If your schematics are hand drawn I would be willing to create an OrCad schematic if you want. PM me if you are interested.
Stewie
Ultimate (?) DIY dual battery controller!
- Thread starter Rando
- Start date
If your schematics are hand drawn I would be willing to create an OrCad schematic if you want. PM me if you are interested.
Stewie
Conrad_Turbo
Observer
I am very interested in more info on this. What do you recommend or what is your circuit look like? I have poked around with microprocessor control in the past and the battery capacity is also dependant on temperature, not just current draw.
Based off this:
I was just tinkering around with battery capacity as well as how long the battery would last under the existing load. I got the coding stuff alright, but just wasn't sure what to use for an accurate measurement of high current levels, without wasting battery power as well.
Last edited:
I am using a 1 mOhm (0.001 Ohm) shunt resistor (rated at 2W) with a fully differential instrumentation amplifier (INA169) to sense the voltage drop and amplify it to the range of my ADC. The hard part about sensing currents in an automotive environment is the stupid common ground. You have sense the current on the high (12V) side, so you are sensing a small voltage drop (1mV at 1A) on high common mode voltage.
In some ways it would be easier to use a hall-effect type current sensor, but you can run into issues with getting sufficient resolution due to the zero-current offset voltage.
The temperature issue is an important one, but it doesn't effect the coulomb counting technique so much. It also has less of an effect when considering capacity (as in Ah) as opposed to strictly state of charge (%). It is true that the open circuit voltage at a given SOC goes down with decreasing temperature, but so does the available battery capacity (in Ah) which somewhat offsets this effect.
The other thing to realize is that the currents are not going to be as high as you expect. Using an 66Ah AGM battery as my house battery, even when at 30% SOC I see a MAXIMUM of ~17A going to the house battery from the alternator, with a stock 90A alternator. There is a tendency to overestimate the current to a house battery and over size the components.
In some ways it would be easier to use a hall-effect type current sensor, but you can run into issues with getting sufficient resolution due to the zero-current offset voltage.
The temperature issue is an important one, but it doesn't effect the coulomb counting technique so much. It also has less of an effect when considering capacity (as in Ah) as opposed to strictly state of charge (%). It is true that the open circuit voltage at a given SOC goes down with decreasing temperature, but so does the available battery capacity (in Ah) which somewhat offsets this effect.
The other thing to realize is that the currents are not going to be as high as you expect. Using an 66Ah AGM battery as my house battery, even when at 30% SOC I see a MAXIMUM of ~17A going to the house battery from the alternator, with a stock 90A alternator. There is a tendency to overestimate the current to a house battery and over size the components.
I am very interested in more info on this. What do you recommend or what is your circuit look like? I have poked around with microprocessor control in the past and the battery capacity is also dependant on temperature, not just current draw.
Based off this:
I was just tinkering around with battery capacity as well as how long the battery would last under the existing load. I got the coding stuff alright, but just wasn't sure what to use for an accurate measurement of high current levels, without wasting battery power as well.
As an aside I do also have a temperature sensor I added to the system, DS1820 on wire digital sensor which I use for measuring ambient temperature. If you were so inclined it would be simple to add another to the one wire bus ans measure battery temperature as well and apply a correction.
wrcsixeight
Adventurer
A couple questions and obseravtions.
Is this 17 amps maximum at idle or at any higher rpm?
Since even a small 66 A/h AGM battery at 30% SOC should be able to absorb a lot more than 17 amps, why is it not. Alternator, isolator, wiring, age of battery?
I have been extremely dissappointed with my 130 amp rated alternator's output. I found a 4 inch piece 14 awg fusible link in the charging path was getting extremely hot when my 2 group 27 flooded house batteries were under 85% and wanted more than 25 amps.
Recently I added a 2 AWG wire from the alternator to my battery switch ignition feed(leaving the existing wiring). I also added another 2 awg wire from alternator to frame.
I have noted an improvement, sometimes vast, except when hot, at idle speed. If my lights are on, and the blower motor, My batteries are supplying 8 to 12 amps when hot idling, but give it another 500 rpm and they are receiving somewhere over 25.
The maximum I saw before the upgrade was a very short lived 62 amps which tapered to 12 in under 5 minutes.
Now I'll see a max 95 amps taper down to ~48 amps at engine RPMs over 1200 and battery SOC under 75%. Hot Idle speed amperage is still pathetic.
I can just feel the extra load of the alternator by the gas pedal on my V8.
Conrad_Turbo
Observer
Ah something like this then: http://www.sparkfun.com/commerce/product_info.php?products_id=9028
This looks quite interesting too for a Hall solution: http://www.allegromicro.com/en/Products/Part_Numbers/0756/index.asp
I'd assume the major current draw would be when first connecting a dead battery to the main battery while the vehicle is running. What type of current spikes have you seen with this scenario? Any other load I would assume would pale in comparsion to the dead battery scenario.
What do you recommend for electronic switching of an isolated battery? I am not a fan of relays.
This looks quite interesting too for a Hall solution: http://www.allegromicro.com/en/Products/Part_Numbers/0756/index.asp
I'd assume the major current draw would be when first connecting a dead battery to the main battery while the vehicle is running. What type of current spikes have you seen with this scenario? Any other load I would assume would pale in comparsion to the dead battery scenario.
What do you recommend for electronic switching of an isolated battery? I am not a fan of relays.
The atto pilot current sensor is where I cribbed my design from. You can get the INA169 as a sample from TI if you want to build it yourself. If you were to use a hall effect sensor, I would recommend an 'open loop' type sensor such as this:http://sensing.honeywell.com/index.cfm?Ne=3025&ci_id=154286&N=3029&la_id=1
That way you don't have to bring your high current lines onto a PCB and can use heavier gauge wire (say 8 AWG).
I really don't see much in the way of current spikes. When I first combine my batteries, I may see 20 -30 A for a couple of seconds, but it fairly rapidly drops down below 20A. I think part of it is the different chemistries between my regular old AC-Delco starting battery and my Diehard AGM house batteries. While I agree that a mechanical relay is pretty poor option, I am still using one and haven't gotten round to adding a solid state system. My plan is to use an array of P type MOSFETS for the switch. They can easily handle 40A continuous with heat sinking and 100A surge.
That way you don't have to bring your high current lines onto a PCB and can use heavier gauge wire (say 8 AWG).
I really don't see much in the way of current spikes. When I first combine my batteries, I may see 20 -30 A for a couple of seconds, but it fairly rapidly drops down below 20A. I think part of it is the different chemistries between my regular old AC-Delco starting battery and my Diehard AGM house batteries. While I agree that a mechanical relay is pretty poor option, I am still using one and haven't gotten round to adding a solid state system. My plan is to use an array of P type MOSFETS for the switch. They can easily handle 40A continuous with heat sinking and 100A surge.
That is the steady state rate of charge at cruising speed (~2000 rpm) after a minute or two with a fairly deeply discharged battery, maybe more like 50% SOC than 30% SOC.
It is a newer battery < 1 year old, and the end to end resistance through the wiring is less than 5 mOhm. I think a big part of it is the difference between the AGM and regular flooded battery.
That being said, I have no problem with this. I don't want to charge my battery at more than about C/4, so 15 - 20A is just fine. Fast charging a deep cycle battery generally shortens its life, and requires the use of huge wire and associated components, which I would like to avoid.
It is a newer battery < 1 year old, and the end to end resistance through the wiring is less than 5 mOhm. I think a big part of it is the difference between the AGM and regular flooded battery.
That being said, I have no problem with this. I don't want to charge my battery at more than about C/4, so 15 - 20A is just fine. Fast charging a deep cycle battery generally shortens its life, and requires the use of huge wire and associated components, which I would like to avoid.
Conrad_Turbo
Observer
Rando, I am starting a preliminary dual battery setup on my vehicle. I don't have a crazy expedition truck...but am simply building up my 2004 Rav4 to be a bit more utilitarian and get a bit further off the main roads.
I'm going to run a dual battery setup in the front, add a 5025 Blue Sea fusebox under the hood and one in the trunk (4ga power wire running to the rear, w 100A breaker under the hood). The batteries will be isolated with a 200A relay (PAC200) and the connection between the two batteries will be controlled via a zBasic microcontroller (because I'm not an electrical guru and I only really know how to program that type of microcontroller).
I am not a fan of relays...but I'll start with controlling that for now and upgrade in the future if needed. But really the relay is only going to be using electricity when there is some form of power being supplied to the batteries (either from the alternator or external battery charger)...so I guess a relay isn't so bad in this case.
I'm going to run a dual battery setup in the front, add a 5025 Blue Sea fusebox under the hood and one in the trunk (4ga power wire running to the rear, w 100A breaker under the hood). The batteries will be isolated with a 200A relay (PAC200) and the connection between the two batteries will be controlled via a zBasic microcontroller (because I'm not an electrical guru and I only really know how to program that type of microcontroller).
I am not a fan of relays...but I'll start with controlling that for now and upgrade in the future if needed. But really the relay is only going to be using electricity when there is some form of power being supplied to the batteries (either from the alternator or external battery charger)...so I guess a relay isn't so bad in this case.
Jay H
servicedriven.org
Thanks for the temperature corrected voltage vs state of charge charts, those help a lot. As I am likely to never see temps lower than -5F and will probably see greater than 110 a lot this helps me significantly. I was thinking low voltage cut off at 11.2V but I will go even higher to 11.5 now.
I think I found something that is cheap and looks good for those of us that cant program micro controllers, its a kit.
http://secure.oatleyelectronics.com/product_info.php?products_id=270
Its only an 80a relay but I cant imagine ever needing more than 40A from a second battery in my set up. I.E running a 6 amp fridge and 5 amps of LEDs at most. I would also hope even fully drained that my second battery would never need more than 20A when charging.
I think I found something that is cheap and looks good for those of us that cant program micro controllers, its a kit.
http://secure.oatleyelectronics.com/product_info.php?products_id=270
Its only an 80a relay but I cant imagine ever needing more than 40A from a second battery in my set up. I.E running a 6 amp fridge and 5 amps of LEDs at most. I would also hope even fully drained that my second battery would never need more than 20A when charging.
landcruising
Adventurer
Would this also work in a 24 volt environment?
Adventurous greetings,
Coen
Adventurous greetings,
Coen
Since working on this project, I have since changed vehicles and battery chemistries, so I no longer use this particular system. I may still have the code, although in all honesty you would do better to write your own code for your particular system. This was very much a tinkerer project, which required a fair number of tweaks to get right, which would be much easier to do if you have full understanding of how your code works.