split charge, voltage sensing

The point is, why not use a plain relay?   (ie, under $10 (£5?) for 60A; $15 (£10?) for 140A etc??) A relay is a relay. It is not a split-charge relay or whatever. FYI - I have just been referred to National Luna isolators. They have what they call a "split charging" system on their voltage-sensing isolators. It is merely a 5-minute delay before the 2nd battery is connected. (So add a 5-minute delay to the charge lamp...) That's what I call a delayed connection. It connects after 5 minutes irrespective of how full the first battery is or was. Links don't work on here (usually). And the last you said was "...to switch the load to the secondary battery". We have not discussed that at all - we have merely discussed (battery) isolators - ie, the secondary loads are already connected to the secondary battery. And those loads and their secondary battery get isolated from the primary battery to ensure they do not discharge the primary battery - that's what the "battery isolator" is about.

See? I am learning all the time.Never underestimate the ignorance of others that dont know your subject ! At least,not mine.Ignorance that is. Yes,I understood your point about using a straight relay,its what I have done before,indeed am doing already,I just gave some extra backfill info on my OP. Sorry the links dont work,didnt know that.The Durite sensing relay I mentioned is sold under many guises all over the world,mainly for marine use.At around £40 it purports to cut in (their words) at 13.3v and cut out at 12.8v,with no connection to the alternator.As you have already said,this seems to be duplicating the action of the alt. indicator light....correct?The other point was another clarification of my OP in that I had thought to use a cheap "sensing" relay rated at 10a to trigger a heavy one at 100a +,costing me half what a heavy duty sensing relay would.But I have understood (some!) of what you say,I merely render this as backfill to my original post. Interesting what you have found on the time delay relay....I had thought along those same lines myself but disgarded for the same reason,ie,theres no telling the state of battery No1 when the time is up. So,just what is it that is inside these devices that say they switch by detecting voltage thresholds?Other than timers?Just what do they consist of ? And how do they work?and why?And the price of eggs?etc etc You are a good natured fellow to put up with my barrage.Thank you for pointing out the difference in the load switch thing.As you can tell,electricals are not my primary area,although I HAVE worked on large outdoor and indoor 3 phase etc,under instruction.I am more at home with mechanical engineering and diesel engines,welding and fabricating,milling and lathing.And shirking off whenever possible,of course.Got to go now and tie down a pig-ark,high winds on the moor where I live and farm,dontcha know.

shirker wrote: 'Spark,if you can tell me how to go about setting up a MOSFET system I would be eternally grateful.

Sorry - I missed that post/reply.... Yes - me too! I've been hoping for the almighty izu to feed me that solution as well.... I amn't too good with tranny & fet etc design, though IMO fets (MOSFET) design for switching should be ...er, simple. Well, simpler than transistors - no base (gate) currents, no "gain" etc. Merely apply a voltage on the Gate (aka #86) greater than the (full) turn-on voltage and it's on. And that voltage is about 5V (as low as 2V...) with a resulting ON resistance between Drain & Source (aka #30 & #87) of milli-Ohms. That's way simplified and with functional equivalence to relay terminals. The G (Gate; like the relay's coil or transistor's Base) is near infinite impedance. Its current in nano-Amps (nA) else micro-Amps (uA). Hence the first rule - include a resistor to pull G to S (source) to ensure it stays off... And I think that's the only rule - except be wary of static discharge, insert a resistor in series with G to limit possible DG currents or whatever, and add DS protection diodes if they ain't built in.    The resistors can be anything from 10 to 1M Ohm, with Rgs probably needing to be 10x the Rg value to avoid voltage divider effects.... Ah - found what I mean... See "An example of using the MOSFET as a switch" at electronics-tutorials.ws/transistor/tran_7.html - it's this diagram... Why didn't I just link that? Rgs to pull down & off; Rin = Rg to limit <whatever>. Anyhow, since that useless izu won't do my work for me, it looks as if I may have to. We both have similar situations... Unfortunately I started to tidy my vehicle's wiring. (Some good Samaritan warned I could get in trouble if inspected. Me: "But but...." But they don't know it is safe - they just see the bird's nest wiring...) Anyhow, some time after my last tidy, I realised my charge lamp was not coming on during boot time (initial start up) nor after stalling etc. Nor were my tested dash lamps (grounded by the L/D+ circuit). That means I wasn't charging.    But I was - my blue-LED dash voltmeter clearly going from 12.7V to ~10V (cranking) up to 14.5V etc. How is that possible? Dunno. But if it works, just leave it. (Or is that if it ain't burgered, leave it?) Sometime later I ran out of fuel. I knew I was low, but the trusty low-fuel light comes on when I have 1/11th of a tank left. So the "fault" means the low-fuel light doesn't come on. Oh well, I won't fall for that again. Until next week. Anyhow, I cannot find the problem. All wiring seems okay, and the various relays including my UIBI are as they always were.... But I know that the disconnection of one relay solves the problem.... Of course I will trace and find the problem. One day. But what better excuse or opportunity to try the MOSFET idea - ie, remove the "load" by inserting a MOSFET? Then after that is done and proven, I can continue to trace the problem... next year, or decade, or (preferably) millennium - ie, as soon as I (need to) get a round tuit. (I have not ruled out terrorism. Nor chanting or casting from a cluey dude in darkest England that is well versed in ancient cultures and hi-tech military...) Trouble is, that gi'damned izu is too lazy to do my work for me. He reckons I should do it. (He doesn't have a life. He ceased forum involvement ages ago! What a loser!) Apologies. It is unusual me to ramble needlessly - I am way too busy for that! It's probably a 5 minute job to wrap the wires around my MOSFET. I won't even bother with resistors... 'cos they'd take too long to organise (2 minutes to get, but wrapping to the FET...?) Pity about the 13-day job to find the MOSFETs I bought a few weeks back because I lost the previous MOSFETs that replaced the... I'd probably never do it for myself (my self-leadtime averaging about 30-years!), but since you are so keen, and obviously desperate... YES - your life depends on it....! A pity it's dark & raining now eh? Remind or kick me in a reasonable time... No promises, but this offer will expire 13 May 2042 (1 year 1 month 1 day plus my DIY-self lead time). [Hey - isn't that the UNIX y2k equivalent - or was that 2032? I know it's already been fixed - I'm merely considering excuses. (If only *NIX people were as greedy as other systems - they could have made a fortune in a couple of decades!)] PS... Remember - if you have many relays as a UIBI, the alt/reg (D+ or L) need only power a single (small) relay. THAT relay then powers the other relays. IE - a "normal" 30A relay with (up to say a) 250mA coil can energise a 30A-coil solenoid, or 30 x 1A solenoids, or 120 more of the same 30A relays (ie, 120 x 250mA solenoids = 30A = the fist or original's contact rating). Yes - we want to replace that first relay with a MOSFET.... (Why 250mA when it could be far less than 250uA!!) And remember too - it's the same principle for any isolator - smart, intelligent, voltage sensing, etc - the first is controlled/switched by whatever "logic"; it energises the others.   Why do we buy more or bigger smart/intelligent isolators for more or bigger connections when we just need to insert more or bigger relays??     [ That solution was so obvious - once I thought of it... (Or after I thought about it... LOL.) ]

Alas I missed your last reply... [ "See? I am learning all the time..." ] (Is my typing so slow?)    That time-delay isolator came from a local dude I am working with - albeit via mp3car.com PMs. Maybe I should cut&paste my verbiage on timers etc. But in summary, why limit what the alternator puts out? Why not get the fastest and maximum total recharge rate? Besides, if the alternator is of limited output, that increases battery life (by reducing recharge current) and still provides the maximum TOTAL recharge possible by recharging both/all batteries rather than limiting to just one. Of course, if that means the main battery flattens over time... But an artificial delay (like 5 minutes) is not the solution for that anyhow. Yes - I like the UIBI solution. Any old relay. Essentially only its current capability is of concern. (But I have self resetting circuit breakers in case it isn't!) And as soon as the system charges, both (or all) batteries are connected. No delays, and no hunting (off-on-off-on) if an under-sized alternator dips for both batteries, recovers for one, then dips when the 2nd is re-connected thence disconnected again.... And no inadvertent paralleling of batteries during cranking because of some disconnect delay. And how do I tell if the alternator isn't handling the battery(s) or load(s)? I have the UIVM (Ultimate Intelligence Voltage Monitor) - namely a dash-inserted $10 LED voltmeter.    From that I can also determine when my cranking battery is flat, or at end of life; if I am over charging or not charging etc. [ Most primitive vehicles had ammeters. As go/no-go electrical status became less important and battery/load condition became more important, the ammeter gave way to the voltmeter which gave far more useful information anyhow. But voltmeters seem to have disappeared yet temp and oil-pressure gauges remain despite their much simpler replacement by alarms etc! How strange. ] Most vehicles in critical or important situations will tend to have voltmeters. Voltmeters are oft considered more important than oil or temp gauges (which are relatively easy to replace with warning lights). The voltage-sensing isolators are merely a voltage sensing circuit. Easy to do in simple systems. But they need to add hysteresis or dead-band (eg; on at 13.3V, off at 12.8V), plus delays against temporary dips (lights on, brakes, indicators; and what about when cranking?) and to prevent fast hunting/oscillation/chasing (on-off-on-off etc). Hence these days they are usually PIC (PICAXE) circuits - ie, simplified micro-processors. The problem is, there is no "best" solution. I have seen isolators that don't disconnect until 12.5V. (LOL! For some reason, they later increased that voltage level - ie, why wait till the batteries have discharged 10% or 20% before disconnecting?) Some have an (optional) extra input to keep paralleled during cranking (eg, BlueSea) whilst others have it to ensure disconnection during cranking (I don't recall who/what). As a wise person wrote earlier in this thread regarding voltage sensing (not me; 'twas above my last...) "this seems to be duplicating the action of the alt. indicator light...". Alas, so well put. I have oft asked, WHAT is the "smart" etc isolator attempting to do? Isn't it trying to determine WHEN the system is charging? Isn't that when you can parallel batteries no matter what age, capacity etc (provided they are all (eg) 12V and lead-acid)? And isn't it when not charging that you want to ensure that at least the cranking battery is isolated so that no matter how loud your audio at the drive-in or early morning but crowded camping place was, or how long you cranked or winched your bogged vehicle or kept your beer and ice-cream cold, that you can still start your freedom machine to get away? Hmmm - what standard feature on many vehicles tells you when you are NOT charging. (Or rather; is +12V when your system is charging?)    I don't know if the voltage sensors were originally intended for non-charge light systems like marine etc "stator" or permanent magnet systems, but people certainly missed the boat when it came to the average car etc! [ It is at this point I would like to neither confirm nor deny that I or anyone else thought of using the charge-Light circuit after her/his/its voltage sensing battery isolator blew up. It has been alleged by some that that is an unsubstantiated quote from the UIBI inventor within my local community. (It is certain else known that it has been done before. I am referring to this local incarnation; this local Eureka moment! But it was after the (latest) rediscovery of concrete.) ] Whilst you Shirker are kind to say that I am a "good natured fellow to put up with my barrage", I'll see you and raise you the same, times two! Meanwhile be aware of the rolling eyes and sighs as they read that. But enough about them. Let's talk about us. You can look up voltage sensing circuits (maybe wiki.. zenor_diode, comparator, operational amplifier...?), but like I said for battery isolators, they need to be rather complex in how they respond to varying voltages - hence computer or PIC/PICAXE programming is more relevant.   (Did I say I like the simplicity of the charge-Lamp controlled rela.... er, battery isolator? And I really do like its recent UIBI name-assignment - whether that be as a reflection, or as sarcasm directed at alternatives...) But you are comprehending fine.... It is easy to get confused (not that my verbiage or expression helps), but I have dealt with others with suitable experience and knowledge that just don't get it! Then there are the frustrations of those that mis-apply - ie, when it is determined that they are talking stator or marine systems when I have responded to and am discussing a typical or normal car/vehicle system. (If only they answered WHY the voltage sensor was better - or why they even considered spending money on a voltage sensor with relay when just the relay itself would do - instead of trying to get me to add further detail to the advantages I already stated! But that falls under experience... I'm FETted out for tonite. I think... Cheers, Peter.

I ain't saying nothing! Just following this where I can, Peter, admitting to rat's nest wiring...blasphemy. The sender unit is where to look for the fuel gauge problem BTW.

The fuel gauge is fine, so too us the low fuel warning lamp. But I have suspected the low-fuel sender - maybe relay(s) have burnt it out - but that is diode isolated anyhow.... Ir still doesn't account for the charge light circuit not turning on the charge lamp (and its diode connected lamps - low fuel, brake-fault) UNLESS I disconnect a particular relay. Everything is as it was (as far as I can tell) and there is still only ONE relay to the charge Lamp. I am certain it is not a connection issue. The next step is re-checking the connections (to where & what) and then impedance or voltage probing... I do think it was something that developed, it didn't just appear after a tidy up etc. Maybe the regulator's solid state lamp switch has been compromised. (Hmmm - did I test-ground the chargeLamp circuit (after isolating from the alternator)?) Ahhhh - maybe next week. I have a full tank. And I can predict reasonably accurately when I'll run dry. FYI - I carry a jerry can, but I suspect a weak (mechanical) fuel pump as it isn't pulling thru after empty too well (yes, both filters are new). I have in situ a pusher electric pump (spare redundancy and all that...) that I intend to Tee into the line, but haven't found any local one-way 8mm fuel valves (a mere $8 each or similar form your country).

I'd also be looking at the 12 volt ignition supply to the instrument panel before the stabiliser if there is one. Track burn or open circuit perhaps?

No. The gauges are separate and they are fine (yes, water & fuel via a double-hetero-metal v-reg of ~8.5V which must be a digital PSU... on off on off...). (LOL) I removed the idiot-light module thinking it may be something there, but no - it behaves exactly as it should - all individually, and the tested ones when alt-L is gnd. [ As if my modifications would prove faulty... I have reconfigured tho the (red) brake-fault (which is merely brake fluid level in mine) is now the (green) 4WD engaged lamp (mine is 2WD), and I fitted the 3-digit LED voltmeter in place of the original brake-fault and an adjacent spare position. I wish that were the problem! ] I was going to test voltages now that I'm back home, but I decided to have brekky before my next outing.... I usually find these problems quickly - or at least diagnose them even if I can't find the wiring or the fault. But I have often written Never Never Ever tidy-up or rebuild a working rat's nest - it will never work thereafter. Granted, I was referring to fabricating and constructing proper (looking) professional PCB circuits - after all, wiring is plain simple wiring. What could possible go wrong? (Especially since I am very conversant with bad grounds and bad contacts, and even suspect that all wires have a discontinuity in their internal smoke.) Now enough about me. And my problem(s?). I'd much rather discuss others' problems!    (My tone leads to too much self mutilation when I attempt to advise myself verbally...)

Last hijack... (maybe?) 5 minutes determined that there is some fault with the alternator - the chargeLamp circuit is NOT grounding - it is high (about 1V below the +12V supply with IGN on, else low (~0.5V) with IGN off. That should not effect the low-fuel circuit - but maybe that is another problem - coincidental of course to ensure maximum diagnosis difficulty. (However we know to always break down into individual sections to test inputs & outputs separately because we assume those coincidence-gremlins to be ever present.) The good thing however is that the alternator charges consistently. I may leave the circuit disconnected to see if that continues - many (or most?) alternators require the charge light (or others) be present for charging to initiate or continue. Not that that is a general test for that alternator - since the open-circuit chargeLamp terminal is near +12V it is providing the tickle current that the charge lamp normally provides. I'll bet it is the alternator's solid state "changeover relay" that has gone shortish to +12V. It's times like this the external Bosch regulator would be welcome - but such alternators alternator would have been replaced several times by now. And they would NOT have charged with a faulty charge lamp circuit!