Lithium Battery Low Voltage Failsafe.

[Philip Middleton]

This is a project I’ve recently started working on.
I’m trying to setup a LiFePo4 12V battery replacement for a normal lead acid car battery. Please note, this is NOT for an electric car, its for a normal Internal Combustion Engine Vehicle. I just want to replace the Lead Acid Battery with a purpose built Lithium Battery with the required under/over voltage protection included.
I’m trying to achieve a complete lithium battery replacement solution for someone to simply pull their lead acid battery out and replace it with a relatively maintenance free safe lithium battery that should last many times longer than a lead acid battery.
I currently have a set of 2 X 4 cells to make up the required 12-14V. So over voltage shouldn’t be an issue.
I also have a set of 2 X 6 SuperCaps to help supply the extra power for starting (and avoid voltage sag when starting).
LiFePo4 Cells are 10Ah each, able to supply 100A each for starting (for less than 30 seconds).
They are arranged in 2 banks of 4.
SuperCaps are 500F, again arranged in 2 banks of 6. I don’t understand Caps and particular Farads well enough to understand how to convert capacitance to Amps available, but my understand is that 6000F available from my SuperCaps alone will start a heavy diesel truck engine by themselves.

The SuperCaps are not an issue, they are (as I understand them) quite robust and will last many cycles taken down to 0V.

The LiFePo4 batteries are a different matter. What I need is a low voltage protection solution so if the total voltage drops below for example 10V (2.5V per cell) the power is cut via a circuit breaker type device.
So, if the Driver leaves the headlights on or the radio going after the engine is switched off, the battery switches off before it damages its self and while still leaving enough power to start the car.

I have noticed that all the circuit breakers I have encountered so far are only based on current draw. That’s no good for my situation. I need the circuit to break on a low voltage situation, not a high current situation.

Can you make a suggestion? I’m not an auto electrician, or an electrical engineer. I have a very basic understanding of electronics and a good working knowledge of auto electrics. So I don’t know if such a device is available that acts as a circuit breaker on a low voltage trigger, or if such a circuit would need to be designed and built. I would be capable of building it, but not designing it.

Any help or hints appreciated.
Thanks… Philip Middleton

 

[ronsimpson]

What you need is a relay. It has a switch and a coil that holds the switch on or off.
What is the maximum discharge current? (starter current)
New idea. Often the heavy starter wires go directly to the battery. So the load through your "relay" will be head lights + radio. So maybe only 40 amps???? If 40A that is much better for the price. I was thinking about 400A.

I think some one makes what you are looking for. Try looking at RV electronics. (recreational vehicle) They often have two batteries. One to start the car and a second one for lights. If the light are on too long there is a box to shut off the power.

Relays come in many types. When there is no power applies to the coil the switch is (NO normally open) or (NC normally closed) Some relays come with both a NO and a NC switches.
This one in the picture is designed for 40 amps and 14 volts. (automotive) You probably can get it at the automotive repair store.

This will not work with out more parts but ...... think about the logic.
If the "coil" of the relay is put across the battery, When the battery voltage is above 7 volts the switch will activate. When the voltage drops to 2 volts there will not be enough power to keep the switch closed and it will open up and shut off your lights.
I know you need different voltages and we can do that (later).

This is a starting pint. What do you think.

 

[MikeMl]

On the caps: C*dV/dt = q = i*t

if you assume the starter draws 500A, let's see how much a 6000F capacitor will drop per second:

dV/dt = i*t/C = 500A*1sec/600F = 0.83V/s. If all you have is the caps, it better start within the first 3 sec or so.

Are you sure that your capacitors are capable of delivering 500A? What sort of terminals do they have?

I have posted (several times) a low voltage cut0ff circuit based on a TL431 on these forums. Search for "TL431" in posts by user "MikeMl" (the last letter in my user name is a lowercase L)

ps: I found one. See posts #9-11 in this forum thread.

With a bit of work we can replace the LED with the relay in Ron's Posting.

 

[Philip Middleton]

RON Simpson
I cant believe I didnt think of that. Thanks.
I'm thinking I might just need a resister to drop voltage enough so that the relay coil is not getting enough power when the battery drops to only about 10V.
Battery internals will need to be setup for one + terminal wired direct to Cells and Caps, wired solely to the vehicle starter solenoid.
A second + terminal wired through a large relay adjusted with a pot to go open at about 10V.
I realise my circuit idea (1 variable resistor) is probably a little too simple as the difference between 10V and 12 or 14V is probably not going to trigger a large relay.
I have a couple of 500Amp relays but They may be too large (the relay coil alone may depleat the battery in a few days or a week, but I'm not sure).
In general, I like it. Its simple and should be robust.
Any ideas about the circuit (or value of the pot)?

Thanks. Philip.

 

[Philip Middleton]

On the caps: C*dV/dt = q = i*t
if you assume the starter draws 300A, let's see how much a 6000F capacitor will drop per second:
dV/dt = i*t/C = 300*1/600 = 0.5V/second. It all you have is the caps, it better start within the first 5 sec or so.
Are you sure that your capacitors are capable of delivering 300A? What sort of terminals do they have?
I have posted (several times) a low voltage cut0ff circuit based on a TL431 on these forums. Search for "TL431" in posts by user "MikeMl" (the last letter in my user name is a lowercase L)
ps: I found one. See posts #9-11 in this forum thread.
With a bit of work we can replace the LED with the relay in Ron's Posting.

Thank you very much for the math. Your right, there's not as much overkill there as I was lead to believe.
the terminals also need to be soldered. I was about to use 10 AWG Gauge silicone wire for each of the 2 X 3000F Banks.
I'll look at the post now, thanks.
OK. let me chew on that for a day.

Thanks heaps. Philip.

 

[Philip Middleton]

A Very quick question Mike.

On the caps: C*dV/dt = q = i*t
if you assume the starter draws 500A, let's see how much a 6000F capacitor will drop per second:
dV/dt = i*t/C = 500A*1sec/600F = 0.83V/s. If all you have is the caps, it better start within the first 3 sec or so..

I noticed the small correction and explination, thanks.
I'm a complete novice here, so I'm probably missing something. but why is C 600F and not 6000F?
Also, I have no idea of what amps the supercaps can supply. I only know they are 500F each, and I have a total of 12 (2 banks of 6).
500F 2.7V Green-Cap EDLC(DB)

 

[MikeMl]

I'm guessing that on a big V8, the starter initially draws ~500A, but the current might drop a bit once the engine is cranking. Probably draws more in cold weather, less in hot weather or on an engine restart.

If the caps are 500F ea, and there are six of them in series, then the effective capacitance per bank is only 500F/6 = 83.3F/bank. Two banks in parallel makes 167F total.

As to the peak current that the cap bank can supply, you would have to check the mfg's specs. Even #10 wire is likely too light.

Ps: I downloaded the data sheet from here:
www.masters.com.pl/files/ds/samwha/samwha-green-cap.pdf

On page 24, it says that there is only a 400F capacitor, not a 500F. The max current that the cap is rated for is 180A. It has an effective series resistance of 3.5mΩ, so it would drop E=IR=180*0.0035 = 0.6V.

This makes it doubtful that a bank of two parallel strings could start an engine without help from a battery.

 

[alec_t]

If you connect the supercaps in series, how do you propose to ensure they are voltage-balanced? From what I read, if a supercap is allowed to become reverse-biased its ESR (internal resistance) increases. You would need a very low ESR for engine cranking.
Lithium and lead-acid batteries require different charging profiles, so the standard alternator and its regulation system would almost certainly require modification.

 

[Philip Middleton]

If the caps are 500F ea, and there are six of them in series, then the effective capacitance per bank is only 500F/6 = 83.3F/bank. Two banks in parallel makes 167F total.

As to the peak current that the cap bank can supply, you would have to check the mfg's specs. Even #10 wire is likely too light.
Ps: I downloaded the data sheet from here:
www.masters.com.pl/files/ds/samwha/samwha-green-cap.pdf
On page 24, it says that there is only a 400F capacitor, not a 500F. The max current that the cap is rated for is 180A. It has an effective series resistance of 3.5Ω, so it would drop E=IR=180*0.0035 = 0.6V.
This makes it doubtful that a bank of two parallel strings could start an engine without help from a battery.

Mmmm.. OK, its obvious to me I know even less than I thought I knew about caps. And I knew I didnt know much to start with.
I need to Halt my project, test out some of the assumptions I have been making.
See if a single bank of caps will start a small v8 (4L). if not, see if two will. if still not, then up the wire to #8 or larger.
Back to basics. See if things work without the safety features. If yes then proceed to adding features. If not then stop or rethink sizing.

This is one of the initial video that inspired me to look into SuperCaps and lithium batteries for my vehicles.

I'm quite familiar with Lithium batteries due to being involved in RC vehicles for many years. (Cars, Boats, Quads etc).
But, Supercaps are new to me. this video shows a single bank of 350F caps (6) starting a 4 cylinder engine.
A single bank of 6 caps (350F each) was able to crank the engine over for about 3-4 seconds before giving out. Engines in good condition start within that amount of time.
I think that upping to 500F and adding another bank should start just about anything (just probably not wiz it over for very long).
add to that 2 banks of 4 10Ah LiFePo4 cells and... well, I'll try it out in the next day or too to be sure.

 

[Philip Middleton]

www.masters.com.pl/files/ds/samwha/samwha-green-cap.pdf
On page 24, it says that there is only a 400F capacitor, not a 500F. The max current that the cap is rated for is 180A. It has an effective series resistance of 3.5Ω, so it would drop E=IR=180*0.0035 = 0.6V.
This makes it doubtful that a bank of two parallel strings could start an engine without help from a battery.

There are plenty of these caps on ebay atm... however that's not to say someone isn't marking them incorrectly
**broken link removed**
I wonder how expensive a capacitance meter/tester is...

 

[alec_t]

I wonder how expensive a capacitance meter/tester is...

If you're considering buying one, check the spec to see what is the maximum capacitance it can measure. Most have a max much less than supercap values.

 

[MikeMl]

[QUOTE="Philip Middleton, post: 1324653, member: 273039"...
This is one of the initial video that inspired me to look into SuperCaps and lithium batteries for my vehicles.

....[/QUOTE]
The guy who made the video doesn't realize that car starters are not rated for more than 10 seconds of cranking followed by a 1 minute cooling off period before cranking again... (duty cycle less than 20%).

A healthy standard OEM lead-acid battery would be capable of cranking that little 4 cyl engine for several tens of minutes( but not continuously, otherwise you would be replacing the starter motor).

 

[Philip Middleton]

[QUOTE="Philip Middleton, post: 1324653, member: 273039"...
This is one of the initial video that inspired me to look into SuperCaps and lithium batteries for my vehicles.

....

The guy who made the video doesn't realize that car starters are not rated for more than 10 seconds of cranking followed by a 1 minute cooling off period before cranking again... (duty cycle less than 20%).

A healthy standard OEM lead-acid battery would be capable of cranking that little 4 cyl engine for several tens of minutes( but not continuously, otherwise you would be replacing the starter motor).[/QUOTE]

Yes, I realize that but decided not to talk about that as that makes comparisons much harder to do and once or twice is not going to kill the starter, just give it a year or so's wear in a few minutes.
The first starting attempt was not using a lead acid battery, it was using 6 much bigger supercaps (2,500F each if I remember correctly) Lasted about the same amount of time as a small lead acid battery would have done (maybe a little less).
Anyway, now setup for testing my banks of Caps and Cells tomorrow. I'll do testing on a small v8 (4L LS400 Lexus engine). Will let you guys know how it goes.

 

[Philip Middleton]

If you're considering buying one, check the spec to see what is the maximum capacitance it can measure. Most have a max much less than supercap values.

Yeah. Noticed that. Most only go from pF up to mF. Not exactly a useful range for supercaps.

 

[MikeMl]

... once or twice is not going to kill the starter, just give it a year or so's wear in a few minutes...

The issue is that cranking for more than ~20 seconds without giving the starter motor a chance to cool down overheats the motor's windings, and you will cook it. Much more likely to happen on a big engine than on the peanut whistle in the video.

 

[Philip Middleton]

The issue is that cranking for more than ~20 seconds without giving the starter motor a chance to cool down overheats the motor's windings, and you will cook it. Much more likely to happen on a big engine than on the peanut whistle in the video.

True enough, but its being used here as a simple quick and dirty method of demonstration. And as you say, small engine not so much of an issue.

 

[Philip Middleton]

I didnt get a chance to test out my supercap banks today, but MikeMl saying that my brand didnt make a 500F supercap got me doing some research to see what I had really purchased.
I found this. Mine look like the Black 4 pin.

So, looks like they range between 280F and 330F, so they are probably 300F - 360F caps that failed, and someone (probably the manufacturer in china) has re-badged them and sold them as 500 cheap, and to hell with everyone who buys them (kinda like the flood of over 4000mAh 18650's that are getting around ebay etc).

Oh well, thought I would share.
Still going to use them, but dont expect as much of them anymore. Wont buy any more of them.