Trying to regulate 13V battery to 12V

For a project, I need to run some 12V components from a 13V battery. I have been searching for a switching voltage regulator that can be either bought or built which will regulate the voltage to 12V at 5A or more. High efficiency is vital as I don't want to avoid requiring heatsinks. I have been been searching for a long time and still don't have anything, although I admit that my limited experience with circuit design is likely a constraint.

Any suggestions to achieve this would be appreciated, but I also wouldn't mind some feedback on what I'm considering so far. The first is using a LTM4613EV. But the trouble I'm having with it is that none of the application examples included in the datasheet do not suit my needs. The closest example from the datasheet is this one which outputs 3.3V. I feel that there must be some way to adjust it to 12V instead but I don't know how.


The other option which I have just come across is to use a LT1083CP, which provides a much simple design:



The datasheet states a Vref value of 1.25, so I can easily find my resistor values according to the equation. My concern with this one is that it is linear, which I believe typically have a low efficiency and require the addition of heatsinks. The datasheet claims a high efficiency although I have trouble determining if it is comparable to a switching voltage regulator.

Any suggestions would be most greatly appreciated!

 

Because of the low difference between the input and the output voltage you need a "Buck-Boost" switcher. See:- http://www.linear.com/product/LTC3780 as an example.

 

There are several buck boost converters that will suit your app.

Chips


The LTM4613 you selected is voltage set by the resistor feedback on pin FCB, for 12V use a 5.23K ohm as per pages 10/11 of datasheet.

 

What kind of battery are we talking about here? Most automotive devices are rated up to 14 volt. So voltage regulator may not be needed.

 

A fully-charged 12V lead-acid battery (automotive or SLA) is ~12.65V at rest (during low current drain). A 5A load will cause the terminal voltage to drop ~1/4 V due to the internal resistance of the battery. The voltage will drop further as the battery discharges into the load. A 30Ah battery will maintain the load at >12.00V for several hours...

If this is in a vehicle with a running engine and alternator charging system, the battery terminal voltage will be >14V with the engine turning at a few hundred rpm above idle. The alternator voltage regulator in modern cars is set somewhere between 14.2V and 14.8V.

If it were me, I would more inclined to slightly lower the operating spec for the downstream equipment, say 11.5V, and then use a "low-drop-out" linear regulator. Switchers cause noise and RFI problems which are usually harder to solve than operating the down-stream equipment at a slightly reduced voltage...

Automotive starting batteries are not designed to power standby equipment, and should not be discharged to less than 90% of their capacity. If discharged even 10%, they should be recharged as soon as possible (within hours), and should be stored fully-charged, and periodically recharged (every two to four weeks) to compensate for self-discharge.

 

I think most of us here would just hook it up and not worry about it. What makes you think you cannot?

The second simplest solution would be to add diodes (10A rating) in series with the load. Each one will drop ~0.7V, so probably just one would probably do it.

 

Thanks for the helpful replies so far! I can see that I should provide a little more information on the project to provide context.
I am part of the University of Manitoba student chapter of SAE, we are designing and building a fully electric race car. I'm working with the low voltage control systems.

The battery in question is an Antigravity AG801. I also would have thought that we could get away with no regulation with the voltages being so close, but the guys who were on the team last year (this is my first year) said that they did not have voltage regulation last year and it caused a lot of problems. Also, we have a couple of engineers who are acting as mentors of sorts, and they seem to find regulation to be necessary.

I like the simplicity of this. But my understanding was that battery voltage drops a little bit as it discharges, so it's not constant. Is that a problem? It was my understanding that a regulator was necessary in order to regulate small variations to a steady 12V.

How do we lower the operating voltage of the components? I was told that they all run on a steady 12V signal.

As for charging of the battery, we are looking into having a DC/DC converter to use the car's main battery pack (300V) to charge our 13V battery. It has also been suggested that we could then not bother with the 13V battery and just run the low voltage components directly off of the 300V pack, but there was some problem with doing that, although I can't remember what it was.

 

That is what you are supposed to tell us.

Picture this: "Let's make a bunch of control circuits that only work on 12.0 volts, fail miserably at 12.6 volts, and use a 13 volt battery to supply them." Why would you paint yourself into a corner like that when most "brain" chips work on 5 volts?

 

You need to ask what the components are, and why they were chosen.

I am very familiar with aircraft avionics and engine monitoring systems. Most of the modern stuff will operate on a wide range of input voltages, typically from 10V to 35V. They do this so that the equipment can be installed in either 14V or 28V aircraft with no changes required. It is also designed that way so that it keeps working in the event of an in-flight failure of the alternator system.

Presumably, your downstream equipment is u-processor data aquisition and controls which runs on 5V. There is likely a regulator that drops 12V to 5V, and it wouldn't care a whit if it was fed 11.5 instead of 12V.

You need a "system engineer" that looks at the entire system, and designs a power distribution system that makes sense at the global level, instead of letting subprojects create a "hard spec" without considering its impact on the total system. Ask me how I know this game...

 

What exactly are you using that necessitates 12v? Everyone here is understandably suspicious of your 12v precisely requirement, and unless there is some hidden factor, I'm betting they're correct that your electronics will work just fine.

You mentioned a primary 300v battery pack and then a 12v automotive battery to run the electronics. You can using a switching regulator to lower 300v to 12v and omit the battery altogether, but depending on your requirements that may not be an option.

 

This flowchart shows the design of the overall low voltage systems as it stands now so you can get an idea of what components we are powering.



There are two separate 12V regulators because I haven't yet been able to find something that would handle enough current for all of the components together, so the plan is to split the components between two regulators. As you can see, we do have a 12V-5V converter, which is already figured out. Unfortunately though, I don't really have any other specifics on the various 12V components since that is someone else's department. All I have been told is that they need a 12V supply. But I will try to get more information on them to see if they will still run within a certain range.

The more I'm thinking about it, the less it makes sense to do what I was trying to do. From what my team lead has told me, it sounded like he wanted a regulator that would step down the voltage when the battery's voltage is above 12V, and step up the voltage when it is below 12V. That sounds to me like two different types of regulators put into one. Is that even possible? Even if it is, I'm starting to agree with you guys that a regulator may not be needed at all.

 

Assuming that the "Battery" on the left is the nominal 6-cell Lead-Acid rechargeable, then here are the things that can be fed directly from the battery voltage with no intervening regulators:

Battery Cooling Fans
Liquid Cooling Pump
Radiator Pump
Contactors
12V to 5V Converter

There are three that I would need to see some more info on:

BMS?
Motor Controller?
IMD?

 

+1
You'll waste time chasing your tail if you cannot get the mile high view.
One thing that jumps out at me from your diagram is that the 12V-5V converter ought to come straight off the 12V battery, reducing the load on any 12V regulator.
I'm assuming overall efficiency is a concern in this application, so forget my diode suggestion. Each 0.7V voltage drop wastes ~0.7/12 = ~6% of the power passing through. It's quick and easy but inelegant and a bit wasteful at low voltage.

 

Yes, a buck-boost regulator exists. You still are not sure you need it.
An important aspect is the minimum and maximum load current. You can do enormous amounts of current with one regulator, but you need to know what "enormous" means for THIS project.
The accuracy and cleanliness of the 12 volt supply is negotiable. Just saying, "12 V' is not enough. You could spend weeks trying to get 12.00 volts delivering 1.2 to 26.4 amps with not more than .001 volt of noise or deviation from no load to full load, and it would be mostly wasted effort.

Go get some data.

 

I don't see anything on your diagram that is 12 volt powered that jumps out to me as being anything remotely fussy about its 12 volt power input.

I think you are over engineering a in theory only problem that most likely does not actually exist regarding what you are working with. Especially if it is running devices that would normally see service in an automotive application.

If it is rated or intended to be used in automotive applications they will typically consider anything between 10 and 18 volts a workable range even if they are rated at 12 volts.

 

The battery is actually an 8-cell lithium ion motorcycle battery, if that makes a difference.
As for the BMS, I don't know much about it other than it can accept an input voltage of up to 16V, so it might not need be to be regulated depending on its minimum input voltage, but I don't know that at this point.
The motor controller is Unitek Bamocar-D3. The datasheet is translated from German and so is a little difficult to understand but it seems to state that it runs off of either 12V or 24V within a 10% range. I have just found out that last year's battery would get charged up to 14V and then depleted down to 7V, so I'm thinking the motor controller will need a regulated source from what I can tell.
The insulation monitoring device is a Bender IR155, which shows an input range of 10-36V with a nominal input voltage of either 12V or 24V.

Good point, that is actually something that has been suggested to the team and that we will likely implement since our 12V-5V converter can receive an input voltage from 6V to 18V.

I have been considering the maximum load current possible already so I think I'm okay there. However, I had not really considered minimum load current since I assumed that a regulator which can handle the max current would also be able to handle the minimum current regardless of what it may be. Is that an incorrect assumption?
and yes sir! I am very much realizing how crucial it is to gather data from others at this point!

 

Just for starters, the LM317L regulator chip must have .005 to .01 amps minimum flow or it will overshoot clear to the supply voltage.

.01 amps is probably not important in this case, but knowing the minimum load eliminates a guessing point for the designer.

 

Something like this might do the trick.

http://www.ebay.com/itm/DC-to-DC-co...930?pt=LH_DefaultDomain_0&hash=item4615814c4a

 

"14V and then depleted down to 7V."

That seems an important clue. Need more battery capacity.
This battery is for control only, correct?

Is that within normal discharge levels for this type of battery?
SLA's certainly wouldn't like that.

 

Yes the battery is for control systems only. But I should have specified that last year's battery which dropped to such a low voltage was a different battery. I think they had made their own battery pack out of 3 cells. I believe this new battery should have more capacity.


Right now I'm strongly leaning towards the LTM4605 for a regulator. It's a high efficiency buck-boost regulator that can give 5A of current which is exactly what I have been looking for. Plus it's cheaper than the 4613 that I was looking at before and the circuit looks simple to build. You have all been most helpful! But I still wouldn't mind additional feedback!