What exactly do I need to leave a pack charging unattended?

[Alan B]

Hey, Alan, how does that work?

Lets say I have 72v pack running drawing 10 amps from the battery, thats 720 watts of power, and I'm at 1/4 throttle cruising at 15 mph. I was under the impression that the controller would convert 72v dc to a 3 phase AC wave at 72v, and use PWM to turn it on/off rapidly to clip the amps to whatever my motor needs to run at that speed.

Am I wrong? is it converting it so 36v at 20 amps to match the KV winding of the motor at that speed? or something else entirely?

Closer to the latter. The PWM is driving a pair of coils in the motor with PWM, and the motor voltage generated will be enough higher than the back EMF to cause the phase current to flow, and the power is mostly conserved. If it exactly matched the back EMF you would be 'power coasting' and there would be very little current flow. That would be the point on the throttle where the motor "just starts" to accelerate. As the motor commutates this "DC" output switches around to the six different current/winding configurations required to run the motor. It is like a buck type DC-DC converter dropping the voltage down and generating increased phase current.

 

[Indubitably]

when sizing a pack, don't forget that 48v10ah is the same at 100v 5ah in terms of Watt hours available. at the same given speed, you should get about the same milage(and have the same weight) on either battery design. and this would be easier on the phase wires as you mentioned, but in reality, you'll set the amps sky high, have high volts and just go WoooHoooo as you haul ass down the street.

Well, with 40ah batteries on a bike application like this all you really can do is run them in series, so higher voltage pretty much does mean more weight, because it means more batteries. Either way I really don't need to go any faster though, from what I understand the motor is wound in such a way that I should be able to hit 50 mph with 48v.

Also remember that changing the pack voltage doesn't change the voltage/current on the motor side of the controller for the same speed/power. It just changes the controller input voltage/current and higher voltages cause lower duty cycle PWM and are a bit less efficient for the controller. Higher voltage/lower current are more efficient for the wiring and connectors. Higher voltage packs also raise the top speed of course and provide the capability of going to higher power (with the attendant possibility of melting things).

To reduce motor current requires higher turn motors. This increases the efficiency of the controller, wiring to the motor and connectors but it limits top speed.

Good to know, I was under the impression that what ever voltage I pumped into the controller was what the motor would wind up seeing. Its kind of a shame that I have a motor wound for high speeds when I would actually prefer something that could more efficiently provide me with torque at lower speeds, considering that there are so many people on the forums looking to build the fastest bike they can. Perhaps I should consider looking around for someone willing to trade. What do you think of the two hub systems? I take it that you essentially need two controllers, and that the controllers need to some how interface with each other to ensure that both motors run at exactly the same speed, but I'd imagine that it would effectively divide the load.

 

[Indubitably]

You know, it just occurred to me that at 92v I'm not all that far off from 120v. Would I be correct in assuming that it would cost considerably less for a widget that directly converts 120v ac to 120v dc rather buying a power supply? I mean, I figure I'm looking at $35 $40 a pop for something like 7 12v 20a or 30a power supplies if I want to fast charge anyway, it seems like I might as well just invest that cash in more batteries if its going to make my charging system lighter, cheaper, and less complicated.

I suppose I'm kind of wondering why 120v isn't a more popular voltage to work with for that matter too. Are there safety concerns that I should be aware of with this approach, if so, are there ways I can conveniently address them?

 

[Njay]

I've seen many people saying that these TS cells don't actually unbalance. Don't use charge equalization, too much heat in a space too small - many EV fires have been caused by charge equalization BMSs. Think about this: if one of the cell fails (voltage very low) during charging, what will it happen?

 

[Indubitably]

I've seen many people saying that these TS cells don't actually unbalance. Don't use charge equalization, too much heat in a space too small - many EV fires have been caused by charge equalization BMSs. Think about this: if one of the cell fails (voltage very low) during charging, what will it happen?

I'm not sure I follow, are you suggesting that I not use a BMS, or are you perhaps speaking with regard to my question about simply rectifying the 120v that comes out of a wall socket rather than running it through a transformer?

 

[Njay]

I'm suggesting you don't use charge equalization and just high voltage cut (stop charging if any cell goes above a certain threshold, until manually reset), on charging, as a measure to increase safety on unattended charging. And constantly monitor the cells individually to detect a bad cell as early as possible and make sure they are not unbalancing. Check evtv.me for lots of details.

 

[commanda]

A 40AH TS pack is not cheap. Do it right and use a BMS. And forget the Ghetto Guerrilla straight off the mains idea. Good way to kill yourself.
And the rectified and filtered dc voltage is sqrt(2) times the ac rms value. So 120 volts ac is 170 volts dc. (approx).

For the record, I've had a 16s x 40AH Thundersky setup running for 2 1/2 years now.

 

[Njay]

I never suggested not using a BMS. "BMS" can actually "mean" several things, and when someone says the word, we don't know *exactly* what it means. From what I've seen around, these TS cells never quite unbalance; they also seem pretty safe judging by experience of others and manufacturer's specs - and however there has been a few accidents (meaning vehicles burned up to the ground) during unattended charging on systems with BMS (http://jackrickard.blogspot.com/2011/03/latest-in-too-long-string.html). As a future "converter" myself, I'll just ditch out equalization on charge and will look carefully into BMS control of the charger, and constantly monitor cells individually (also known as BMS - Battery Monitoring System). Check evtv.me and posts on the Internet about burned up vehicles.

 

[morph999]

Thundersky's like to be charged to about 3.75v .

 

[commanda]

Njay,

Don't use charge equalization

I guess there could be a subtle distinction there.

Yes, poorly designed/implemented charging/bms can be a serious problem.

It is my considered opinion that any charging solution which relies on the voltage out from the charger matching the batteries voltage is ultimately doomed to fail.

My solution relies on controlling the current. Bulk charge current is limited by the Meanwell power supply, conservatively at about 60% max power. Backup bulk charge current limiting is in the bms, and set slightly higher.
Once any cell reaches it's fully charged voltage, the bms limits the charge current to about 1/2 amp, until all cells are full. It then runs another 10 minutes before switching off the current entirely. A current monitor on the Meanwell detects the lack of charging current, and switches off the mains supply.

The bms current sources have a NTC (negative temperature co-efficient) characteristic, so rising temperature decreases the current.

The Meanwell cooling fan runs full time. The bms cooling fans run full time. The bms also has plenty of air space around it.

 

[Indubitably]

Ok, so the impression I'm getting here seems to be that the only real way to be deal with charging safely is to charge fast enough that I can actually hang around to keep an eye on the pack in case something goes wrong. So maybe what I need to be looking at here is a quick charge solution. Pre-made chargers are prohibitively expensive, so I'm guessing meanwells or meanwell-like power supplies are my best bet. For simplicity's sake, I'd prefer to have fewer but higher voltage power supplies strung together, but if I break it down to watts / dolar it seems like 12v at 30 to 40a supplies are definitely the sweet spot (I'm guessing that 12v supplies just happen to be much more common for some reason) so my plan is to string about 7 of these these in series to get the 90 or so volts that I need.

Commanda: Is Mini-BMS one of these apparently "doomed" bms solutions of which you speak? I haven't heard any complaints about it yet, but if the bms I'm using has some fundamental limitations, I want to be aware of it none the less.

 

[commanda]

Indubitably,

I went back to the site of your bms, and read the user manual carefully.

They talk about a unit with shunting option, but don't give any details.

The standard setup connects the charger directly to the batteries. It then switches off the charger when the first cell hits HVC.
This will achieve your first aim of charging unattended.

Down side is, you won't know how well balanced your pack is till you hit LVC. And since Murphy's law has not been repealed yet, this will occur at the least opportune time, when you are farthest from home, & late for an important meeting, etc.

This is what I would do. Get a single cell charger. When you first assemble this thing, charge till it cuts off at the first HVC. Then use the single cell charger to fully charge every cell. This gives you your starting point.

Then, periodically, check with a meter the individual cell voltages, at or near the point where the charger switches off. Typically, the cell voltages will remain below 3.3 volts until the last few percent of charge. If ,at the time of testing, all the cells are in the 3.3 - 3.6 volt range, call it balanced near enough. When you start to see one or more cells below 3.3 volts, break out the single cell charger and top them off.

You could, of course, get 3 cell-log units from HK, and wire up a loom with dangling plugs. This makes it easy to periodically check (& is what I've done).

As for how much charging current you need, this depends on how you use your vehicle. I typically travel to work & back home, using about 25% of my capacity. I then have all night to recharge. So I don't have any compelling reason to charge it as quickly as possible. Your situation may be different.

Now, stop dilly-dallying & get the thing on the road. The grin will be worth it.

Amanda

ps: having some sort of meter which records AHrs drawn from the battery is a really useful add-on. I use the Doc-Wattson, but there are a number of similar alternatives.

 

[Indubitably]

Indubitably,

I went back to the site of your bms, and read the user manual carefully.

They talk about a unit with shunting option, but don't give any details.

The standard setup connects the charger directly to the batteries. It then switches off the charger when the first cell hits HVC.
This will achieve your first aim of charging unattended.

Down side is, you won't know how well balanced your pack is till you hit LVC. And since Murphy's law has not been repealed yet, this will occur at the least opportune time, when you are farthest from home, & late for an important meeting, etc.

This is what I would do. Get a single cell charger. When you first assemble this thing, charge till it cuts off at the first HVC. Then use the single cell charger to fully charge every cell. This gives you your starting point.

Then, periodically, check with a meter the individual cell voltages, at or near the point where the charger switches off. Typically, the cell voltages will remain below 3.3 volts until the last few percent of charge. If ,at the time of testing, all the cells are in the 3.3 - 3.6 volt range, call it balanced near enough. When you start to see one or more cells below 3.3 volts, break out the single cell charger and top them off.

You could, of course, get 3 cell-log units from HK, and wire up a loom with dangling plugs. This makes it easy to periodically check (& is what I've done).

As for how much charging current you need, this depends on how you use your vehicle. I typically travel to work & back home, using about 25% of my capacity. I then have all night to recharge. So I don't have any compelling reason to charge it as quickly as possible. Your situation may be different.

Now, stop dilly-dallying & get the thing on the road. The grin will be worth it.

Amanda

ps: having some sort of meter which records AHrs drawn from the battery is a really useful add-on. I use the Doc-Wattson, but there are a number of similar alternatives.

Yeah, I actually picked up a 6a single cell charger a while and a turnigy watt meter widget a while back. The 6a charger takes about 6 hours to finish charging a cell, so generally speaking I will leave it over night to do its thing. It does get a pretty hot for a while, and I'm a little bit apprehensive about leaving a charger on the battery without an active bms but I suppose since its a "smart charger" and I'm charging cell by cell, its pretty much the same as having a bms on there anyway. At this point my next step will be to track down some sort of a battery testing widget to make sure that they can all hold a load (any suggestions would of course be appreciated). I bought them about a year ago and they pretty much sat on a shelf the whole time, so I'm guessing some will need to be replaced before I put the pack into action. Most seem to be charging to about 3.8 or 4v then falling back to something between 3.36 and 3.38 the way they should, but a few that have been hanging around 3.41 or 3.5 for the last week are making me suspicious. Hopefully its just surface charge or something that will bleed off as soon as I apply a load, but I'd like to be sure before I get everything bolted onto the bike and rolling.

At any rate, as I understand it, the BMS comes with shunting standard, and you can snip off some diode or resister or something to disable the shunting function if you so not want it. I figured shunting was one of the main resons to have the BMS in the first place, so I had planned to leave it intact.

 

[Gordo]

I have 20 40ah TS cells in my scoot and after 4-5 cycles 3 of them get lazy and go low. So I use a single cell charger to bring them up before charging the pack. I have purchased 3 replacements and tried to get them all equal before installing them. This proved impossible either in series or individually, so I put the 3 in parallel and by using a large ceramic adjustable resistor to adjust the voltage, left them charging at 3.85V for days. Now they have been separated for many weeks, the individual voltage is still not equal. 3.65v, 3.55V and 3.55V. Next plan is to put a BMS on the 20 cells for charging only.
There are many EV cars and trucks in this area with TS cells which have been running for years with no BMS. There have been a few single cell failures, but far fewer than when they tried various BMS. With the BMS, when one failed, it took the cell with it. One fellow, who did not have quite enough range, decided to risk overcharging rather than running his battery too low. He has been charging to 4.10V for many months now and has had no problems. Perhaps he is shortening the life of the pack? Time will tell.

 

[commanda]

Most seem to be charging to about 3.8 or 4v then falling back to something between 3.36 and 3.38 the way they should, but a few that have been hanging around 3.41 or 3.5 for the last week are making me suspicious. Hopefully its just surface charge or something that will bleed off as soon as I apply a load, but I'd like to be sure before I get everything bolted onto the bike and rolling.

I think you've got that backwards. The ones that fall back quickly are the ones to be suspicious of. A good cell will hold its surface charge for a good long while.

 

[Indubitably]

Most seem to be charging to about 3.8 or 4v then falling back to something between 3.36 and 3.38 the way they should, but a few that have been hanging around 3.41 or 3.5 for the last week are making me suspicious. Hopefully its just surface charge or something that will bleed off as soon as I apply a load, but I'd like to be sure before I get everything bolted onto the bike and rolling.

I think you've got that backwards. The ones that fall back quickly are the ones to be suspicious of. A good cell will hold its surface charge for a good long while.

"Quickly" might be a bit strong of a word, it still takes them several hours to hit that range, but there are a few that hang pretty much indefinitely at a higher voltage. To be honest, I don't really know that there is necessarily anything wrong with the cells that are at 3.41 and 3.5v, its just that I have heard of damaged cells hanging at higher voltage so I'm a little weary.

 

[AndyH]

Most seem to be charging to about 3.8 or 4v then falling back to something between 3.36 and 3.38 the way they should, but a few that have been hanging around 3.41 or 3.5 for the last week are making me suspicious. Hopefully its just surface charge or something that will bleed off as soon as I apply a load, but I'd like to be sure before I get everything bolted onto the bike and rolling.

I think you've got that backwards. The ones that fall back quickly are the ones to be suspicious of. A good cell will hold its surface charge for a good long while.

"Quickly" might be a bit strong of a word, it still takes them several hours to hit that range, but there are a few that hang pretty much indefinitely at a higher voltage. To be honest, I don't really know that there is necessarily anything wrong with the cells that are at 3.41 and 3.5v, its just that I have heard of damaged cells hanging at higher voltage so I'm a little weary.

Commanda wasn't talking about the bleed down speed. The point is that the ones that stay higher are the GOOD ones and the ones that bleed down are the ones that have been damaged through either overcharge or overdischarge.

I've presented two years worth of logged data from a 21S TS 60Ah pack with and without proper management as have others here. Bottom line - If one wants to buy cells, skip the BMS...and I mean management not monitoring here.

 

[Njay]

I've presented two years worth of logged data from a 21S TS 60Ah pack with and without proper management as have others here. Bottom line - If one wants to buy cells, skip the BMS...and I mean management not monitoring here.

Is that data available?

 

[AndyH]

I've presented two years worth of logged data from a 21S TS 60Ah pack with and without proper management as have others here. Bottom line - If one wants to buy cells, skip the BMS...and I mean management not monitoring here.

Is that data available?

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=8675&hilit=+xm5000li
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17035&start=45#p251105

Looks like they need work - the picture problem hit them hard...

 

[Indubitably]

Yeah, after poking arround on the inter-webs I've come to the conclusion that I am simply not qualified to weigh in on the issue of exactly what constitutes a healthy voltage for these batteries. There seems to be a lot of conflicting opinions on different forrums coming from any number of individuals who appear to be quite knowledgable. At any rate, for now I'm just going to do the load test to make sure that the batteries hold a charge and leave it at that. If any of them are dead they will be replaced, otherwise I'm just going to assume that they are good-enough.

 

[Alan B]

Best way to characterize them is to do a discharge curve and especially look at the amp-hours.

 

[neptronix]

Yeah, after poking arround on the inter-webs I've come to the conclusion that I am simply not qualified to weigh in on the issue of exactly what constitutes a healthy voltage for these batteries. There seems to be a lot of conflicting opinions on different forrums coming from any number of individuals who appear to be quite knowledgable. At any rate, for now I'm just going to do the load test to make sure that the batteries hold a charge and leave it at that. If any of them are dead they will be replaced, otherwise I'm just going to assume that they are good-enough.

The difference is in how safe you want to be.

Near the end of the charge, around 5%-15% or so for lipo, the problem is that the small differences in the cells start to rear their ugly heads.

Out of the 13 lipos i tested, this one was the worst:

If i discharge this pack to 3.6v per cell, i won't risk this bad pack having cells that go under 3.0v.

Notice that after about 3.65v, the cells start going off a cliff on this pack.
That's generally the consensus on all lipo packs i've tested.

The 'cliff' is where the cells get disbalanced. If you stop discharging when you hit the cliff, you can stop worrying about cells going below 3.0v unless you have a real stinker, which would need a cell replaced anyways!

You end up using less AH of the pack but the side benefit is that shallower discharge cycles will extend the life of the lipo far past the "300 cycles" number that hobbyking quotes.

If you have a charger like an iCharger/hyperion or a celllog 8s, find out for yourself how these lipos behave in real life!

BTW if you are not using lipo, the same applies to using lifepo4 without a BMS. The lower end of many lifepo4 batteries i have seen also have a cliff where they get disbalanced. That cliff is just longer. The same rules apply but the voltages differ.

A good charger will allow you to plot out a discharge curve like the ones i posted above. That includes the hyperion and iCharger line, as well as many others that are not as popular on here. a pair of Celllog 8s's will do it too; they have a USB port and act like an iCharger when you hook them up to the computer and view the info in logview.

 

[AndyH]

Yeah, after poking arround on the inter-webs I've come to the conclusion that I am simply not qualified to weigh in on the issue of exactly what constitutes a healthy voltage for these batteries. There seems to be a lot of conflicting opinions on different forrums coming from any number of individuals who appear to be quite knowledgable. At any rate, for now I'm just going to do the load test to make sure that the batteries hold a charge and leave it at that. If any of them are dead they will be replaced, otherwise I'm just going to assume that they are good-enough.

Save yourself a lot of time and do a quick internal resistance check first. Low=good, high=bad.

It's best to check at about 50% state of charge, but testing on top works just fine for this culling process. Charge them all up with a single cell charger and let them sit a few hours to rest. Check the open circuit cell voltage, then check the voltage when you connect a big resistor or headlight bulb or some load. Doesn't matter what you use as long as you use the same thing for all the cells. More voltage drop means higher resistance.

Spot the bad cell from the loaded voltage: 3.1, 3.1, 3.1, 2.6, 3.1 3.1


Good hunting!

 

[Indubitably]

OK, so, load testing was carried out on batteries fully charged by a powerizer 3.8v 6.0A lifepo4 single cell charger, one battery at a time, using a halogen headlight, and a cheap "cen-tech" multi meter to test the voltage across each battery's terminals (first unloaded then loaded). Max unloaded voltage was 3.38v, min unloaded voltage was 3.35v, max loaded voltage was 3.37v, and min loaded voltage was 3.34v. Most cells displayed a difference of .01v between loaded and unloaded phase, with a few measuring in at a difference of .02v. All batteries that had previously appeared to be "hanging" at voltages higher than 3.8v appeared to drop into the 3.5v to 3.8v range within a min or two, and afterward appeared to behave like other cells in the pack.

Basically, as far as I could tell, none of the cells were dead, and the internal resistance of most all cells was nearly identical.