Bear in mind that you are only 'estimating' load. An exact calculation isn't necessary, in part because an exact calculation rarely holds true to life, so in effect, it is still just an estimation... and, in part because you will incorporate a margin of error to allow for fluctuations in load and to ease stress on your alternator. Understand that at times an exact calculation is useful, it just isn’t necessary for this particular exercise.
Excluding the winch, you guesstimated 388 watts in aftermarket accessory load - round that up to 400 watts. In addition, you estimated 65 amps maximum load for factory accessories. So...
400 watts / 12 volts = 33 amps
and...
33 amps + 65 amps = 98 amps, so ~100 amps of accessory load w/o the winch.
A 140-amp, three-wire alternator should be adequate for your purpose, requiring a 250+ amp diode isolator, such as a Cole Hersee 48161. But, diode isolators with that high capacity are pricey.
[Sidebar - Specifying a diode isolator: A two-battery isolator contains two silicon diodes; one diode for each battery connection. Each diode is rated to carry only one-half of the total current for which the isolator is rated. So, a 200-amp isolator can pass a maximum of 100 amps to each battery without fusing one or both diodes. If the charging system is capable of producing greater than one-half of the isolator rating, you run the risk of destroying the isolator. Fortunately, charging system output is virtually always apportioned between the connected batteries, so you will never see 100% of the charging current flowing through only one diode. The notable exception to this is if one battery is disconnected or removed from the system.]
Simple relay isolation that mirrors the factory wiring will be significantly lower in cost. For example, you might use a 225-amp, Cole Hersee 24812, continuous-duty solenoid with a variation of the factory circuit that Vile linked. Or, if you prefer the idea of a diode isolator in spite of the high cost, you could selectively shunt the diode isolator, connecting the batteries directly together during crank and heavy winching, using that same Cole Hersee solenoid. There are a lot of possibilities, depending on what you’re after.
Regardless, with a 140-amp alternator, use 4-gauge copper cable and an 8-gauge fusible link for the alternator charge lead. Use 1-gauge all copper cable without fusible links for battery-to-battery and battery-to-starter connections. All terminal ends and splices should be securely crimped, soldered with 60/40 rosin core, and double-wall shrink-sealed.
------------------------------------
You posted that your winch will draw ~370 amps under full load. Unless your crawling difficult courses regularly, winch pulls should be rare - not the typical routine. Still, when it occurs, a 400-amp draw is a high current burden to sustain from a single battery for very long. So, for greater flexibility and performance, as mentioned earlier, engage both batteries while winching heavy loads. Just be sure to connect the winch along with your other add-on accessories to the auxiliary battery, so they are generally isolated from the primary. If you readily want to be able to disconnect your winch from battery power when it's not in use, install quick-connect plugs that are appropriately sized to the cables. All auxiliary loads should be fused.
Consider an 800 CCA, deep-cycle, dual purpose, absorbed glass mat (AGM), marine battery for auxiliary (such as an Optima blue top) and an 800 CCA AGM (Optima red top) for primary cranking. The AGM batteries should yield increased service life over conventional open cell lead-acid batteries.
That Sterling Power PSR252 marine isolator looks like a great choice. Call them to verify, but it looks like they rate their isolators per bank. You might be able to upgrade to a 240-amp alternator in the near future and continue to use that same isolator.... :o
When I posted, "...you could selectively shunt the diode isolator, connecting the batteries directly together during crank and heavy winching,..." I meant connect the relay, or appropriately rated battery disconnect switch, between the battery posts of the isolator, so that the isolator and relay (or switch) are in parallel. This would connect the two batteries directly together on demand without disconnecting the alternator from the circuit. You could control the relay by a lighted switch mounted inside the cab. You could install a simple warning buzzer as well if you wanted the additional reminder that the batteries are in "Boost Mode."
Every configuration presents pros and cons that force the application of good judgement. Good judgement comes from knowledge and awareness. When planning a system, try to envision any hidden consequences of your design. Do your best to predict any potential weakness in its function. Try to allow for the odd, unintended contingency. Within the context of your project, its intended purpose, and the goals for your new system, ask yourself: “What would occur if ___(something bad)___ happened?” ...and... “What if, while I'm in the jungle, I suddenly realize I also need the system to do this: _______________ ?” ...and... “What if I decide to expand the system? Where will I tap in?” ...then build accordingly. For instance, if wire “X” chaffed against the sheet metal, is the circuit fully protected? This can be a bit more complex when integrating multiple power sources. That’s why both ends of the factory installed cable leading to the starter from the auxiliary battery has fusible links (referring to the link posted by Vile). There is limited space to connect cables to a battery. Do you need to incorporate a junction block off to the side? Where will you mount it?
--------------------------------------------
Following is a little more information to help you dial in the batteries and installation for your project.
Battery Overview:
Batteries do wear out! A fully charged lead-acid battery produces 12.6 volts across its terminals. Conventional lead-acid batteries are designed to cyclically discharge and charge between 65% and 100% state-of-charge without incurring abnormal damage or sacrificing lifespan. Discharge below 65% state-of-charge (~12.35 volts) is referred to as deep-cycle discharge. A conventional lead-acid battery will not tolerate repeated deep-cycle discharge without a noticeable loss of lifespan. When electrical load exceeds the reserve capacity of a battery, or the combined reserve capacity of parallel connected batteries, the battery enters deep-cycle discharge.
Undercharging accounts for nearly 40% of battery failures. Undercharging results from: (1) insufficient charging voltage actually reaching the battery, and (2) insufficient charging time for the battery to fully recover following discharge. The effects of undercharging can be greatly exaggerated by chronic deep-cycle discharge caused by overuse of electrical accessories, and by parasitic draws over long periods of battery non-use that result in sulfation (an impermeable coating that forms on the plates, limiting chemical activity and diminishing capacity).
Although not as common, overcharging and high heat also ruins batteries! So does freezing the electrolyte by prying loose internal components and cracking cases. Contrary to popular myth, batteries will not discharge by sitting on a cold concrete floor. However, when charging out of the vehicle, it's best to set batteries on a thermal insulator (wood block), so the internal temperature of the battery remains uniform, since temperature directly impacts the rate of chemical activity inside the battery.
Abnormal chassis vibrations and impacts, common with off-road trucks, can damage batteries internally by causing conductive materials to break loose, effectively decreasing plate surface area and causing internal shorts that diminish battery capacity. In addition, exceedingly stiff battery cables, when not properly secured, can wrench battery terminals loose from the case, resulting in electrolyte leakage, terminals separating from the plates and, occasionally, battery explosion from internal arcing.
Maintenance:
Batteries generally fail over a period of time. Neglect and poor maintenance, including improper maintenance procedures, are the underlying causes of most battery failures. Batteries require maintenance! Even “maintenance free batteries” require maintenance. At the very least, batteries should be kept clean, rinsed with clear water and scrubbed with a plastic-bristle brush. Never use chemicals to wash a battery case or neutralize acid-salt encrusted battery posts. Physically remove hold-downs to scrub away encrusted salts. Chemicals "find" a path into the cells through the vents and around imperfectly sealed posts and pads. If not completely rinsed away chemicals when moist are as conductive as electrolyte, resulting in parasitic draws. Disassemble, then scrub and clean battery posts and pads with a steel brush made specifically for that purpose. Keep all of the cable ends and terminals thoroughly clean of encrusting salts. Liberally coat battery connections, cable ends, and connecting hardware with viscous, non-conductive white grease to seal out penetration by the electrolyte. Use of the fibrous acid-neutralizing battery post rings available from the local auto parts store works, but check the rings and their effectiveness often.
Periodically remove the cell caps and check cell electrolyte level. Electrolyte levels should be consistent between cells and touching the bottom of the split rings. Grossly fluctuating fluid levels between cells potentially indicates a developing battery problem. If the tops of the plates are exposed, you procrastinated too long and probably have some battery damage. You'll need to add water, slow charge, and then test the battery. Add only distilled water when necessary to raise cell fluid levels. Never add supplemental electrolyte, battery rejuvenator, or any other chemical to a battery. Check battery mounting and ensure that it holds the battery secure and level.
Choosing the Right Battery:
Select a battery that properly fits the tray, or modify the tray, so the battery sits level and can be stably and tightly secured. Ensure that the terminals are a safe distance from metal hold-downs, brackets and any surrounding metal.
Select a battery with sufficient cranking capacity for the application. The CCA (cold cranking amps) rating is the maximum current a fully charged battery can deliver at 0° F over a period of 30 seconds without individual cell voltage dropping below 1.2 volts. CCA describes a battery’s ability to crank.
Make sure the battery can adequately supply the typical engine-OFF electrical load of the vehicle. Reserve capacity is the time in minutes that a fully charged battery can maintain a 25-amp load without individual cell voltage dropping below 1.75 volts. Reserve capacity describes a battery’s ability to supply all electrical loads while acting as the primary power source. The greater the reserve capacity of a battery, the longer it will take to discharge. When running multiple “parallel connected” batteries, match the reserve capacity ratings to avoid chronic differential discharge between batteries.
CCA and reserve capacity are distinctly different ratings. CCA is important for determining the amount of load you can place on a battery. Reserve capacity gives you an idea of how long the battery will support its load.
Cabling:
I touched on battery cable gauges in a previous post. Onto battery ground details and winch connections. Ground the auxiliary battery directly to the primary battery (negative to negative) then ground the primary battery to the engine. Routing the ground in this fashion, to a common point, will eliminate potential voltage loops and help avoid electrolysis. It also lessens the number of large and weighty cables dangling off the motor - one less thing in your way while working and showing off. :)
Run the winch power leads directly to your auxiliary battery. Use the winch manufacturer's recommendation for circuit protection. Run all other accessory power off of a jumpered 3/8" (10 mm) junction block with cover mounted adjacent to the auxiliary battery. If the block is more than two feet away, protect the jumper with a fuse link at the battery connection. Remember, fusible wire is always four wire gauges smaller than the wire it protects. So, a 4-gauge jumper will use a six-inch length of 8-gauge link.
Way too many words for such a simple thing. :'(
Over-specifying electrical cable for a given application doesn't hurt - it just doesn't benefit. Because of the high cost of good quality cable, the benefit-to-cost ratio tanks if you specify much beyond the actual need. Then there is the fact of introducing unnecessary weight, cable stiffness, and bulk - none of which benefit the end result.
The standard approach to wire selection is to decide on the amount of voltage loss that is acceptable over the length of the cable, and then use the appropriate formula and table (or quick reference chart) to determine suitable cable diameter to carry the anticipated current load. However, using commonly available charts is risky if and when the chart doesn’t clearly state the percent voltage loss upon which it is calculated. An engineered voltage loss of <2.5% is generally ideal, balancing cost and performance. If you're able to secure a 2.5% quick chart, laminate it for future reference.
Virtually all electrical cables have a unique resistance per linear foot of length. Hence, for any given cable material and cross-section, the longer the wire run, the greater the inherent resistance of the cable. Because of the relationship between voltage, current and resistance, the greater the current that is demanded by an electrical appliance or load, the more significant the inherent resistance of the wiring becomes with regard to that demand.
Current varies directly with voltage, but inversely with resistance. That is to say, as voltage increases, current increases; as resistance increases, current decreases. Ideally, 100% of the supply voltage should be available to power the load. By design, the resistance of any electrical appliance is the dominant resistance in its circuit. But, as cable resistance increases, the ratio of cable resistance to appliance resistance shifts, and more voltage is consumed by the wiring, rendering less voltage and less current to power the load. To manage acceptable voltage loss across a wire run, the cross-section (gauge) of the wire must be chosen with both the maximum anticipated current load and the total wire length in mind. The further a cable runs, the greater its required cross-section to minimize voltage lost to the cable.
A secondary problem that sometimes plagues electrical systems is the heat generated off the wiring when wires are of insufficient gauge to safely power the load. When a wire becomes hot, it's because the wire is consuming voltage and dissipating the energy as heat. The heat dissipated by a wire in watts is equal to the current in amperes flowing through the wire, times the voltage being burned by the wire. A "hot" wire is a wire overburdened by current flow (insufficient gauge). For example, a 250-amp starter current flowing through a battery cable presenting a 2.5% voltage loss (0.32-volt drop) is dissipating 80 watts of heat across the length of the cable. Decreasing the cable size to create a voltage loss of 15% (1.9-volts drop) increases heat dissipation to ~473 watts! At some point, the insulation will melt!
Additionally, for each linear foot of distance a cable must run between the power source and its load, the actual current path is roughly double, because of the added length of the return or ground path. In cases where the ground path is actually wire, incorporation of the ground wire length is important to overall cable selection. In the usual case of a steel frame, although steel presents greater electrical resistance than copper or aluminum, the cross-section of the frame in any given application is generally adequate not to be of significant concern beyond corrosion-free connections.
With regard to momentary high-current circuits, such as engine crank circuits, it’s common industry practice to specify cable gauge based on between 5% to as much as 15% voltage loss during periods of maximum current transfer (i.e., while cranking). Meaning, a cumulative loss of as much as 0.9 volt across the combined length of starter cabling (positive cable plus ground cable) is not unheard of while cranking. In fact, diagnostic service criteria specifies a maximum of 0.5 volt loss across each starter cable while cranking, or ~4% voltage loss per cable, for ~8% combined total loss.
----------------------------------------
I consider greater than ~8% voltage loss across high-current cabling to be excessive. But then I generally spec battery cables and wiring for no more than ~2.5% voltage loss, based on 115% current flow (I incorporate a 15% margin). For a starter that draws 240 amps on a hot day with a combined cable length of about 7 feet, that equates to 2-gauge copper cable. Roughly doubling the combined run from the bed (~16 feet total), equates to 3/0-gauge copper cable. Calculations are based on the formula:
cmilmin=40(Cmaxd)
...where cmilmin is the minimum acceptable cross-sectional area of the wire in circular mils (one circular mil is equivalent to the cross-sectional area of a cylindrical wire possessing a diameter of 0.001 inch or 1 mil), 40 is a constant with the units ohm-cmil per volt-foot, Cmax is the actual or anticipated maximum current draw in amperes, and d is the total cumulative length of wire run in feet, including the return run. The calculated cross-sectional area is then looked up in a table relating wire cross-sectional area in cmils to AWG wire gauge.
The quick reference chart is easier!
I know... I know...
I was taught to assemble battery cables the same way you were, and it worked most of the time. I agree, 60/40 tin/lead rosin flux solder is a good choice. Silver solder also works, but requires more heat.
This is what I learned through observation and experience: (a) Preheating large-gauge copper cable with a torch (which is pretty much necessary to supply adequate heat, since the cable itself acts like a sink) causes the copper to anneal and oxidize. Annealing weakens the metal; oxidation retards solder penetration and adhesion in spite of the flux. (b) If you don't assemble a lot of cables to maintain decent technique, you run the risk of a cold solder joint because you're stabbing a too cool cable into a pool of molten metal, quenching the pool before the solder wicks up between the strands, resulting in a poor connection. (c) Having repaired & wired many commercial vehicles, I observed high draw and connection problems that actually melted solder, allowing cables to separate from soldered only terminals. Not good when its the starter end of a four battery set! :o
I discovered that in order to lessen the potential for subsequent service problems, when I securely crimped terminal ends to cables to create a strong mechanical connection before soldering, the crimp would support and retain the cable secure in its terminal, so that the solder merely had to penetrate and seal the electrical connection from unwanted penetration by air, moisture and contaminants, to maintain good electrical conduction over the life of the cable. Heating the outside of a crimped terminal shields the copper strands from direct flame while quickly and evenly transferring heat to the cable. When the terminal and cable are sufficiently heated, flux core solder melts and wicks into the connection, filling the voids and sealing it. When connecting large parts together, it's best not to rely on solder alone to retain those parts, especially if they might be subjected to subsequent high heat cycles.
But, ultimately, as you posted, whether you crimp, crimp and solder, or just solder is individual choice. All we can do is provide the data.
-------------------------------
Can't say that I ever had solder tears on my face, that must of hurt! ???