Civic alternator shut off, the easy way

[steffen707]

Well I've spent hours testing today and i've confirmed and refuted some theories with the Honda alternator.

My test setup:
96 civic hatch, with 96 civic alternator on a 92 d15z1 engine. I think the obd1 civic alternators operate the same way as the obd2a civics, but I can't be certain as I didn't test it.

I had a basically dead car battery (so dead that it wouldn't run the fan or light up any lights) and hooked that up to the battery cables. I then used a known good working battery to jump start the car for the beginning of each test. As soon as the car started I disconnected the jumper cables to isolate the good battery from the bad one.

The car was brought up to operating temp.

Note: When its up to temp when you start the car, the engine doesn't rev over 2000rpm. This is an important thing to remember, because when you test the voltage at the field winding with the car off there is .01volts, when you start the car its .05-.3 volts. When you rev it past 2000rpm it now reads 3-5volts and even if you let the car idle, it will stay at 3-5 volts.

There are 5 wires at the alternator. 1 white "B" charging wire, 1 yellow/black "IG" field winding exciter wire, 1 white/green "C" charging control signal wire from the ecu, 1 white/blue "L" charging warning lamp wire, 1 white/red "FR" charging rate wire from the alternator to the ecu.

I disconnected these wires directly from the 4pin plug at the alternator for guaranteed results.

Don't try pulling the alternator fuse because the manual, says it controls the alternator, vss, eld unit, evap purge control oxygen sensors and tcm. That's for a 96 civic anyhow. Its much better to disconnect the "IG" wire at the alternator.

TESTS:

1. First test was unplugging all 4 wires together. The car immediately died after disconnecting the jumper cables, but wait..... If you rev the engine past 2000rpm before disconnecting the jumper cables, the car would run very poorly. headlights would flash, the fan would pulse, it is a not drivable but actually charging at a super small amount.

2. Disconnect all 4 wires, but ground the "C" wire, The car immediately died like in 1, but if you revved it past 2000rpm before removal of the jumper cables the car idled, but was more rough than in 1. This is because grounding the "C" wire tells the alternator to charge at the lower rate.

3. It has been thought you can simply disconnect the "IG" wire and disable the alternator. This resulted in charging the battery and providing enough power for lights, fan stereo, ect. The car idled just like normal and you could drive the car just fine. This didn't work.

4. Another ecomodder California98Civic http://ecomodder.com/forum/showthrea...8-a-20513.html told me that he found out that if you disconnect the "IG" and the "L" wire this disables the alternator. He thought that the "L" circuit was providing enough power to energize the field windings. Well he was right. I tested this by starting the car with both "IG" and "L" disconnected and it operated just like test 1. I also tried starting the car with just "IG" disconnected, removed jumper cables, the car ran fine, then disconnected "L" on the fly, the car imediately starting running super rough like test 1.

5. I read that if you soft ground the "IG" wire via a light bulb this will disable the alternator. With just soft grounding "IG" the results were the same as test 3.

6. With disconnecting "L" and soft grounding "IG" i got better running characteristics and presumably better charging than in test 1. (I can't remember if it was like normal running, or just smoother test 1 running, sorry. This isn't a good option anyway.

People should either do 1, 2 or 4. 1 and 4 are almost the same, except with 4, the "C" wire is still connected to ecu. I feel however that 2 is the best.

So what did I end up doing? Well you can disconnect the "L" wire anywhere between the alternator and the ELD. So for ease of turning back to stock i simply took the charging bulb out of the dash and got the same results as if you disconnected it at the alternator.

I also used a dpdt switch in the UP SETTING to switch "C" to ground, and simultaneously switch off the "IG" wire. This is the same as test 2. Except that the bulb is out instead of the wire at the alternator for the "L" wire.

In the DOWN SETTING the dpdt switch will reconnect the "C" and "IG" wire to their normal connections, allowing full use of the alternator in normal configuration. The only thing is that you will never see a charging problem because you took the bulb out; however, if you use a voltmeter to monitor the battery as most would be doing, you will notice when something isn't right. We don't need no dumby lights.

This DPDT switch also has an OFF setting in the middle. In OFF mode, you get the same results as test 1 except that the "FR" wire is still connected. I didn't notice any differences in any of the tests with or without "FR" connected.

So how did the MPG compare?? I picked out a flat 35mph road and picked out 2 mailboxes 1 mile apart. Car was up to temp, car was already cruising at 40mph before I got to the first marker to eliminate variation with acceleration up to 40mph. Once I passed the marker I reset mpguino to calculate just the mile long MPG. I read what mpguino said at the mile marker and then slowed down after.

I calculated from A-B and B-A, I tested with high beams on, fan on full blast, A/C off, and rear defroster on. I did the alternator on tests with a fully charged battery first to eliminate any extra load from trying to charge up a low battery from the other 2 test types. This was a high load test, i didn't have time to also test no load, but plan to do that sometime this week.

Results A-B Full load:
Alternator On normal: 70.7, 68.8, 68.1 = 69.2 average
Alternator Test 2 setup: 73.4, 76.7, 73.6 = 74.5 average
Alternator all 5 wires disconnected: 74.9, 74.2, 74.5 = 74.5 average

Results B-A Full load:
Alternator On normal: 62.8, 61.5, 62.2 = 62.16 average
Alternator Test 2 setup: 68.7, 69.7, 69.8 = 69.4 average
Alternator all 5 wires disconnected: 69.2, 68.8, 69.0 = 69.0 average

Alternator test 2 setup and completely disconnected (except belt) resulted in almost identical results. I truly think that test 2 setup hardly provides much charge, and thus puts very little mechanical load on the engine.

In the A-B test i saw a 7.65% gain in mpg, and B-A was a 11.6% gain.

I'm going to post these results on the first post of the thread as well.

[steffen707]

These are my original results, but it appears that the alternator was still charging just in low mode.

***UPDATE*** It is thought that this method just puts the alternator into a 12.5volt charging mode. I will try to test this further and report back with results. Don't be convinced yet that this is fool proof.

This is a half how to, half my results from testing this mod thread.

I have a 96 civic hatch and swapped a d15z1 into it. After reading about people disconnecting their alternator I decided to give it a try. if you go to pdftown you can download the service manual for your honda. From this site Timely Topics -- Your Vigor's Charging System The diagram they use is almost if not identical to my civic's diagram in the manual. I learned the alternator charging control wire which is green/white for 96 civic) and sends reference voltage from the alternator to the ecu. If the ecu grounds this wire, the alternator stops charging. Also, with this method the charging light indicator on the gauge cluster still works. You don't have to get out of the car and use a huge 100amp switch or relay on the charging wire like some people have done. You also don't have to run wires through the firewall. The alternator charging control wire for my car goes straight from the alternator to the ecu. i tested this with an ohm-meter. I simply cut the wire about a foot from the ecu and ran extensions from both of these ends to my switch mounted next to the steering wheel.

I used a dpdt switch from radio shack to connect the alternator charging control wire to the ecu and also to ground (they didn't have a spdt switch or I could have just used one of those). Now I can switch it to normal operation or "off via grounding" on the fly. (See http://en.wikipedia.org/wiki/Switch for info if you don't understand what dpdt/spdt switches are)

The switch works and the voltage drops right away. I’ll post pics in a day or so.

By the way if you use a similar switch, the middle part is “off” and nothing is connected. When you do this the alternator is still charging because it doesn’t see ground on that wire; however, if you read that whole page on the acuravigorclub website if says the ecu uses the ELD and the FR signal from the alternator along with other engine sensors to control the idle. Well when I set it to off it was still at 14.7volts at the battery; however, the idle speed dropped like 200 rpm. But when the wire was grounded it was 12.7volts but the idle was the same as if the wire was still connected to the the ecu. ODD....

Here are my results of a ½mile test with no electronics on except radio, 40mph steady speed, 54degrees outside, car weighs 2460 with driver and full tank, used mpguino and got car up to test speed and reset the mpguino trip odometer at the same sign and checked at the same sign.

73.8mpg - alternator on
74.8mpg - alternator off - 1.355% better

Same tests were run but this time with headlights on, A/C on full, rear defrost on.

56.1mpg - alternator on
63.45mpg - alternator off - 13.101% better

The first test was interesting because the difference in mpg was so small that a variation on my driving could have caused the difference, also I should have had my multi-meter connected to the battery so i could see if the alternator was even charging. I will probably re-test this sometime soon.

The 2nd test appears to corroborate the results of other people. Though I understand this was only a 1/2mile test it was more of a proof of concept for me and my car.

Next i tried some highway testing. Same limited electrical load and full electrical loads as before, 30minutes after first 2 tests. Car was brought up to 65mph on the speedometer (62.5 on mpguino) and i drove a 2.5mile course from A-B and then B-A and recorded results.

A trips (more downhill than B trips on same highway stretch)
56.6mpg - 62.7mph average - no load, alternator on
57.21mpg - 62.8mph average - no load, alternator off - 1.077% better
48.2mpg - 62.66mph average - with load, alternator on
50.14mpg - 62.90mph average - with load, alternator off - 4.024% better

B trips (lots of uphill, and my car is heavier and less aerodynamic than the civic VX from which the motor came from and thus it drops out of lean burn more easily than I would like, especially on hilly terrain at speed)
38.8mpg - 62.4mph average - no load, alternator on
39.7mpg - 62.5mph average - no load, alternator off - 2.319% better
38.3mpg - 62.5mph average - with load, alternator on
39.7mpg - 62.4mph average - with load, alternator off - 3.655% better

Here are some of my thoughts on the results.
1. With alternator off there was always a gain in mpg. I think this would carry over to full tank results as well.

2. with my vehicle the short run highway testing resulted in only a best of
4% better. I would like to do some long range highway tests to see if there is more of a gain or not.

3. The short run testing with load results was amazing, and with test results like this I ‘m curious if there was a fluke. More short run testing needs to be conducted to verify this. My car gets awesome gas mileage at 40-50mph, and I should re-conduct these tests without A/C, but instead just max blower on, as the A/C sucks a lot of power from the motor.

4. Though with no load my results at low speed were similar, I have to feel that over time and distance the battery will need to be recharged more (obviously) and when it needs this the alternator will kick in and then the MPG difference will be relevant again.

5. If a lot of my driving will be in town then this mod might be worth doing depending on charger/battery cost, cost of fuel, charging cost(120volt from house), and years for payback.

6. As for my initial highway results, it would not be worth the amount of effort or cost (for me). Though I agree that every little bit helps, the inconvenience of having to plug the car in to charge the battery, the cost of battery, the risk of being stranded due to user error, possible decrease in battery life, ect would not be worth a sub 5% increase to me.

7. I don’t pulse and glide or engine off coast, I don’t know by what amount that would change the percentages, but for these tests i was concentrating on staying at a steady mph, and less on feeling the gas pedal and trying to maintain MPG by slowing down up hills or accelerating in anticipation of hills like I normally do. A few 30-100 mile trips need to be conducted for me to determine if I should or shouldn’t persue this long term.

8. I would like to test the results with the alternator turned off against taking the alternator belt completely off to see how much actual drag there is. Lots of people say "its hardly any drag", but I havn't seen any actual back to back, ABA testing.

9. I did some in gear coast-downs and while getting 999999mpg doing so I switched from alternator on to off and you could feel the car slowing down faster with it on than off, so i know the switch was working. Also, i tried a few full throttle in 2nd gear tests getting up to highway speed and when you switched the alternator off it was like a subtle "vtec kicking in", you actually got a boost in acceleration. It was fun to flip it on and off several times on the same acceleration run to feel the car bog, then accelerate with the switch. By "bog" i mean you could feel the drag, it wasn't a ton, but was definitely noticeable.

Hopefully this information helps some guys. if you have any questions feel free to ask, I hope i can help.

This first picture is my mess of wiring for my obd1engine to obd2acar conversion.


This one is of me holding the green/white alternator control wire from the engine harness.


I taped the lead to the middle wire to test for continuity to the alternator plug.


Picture of the ohms while testing.


Picture of what pin on the alternator plug is the alternator control wire.


This picture is with the alternator switch in the up position to run as normal.


This is when you switch to the down position to ground the alternator control wire.

[Gasoline Fumes]

I've been playing around with this too. Does the alternator actually shut off? I got the impression from my service manual that it just reduces the charging voltage to 12.5v. In other words, the battery won't be drained by running at the 12.5v mode for extended periods. I know the voltage drops fast during engine off coasting with the headlights on. I can run a lot of accessories with the 12.5v mode and the voltage seems to stay there.

I have a voltmeter in my Civic and grounding the control wire does drop it to 12.5v from 14.5v. But not always. Sometimes the alt doesn't respond and sometimes it will drop to 12.5v and then eventually jump back up to 14.5v. Have you seen anything like that or is my alternator just weird? Or does the alt have the ability to override the control wire signal?

[Tesla]

steffen707,
Excellent report, many thanks, you're a champion.
I have been pondering this for a while with my vehicle, so through all you're testing no warning or alternater lights, that's great.
It looks like you have confirmed a positive gain of 2-3% if this was used constructively, i.e. on downhills and other times when brakes are required, this could come out as an overall gain of 1%, although it doesn't seem like much add a few of these types of adjustments together and the $'s start to add up, or as I am thinking add say a 50w solar panel to the system and there may be no need for alternator or home plug in 90% of the time.
Good work, look forward to hearing more of your progress.

[kingsway]

You raise a thought that I have also had. Compared to modified driving, like P&G, savings like these are so small you really have to wonder if it is worth the trouble...

[steffen707]

Originally Posted by Gasoline Fumes

I've been playing around with this too. Does the alternator actually shut off? I got the impression from my service manual that it just reduces the charging voltage to 12.5v. In other words, the battery won't be drained by running at the 12.5v mode for extended periods. I know the voltage drops fast during engine off coasting with the headlights on. I can run a lot of accessories with the 12.5v mode and the voltage seems to stay there.

I have a voltmeter in my Civic and grounding the control wire does drop it to 12.5v from 14.5v. But not always. Sometimes the alt doesn't respond and sometimes it will drop to 12.5v and then eventually jump back up to 14.5v. Have you seen anything like that or is my alternator just weird? Or does the alt have the ability to override the control wire signal?

UGH! I didn't have my voltmeter hooked up while trying all the tests. I was too lazy to find a positive and ground and tap into them, but it looks like I need to confirm or deny your results that the alternator will still flip to 14.5 volts. I was really hoping that my results were good and that it fixed this troubling problem for civic owners.


You could be right about this. How would I go about checking to see if the alternator is actually off, and not just charging at the low setting? This is frustrating but I like to try and be thourough. I suppose I could try my tests again and just unplug the 4pin plug and main charging wire from the alternator between tests.

[steffen707]

Originally Posted by Tesla

steffen707,
Excellent report, many thanks, you're a champion.
I have been pondering this for a while with my vehicle, so through all you're testing no warning or alternater lights, that's great.
It looks like you have confirmed a positive gain of 2-3% if this was used constructively, i.e. on downhills and other times when brakes are required, this could come out as an overall gain of 1%, although it doesn't seem like much add a few of these types of adjustments together and the $'s start to add up, or as I am thinking add say a 50w solar panel to the system and there may be no need for alternator or home plug in 90% of the time.
Good work, look forward to hearing more of your progress.

Thanks Tesla, as others have pointed out, this might not actually be working as efficiently as a true alternator shut off should be. I will have to do more testing and report back before we go jumping for joy.

[steffen707]

Do you guys think if I disconnected the alternator charge wire from the battery, put a big fuse between the battery negative terminal and connect that to a light, and then connect that to the alternator, that would let me know if the alternator is actually charging or not?

[HAHA]

Another factor as the voltage drops is that the ECU adjusts the fuel injector timing to compensate. This may affect efficiency and also the calibration for Scangauge or similar systems. I tried to find out once if Scangauge did calculate fuel correctly based on injector timing. I think it may do so but I'm not 100% sure.

Since you are dealing with rather small differences here, you may have to perform a dedicated experiment to quantify these effects. If your instrumentation calculates correctly when the ECU/injectors run on low voltage, it gets easier. Then you only need to find out if the engine works better or worse with normal/low voltage. If there is a degradation, you may be able to find a workaround, eg. extra batteries to keep voltage up.

At least some of your test runs indicate that your engine may degrade with lower voltage.

[steffen707]

Originally Posted by HAHA

Another factor as the voltage drops is that the ECU adjusts the fuel injector timing to compensate. This may affect efficiency and also the calibration for Scangauge or similar systems. I tried to find out once if Scangauge did calculate fuel correctly based on injector timing. I think it may do so but I'm not 100% sure.

Since you are dealing with rather small differences here, you may have to perform a dedicated experiment to quantify these effects. If your instrumentation calculates correctly when the ECU/injectors run on low voltage, it gets easier. Then you only need to find out if the engine works better or worse with normal/low voltage. If there is a degradation, you may be able to find a workaround, eg. extra batteries to keep voltage up.

At least some of your test runs indicate that your engine may degrade with lower voltage.

Well I don't use scanguage because the d15z1 and civic VX are obd1, so i'm using mpguino.

I would have to do long term testing if what you think might happen is true, that the injectors fire slower, or the ecu compensates on the injector pulse width.

I don't see how my test results indicated degradation though, because in every test I got better gas mileage with the alternator off.