Troubleshooting lean condition on turbo H22

[Raf99]

Check for corrosion on ground and power cables, i know it can cause weird symptoms and it would make sense to build up over time.

Ya. I don't allow rust on my car. But all grounds have been checked.

[gloryaccordy]

Do CBs have voltage regulators separate from the alternator?

[Raf99]

Do CBs have voltage regulators separate from the alternator?

No. And i was reading about C-wire Alt cut thinking that would explain the battery voltage dropping, but this was only used in S100/S200. In S300 we use the ELD disable function to accomplish the same. I wonder if that feature is simply not working as in the last graph the car running at 12.4V when driving is simply running off battery. But this only happens when the car is moving above a certain mph. So weird..., i wish i had a prior datalog to compare to.

[sonikaccord]

End of the run. I'm not seeing any reason for this battery fluctuation.



I'm fresh out of ideas on this. Let's keep in mind I haven't changed anything with S300, just moved the fuse box back and throwing new parts at it.

For the first graph:
Everything looks good before 10:05. Batt Voltage is normal, AF and S.Trim are doing their job.
Then we let off the throttle at 10:05, and our AF goes lean, with normal batt voltage. (Injector cut? Doesn't the ECU cut fuel when coasting or the throttle is closed?)
We give it a few throttle blips and on the second blip we see the batt voltage go low. The first throttle jab should have dropped the Batt voltage as well...condition may be rpm dependent?
On the third blip the AF goes rich and our Batt goes down to I guess 10v there?. S.Trim here is attempting to lean it out, but the rpm just drops to 200. Batt voltage is still low.
Here is the interesting thing, the AF drops like a brick and so does the Batt voltage. The ECU is dumping fuel for a good 3 seconds or so, then a throttle blip saves it.
Goes lean for about 7 seconds, and again BATT and AF drop.
Then the throttle gets a few more jabs, and AF responds nicely, BATT voltage is normal, then everything stabilizes again.

Ok, let's get our correlations togther.
A severe battery drop and rich condition occur at the same time.
A closed throttle causes a lean condition for a bit...which may or may not be normal. As I look at the other throttle blips, the time that the AF stays lean seems to be consistent at around 3-5 seconds. If the lean condition continues past that, the ECU starts to increase S.Trim to try and bring it down.
When S.Trim goes up, Batt goes down.
There is a pattern there! When the ECU wants to deliver more fuel, you get a Batt drop. The lean condition is normal when it is within the 3-5 second range, but when it extends beyond that, your ECU richens the AF and your Batt goes down.

I'm sure you've come to that conclusion already. So what I recommend doing is probing the Injector voltage supply line (I think it's the same line as the ECU is on) and see what is happening there. Also, check your resistor box and injector resistances. If they are presenting a high load (low resistance) to the ECU, then that may cause a voltage drop internal to the ECU. A bad harness pin will also cause a voltage drop so make sure your connectors are tight.

If all checks out ok, force the alternator to charge at 14v at all times using the ELD bypass technique and see if the batt voltage still drops.

[Raf99]

.........................

I'm sure you've come to that conclusion already. So what I recommend doing is probing the Injector voltage supply line (I think it's the same line as the ECU is on) and see what is happening there. Also, check your resistor box and injector resistances. If they are presenting a high load (low resistance) to the ECU, then that may cause a voltage drop internal to the ECU. A bad harness pin will also cause a voltage drop so make sure your connectors are tight.

If all checks out ok, force the alternator to charge at 14v at all times using the ELD bypass technique and see if the batt voltage still drops.

Yep, we are on the same page.

How do I check these things? (injector supply voltage and injector resistors)
What pins am I checking on what harness? The pins for the injectors?
And disabling ELD in S300/Hondata should force constant 14V should it not?

[sonikaccord]

I'm at work so I can't post schematic screenshots right now. I'm going to try to explain this in a step format

Checking voltages:
The YEL/BLK wire supplies power to the ECU and Injector Resistor Box. It goes through the main relay and starts at the U/H Fusebox. We will check these lines starting from the source. Voltmeter should be set to 20 VDC.

1. Check voltage on both sides of FUSE #28 in the underhood fuse box. Voltage should read the same on both sides. Write this voltage down as it should remain the same at all the points that we are going to check.

2. Check the voltage at the YEL/BLU wire of the main relay. Relay should be plugged in for this test.

3. Turn the ignition on, don't start car, check voltage at the YEL/BLK wire of the main relay. Relay should be plugged in for this test.

4. Go to the ECU, check the voltage on Pins A25 and B1 (IGP1 and IGP2, respectively).

5. Go to the resistor box, check the voltage on the YEL/BLK wire. You can unplug the connector for this. Reconnect it after you finish.

6. Go to the resistors, check the voltage on the RED/BLK wire. They should be equal across all resistors, and a bit lower than the fusebox voltage.

7. (optional) You can check the voltage on the other pin of the injectors, the one that leads to the ecu. They should read the same across each injector.

You can check injector resistance while the clips are off. Ohmmeter set to 1 Ohm setting. All should be equal across.

Now if your meter has a min-max feature, you can set it to MIN and hook it to any YEL/BLK line. Then see if you can get your BATT line in Hondata to drop, then see if it matches the MIN value on the meter. If it matches, your problem is outside the ECU, if not, then it will be internal to the ECU.


For the ELD, it looks like Hondata has it set to 3.81V all the time. According to the thread I posted earlier, an ELD voltage of 2.1v will put the alt in full output mode. You will need an 820 ohm resistor connected across the ELD. Or can you adjust ELD voltage in Hondata? I don't think simply disabling it will do the trick here...you could try it and see what happens anyway.

[Raf99]

Man, much appreciated. If you remembered those wire colors from the top of your head that's crazy. I'll tackle this tonight. Have to take the front bumper off to get to the injector box. And I only have a cheap multi-meter, maybe time to invest in a better one.

[Raf99]

Ok. I was doing this last night. Finished up, had everything written out. Hit my mouse pad by accident which clicked a link and I lost everything i wrote. haha. Charged the battery back up and now it's the AM. Morning.

I'm at work so I can't post schematic screenshots right now. I'm going to try to explain this in a step format

Checking voltages:
The YEL/BLK wire supplies power to the ECU and Injector Resistor Box. It goes through the main relay and starts at the U/H Fusebox. We will check these lines starting from the source. Voltmeter should be set to 20 VDC.

Battery in trunk currently - 12.8V

1. Check voltage on both sides of FUSE #28 in the underhood fuse box. Voltage should read the same on both sides. Write this voltage down as it should remain the same at all the points that we are going to check.
12.8V

2. Check the voltage at the YEL/BLU wire of the main relay. Relay should be plugged in for this test.
12.8V (no lie, it took me over an hour to get that main relay out. There should be service awards for those who can do it in record time. Not sure I want to put it back in the OEM spot)

3. Turn the ignition on, don't start car, check voltage at the YEL/BLK wire of the main relay. Relay should be plugged in for this test.
12.4V

4. Go to the ECU, check the voltage on Pins A25 and B1 (IGP1 and IGP2, respectively).
12.2V (fuse #23/batt now showing 12.3 (car on for a bit) Swear when i was doing this last night A25 had 12.2V and B1 had 0.01V. Today it changed. Tripple checked last night too!

5. Go to the resistor box, check the voltage on the YEL/BLK wire. You can unplug the connector for this. Reconnect it after you finish.
12.2V

6. Go to the resistors, check the voltage on the RED/BLK wire. They should be equal across all resistors, and a bit lower than the fusebox voltage.
12.2V on all resistor wires

7. (optional) You can check the voltage on the other pin of the injectors, the one that leads to the ecu. They should read the same across each injector.
12.2V

You can check injector resistance while the clips are off. Ohmmeter set to 1 Ohm setting. All should be equal across.
Ahhh. these things come off...

Now if your meter has a min-max feature, you can set it to MIN and hook it to any YEL/BLK line. Then see if you can get your BATT line in Hondata to drop, then see if it matches the MIN value on the meter. If it matches, your problem is outside the ECU, if not, then it will be internal to the ECU.

Negative. cheap multi-meter
For the ELD, it looks like Hondata has it set to 3.81V all the time. According to the thread I posted earlier, an ELD voltage of 2.1v will put the alt in full output mode. You will need an 820 ohm resistor connected across the ELD. Or can you adjust ELD voltage in Hondata? I don't think simply disabling it will do the trick here...you could try it and see what happens anyway.

Ummm............. i ahh feel..... like an idiot right now. And i know it won't help but hmmm. Look at screen shot #5, page1 of this thread. ELD.......

[sonikaccord]

Great work! It looks like all the connections are good so we can rule that out. Well, maybe not the ECU one... Could've just been a fluke (see what I did there?).

Wait. Did you check the injector voltage while they were connected? That's fine since everything looks equal. I wouldn't worry too much about that.

Let's go back to the ecu. That's the only input that is battery voltage related that the ECU can see. There is the back up (radio fuse) too, pin D1. Can you probe A25, B1, and D1 while the car is acting up. We need to see if the ECU is telling the truth about our voltage. If it's reporting a low voltage when everything is OK, I'd try to get another ECU to swap out. I have two, but they are auto ECUs.

Leave the ELD enabled, lol, for now

[Raf99]

Great work! It looks like all the connections are good so we can rule that out. Well, maybe not the ECU one... Could've just been a fluke (see what I did there?).

Wait. Did you check the injector voltage while they were connected? That's fine since everything looks equal. I wouldn't worry too much about that.

Let's go back to the ecu. That's the only input that is battery voltage related that the ECU can see. There is the back up (radio fuse) too, pin D1. Can you probe A25, B1, and D1 while the car is acting up. We need to see if the ECU is telling the truth about our voltage. If it's reporting a low voltage when everything is OK, I'd try to get another ECU to swap out. I have two, but they are auto ECUs.

Leave the ELD enabled, lol, for now

Couldn't I also just hook up a volt gauge in the cabin to verify volts are dropping?

So today I finished putting the car all back together and went for a short drive around the block. I disabled ELD to see if it would drastically change how the battery is responding (or the ALT output). And there are still some issues here.... i think. No drastic drops compared to before but...

It still appears that over the course of a 11min drive the battery voltage is dropping. We start off with 13.96V ....


And by the end of a 10min drive we are at 13.4V. A steady decline...


And near the end of the drive, sitting in the driveway a blip of throttle resulted in the battery dropping again to 12.4V. I have no explanation for this and it also implies that we are running off battery right now. No ALT.


Same throttle blip before the start of the run


Driving home down the street, pulling into the driveway


The weird transition from smooth to jagged on the battery line when I am on the rpms. Once jagged the battery never returns to smooth again, even at idle. I think this is the problem.....



So I had to disable the ELD. Having the car running at 9V can't happen, and by disabling this it appears we fixed that. But the voltage is dropping over time. And when I blip the throttle / come to a stop / the car acts different at the beginning of the run compared to the end. At the end we start to see battery drops for no reason, we see erratic battery fluctuations. If I have time tomorrow after work I'll take it for a longer drive. I think if I went for a 20min drive the battery would be down to 13V by the time I got home with a ton of lean conditions; same old song and dance.

[sonikaccord]

Couldn't I also just hook up a volt gauge in the cabin to verify volts are dropping?
..........

So I had to disable the ELD. Having the car running at 9V can't happen, and by disabling this it appears we fixed that. But the voltage is dropping over time. And when I blip the throttle / come to a stop / the car acts different at the beginning of the run compared to the end. At the end we start to see battery drops for no reason, we see erratic battery fluctuations. If I have time tomorrow after work I'll take it for a longer drive. I think if I went for a 20min drive the battery would be down to 13V by the time I got home with a ton of lean conditions; same old song and dance.

You could hook it up in the cabin to get a rough look at what is happening, but hooking it up at the YEL/BLK wire at the ECU is going to give the best readings.

Also...how does the car act when it's running in open loop? The lean condition only happens when the car is warmed up? Can you force the car back into open loop when its warm?

[Raf99]

You could hook it up in the cabin to get a rough look at what is happening, but hooking it up at the YEL/BLK wire at the ECU is going to give the best readings.

Also...how does the car act when it's running in open loop? The lean condition only happens when the car is warmed up? Can you force the car back into open loop when its warm?

I can connect a volt gauge to the ECU or even Fuse 28.

I can force the car into open loop and stay there. I did this before but a lot has changed since then. The flip between closed and open is when the vehicle starts moving right? And yes, issues only happens when car is warmed up. It will be interesting to see how the battery reacts in a open loop. Give me some time on this....

[Raf99]

Well.......... that was interesting. Took Sally for a 30+ min long drive and it like the good ole days. Left closed loop on, ELD disabled, connected a volt meter to ECU fuse. When I started the car the gauge read 12.2V from messing around the other day. After 10min it read 12.7V. So the alternator is charging the battery.

Oh but the drive. I could tell it was back to her old self. No back firing when off the gas coasting, no lean conditions, no battery fluctuations, nothing. I was even doing extreme things like having the lights on, blower on, off the gas!/on the brakes, trying to make it stall. No stall.

Lowest the voltage got was 13.2V according to SManager. The gauge I installed in the cabin was at 14V the whole time and dropped to 13.9V for the later half of the drive. No big deal. 13.2V is at the end of this graph....


We even got to have some fun! Ooooo Sally, don't get me excited and let me down when I put you all back together. Injectors maxed to the limit.



And guess I can show you how I roll around with a fuse box, ECU, gauge on the floor and the laptop.


At the end of the drive the voltage in Smanager was 13.4V. Idling like a champ.


Now............ what did we do? Nothing. haha. Dam. We played around with some wires, disabled ELD (still not sure how that got enabled), and that's about it. Soooo.......... still not sure how it fixed itself. I suspect I'll go for a drive tomorrow and it will be the opposite of today.

[sonikaccord]

She is such a tease bro...I guess take note of the weather and how the stars were aligned today too.

[Raf99]

Ya really eh. And what alot of people dont realize is this is the stuff you have to deal with when you mix a turbo setup with a 25 year old car. But dam!! I owe you a big thank you. So thank you. I learned a bit which is great. I'd still like to learn tuning the car.
Now a question for you. Do you think anything will change using the fuse box extension I made? This puts the fuse box behind the back seat and lengthens the wires by 9'. Using proper gauge wire of course. And the fuse box gets its power from the battery in the trunk rather than the engine bay.