EV managment module

Chat about all things battery in here.
Need to know what type to use or size or capacity then again place your thoughts here
User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Postby Jeremy » Fri Aug 15, 2008 4:05 pm

I appreciate your views, Paul, but it does depend very much on how the cells behave under load, the application vehicle type and the understanding of the operator. In my own case, this is for my own vehicle which will be carefully operated to avoid cell damage. I've incorporated a meter that reads total watt hours and amp hours used since last charge, plus system voltage and instantaneous current, to allow me to see just how much is left in the battery. It's also a short distance commuting bike, so will rarely exceed half its maximum range anyway - the LVC is really just a last ditch warning system.

Additionally, I'm using cheapish cells with a fairly high internal resistance. One consequence of this that there is a strong likelihood that cutting the current in half will bring the cell back up to the safe zone. Obviously, ignoring the warning to stop risks cell damage, but then that would be a very stupid thing to do, wouldn't it?

I could build in a more linear LVC, but it was much simpler to just use the half speed input on the controller as a warning mechanism. Were I building something to be driven/ridden by others then I'd almost certainly go for a more comprehensive protection system. As with all things, it's a matter of using the most appropriate solution for the particular application - I don't see that there can be a "one size fits all" solution that suits all EVs.

Jeremy

User avatar
EVguru
Posts: 305
Joined: Fri May 25, 2007 5:17 pm
Location: Luton
Contact:

Postby EVguru » Fri Aug 15, 2008 5:18 pm

In my own case, this is for my own vehicle which will be carefully operated to avoid cell damage.


If I had a pound for every time I'd heard that........ :lol:

The problem is that people say, 'well so and so's been getting away with it, so can I' and it's not true for them.

I've been told I was exagerating the care with which Lithium cells needed to be treated and the systems needed to protect and get the best life from them. The same person later told me that Lithiums were unreliable and unsuitable for the amateur user (they had destroyed quite a few cells).

I've been trying to get people to use decent protection systems and instumentation for years, on lead acid! Before that it was using a proper charger. Some people thought that anything more sophisticaed than a variac and bridge rectifier was unnecessary complication and cost. Before that it was electronic speed controllers 'Don't use one of the them, they're inefficient and wreck brushes'.

I HATE being right about other peoples disasters, it's just depressing.

But I do reserve the right to say 'Told you so!'.
Paul

http://www.compton.vispa.com/scirocco/
http://www.morini-mania.co.uk
http://www.compton.vispa.com/the_named

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Postby Jeremy » Fri Aug 15, 2008 5:46 pm

EVguru wrote:But I do reserve the right to say 'Told you so!'.


I can assure you that I absolutely will not give you that opportunity!

It's worth remembering that we are not all numpties who have a total disregard for the physics and chemistry that underpins our projects............

Jeremy

User avatar
ChrisB
Posts: 4657
Joined: Fri Jul 14, 2006 11:33 am
Location: Hampshire on the Southcoast
Contact:

Postby ChrisB » Fri Aug 15, 2008 11:43 pm

EVguru wrote:I've been told I was exagerating the care with which Lithium cells needed to be treated and the systems needed to protect and get the best life from them...............


One reason I'm still very very nervouse of this Li tech.

Yeah they are light weight, pack a very good punch, and appear to be ideal for EV use, but wow look at all the control mech's you need in place to stop them ending up being door stop moduals :roll: :cry:

I've been following all the bms threads with interest, but goodness me what a load of faffing about :shock:

ChrisB
I reject reality and substitute my own !!!!!!

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Postby Jeremy » Sat Aug 16, 2008 7:07 am

To be fair, it's not just lithium cells that need some form of battery management. Nickel chemistry cells can be irreversibly damaged by being allowed to reverse in a series string due to one cell depleting ahead of its colleagues, so some form of BMS is useful. Even lead chemistry cells can be killed by abuse, it's just that they tend to give a fair warning of depletion by running out of puff. A BMS is nevertheless a useful addition to a lead acid battery pack, as it would reduce the incidence of cell failure.

The snag with lithium cells is that they give very little warning of being discharged. They maintain a fairly flat terminal voltage, and are able to continue to deliver high current, right up to the point where they are discharged. This is good, in terms of increased usable capacity, but provides no warning that a cell is about to die.

Depending on the application, there really is no need for a complex BMS. In many ways a decent battery "fuel gauge" is all that is needed, provided there is a small reserve capacity in the cells to allow for slight balance errors. The LVC circuit I'm using is very simple indeed, just a small handful of cheap components per cell. It just gives an indication when any cell in the pack drops below a safe terminal voltage. It's up to the operator to take the hint and stop the vehicle at this last ditch warning, but then if he/she has been watching the "fuel gauge" then they will already know that they've been chancing there arm for the previous few miles, anyway.

To be absolutely certain of protecting the battery pack I could have opted to use the LVC signal to turn the main contactor off, perhaps after a few seconds of half power, so as to provide a safe warning (cutting power completely could be hazardous in some situations). Were I building this as a product I'm pretty sure this is the approach I would take, if only to prevent numpties from trying to continue to use power from a depleted battery.

Jeremy

User avatar
EVguru
Posts: 305
Joined: Fri May 25, 2007 5:17 pm
Location: Luton
Contact:

Postby EVguru » Sat Aug 16, 2008 8:13 am

Even lead chemistry cells can be killed by abuse, it's just that they tend to give a fair warning of depletion by running out of puff.


Optimas, along with many AGM Lead acids, don't give much warning. They have quite a flat curve and a sharp knee. When use in a high voltage system like my Scirocco, the situation was even worse. The performance under 30mph wouldn't really change until the total voltage had fallen from a nominal 192 volt down to around 96 volt.

but then if he/she has been watching the "fuel gauge" then they will already know that they've been chancing there arm for the previous few miles, anyway.


If they've been watching the fuel gauge.

I don't know about you, but I hear stories from time to time about people running out of fuel becuase they simply forgot to put any in! Almost every person who calls me on their mobile seems to be complaining they haven't got much charge left.

To be absolutely certain of protecting the battery pack I could have opted to use the LVC signal to turn the main contactor off, perhaps after a few seconds of half power, so as to provide a safe warning (cutting power completely could be hazardous in some situations). Were I building this as a product I'm pretty sure this is the approach I would take, if only to prevent numpties from trying to continue to use power from a depleted battery.


Cutting the main contactor under load can in some circumstances result in a blown controller. Ideally you want a soft cut as I explained before. In fact I've realised you can do it with a simpler system. Connect the output of an optocoupler across the throttle input. As the input current increases the output resistance will drop and the throttle input reduce. That's pretty much the same as Cedric's sysytem. If the input to the optocoupler was driven by the digital low voltage signal, you would likely get a nasty vibration, but an RC network with an appropriate time constant should deliver a reasonable soft cut.
Paul

http://www.compton.vispa.com/scirocco/
http://www.morini-mania.co.uk
http://www.compton.vispa.com/the_named

User avatar
Jeremy
Posts: 472
Joined: Thu Jul 31, 2008 6:35 pm
Location: Salisbury

Postby Jeremy » Sat Aug 16, 2008 12:56 pm

I'm not sure that it's ever truly possible to design out the "numpty factor". No matter how many safeguards you put in, I'm reasonably sure that a truly creative numpty would find a way around them. It could be argued that creating safeguards just adds to the problem, by allowing complacency. This is similar to risk compensation, where the increasing use of safety devices tends to cause people to take greater risks. I've long felt that a 6" spike sticking out of the centre of every car steering wheel would be an extremely effective way of inducing safer driving............

Whilst I'm sure that the analogue system would work as you describe, it would be challenging to get linearity over the normal operating temperature range. Optocouplers normally have a pretty horrible spread of both transfer gain and dark current with temperature. Anyway, the cheap and dead simple LVC sensor I'm using is a digital device (it's the TC54, which draws just 1uA from the battery) so I don't have the option to use analogue sensing.

I think it's tempting to get drawn into thinking that a complex BMS is essential, when in reality it may not be at all. As with everything, it's a trade between the risk of damaging a cell or two and the cost and complexity of the BMS that might have prevented it. If you make the BMS too complex, then the chances are that it may be a BMS problem that leaves you stranded rather than a battery one - the relative risk of failure of both the cells and the BMS needs to be taken account of, particularly if the BMS is a homebrew design that may not have the reliability of a well-tested commercial product.

I'll stick with my simple LVC as I am absolutely certain that it will do all I need, at least for now.

Jeremy

User avatar
Peter Eggleston
Posts: 162
Joined: Tue Apr 10, 2007 9:12 pm
Location: Tyne Valley Northumberland

Postby Peter Eggleston » Sat Aug 16, 2008 4:23 pm

And I will stick with my variac and bridge rectifier as I have for the last 6 years without toasting any batteries.
Peter

User avatar
EVguru
Posts: 305
Joined: Fri May 25, 2007 5:17 pm
Location: Luton
Contact:

Postby EVguru » Sat Aug 16, 2008 10:15 pm

Whilst I'm sure that the analogue system would work as you describe, it would be challenging to get linearity over the normal operating temperature range. Optocouplers normally have a pretty horrible spread of both transfer gain and dark current with temperature.


To start with the low cell signal is in the low state.

The voltage on the RC filter is zero and the current through the optocoupler LED is zero.

The optocoupler output dark current is not high enough to lower the throttle input signal from the common 5k pot or voltage singal from a hall effect throttle.

The low cell signal goes high, either due to state of discharge, or excess current draw for the circumstances (state or charge, cell temperature, etc.).

The capacitor in the RC network starts charging and when the voltage is high enough the optocoupler LED starts to conduct.

The optocoupler output starts to conduct and starts to pull down the throttle input.

The longer the low voltage system is present, the higher the voltage on the RC network and hence more current through the optocoupler.

Eventually the reduction in throttle input will relieve the load on the battery and the low cell signal will return to the low state.

The current through the LED will drop and the output will reduce the amount it is pulling the throttle input down.

If the low cell signal stays low, the throttle input will eventually return to normal.

More likely the low cell signal will alternate between high and low the the throttle input will vary at a frequency determined by the time constant of the RC network. This is the only difference between the fully analogue system (like Cedric's) and could even be seen as a useful additional warning. It can reduce the current to zero without ever causing an instant loss of power.

I don't know how the 'half speed' input on the Alltrax works. Does it reduce whatever input there is by half, or does it simply limit the maximum PWM duty to 50%? If the latter is the case, then it won't necesarrily limit the current.
Paul

http://www.compton.vispa.com/scirocco/
http://www.morini-mania.co.uk
http://www.compton.vispa.com/the_named

GregsGarage
Posts: 870
Joined: Tue Apr 01, 2008 5:27 pm
Location: Galashiels, Scottish Borders
Contact:

Postby GregsGarage » Sat Aug 16, 2008 10:27 pm

Lots of good ideas being discussed. 8)

My idea is to have the EV management module (must come up with a catchier name) sit between the throttle switch input and the controller (or KSI terminal on some controllers). It will receive the idle switch (or ignition) signal and then send the signal to the controller as long as all conditions are met. If a low voltage condition occurs, it will first alert the driver through a warning light and buzzer. It will also start a timer, 5 seconds is probably enough, that if the low voltage condition hasn't cleared within, it shuts off the idle signal going to the controller, which will shut down the controller. The driver will then know that unless he backs off the throttle, the vehicle is going to be shutdown. It is not limited to just battery voltage though. I have inputs for other things like temp snap switches, charger connected, regen, etc. I am also thinking about seeing if I can add a rpm input to prevent motor damage if over revved.

Concerns were expressed about reliability. In the event of a failure of the system, it can be easily bypassed with a simple switch that directly connects the idle switch to the controller.

It was also pointed out that you should have an indication about charge left in the battery pack. I am using a e-meter to keep track of amp hour consumed and have set the bar graph to show empty at 80% discharge.

Some e-meter type gauges also have a low voltage alarm connection. This could be connected to my module, enabling it's use on lead acid packs.

On a separate note, I have decided to abandon the use of my Ping BMS modules, and use Peters slave boards for my initial testing. The board I shorted out is a bit worse than I thought, and I have plenty of other things to be getting on with while waiting for Peters boards.
Greg Fordyce

Daewoo Matiz
http://www.evalbum.com/4191


Return to “All things battery related”

Who is online

Users browsing this forum: No registered users and 22 guests