How to Install an alternator cut-off ("kill") switch on a 1998 Honda Civic

[OneOfFew]

Originally Posted by California98Civic

Here is the wiring diagram for the charging system, taken from the Honda Service Manual for the 1996-1998 Civics. You'll see that the large gauge white wire running from the alt to the battery also has a connection back to the regulator, seemingly still inside the alternator. I think that's the culprit, here. What do you think?

I think I have proven your hypothesis true. I created a circuit that disconnects all 4 wires on the connector and is STILL generates power. The voltage is unstable but it creates power, up to 25 amps of power at 12-13v. I used a ~70amp load tester with the alternator circuit disconnected after engine start. I was very surprised. That connection to power internal to the alternator is keeping the regulator running, as you thought.

I'll post some details in a few.

[OneOfFew]

There are 2 series of tests below. The first adds a small load to the alternator circuit attempting to "pull down" the alternators self sustaining power. The second is a voltage/load test to see if the alternator can produce amps when the connector is unplugged after the engine was started.

This is all in attempt to create a actively switched alternator circuit that can be enabled and disabled while driving.

I added text to some images to best explain what is happening.

Test series 1
Details: This test is based off of unplugging the alternator power wire, the blk/yel wire on the alternator connector. This would simulate a switch turned on/off on the blk/yel wire after the car had been started and charging under normal conditions.





*Image below was corrected with proper image for caption*




Test series 2
Details: This test simulates unplugging the connector on the back of the alternator while the engine is running and after it had been charging normally. It disconnects all 4 wires on the back side of the alternator only leaving the alt. output wire intact.


*Image caption updated, video of condition added.*
---->Video<----




Next tests will be done with a 70amp relay on the alternator output wire in conjunction with the plug disconnect.

[turbothrush]

Great work. Thanks for posting. Can't wait to see what you come up with

[California98Civic]

This is great. To be clear, the first set of photos shows that the resistor on the alternator circuit does in fact take the juice out of the field windings. In the second set of pics, you did not use the resistor on the alt circuit and the alt continued powering but at a reduced rate, correct?

Originally Posted by OneOfFew

I think I have proven your hypothesis true. I created a circuit that disconnects all 4 wires on the connector and is STILL generates power.

Also, this is after the car is running, no? If you start the car with the alt unplugged the field windings do not fire-up, no?

[OneOfFew]

Originally Posted by California98Civic

This is great. To be clear, the first set of photos shows that the resistor on the alternator circuit does in fact take the juice out of the field windings. In the second set of pics, you did not use the resistor on the alt circuit and the alt continued powering but at a reduced rate, correct?

Yes, you are correct for the first tests. Only a resistor was added to the one disconnected power wire. The second test involved the 4wire disconnect circuit. It would basically unplug the 4 pin connector without me having to physically unplug it while the car was running. I included more detail on the tests and even a video of the effect of the alt. 4pin connector unplugged after the engine is started.

I did not actually to see if it still generated power/amps if demanded like I did in the second test, I only tested the voltage. I can do a test with the load bank while the alt blk/yel wire is loaded with the 6.6ohm resistor Tuesday. The equipment i used for these tests is at my place of employment being the O-scope and battery load tester.

Also, this is after the car is running, no? If you start the car with the alt unplugged the field windings do not fire-up, no?

All of my tests were disconnects after the engine was running. I did a quick retest of Battery voltage before startup and Alt. 4pin connector unplug then startup. It shows the alternator will not produce power until it is given power. After that it cannot be turned off by unplugging the 4pin connector as shown in the other tests.

[OneOfFew]

I can finally say I have a 95% solution and I think I'm going to settle with that.

I'm going to short-story this since it took me HOURS of diag. to figure it out.

The answer is...

Step 1: Follow California98Civic's guide on page 1.
Step 2: Disconnect the white/green wire on the 4pin alternator connector
Step 3: Ground the alternator side of the white/green wire through a low ohm resistor. I used 6.6ohm.

To re-energize the alternator
Step 1: remove grounding resistor and reconnect white/green wire to harness
Step 2: reconnect black/yellow wire to factor harness

Both of these can be easily done simultaneously with a dpdt switch.

This method will reduce the alternator voltage to battery level and keep it from putting out significant amps at the lower voltage.

I did a load test with this setup and there is no audible loading of the engine from the alternator and no real current coming from the alternator to the battery.

From my testing today it appears that the ECM grounds this wire when the car is under acceleration. I'm not sure why since it can still put out A LOT of amps at this lower voltage when load tested. The trick is disconnecting the power wire to the regulator at the same time, the Page 1 instructions California98Civic walks you through. The alternator is in what appears to be a "standby mode" for lack of a better word. The alternator is still communicating to the ECM but the ECM can't communicate back. You're giving the alternator a false signal. I'm not sure if this alone has a negative effect since I don't have the instrumentation to test it, a ScanGauge. The voltage appears to be very stable under this condition.

I don't know if this effects the alternator negatively over any length of time.

****Try mod at your own risk.****

Description for videos
Scope Traces
Yellow: Voltage regulator power, blk/yel wire.
Green: Battery light, whi/blu wire.
Red: ECM control, whi/gre wire.
Blue: Alt signal, whi/red wire.

Alternator signals while operating normally with 70amp load.
Video

Alternator signals while accelerating from a stop. Notice the RED trace.
Video

Alternator with whi/gre wire resistored to ground. (just as Honda does in acceleration video) Listen for the engine load down when a 70 amp draw is put on the battery. You can see charging amps on meter to the right. AmpMeter scale is 000.0amps.
Video

Same test as above and with blk/yel wire on alt disconnected. 70amp draw on battery. No audible loading of the engine, no significant measured load on the ampmeter.
Video

Tomorrow I'm going to wire this up in a driveable way. I'm going to be using a petal switched circuit that has a build in hold delay (prevents rapid on/off of circuit). I'll upload the schematic tomorrow for those that are interested in this method of control. It cost me about $15 worth of parts from radioshack.


insignificant note
I also wanted to mention why I did not test the alternator with the output wire (the big wire) disconnected as mentioned in previous posts. I thought it would not be best for the alternator to disconnect this wire. As most of you probably know when not connected to a battery alternators make very crude 14vdc. The battery dampens the spikes so you get smooth 14v. Without the battery on the circuit the alternator would have been powering itself (the regulator) with this rough power. I did not think it was a reliable long-term solution.

[California98Civic]

This is a remarkable set of tests. Thanks.

I'm sure the switch will work, having watched how you have tested and developed it. What a great example of careful modding this is! When you have reached your conclusion, I'll edit the OP to link down to your final alternate build. Or you might want to start a separate thread that walks a modder through a build and links to these test of yours.

Originally Posted by OneOfFew

I don't know if this effects the alternator negatively over any length of time.

In my still limited experience, the alt is fine after several months. But I don't switch on/off often at all... I run with it off all day unless I am confronted with a long drive or night driving, then I will turn it on more often or continuously.

Originally Posted by OneOfFew

insignificant note
I also wanted to mention why I did not test the alternator with the output wire (the big wire) disconnected as mentioned in previous posts. I thought it would not be best for the alternator to disconnect this wire. As most of you probably know when not connected to a battery alternators make very crude 14vdc. The battery dampens the spikes so you get smooth 14v. Without the battery on the circuit the alternator would have been powering itself (the regulator) with this rough power. I did not think it was a reliable long-term solution.

I think a guy on here a couple years ago fried hIs Metro alternator this way. Good thing to avoid!

[brucepick]

Nice work!

A couple notes.

I thought I was pretty clever with my extra junkyard female alternator connector, but I was never able to physically get my fingers in there to connect it to the car's wiring. I think it could be done though.

After finding ~14V from the alternator, after disconnecting the 4 pin connector at the alt, I gave up on trying to figure out what I think you finally did achieve - a way to switch the alt on and off while driving, by managing two of those four wires.

Kudos to you, for keeping after it and for giving us a nice thorough writeup of what you found with photos and even videos. Great job!

What I did & why:
I based my alt disconnect on disconnecting the alt's output cable. I ran a 4 ga. cable from the alt to the battery + terminal, where it is connected via a hand-screw type battery disconnect switch. They are built for the smaller diam of the ground terminal so it needed some filing to fit the + post. Obviously you can't disable it while moving but my trunk-mounted lithium battery pack is set up to run 1-3 hours on a full charge. Only rarely have I had to switch the alt and the standard (front) lead acid battery back into the circuit to complete a drive.

Since my main battery is a lithium pack it won't optimally be charged by the alternator anyway. If I need to drive more than what the pack can handle, I need to pull over and change three switches, including the alternator disconnect that's under the hood. That done, the lead acid battery and the alternator will work as originally designed. Normally my lead battery only powers the starter and is recharged at home every few days, so it's not deeply drained.

Anyway, REAL NICE job on figuring out how to disable the alt from the driver's seat. Should be a real help to anyone else doing this kind of mod on a Gen 6 Civic.

[meanjoe75fan]

Having driven a vehicle with worn alt brushes, and observing (via lights, fan speed,etx) system voltagw fluctuation with intermittant alt ops...

...what about a system that mechanically lifts the brushes off the slip rings?

[California98Civic]

Originally Posted by meanjoe75fan

...what about a system that mechanically lifts the brushes off the slip rings?

I missed this question, sorry, but I also don't know the answer. There is a great alternator mod thread on here with some of the best experienced technical folks on it... try posting the idea there?

AN UPDATE: A note: lately I have noticed less than optimal performance from my three month old Odyssey deep cycle battery. The Odyssey people are helping me with charging issues. I'll report back when I learn more, but it seems my 6amp charger and charging technique may have been inadequate and that the situation is entirely reparable. MORE LATER...