Actually 10% (or even more) is quite plausible in small cars, especially going down the highway. Remember, the smaller (and more fuel efficient) the car is to begin with, the fewer HP you really need to use to maintain highway speed. As a result, a one or two HP drag from the alternator, could easily translate into 10% (or more) of the total HP that has to be generated by the engine (to both power the alternator and keep the car moving down the road).

BTW: No, I haven't fully disconnected my alternator either (due to the issues involved with doing so). However, I have taken steps to lower my electrical usage in my CRX (energy efficient LED car lights, don't use accessories when not needed, adjusted my electric radiator fan to run less often, etc) , and such steps have produced very noticeable (if not huge) gains in my FE (in my case, I seem to be getting a 5%-10% gain in my FE as a result of my electrical savings efforts). And since disconnecting the alternator completely is (at least from a theoretical standpoint) equivalent to bringing your electrical usage drag (on the engine) to zero (vs simply lowering it, as I have done with my electrical savings techniques), it stands to reason that fully disconnecting the alternator should show even more gains than I've achieved with my electrical power saving approaches.

So this approach is a very sound way to improve FE some. OTOH how much it will help, will generally depend upon the car (some are already better in this area than others). And even in the cars where such an approach helps a noticeable amount, it's still a question as to if the cost and effort of doing this mod is justified by the fuel savings. And the later reason is why I don't (at least not yet) fully disconnect my alternator, but instead just take steps to lower electrical usage in the car (which gives me part of the benefit of a full alternator disconnect, without some of the hassles associated with a full alternator disconnect).

of everything ive studied and read 10% mpg gain is a very realistic number. i think it maybe more or less depending on the car but thats an average. i think older carbed cars would respond much better to this since you can run the battery for longer with no alt but even then modern cars are pretty energy efficient. of course this would probably a mostly summer thing with me since i would be driving at night in winter so i better get cracking. probably get everything sorted and in in the next few weeks.
my new plan after doing a bit more research involves running the field wire to an illuminated toggle switch and to my brake light. im thinking itll act abit like regenerative braking. you want the car to slow down if your on the brakes anyway so the alt should add a bit of resistance to the motor. not much gain really but thats the best time i can think of power on the alt for as little mpg lose as possible. the rest of the system will consist of a voltage meter in dash for alt output and one for battery state. seems like 12v is as low as you want to take a car battery before you start killing it.
i know from my recent endeavors of getting rid of the a/c and power steering the net gain was nearly 3mpg for each on. so the gain was around 6mpg for the two but im still working on more backup data to prove it but i know it does, just need the numbers to dispute the claims of others. it would stand to reason that the alt would be the same. if you spin them all with a load the alt is one of the hardest to turn. the power steering is very easy until you turn the wheel and the a/c is a bit hard but the clutch cycles it on and off.
i would love for someone to do this ahead of me. i dont doubt the gain and i dont doubt the way to do it, i just cant start changing things now. ideally id like to see a bunch of data from others on various cars to get an average of findings. im also gonna start looking into reducing my electrical load in the mean time. ive already ditched my amp and subs so i guess im looking at bulbs. ill get started on that but again i cant put them in until after my currecnt trials and after my trials on the alt disconnect. by then i should be able to have an average tiem i can run the car as is on battery alone then compare to low load electric car. im figuring around 20 min of highway driving in the day before i have to reactivate the battery, probably around 15 min of light city driving due to brake light usage. ill also see how the small solar panel pans out. im not figuring on it doing much but it free since i already have it.

I like the idea of hooking up the brake light to the field control. Even better would be if you could make it switch on during DFCO, but it could be difficult to detect DFCO.

BTW: As someone who has converted most of his car lights (in the CRX, lessor conversion has been done to the other family cars) to LED modules (except for the headlights, as there aren't street legal LED headlight replacements yet), IMHO one of the the best "bang for the buck" areas (i.e. most power saving benefit, for the least cost/hassle of the LED modules) is to start with the small wedge based lights that most cars have (the even smaller "74 bulbs" that many dashes use for dash lights are even a bigger "bang for the buck", power savings wise, but "pulling the dash" to get to those little suckers is often more of a PITA than many want to tackle for their first LED replacement task). And after trying various small wedge modules, I'm finding that the $5 modules from https://www.superbrightleds.com, that look like a small circuit board, seem to make an excellent (and very bright) replacements for a wide range (and even size and wattage) of small wedge based car bulb model numbers. And if you do follow my advice on this style LED module, (as with all LEDs) be sure to match the module color to the car lens color (i.e. red module behind a red marker lens, amber module behind a front amber marker light lens, and only use the (warm) white module for clear lenses such as the license plate lights use for example). Picture below:

https://www.superbrightleds.com/produ.../wled-whp6.jpg

NOTE: As someone who has converted most of my car lights to LEDs, I naturally know that buying some of the big ($24+) Luxeon LED modules (for the main brake/turn/tail lights) will ultimately be needed if you want the full power savings of getting all the lights converted (while still maintaining better than OEM brightness in those "lights"). And before you do that, you will likely have to convert your existing flasher to an electronic one, or you will likely have a "too fast" flash rate (as most OEM flashers will flash the bulbs fast, due to them thinking a bulb is burned out, if/when you have the lower current draw of LEDs).

However, you can easily do a project over time, starting with "the low hanging fruit". And (as I've already mentioned) I've personally found those $5/each (very bright, full 360 degree) 6-LED modules to be some of the best deals out there in very bright LED "bulbs" that will replace virtually any small (12v) "wedge based" bulb you might otherwise be using in the car (which with many cars includes not only "marker lights", but also things like trunk and glove box lights, and even things like the bulbs in the "high brake light"). So IMHO they are a good module to get you started (just follow my advice and don't put a white module behind a colored car lens, even if/when the original incandescent bulb was "white" in that situation, as you will get much better light output if/when you match the LED module color to the color of the car lens that LED module is behind).

But of course, all cars are different, and so while I've had good luck with that style LED module (in fact, I'm about to order a few more for the family cars that I haven't yet converted all the small "wedge bulbs" over to LEDs), YMMV. i.e. I can't promise that those modules will work as well for you as they have for me. All I can do is relate my (positive) personal experience with them when used in multiple family cars.

Edit/addition:
If you still have a "high brake light" that uses real bulbs (the high brake light was often the first light car makers used LEDs for, but some older cars still use incandescent bulbs for this function), don't forget to check which style "bulbs" it uses, as many such "brake lights" actually use (somewhat bigger and slightly higher wattage) bulbs that still use a small wedge base. And if your car has such "bulbs" for its "high brake light", those $5 modules (obviously you'll want the red version of that module for brake lights) IMHO make excellent replacements for those "high brake light" bulbs. Because, not only are those modules (as is pretty much true of all LEDs) able to turn on a fraction of a second quicker than a stock incandescent bulb (which for a brake light can actually be a minor "safety feature", as it gives the person behind you just a little more reaction time), but at least in my experience (and I've now got that module in the high brake lights of three different family vehicles) the red version of that module (when used in a high brake light containing the normal red lens) usually has an effective brightness GREATER THAN the stock incandescent bulb it was replacing! So you save power vs the stock bulbs, get slightly faster turn on of the lights, and also get a replacement that is brighter (and easier to see) as well. Win-win-win in my book.

im not sure that can be done. i think the brake light thing is as close as you can get. later on ill probably rig a system to do everything automatically. once the battery has a decent preset charge itll shut it self off until the battery hits the preset low volt or if the brakes are applied . i cant see a better way of doing it. im not sure it can be done with dfco, if my car even has it. although if dfco cuts power to the pump completely a simple light could tell you its deactivated then a signal would go to a controller that switches on the alt.

its gonna take me a while to figure out that led light thing. i wouldnt even know were to begin when ordering. ill have to go out and do a light census.

I've been holding my brake pedal at about 1/4" while DFCO'ing to traffic lights for fear that I might get rear ended otherwise since I live in a pretty urban area. I also remember seeing a module in the JC Whitney catalog many years ago that would hold your brake lights on for up to 5 or 6 seconds after the brake was released. I may consider tracking one of these down. I think what I may do is take the module and wire it up to a momentary on pushbutton switch on the bottom of the dash. Pushing the button would light the brake lights for 5 or 6 seconds, then they would go out.

-Jay

As mentioned in my previous post, I personally would suggest you start with a few smaller wedge based car bulbs, to get your feet wet.

As to the car bulb census, you will most certainly have to walk around the car to count bulbs. But the following web site actually does a pretty good job at identifying the type of "bulbs" that most cars call for. Now, naturally (since that site if trying to sell you their incandescent bulbs) you won't buy your "bulbs" from them, but they are still a nice resource to tell you which bulb model numbers you are trying to match (but they aren't such a good resource for giving you an exact bulb count, as they won't for example tell you how many "marker lights" you have, just which "bulb" each "marker light" uses):
https://www.sylvania.com/ConsumerProd...lacementGuide/

And once you have the general OEM bulb "model numbers" of your bulbs (by a combination of that web site and a physical count of the number of "bulbs"), you can use this cross-reference to get an idea of which types of "bulbs" can be replaced with which LED modules (even though they are "technically" different wattage/color/etc OEM bulb models):
https://www.superbrightleds.com/bulb_cross.htm

NOTE: The "WLED series" (i.e. small wedge based bulbs) are pretty much ALL interchangeable (except for LED color, which you will always want to match to the color of the car lens the "bulb" is behind), no matter what the original OEM bulb model number is! Yes, there are technical differences between the different (small wedge based) bulb models, but those differences are things like how many watts/lumens the bulb is, what color it is, etc. And most of that is not material info when converting over to LED modules! Which is why I'm now thinking the best approach is simply to use those $5/each 6-LED modules (again, match the module color to the car lens color) I mentioned in my previous post, for pretty much all 12v small wedge based "bulb" needs. The reason I'm now thinking this, is that those modules are pretty much the brightest (and most wide-angle) of any small wedge based bulb replacement I've found (and at $5/each, they are still not that expensive of a replacement LED module).

NOTE: My experience is that you do get what you pay for with LED modules. Try to save a few $$$ on the module cost, and your "cheaper module" will likely be less bright, more narrow angle field of view, or both (and as a result, you may eventually replace it anyway). So IMHO the extra cost for getting the brighter (and wider viewing angle) modules is often well worth it in the long run.

Your owners manual should tell you what bulb numbers are used in your vehicle. I know mine always have, but maybe that's a GM thing... I've never owned a non-GM vehicle.

Most cars will turn on their brake lights before you have done enough pedal pressure to actually start braking (and with many cars you can even adjust the sensitivity of the brake light on switch connected to the brake pedal). And if you experiment somewhere you can see this light (or temporarily wire up an extension bulb to the brake light circuit, so that you can see the "brake lights" while driving), you can quickly learn how light of a touch "on the brake" will engage the brake light in your car.

And FWIW another thought I had (which I haven't yet gotten the energy to do to my CRX, but is still on my "to do list"), was to simply wire up a small (red) dash LED to the brake light circuit, so that I can always (easily) see (while driving) if/when the brake lights are on. Yes, I know if I'm braking a lot that my brake lights will be on, but when I simply want to press them just hard enough to engage the lights, the exact pedal pressure is not so clear...

Sounds clever. However, I could see that also getting you "in trouble". For example, you are starting to move forward from an intersection, and yet your "brake lights" are still on for a few seconds (confusing the driver behind you, and possibly even causing undo attention from a cop wondering why you are "riding your brakes").

So IMHO a better choice would be simply another (only on while being pressed) button, that is positioned somewhere that is easy to press when you want to, which only turns on your brake lights (while you press it) and otherwise does nothing. Ideally, you would want this mounted somewhere that it is easy/trivial to press when desired, but otherwise not in your way. For example, maybe you could mount it somewhere on the steering wheel itself (or maybe on your stick-shift if you have a manual)? Or perhaps a floor mounted button would work good, as you don't exactly need your "brake foot" (and therefore it should be "free" to engage the brake lights) if all you are doing at the time is DFCO (and you just want to alert the drivers behind you to the fact that you are slowing down).

In any event, those are some thoughts I had, and you are free to take this advice or leave it as you see fit. Either way it costs you the same amount (i.e. free). ;)