst) thats still only 20% of the batteries potential.

Should I be somehow stress testing my cells before assembling them into a pack? maybe run them on 5amp discharge and see if they discharge the same mah as a 1amp? I know 5amp isn't anywhere near the 23a these particular cells can handle, its the max my charger will do. Also it's more my bike will be drawing from it if I stay with a 1000watt hub.

I've heard of the isolating problem before. The setup I want is to be able to just plug the bike in and walk away. No button pushing, like a bought one ;) - I am aware that you should be present whilst charging, I meant it more as an expression.

Wow, photobucket isn't free anymore, been a while since I've used it... looks like I need to find another provider for my online images.

So what I was assuming was right in that the meanwells are simply a PSU much like a server psu with a CV and CC, there are no 'smarts' behind it. The bms is what does the discharging, and the bms is wired up to switc the psu on and off at the end of the battery charge cycle. How often does the bms do this in order to balance the cells correctly? seems rather inefficent to turn on and off a large psu to put a few tiny mah in at the end if it does it often. Or does it only do it like a couple times? I think I'm not quite understanding the process completely, does anyone have a link or a video which explains it?

I'm hoping that I'll be able to find enough good matched cells to make a 120cell pack. He can get me more batteries in time too if I need them. I'd rather not spend $600+ dollars on a battery if I can save the money.

User avatar
amberwolf   100 GW

100 GW
Posts: 25457
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Charging setups for high voltage packs

Post by amberwolf » Apr 15 2018 2:46am

Danzzigger wrote:
Apr 14 2018 9:25am
I've charged 3 cells so far, and I've got a bit of variation, but I wasn't sure whether it was cause of my charging process. my results were 1785, 1884 and 1940. All from my 20R cells (2000mah)
If you test them at higher currents you may see more variation, or just less capacity in total.

One thing you should keep in mind aobut used cells is that you cant know anything about how long or how hard theyve been used, so even the ones that are still good in packs might have 100 charge cycles left or they might have a dozen, before they start to wear out more quickly.

The bike itself is not going to be overly powerfull if I end up going for the qs205 I have a odd 11kw of potential discharge at my disposal, if it only needs say 2000watts at peak (burst) thats still only 20% of the batteries potential.
I think there might be a couple of issues with some of your math ;) unless I ran across wrong numbers in a search for the Samsung 20R cell data. In your OP you mention a configuration of
20s 6p 74v-84v 9ah
which is a bit confusing in itself. 20s would nominally be 74v, for 3.7v nominal cells, and 84v would be full voltage. But 6p for 2Ah cells is 12Ah, not 9Ah. If its 6p, then if you used the max continuous current (whcih will probably cause a lot of voltage sag and heat) then assumign its around the nominal voltage at that point, and not sagging below that, 74v * 6p * 22A = around 9700W max. If its not 6p / 12Ah then those numbers change too.

You might get more watts if the sag is less, but given these are used cells of completely unknown condition (at this point) Id personally expect at best to get half of what theyre rated for. If you use the higher currents, Id expect more sag, and Id expect much less capacity while used at the higher currents.

Also, since teh pack (for 12Ah) is nominally 888Wh, if you actually used that max wattage a lot the pack might only last 5-10 minutes to empty. ;) And thats all assuming brand new cells, rather than used ones of unknown lifespan/capabilities.

Remember that the motor doesnt limit the watts--the controller does. So the motor being a 2kw motor doesnt mean anything if the contorller is different. Some are programmable, some are not, so you may be able to change the power or current limits to adapt to your motor and battery power handling capabilities.

Also, many motors can take several times their rated wattage for short periods. See threads like the Definitive Testing of Heating / Cooling hubmotors by Justin_LE, for some data on that sort of thing.

I dont know which brand would be better for you; theres a lot of them out there and sometimes generic stuff does the job just fine, and sometimes you need the features in a pricier brand.

Im using generic stuff right now, but have been collecting parts to build my own pair of Lebowski controllers for the past year and a half or so, as these are programmable and have some nice features I want (especially the way braking works), and are tunable to work with specific motor properties fro better efficiency and performance. (the phaserunner from Grin is, too, but I cant build those myself, and theyre way out fo my price range).

Should I be somehow stress testing my cells before assembling them into a pack?
You should test them under the same conditions you expect to use them in. If youre ever going to run them at higher currents, you should test them at those same currents so you know which ones can still handle it and which cant, so you know which ones to not put in your pack. ;)

For testing methods, your best bet is to look at the other recycled-cell build and testing threads. DrkAngel has at least one thread wiht various testing info in it, and there are others around too. Probably 18650 and either used or recycled or other similar search terms would find them, looking in just title of threads, display by topic, or maybe even just in the first post.



looks like I need to find another provider for my online images.
Just use the Attachments tab below the text entry box in your posts. ES hosts teh images, so anyone that can see the post can see your pics, unlike right now when probably no one can see them (I cant see any pics at all your posts in this thread, for instance). Also, the images will always be here in the thread (which wont be true if whatever external site you hostimages on goes away or does shitty things like PB did).
So what I was assuming was right in that the meanwells are simply a PSU much like a server psu with a CV and CC, there are no 'smarts' behind it. The bms is what does the discharging, and the bms is wired up to switc the psu on and off at the end of the battery charge cycle.
Well, the BMS doesnt switch the PSU on or off. It just disconnects the incoming voltage to the BMS (from any voltage source, regardless of what it is, charger or PSU or solar panel, or even another battery or even regen on a controller), if the voltage on any cell goes above the cell-level HVC.

As far as PSUs vs chargers, and discharging, neither one does any of that. At least, not for non-RC type chargers (those are the only ones commonly found that do the job of both charger and BMS).

So a designed-to-be-a-charger unit like the Cycle Satiator or a common ebike charger, etc., may have end-of-charge termination, where it shuts off its own output voltage once current drops below a certain point. (it also may not do this, as it might be designed to continue trickling current for a BMS to do balancing with). And some of these chargers taht do shut off below some current draw, will restart if they detect a low enough voltage on their output (from a BMS reallowing input), and some will never restart until they are disconnected from both AC power and the battery, and then replugged into both. Ive had some of each of these kinds.

The PSU simply never shuts off its output voltage, so current could continue to flow if the battery never reaches the PSUs max output voltage (for instance, if the batterys full charge voltage as it ages drops below the voltage you have set the PSU to). But if the PSU output voltage equals the full charge voltage of the battery assuming its well balanced, then current will still drop to essentially zero once it does reach that full charge. Also, with the PSU, and a BMS, if the BMS needs to balance then the PSU will always let the BMS have the current it needs to do taht with, where some chargers dont (if they are the no-restart type).


How often does the bms do this in order to balance the cells correctly? seems rather inefficent to turn on and off a large psu to put a few tiny mah in at the end if it does it often. Or does it only do it like a couple times?
It happens as many times as it takes to balance the cells, and that depends on the cells themselves. WIth brand new cells that are all well-matched to each other, that are not used near (or beyond) their performance limits, theyll probably stay well-balanced until they get pretty old and worn-out, and the BMS wont have to do much, if anything, until that aging happens.

But with unmatched cells of varying capacity, internal resistance, age, and other performance differences, cells may become very unbalnced on every single discharge cycle, and require a lot of balancing each charge cycle, so it could take hours (or days, though thats unlikely) to balance such a pack on every charge cycle.


RC chargers do this same thing, though you dont see it happening from the outside because its all one unit (BMS and charger in one box).
I'm hoping that I'll be able to find enough good matched cells to make a 120cell pack. He can get me more batteries in time too if I need them. I'd rather not spend $600+ dollars on a battery if I can save the money.


Remember that if you have the space, you can also just parallel a lot more cells to make up for the failings of unmatched or poorly-performing cells.

Danzzigger   1 µW

1 µW
Posts: 3
Joined: Mar 30 2018 9:53am

Re: Charging setups for high voltage packs

Post by Danzzigger » Apr 15 2018 6:16am

Pics fixed, gonna do some more research on DrkAngel threads, thanks heap for that suggestion. Wow he's accumulated a lot of information!

Sorry for the confusion, I was talking about the 1.5ah batteries, as they're the ones I have 200+ of. I looked up the specs on them and they're 23a discharge which is equal to 15.3C I thought.

6 x 23amps = 138amps
20 x 3.7 = 74volts
138 x 74 = 10,212watts. at 84v its roughly 11kw.

Thanks again. I'll do some more research when I get home tonight.
Saw the thing on bulk charging, I'll just use my turnigy 400watt charger to parallel charge a single heaps of them then.
Unless there's a alternative way of leaving the batteries in their current setup 5s 2p and 7s x 2p? for speed of testing? then they remain welded together and I can do the discharge test? Anyway I should really read up before asking all these questions, it's probably already answered in his threads / posts. I did glance quickly over it before posting this.

User avatar
amberwolf   100 GW

100 GW
Posts: 25457
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Charging setups for high voltage packs

Post by amberwolf » Apr 15 2018 3:26pm

Danzzigger wrote:
Apr 15 2018 6:16am
Pics fixed,
I still dont see any; looking at a quote of your post they appear to be linked from a google site, which doesnt work right everywhere (like most image sites). Some poeple might be able to see them, others might not.

I highly recommend attaching all your images to the post itself, so that anyone who can see the post can see the pics.

Anything on an external site depends on that site cooperating with the site linking to it, the browser the viewer is using, and all of the other stuff that varies between viewers and sites. And it will go away at some point, leaving a thread useless if it depended on those pics for any of its information. (theres already a lot of those trashed now-useless threads here on ES, many of them simply from one single sites single change (PB))


Sorry for the confusion, I was talking about the 1.5ah batteries, as they're the ones I have 200+ of. I looked up the specs on them and they're 23a discharge which is equal to 15.3C I thought.

6 x 23amps = 138amps
20 x 3.7 = 74volts
138 x 74 = 10,212watts.
Ok, I mustve gotten confused somewhere in there about which cells you were using. :)
at 84v its roughly 11kw.
Keep in mind you wont have 84v while drawing that max current--voltage will sag, possibly a lot, so its safer to calculate peak power at the nominal voltage. You can look up the manufacturers cell-specific discharge curves to see what voltage drop to expect at various current draws.

Post Reply