Help me kill my alternator - it is stubborn and won't die
 

For the past few days I have been attempting to do an alternator disable on my 2014 Mitsubishi Mirage. I'm going to do it electrically, not unbelting, or removing the alternator. I've done this on several vehicles, but this one is proving to be a bit of a challenge!

Here is the system description of how it works, and diagrams to go along with it for reference.

https://ecomodder.com/forum/attachme...1&d=1535561282

https://ecomodder.com/forum/attachme...1&d=1535561301



My first thought was to cut and put a switch on the FR (field regulator) wire. I did this and it did stop it from charging... kinda. As soon as the engine revved up too high, it would start charging (self exciting I am guessing).

So, version two. I figured that killing the S (sense) wire would do it. However, I couldn't be more wrong. Instead of defaulting to not charging at all, the alternator blasts on at 15V.

I am thinking that my next step will be to install a dc-dc converter that takes 12V and converts it to 15V. I will send this 15V signal to the sense wire when I want the alternator to turn off, and let it see normal battery voltage when I want it turned on. The service manual specifically says (see above):

So, I am hoping that 15V will do it. I have already ordered a tiny $3 dc-dc converter to do this job.

However, if you see a simpler way to do this, I am all ears!

Just thinking now... if I cut the G wire, the regulator won't have any ground reference unless its grounded to the alternator casing (which is quite possible). So, I wonder if cutting that wire would do it?

My Acura has a 4-wire harness going to the alternator. I don't remember what all those wires do now, but I do recall I had to disconnect all of them to kill the alt. Have you tried unplugging all of the control/sense wires to the alt?

I bought a 4 circuit switch and used wire harness from Ebay (so I don't have to slice my original harness) for the purpose of wiring in the dash switch, but never got around to it.

Why not make an extra, 14.4V battery pack to boost system voltage to where the alt doesn’t charge? Like the guy with the virtual alt delete? Would save you the cutting and splicing...

Thanks Redpoint. Mine is also a 4 wire as you can see. I may try the ground next before my dc-dc converters get here.

19bonestock88, the 14.4V pack is a great idea. It would definitely work nicer than just the alt delete. However, I'm really not looking to spend on a pricey lithium battery at this time.

This is what I do with my Civic. I had the problems you had. I had to disable all wires into the alt or the unit did not entirely quit. One problem remained, however, once I turn the unit on from the dash, the only way to turn it back off is by cutting the engine. I decided to live with it because of how often I do EOC, but a better way is certainly available if I figure out how to draw off the extra power keeping the field windings charged.

A 4 pole double throw switch would probably be the easiest option, one switch to disconnect all 4 wires.

Great info. Thanks! :thumbup:

Yeah, its sounding more and more like the dc-dc converter route is the way I'll go then. I can just fake the signal to the voltage regulator and that should kill the power generation pretty effectively. I'd rather not cut a bunch of wires if I don't have to.

Another way is to get an extra OEM connector and plug it into the alt on one and and your car's wire harness's connector on the other. That way you do not cut wires on your car's harness.

I like the DC/DC converter, tho. Smart, if slightly more costly.

I hope it works out. I just got the dc-dc converter in today. I should be able to get it installed this weekend. Its just a very tiny chip based converter, it can only handle 1W of load. But, it should work great for this signal level stuff. I tested the sense wire on my alternator and it pulled 1mA.

I am also hoping to add a few transistors in as switches so I can enable the alternator on demand for some regen braking capability.

However, first steps first. Today I just unplugged the connector on the alternator. Yep, that worked haha.

You could put resistors in so that the double throw position grounds the 4 circuits through the resistors. I'm not sure that's what is needed, but there is no reason for the switch to be double throw otherwise.

That's the route I will go with my Acura. I bought an extra harness for a couple bucks on Ebay.

Alright, here is the proposed setup. Its pretty simple, the dc-dc converter, and a SPDT switch to flip flop signals from actual (normal alternator operation) to false (kill alternator output).

https://ecomodder.com/forum/attachme...1&d=1535939253

Looks smart, help the inexperienced understand. I assume the red represents the entire set of maybe four signal, power, and grpund wires going into the alt? Also, I found my thread on the switch for my car. Very detailed, and in some pleaces, like on page six, people show up with their own experiements and data:
https://ecomodder.com/forum/showthre...tml#post286419

The red wire is only one wire. It is the "S" wire or sense wire. The service manual says that if the sense wire goes above approximately 14.4V, the current to the field coils is shut off. So, I am using the dc-dc converter to send 15V to the sense wire.

Alright, the dc-dc converter seems to work great. I just installed and tested the circuit out. I made a quick little video that explains it fairly well I think.

https://www.youtube.com/watch?v=zs3fY1rx5Hs

You gonna want that switch inside the cabin, tho! Haha.

Haha, quite true. This is phase 1. I just want to get a quick setup to plug into the fuse holder and it'll be operational. With my Metro, I just disconnected the big battery wire. I installed a relay to be able to hook it back up, but I never wired power to the relay to turn it on. It was never an issue.

Phase 2 brings the wires into the cabin and I want to tie into the brake light to reactivate the alternator for a regen braking setup.

Phase 3 is modifying the sense signal even further to crank up the alternator output to get a bit more regen out of things.

Cool. Add a phase four with capacitors to absorb the regen energy better than the lead acid battery can??

Perhaps, if I ever even get to phase 3! Plans are easier (and IMO funner) to make than doing the actual work haha. Just ask Darin all the hair brained stuff I've told him over the years.

I've been wanting to do the regen thing for a long time though, so I'll probably at least get that far. However, the returns greatly diminish after phase 1 of the project. Its more of a 'neat factor' kind of thing after that.

Funny enough, I've done the exact same thing to my car, but never noticed any real mpg difference from switching the alt.

I've got a switch in the cabin so I can manually switch the alt while I'm driving, so I'd switch off while accelerating or holding speed then back on while braking.

If I switch the alt off while idling the revs alter and load reduces, but if I switch while cruising at a steady speed displayed mpg drops very slightly (according to torque pro).

With the alt off my voltage drops pretty quickly, which made me wonder if the dc-dc converter was adding to the drain on the battery and effectively negating any gain, but I've got two dash cams and glow plugs burning away too... at this point I got bored and moved onto some other project...

:turtle:

It really only works well if you run off a deep cycle and basically never turn the alternator back on (except for regen situations). If you turn the alternator back on in normal driving it is going to turn on at pretty much full blast and that will hurt your fuel economy.

Very elegant solution indeed.

What kind of mpg gain will you get from this? I am assuming here that it will be more significant than your older car since the 2014 is more electronic and has a slightly increased power consumption.

Liters / 100km or hour (or gal/mile) units would be much appreciated.

I was intending on continuing down this path as well...I picked up a cheap charge controller from ebay for a couple of bucks...it will switch on a relay when the voltage gets too low, and switch it back off once it reaches another voltage. It's adjustable. The point of it is if you aren't watching closely and your battery gets drained, it will kick the alt in automatically for you and save you from needing a jump start.

You could modify the system just a little bit and feed it 9V or so so you charge the batteries faster.

The rough calculations I did show that disabling the alternator reduces engine load by ~8%. I expect a 5-10% gain in fuel economy which is what the ecomodder wiki page shows to be historically true for most alternator delete mods.

So in the range of 0.3 l/100km or 0.3l/hour

%10 on a car that gets 10mpg would be impressive. :)

Disaster struck last night. My dc-dc converter got toasted I am pretty sure. The input voltage range for it is only 10.8-13.2V. Well, I think my switch (which actuates with almost no physical input) got bumped or jostled and the alternator turned on bumping the voltage up to 14.4V. Then, it got turned back off which exposed the dc-dc converter to more than 13.2V as the battery voltage slowly fell. And, while it can take a 100ms spike at 18V (which I had hoped would be enough), it stopped working. So, I am thinking that I have two choices:

1) Create a protection circuit so the dc-dc converter never sees above 13.2V.

2) Use a different dc-dc converter.

I'm partial to #1 because I bought two dc-dc converters just in case I fried one. However, I'm thinking #2 may be a better option. A voltage divider will not work for the spread of voltages I'm looking at. I did the calculations. I'm not aware of any other simple circuit that can keep the voltage to the dc-dc converter in the necessary range. I'm not looking to make this complex.

On the upside, this would have happened anyway once I figured out the regen side of things because that would have been intentionally turning the alternator on and off. So, the dc-dc converter has to be able to handle this scenario if I am going to go through with that phase of the modification.

Add diodes to its input.

Not a bad idea. Lets do the math though. The alternator's max output is about 15V. We have to drop 15V down to 13.2V. That would take three .7V drop diodes which would create a 2.1V drop. But, lets say I'm now running on the battery. The lowest my battery voltage to the dc-dc converter can be is 10.8V. With the diodes in place, that means my battery voltage can't drop below 12.9V. If it does, the dc-dc converter won't work anymore. 12.9V is higher than the resting voltage of a 12V lead acid battery, let alone a loaded one. So that unfortunately won't work.

Then get a voltage converter to feed your voltage converter.

My current thought that would probably be the most robust is to use a linear voltage regulator like a 7805. These are extremely common chips. The one I linked to can handle up to 35V. I would need to go to one that puts out 5V though due to how they work. That means I would also need to get a 5V to 15V dc-dc converter instead of my 12V to 15V converter. However, this would pretty much ensure nothing would fry. That being said, there is one of these built into an arduino which is the programmable micro controller that I'll be using for phase two anyways. So, I wouldn't even need one if I got the new dc-dc converter.

You can convert a 5v linear regulator to an 11v linear regulator by voltage dividing it's reference between it's output and ground.

Regs are cheap, just get the one that fits the purpose.

Do you have any idea on how much current is drawn?

The arduino ones can handle up to 800mA if i remember correctly. (Which also has to feed the atmel microprocessor).

I did test the sense line current draw. It was 1mA.

For the phase 3 "E-brake"(electron brake) a higher voltage capacitor with enough capacity to hold 5-10HP of eletrons thrown at it when on the brakes .@13.7v × 120A= 1644W/2.2HP ( × the inefficiencies of the charging system ~50% ) ~5HP@13.7VDC with a 24v cap @120A you can increase the strength of your e-brake closer to that 10HP strength .

I live in "Grade Territory" (long steep mountain climbs) my prior thoughts on this E-brake involved a power dissipation system,dump load for when I run out of storage. I was thinking maybe a Jacob's Ladder big enough to dissipate 25-50HP. Call it a Jake Brake.

Phase 6, do the same with the AC. Blower on full, ac compressor activated.

I am now commuting (on a coffee break) and I do this manually on down hills where I have to decelerate. It works quite well. The blower motor is a fairly good load if you want to dump some electrons.

If anyone knows how to increase the size of the AC refrigerant please let me know.

A bit earlier this weekend I finished up the wiring for the charger. The 110V lines got blade connectors while the 12V lines got bullet connectors. I then stuffed the charger behind the center console.

I have a new 5 to 15V dc-dc converter on order for phase 2 of the project.

https://mirageforum.com/forum/attach...0&d=1536535046

https://mirageforum.com/forum/attach...1&d=1536535112

Whilst waiting for the new dc-dc converter to arrive. I should probably think up how I am going to do the regenerative braking thing.

Ideally, we really only want the alternator turned on during a certain set of conditions:

1) The alternator is spinning.
and
2) We are slowing down.

I haven't really found out a great way to do this yet. Initially I was thinking that I would just take a signal from the brake lights. When the brake light would go on, the alternator would turn on. Unfortunately, this really doesn't actually ensure either of the two conditions are met. It simply means the brake is on. I could be sitting at a stop light, or starting the car on a hill. In both of those cases I don't really want the alternator to turn on. However, it does make enabling the alternator a simple affair. You get brake light signal, you turn on. Pretty easy. The complicated route would include watching the speed sensor to ensure you're actually slowing down, and probably watching the tach signal to make sure the engine is running. I really don't feel like doing all of that though haha.

So, for now, I think I am just going to hook it up to the brake light and see how that works. I can always tweak it in the future. But, if it works, there is no reason to make it more complicated.

As always, if you have suggestions I'm all ears!