oldspark battery bank w/possible solar

Keep in mind that inverter outputs cannot be paralleled. I can't help with a battery since the final load & reserve time hasn't been determined. But that would only be a size recommendation - you'd best seek advise from others wrt local supplies. We have no WallMart here and all I'd suggest is a particular flooded 110AH else Deka AGMs (aside from my own recommendations for Yuasa). Certainly Optima & Odyssey doe not fare well here. I hear that stateside, Kinetik are a good AGM. BTW from earlier - you would only run a generator whilst you are there and have the extreme loads like hotplates etc, else need to charge the batteries. And all fridges should have thermostats - how else do they regulate temperature?

I'm aware that normal inverters can't be put in parallel unless specifically designed to do so but I will do some research on the ones that are designed to be stack-able. If it doesn't require a total reworking of the circuit then I would consider it. I have some basic ability to modify simple things in a circuit otherwise I would just run multiple inverters. I think I want to shoot for the equivalent capacity of four of the Trojan batteries so roughly 450-525AH @ 12v. The A/C is no longer a factor and I will take your advice on the cook top using propane. That leaves mainly the computer and mini fridge along with small drain equipment such as cell phone charger etc.

You could save a ton more power if you grab a cheap laptop from craigslist and run off about 30 watts instead of 350. I did this recently and absolutely love it. My big overclocked gaming rig is nice, but the overwhelming majority of use is for internet or homework and an old laptop is perfectly suited for that. Keep the desktop for when you need it, but use the laptop for economy. Another bonus is never worrying about the laptop loosing power or faulting from a hiccup when you turn on something else power-hungry ... the battery is a built-in ups backup system. All this talk about matching batteries, series and parallel wiring, 6v and 12v makes me wonder why it's worth worrying over. I operated and maintained two broadcast vans (tv news style, mircowave antenna on a mast, etc). They used a setup very similar to what you want to build. We had a pair of marine deep cycle batteries with a 1kw inverter, and a 2kw generator for AC power. We also had a 40amp battery charger for shore power. All AC circuits had breakers, and all 12v DC circuits had fuses. Both 12v batteries were wired in parallel. We had a single battery fail twice (over 6 years), and neither time did it affect anything else in the system except for lowering the runtime of our gear. One other consideration. Try to avoid as many transitions from DC to AC and so forth. If your batteries run the inverter, you have a loss of efficiency. If your desktop computer feeds off that AC inverter and converts back into DC, you have another loss. But if you have a laptop with a 12v car adapter, there are no conversions. Also if you have a fridge with a 12v adapter, there are less conversions. If you can use 12v lighting, so much the better. Malcom: "This is the captain. We have a...little problem with our engine sequence, so we may experience some slight turbulence and then...explode."

Jayne: "We're gonna explode? I don't wanna explode.

A laptop would be nice but that's another added expense and while the need for the computer will be school, it will also serve as my entertainment.  The only thing that I do that bogs down my computer would be 3D magnetic modeling software, basic 3D CAD (SketchUP) or playing a game while having about 3 dozen browser tabs open in the background and music going.  The summer is going to be extremely tough financially since there is very limited financial aid available, this is the last year they will do Pell grant and that's all that is available beyond getting my tuition covered.  This wouldn't be such a big issue except I won't find out if I get any of the Pell grant until after school starts and I wouldn't receive any of that money until mid semester at the earliest.  I have a lot to do to the van especially considering the one I wanted is no longer available. 

For now I am going to try and under volt the cpu and gpu to bring the power down and I am considering getting a new monitor because the one I have is small and on it's last leg, if I get a new one I'll seriously consider an LED one for the low power consumption. 

As time goes by I plan on adding individual voltage monitors for each battery and create my own charging circuit that can isolate each battery and charge each one individually and appropriately but that will have to wait until I find some people at school that could help like the robotics club guys.

It's good to hear that you have implemented a system similar to what I want to do and it worked out for you pretty well, I hope I have the same luck!  I stopped by WalMart to check their deep cycle batteries but couldn't find any and they don't list them online so short of calling stores or going in to check I'll have a hard time finding them.  I really would prefer to buy them there simply because of their return policy. 

As far as the extra power conversions I hear what you're saying, if the laptop had a special built power pack that would be best, eventually I want to build a more efficient DC/DC psu but again that's down the road.  A 12v "adapter" for a fridge wouldn't have any benefit if it's used on a 110/120v fridge, that's essentially what the inverter is already doing and from what I've read and been told the purpose built 12v fridges aren't great.

I will take your advice on the 12v lighting, I've been thinking of LEDs. 

Something cool I found out from the guy who has this other van I may get is there is apparently quite a bit of room under the van along the driver side, enough to make a slide out rack for extra batteries!

Thanks for the reply and the useful info and suggestions!

Oldspark - This van that I may now end up buying doesn't seem to have a charge lamp indicator, what is your suggestion on setting up the charging system for extra batteries? Find a wire on the steering column ignition that only comes on when the van is running and use that for the relay?  If I can find one like that do I need to use that to power a smaller relay to power the actual switching relay?

teenkertoy wrote: We had a single battery fail twice (over 6 years), and neither time did it affect anything else in the system except for lowering the runtime of our gear....

teenkertoy - how long did it take you to figure out that one battery was bad? The problem is that since both batteries are at the same voltage, the good battery is dragged down and hence fails for the same reason as keeping batteries flat. The drag-down may not be bad if it is merely a capacity drop (ie, the bad battery may still have its normal voltage), but if it's a collapsed cell, then the good 12.7V battery will reduce to 10.7V... The longer that exists, the more the damage. (Then there are also high-discharge rates and - especially for AGMs - the possibility of thermal problems and gassing.) meltmanbob - as I wrote, I spent nearly $1,000 on another fridge because that was cheaper than not spending it. That $1,000 saved well over 5A of consumption at 12V - ie, 60W which meant another 100W panel and another battery (for a mere 12 hour reserve time). If you don't have a charge lamp or similar signal (but alternators usually do, even if newer ECU-interactive DP types), then a voltage sensing system is required.    That's unless you are happy with manual control in which case I suggest the ACC circuit unless you do want the batteries paralleled during cranking (in which case the IGN circuit). But keep in mind that with IGN or ACC on and one battery goes flat, they all will - hence the desire for the automated isolating systems I have described.

Oldspark - I understand the fridge situation but I don't have that much to spend. I'm going to try and pick up a small used chest freezer and convert it, I think I had mentioned and even linked a pdf article talking about doing this so that it only consumed roughly 100w/day. Regarding the charging system the guy said it was a 60 amp alternator and when I started the van I didn't see a dash lamp indicating when it was running on the battery or charging and running electrical off of the alternator. I think I got the IGN and ACC wires mixed up but I believe the description I gave was for the correct wire, just switched the labels. Seems like either I missed something earlier in the thread by not getting on the same page or I'm just confused by the additional information. The way I thought I understood it was this, charge lamp indicator is the same as the battery symbol indicator lamp on the dash. This is normally on unless the engine is running and only turns off at the end of cranking when the engine actually starts and all electrical load is now being carried by the alternator. The only other time it is not on is when the key is not in the ignition in any position but it is only not on the dash, essentially it is on but the lamp in the dash itself won't come on because the key has to be turned at least one position forward or backwards to let power flow through the dash lamp. Now that I think about it this sounds as though there is a circuit from the battery to the dash lamp with 2 relays and is open. One relay is tied to the ignition key position so when you turn the key to some position other than off and haven't started the vehicle, it pulls the circuit closed and the dash lamp comes on. The second relay is tied to the alternator indicator but is closed which still allows the dash lamp to be on. When the car is started, assuming the alternator is not malfunctioning, the alternator starts up and powers up that wire and pulls the circuit open. Does this sound right? If I thought about it some more I could probably figure out the same thing for the battery but I'm assuming it's very similar and I'm assuming that the description I just provided is accurate on principle but it's just speculation, let me know if I'm close or way off!

Yes... The charge lamp as used by the UIBI (Ultimate Intelligence Battery Isolator - aka a relay energised by the charge lamp - the same wiring as many electric fuel-pump relays in carbureted cars) comes ON when ignition is first turned on. It extinguishes when the alternator is charging - ie, after cranking and starting. It will also come on if the engine stalls - ie, the alternator is no longer charging. It is that circuit that is used for the UIBI. The lamp only comes on when the IGN power is on. That's because the lamp's +12V side is the IGN +12V which usually powers gauges and some other dash/warning lamps (oil, temp, brakes, etc). The lamp's other side goes to the L = chargeLamp circuit of the alternator. (This L-terminal is called D+ on "single-wire" alternators. Note that "single wire" does not include the standard HEAVY +12V to the battery nor the GROUND path through the alternator chassis to engine to ground etc.) So, one side of the charge lamp is +12V. The other side goes to the alternator (L or D+). That L circuit is grounded when the alternator is NOT charging. Hence the charge light is lit - it has 12V across it. When the alternator charges, its L circuit goes to +12V, hence +12V on BOTH sides of the charge lamp, hence 0V (zero Volts) across the charge lamp so it is therefore extinguished. It's like the L terminal is connected to #30 of a SPDT relay (2-way aka changeover relay). The Normally Closed contact #87a of that relay is ground. The Normally Open contact #87 is +12V. When the alternator charges, it energises that relay and swaps the L chargeLamp from #87a ground to #87 +12V. I was searching for some simplified diagrams for the above, but I keep finding the same old "system" circuit from older the12volt posts - eg, trailor alarm wiring (6th reply); adding a second battery (page 5, center); and stop starter motor once engine fires (page 2, center). Some of the surrounding text might explain things better. And that "Regulator" depicts the changeover-relay type of alternator L/D+ circuit I have tried to describe. There is a modern version of that diagram somewhere - it's the same but with the Aux battery moved to the LHS to highlight that the only thing added to the vehicle's wiring is the relay - the rest is standard vehicle stuff along with the additional Aux battery and its +12V cable and 2 fuses. FYI - the same charge lamp circuit is often used to test certain dash warning lights when the ignition is first turned on - eg, brake fault, low fuel etc (but not high-beam, handbrake, flashers, oil pressure etc). On older vehicles those tested lamps will be on whenever the ignition is on and the engine is not running (ie, the alternator is not charging). On newer vehicles they might NOT come on after an engine stall because they use smarter dash-lamp management - only the charge lamp will light. But I'll pause here for a while before explaining some further considerations - like if the L or D+ circuit is NOT powerful enough to energise the UIBI/relay).... Until then, see if this reply clarifies anything. And BTW - that relay could be a small or normal relay (15A or 30A) that energises bigger or more relays eg, one relay for each auxiliary battery.

Ok cool, I thought I had the right idea but I hadn't stopped to really think about it until then. I understand having to use a relay to trigger another, you mentioned that in another thread, something about the alternator L/D+ putting out a max of 250ma I think. If the relay that actually handles the power to the batteries requires a larger current to close then I would need to use another relay that could be triggered by the alternator and use that one to trigger the larger relay. You mentioned separate relays for each battery, I'm sure there is a way to detect the actual voltage on each battery and use that to trigger the relays. Essentially they would have 2 relay conditions to meet, engine running signal from the alternator and then a voltage ok signal from some sensor. The question is can it be done easily with an analog circuit or does it require using micro controllers and programming. Any thoughts on this? Also correct me if I'm wrong but wouldn't that only potentially protect from cell voltage degradation and not capacity degradation? I wonder how you would keep track of capacity degradation, I'm assuming that's much more involved. I guess the other part to voltage degradation is that you wouldn't be able to tell until you tried to charge it completely. In a situation like mine it could mask itself as not having had enough time to charge all the way and I might miss it because they would be constantly in use...

Bingo! for the first part.... Not that L/D+ is necessarily 250mA etc... Older external regulators used relays that could supply several Amps. Internal regulators might supply a few Amps, but maybe far less - after all, they only have to "sink" highish currents - ie, GROUND a charge lamp and maybe other tested lamps (maybe 4 x 250mA bulbs = 1A etc), but they don't have to source high currents per se. And then there are the very new alternators that might be designed to signal CPUs/ECU/EMS or LEDs etc, hence mere uAmps or mA.... And if overloading that L/D+ circuit ruins it.... (FYI - mine is currently faulty - my charge lamp and tested lamps do NOT come on. However, the alternator still charges fine. Many alternators MUST have a charge lamp (else those tested lamps in parallel) to provide a trickle/tickle current to ensure alternator charging - not that that means that circuit has to work.... Maybe mine has blown but still gets trickle current, or maybe mine does not need the charge lamp?) (BTW - I'm not sure what blew mine, but I have also lost 2 other circuits, though I think that is mere coincidence. READ: I recently started to tidy up messy wiring. That IMO should NEVER be done if it works fine... If it works, leave it - ignore other peoples derogatory comments!) HEY!! Stop conning me into un-pausing my aforementioned pause to discuss "weak" L/D+ circuits! But yes - totally correct - a small relay to power a larger relay and then larger or more relays... As to voltage sensing other batteries, I come back to "WHY?". You cannot test the ACTUAL capacity of a battery from its terminal voltage. An old battery with 5% of its original capacity might still be 12.7V at its "now" full capacity - it will simply crash thru 12.6, 12.5, 12.4....11.7, 11.6V about 20 times quicker than it once did. (To quote my old izu guru... "....(in practice) the only way to measure a battery's capacity is to discharge it with the desired load. And that does not mean it will do the same next time..." (ie, it could fail overnight). He then explained why critical batteries were therefore replaced after a certain time else a certain number of discharges. EG - critical UPS (Uninterruptable Power Supply) batteries with (say) a 10-year design life or warranty were replaced after 5 years - and even sooner if they had been used more often than intended! Yes - there are impedance measuring techniques, but they are too expensive and specialised for common use, and still merely a model.)      So back to voltage sensing... Why? If your system is charging, then connect your batteries and charge them. If you are not charging them, then isolate them to keep them independent UNLESS you want some or them to remain connected for whatever reason (ie, share a heavy load or provide longer reserve time). And if you can find a TRUE priority charging system (ie, based on the actual charge state of the batteries - not merely a "let's wait a while before connecting"), then I will ask why do you want to take longer to get maximum charge into all the batteries?    Determining NOT to connect a battery because it is too flat or faulty is another issue. That can be managed, but it is very complicated - it really requires data-logging etc. EG - you have 2 aux batteries at (say) 10.3V. One is fine but flat and hence wants recharging ASAP to minimise damage (ie, extend its life). The other is fully charged but one cell has collapsed.    How will your system decide not to connect the collapsed battery? And do you want to connect the other good but flat battery - what if its (initial?) charge current is too high? (What is "too high"?) IMO the best (and only practical) intelligence is YOU. All you need is a voltmeter. And maybe some suitable fusing - though that might assume your recharge current is higher or equal to your max discharge current - unless you want to have current sensing that controls a relay/breaker (ie to trip at 20A charge current or 100A discharge current)? Other than the voltmeter, INSPECTION. Feel for heat; check for gassing; measure rested open-circuit voltages (for collapsed cells) - ie, not with surface charge present. (Hence temperature alarms - a possibility but they must sense temperatures relative to ambient temperatures. Chargers might use temp sensing where a sudden increase in temp means the battery is fully charged - but that is when the charger is charging batteries... Batteries discharging into a load, in parallel...)    FYI - I have a 3 digit LED voltmeter in my dash. It sits across my battery: ie, -ve wire to battery-, and +ve wire to battery+ (although that is thru a relay which is IGN controlled but could be manually controlled or timed etc). I don't monitor my UIBI connected aux battery because it is not in use - it merely sits & gets charged. I place a DMM across it now & then. The dash voltmeter would alert me if either battery took very heavy current. (75A alternator; 2 x 38AH AGM batteries that typically accept up to 45A at ~14.6V after cranking.) I do have a great and cheap little LED voltmeter that could measure 5 or 6 batteries. Intended for LiPo batteries in series, it requires one battery across its main terminals (GND and Batt#1) to power it. It scans if other batteries are present (Batt#2, Batt#3 etc) and if so, displays their voltage With Respect To the main Batt#1 voltage. Keep in mind it's intended for a series chain of LiPO batteries (about 4V each) but it handles up to about 24V. But I haven't yet tested to see how it handles a -ve drop; ie, imagine 4 batteries with common ground at 12.4, 12.5, 12.0, 12.7V. The 12.4 is Batt#1 so it displays 12.4V. It then shows 0.1V for Batt#2 at 12.5V. Will it show -0.4V for Batt#3 (Batt#1-#3 = 12.4-12.0 = 0.4V) for Batt#3? It should show Batt#7 (12.7V) as 0.3V. It's not ideal, but for a small unit costing a $few that can measure 5 or 6 voltages (I forget), I thought it worth getting a few. Certainly handy for measure voltage drops along a path - eg, from alternator to aux battery or audio amp, power it (Batt#1) from the far end (amp or aux battery) and then successively connect Batts #2, #3, ... #6 inputs back to the alternator - ie, Batt#6 to alternator B+; the others at fuses or junctions in between. You would then see the "end" voltage (say 12.4V) and then all the voltage drops for each segment to the 14.4V alternator (displayed as 2.0V or 2.00V = 14.4-12.4V). THe target being to reduce each segment to 0V, hence te amp or aux battery equals the alternator voltage (so 14.4V, 0, 0, 0, 0, 0). Woops - too complicated? And totally stupid - except for dorks like me that do things on the cheap (for multiple applications).    A practical equivalent would be dedicated voltmeters, else a voltmeter that is switched between different points with some label that indicates what is being measured.    IMO that is a PIC or PICAXE or uPC application with a suitable display - maybe an LCD with "Voltage drop from Aux fuse-to-battery+ is: 0.13V. And add datalogging. And an ammeter to log "0.13V @ 10.0A" and later "0.26V @ 20.1A" (hence R = 26mΩ)... ... (at 18May11 01:20; 13.7°C; altitude 132m; ...) That's if you want to get into programming... I have an Arduino that could do it - probably the cheapest way to do this sort of stuff... Ready for another pause?

Wow long post, I haven't read it all but I thought I'd start my reply and go as I read. Funny thing about wiring... I cut and extended pretty much every wire in my mustangs wiring harness... all at once but I labeled each one... not a single problem and it's been over a year and a half As for the voltage sensing, I wasn't saying that it was a really great thing just that it could only possibly show you if a cell has gone bad but like you were saying in your posts, essentially knowing the voltage is only part of the picture. So I'm with you on the fact that you could have a normal voltage reading of 12v+ and still have a bad battery when it comes to capacity, I guess what I was getting at is ideally you would want a way to monitor voltage AND capacity to really have a better idea of how your batteries are doing. Ok back to reading the post... Ok whew...! After reading that I want to add to what I was writing earlier, the voltage sensing would be meant only to detect a battery with a failed cell, I'll leave that at that since I think we're on the same page now! I understand what you are saying about how essentially it is difficult or rather a very involved process to automate bad battery detection especially when you have many used together. Your example of 2 batteries with one having a bad cell and one being good but flat is exactly what I'm wondering about. Besides checking them on a regular basis are the only realistically effective ways to use a micro controller? Everything I can think of would require data logging as you mentioned. With data logging you could track how fast the voltage is dropping on each battery which I'm assuming could be a pretty good indicator of a "bad" battery and by bad I mean one that at the very least is significantly under performing compared to the rest. I guess you would also have to track the power draw to extrapolate current draw or monitor current draw directly but they should be important in factoring in the expected efficiency otherwise a high draw run when compared to a low draw run would make possibly all of the batteries to appear "bad," essentially it would give perspective. I'm not familiar with programming but that is part of the reason I am going to study electrical engineering. At least at the junior college level I will be taking all of the C programming language classes this fall and spring and I am going to try and have time to get into the robotics club since the engineering club there is kind of a joke in terms of kids that actually know how to put all that book smarts to use. Before I forget, where can I find one of these voltmeters you have mentioned that can track multiple batteries? My reasoning for eventually wanting to create a "smart" charger would be mainly for bad battery monitoring and charge them more efficiently. Maybe that's stupid or pointless but here is what I see being possible in an ideal situation: Track engine RPM to calculate alternator max safe current output Track individual battery voltages while operating and charging - bad cell detection Track power and current draw individually - reducing capacity detection, limit or cut off batteries if drawing too much from alternator Isolate batteries individually for charging that way if you end up not having enough time to charge all of them fully you won't have the less charged ones pulling the more charged ones down to where their voltage is not useful. I'm sure I could think of more and I'm sure most of that would be a waste of time but that last one made me think of another question. Here's the scenario - 1 starting battery, 1 or more reserve batteries. At least 1 if not more of the reserve batteries are significantly discharged needing say 2hrs @ c/10 charge rate. Start the vehicle which goes fine because the starting battery is fine, engine starts and reserves are connected to charge. Question - Do the reserves start drawing most of the power from the alternator ie starve the starting battery and keep it from recharging from just starting the vehicle? Also do the reserves draw power from the starting battery to try and reach equilibrium and only once that happens can the starting battery actually start to recover it's charge? To me it seems like at the very least the starting battery would be neglected until the reserves get closer to it and possibly get sucked down to their level. There is something I'm obviously missing and don't understand about this situation so I'm just trying to clarify! Oh the other thing about an ideal charger would be to sense when a battery should go to a trickle charge so a slower charging battery doesn't end up over charging one that's already charged. Again I'm speaking of what seems ideal, I don't have the knowledge or experience yet to know if that is actually practical or reasonably possible :) On a side note I did want to say thank you for taking so much time to explain a lot of this stuff, most people don't both with such constructive conversations especially when educating someone who has lots of ignorant and big ideas!