Honda D Series Forum banner

1 - 16 of 16 Posts

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #1 (Edited)
Engine: Supercharged D16A6
Tuning: Turboedit NG63
Fuel pump: Walbro 255
Injectors: Injector Nation Bosch EV14 (600cc)

Background:

I was previously using 450cc DSM injectors. One of the injectors stayed stuck open while on the dyno, flooding the engine with fuel. This prompted me to buy the new IN600 injectors.

Note - I did delete the OEM resistor box by crimping the four black wires and the single red wire together.


Issue:

During cold starts, the engine idles at around 14-14.7:1 AFR like it should. However, after the engine warms up completely, the AFRs lean out to around 16.5 - 18:1. I know that this isn't a sensor issue because the engine tone changes and it begins to stumble.

So far I've done the following to troubleshoot:

-Verified that the fuel level is above 1/2 tank and that the fuel pressure is the same during both hot and cold idle.

-Added 2-4% fuel to the idle cells.

-Datalogged when the issue is occurring.


Cold idle - normal AFR

Code:
Time	RPM	MAP	A/F	TPS	VSS	IAT	ECT	ELD	Baro	Row	Col	Gear	Pulsewidth	Ign. Adv.
														
20.453	986	-17.48	14.61	1	0	115	149	2.55	14.54	4	3	0	3.41	18
20.671	986	-17.48	14.64	1	0	115	149	2.55	14.54	4	3	0	3.41	18
20.89	987	-17.48	14.67	1	0	115	149	2.55	14.54	4	3	0	3.41	18
21.125	987	-17.48	14.55	1	0	115	149	2.55	14.54	4	3	0	3.41	18
21.359	987	-17.48	14.52	1	0	115	149	2.53	14.54	4	3	0	3.41	18
21.609	987	-17.48	14.54	1	0	115	149	2.53	14.54	4	3	0	3.41	18
21.859	993	-17.48	14.54	1	0	115	149	2.53	14.54	4	3	0	3.41	18
22.125	993	-17.48	14.54	1	0	115	149	2.53	14.54	4	3	0	3.41	18
22.39	993	-17.48	14.55	1	0	115	149	2.63	14.54	4	3	0	3.44	18

Hot idle - Lean AFR

Lean

Code:
Time	RPM	MAP	A/F	TPS	VSS	IAT	ECT	ELD	Baro	Row	Col	Gear	Pulsewidth	Ign. Adv.

0.265	822	-16.9	16.29	1	0	113	205	1.35	14.54	3	3	0	3.48	18
0.375	822	-16.9	16.29	1	0	113	205	1.35	14.54	3	3	0	3.48	18
0.484	822	-16.9	16.29	1	0	113	205	1.33	14.54	3	3	0	3.49	18
0.609	822	-16.9	16.32	1	0	113	205	1.33	14.54	3	3	0	3.49	18
0.734	822	-16.9	16.32	1	0	113	205	1.33	14.54	2	3	0	3.49	18
0.875	822	-16.9	16.39	1	0	113	205	1.18	14.54	2	3	0	3.54	18
1.015	822	-16.9	16.38	1	0	113	205	1.18	14.54	2	3	0	3.54	18
1.171	836	-16.9	16.38	1	0	113	205	1.18	14.54	3	3	0	3.54	18
1.328	836	-16.9	16.29	1	0	113	205	1.33	14.54	3	3	0	3.45	18

Even leaner:

Code:
Time	RPM	MAP	A/F	TPS	VSS	IAT	ECT	ELD	Baro	Row	Col	Gear	Pulsewidth	Ign. Adv.

0.703	828	-16.12	17.95	1	0	127	200	1.37	14.54	3	4	0	3.72	16.92
0.843	828	-16.12	17.89	1	0	127	200	1.37	14.54	3	4	0	3.72	16.92
0.984	829	-16.12	17.9	1	0	127	200	1.37	14.54	3	4	0	3.72	16.92
1.14	829	-16.12	17.9	1	0	127	200	1.37	14.54	3	4	0	3.75	16.92
1.296	829	-16.12	17.71	1	0	127	200	1.37	14.54	3	4	0	3.75	16.92
1.468	813	-16.12	17.8	1	0	127	200	1.37	14.54	3	4	0	3.75	16.92
1.64	813	-16.12	17.8	1	0	127	200	1.37	14.54	3	4	0	3.66	16.92
1.828	813	-16.12	17.73	1	0	127	200	1.24	14.54	3	4	0	3.66	16.92
2.015	813	-16.12	17.76	1	0	127	200	1.24	14.54	3	4	0	3.66	16.92
2.218	813	-16.12	17.77	1	0	127	200	1.24	14.54	3	4	0	3.66	16.92
*The TPS reads 1% because of adjustments I made to the idle stop screw. Need to fix this eventually.

I do have some IAT ignition compensation configured that pulls timing which is why you'll notice the reduced timing in the last datalog snippit:




There's also some minimal ELD fuel compensation which I did not configure but figured it was normal:




If you look at the "good" datalog and the first "bad" datalog, you can see that the injector pulse width and the ignition advance is almost identical which makes this all the more confusing to me. Obviously the idle RPM is slightly higher during the cold idle and I'm hitting a different row because of it but again, the fuel and the ignition values are nearly identical.




This seems to be isolated to idle and possibly low RPM/high vacuum conditions. WOT pulls are unaffected and the AFRs are around 12:1 as I expect.

Any ideas? I'm really not sure where to go from here since the injector pulse width and ignition advance are so similar during the normal and lean conditions. I suppose I can remove all of the ELD fuel compensation and try that out.


ANY input is appreciated. I'm stumped...
 

·
Registered
Joined
·
5,880 Posts
I know you stated the engine tone changes and it stumbles, but how does the exhaust smell when this occurs?


When its hot idling, if you have a rubber fuel return line, pinch the hose about halfway thin with an adjustable vicegrips. allow some flow.

now smell the exhaust again, and listen for the engine tone.


did it help?



If it does, perhaps you need to verify the coolant temp sensor that talks to ecu is giving a clean signal when hot.
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #4
Thanks for the input guys. Joe, I’ll modify the bin as requested and go datalog now.

Matt, I’ll give that a shot, thanks. I have a fuel pressure gauge so I’ll be able to see if the pressure increases when pinching the hose.
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #5 (Edited)
OK so this seems like it’s at least partially related to ELD voltage and battery voltage.

I created a bin with no fuel/ignition compensation values and set it to datalog every second.

ECT - 159 - 207° F (Radiator fan off)
ELD - 2.6 - 2.78v
Batt - ~14v
AFR - 14.2 - 14.6:1

ECT - 205 - 207° F (Radiator fan on)
ELD - 1.7 - 1.9v
Batt - ~13.3v
AFR - 15 - 15.3:1

ECT - 205 - 207° F (Radiator fan & headlights on)
ELD - .78 - .8v
Batt - ~13v
AFR - 16.2 - 16.6:1

I've attached the datalog and the bin I used. The log file is really long because I wanted to capture the entire warmup process. Row 845 and beyond is where I started to turn the lights on while the rad fan was going and I began monitor battery voltage.

I’m not entirely convinced that this is the entire story though. For one, I did have some minimal ELD compensation configured before (shown in the first post) which was not present in these datalogs. Secondly, the 3rd datalog snippet I posted in the original post showed a leaner AFR with a higher ELD voltage. Finally, the headlights weren’t on during any of the previous occurrences and the lean condition persisted through the radiator fan cycling on and off.
 

Attachments

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #6 (Edited)
So I played with the ELD settings tonight and I was able to get the car to idle better without going extremely lean. Unfortunately, the ELD compensation required about 10% fuel enrichment at an ELD reading of .8v.

Of course this means that basically everywhere else on the fuel map, where I've already dialed in the AFRs, the fuel will also be enriched by that same 10% when there's a high electrical load.

So I suppose I have a few options...

-Significantly minimize the ELD compensation and enrich the idle cells by 6-8% and deal with an idle that fluctuates between 12.5:1 - 15.5:1 (estimate) based on the varying ELD reading.

-Set the idle to around 1000 RPM which would provide a longer pulse width.

-Try to use the smaller 400cc bosch EV14s which would require a higher pulse width at idle and should idle better.

-Shell out the cash for peak-and-hold injectors like RC 440s.


I will play around with it more tomorrow and see if I can't get this going without switching injectors.
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #7 (Edited)
Ugh, even with an 8-9% ELD comp, idle at 1200, and enriching the idle cells by 4%, I still saw AFRs approaching 17:1 at idle.

The part that still confuses me is that cold idle, even with a high ELD load, doesn’t show these crazy lean situations.

About ready to throw in the towel on these injectors and just break down and buy some RC peak and hold injectors.
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #9 (Edited)
Thanks Joe. I think the stuff I posted is really only going to show the typical ELD behavior you’d expect to see. The issue with the AFRs in the 17s seems to only present itself after I’ve driven the car. I can leave it idling for 20+ minutes and let the fan cycle on a dozen times and it’ll never go into the 17s. But after I drive the car, even with a bin that has aggressive ELD compensation, I’ll see the AFRs creep into the high 16s/low 17s

I’m beginning to think this may be related to the alternator output voltage. If I just let the car warm up in the garage with the rad fan, headlights, and defroster on, the voltage at the battery seems to always stay above 14v

If I drive the car and make a few pulls, I start seeing the really lean AFR at idle. When I checked the battery voltage while the issue was occurring, it was at 13.8v.

Who knows. Maybe the alternator voltage regulator is doing something funky and the voltage to the injectors is just low enough to degrade their performance. I am over driving the alternator with my pulley setup but it’s also a new reman denso unit.

Tough to say. Maybe I’m chasing a red herring with the alternator voltage stuff. It’s definitely frustrating
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #11 (Edited)
The ELDs are consistent before and after. One thought I had was to check the injector voltages before and after.

Since all the injectors are fed from the same circuit, Im wondering if I can check the voltage with a multimeter using a chassis ground. Not sure if this will affect the power to the injectors but I wouldn’t think so?
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #12 (Edited)
OK so this appears to be directly related to the voltage I'm reading at the battery.

Cold start idle
--------------------

AFR = 13.8 - 14.1:1
ELD = 2.5-2.7v
Batt Voltage = ~14.3v


Warm idle
--------------------

AFR = 14.7 - 15:1
ELD = 2.5-2.7v
Batt voltage = ~14.1v


Hot idle (Fan and Headlights on)
--------------------

AFR = 16.3 - 16.5:1
ELD = .8v
Batt Voltage = ~12.9 v


ELD compensation adjustments don't appear to help but I’ll test again tomorrow.


I tested the ELD per the FSM and it checks out.

Can anyone comment on whether or not these voltage values seem normal ?

I believe they are so this may just be a a case of the injectors not being well suited for an OBD0 ECU where the dwell time cannot be adjusted

I have a feeling that the only solution is to switch back to low impedance injectors which have far lower dead times at lower voltages
 

·
Registered
Joined
·
5,880 Posts
my 95 accord has a bad eld, i unplugged it and get full voltage all the time.

could try that and see if hot lean goes away
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #14
Thanks Matt. Interesting idea that I hadn’t thought of.

My ELD tested good but I may try this.

Hoping someone can provide some input regarding the battery voltages. Really just need confirmation as to whether or not seeing 12.9v with the rad fan and headlights on is normal.
 

·
Registered
CRX ECODiesel Ram
Joined
·
1,036 Posts
My 89 crx with d16a6 and obd2 alternator, hondata p28 and working eld shows a similar drop. I used peak and hold injectors with the obd0 ecu and switched/converted to saturated injectors with the p28. I am wondering if the injector pulse characteristics as produced by the ecu are an issue. Voltage waveform for p/h vs saturated is different enough that i wont mix ecu and injectors.

Might already know all this but it's helpful http://performancefuelsystems.com/InjectorCompattibilitywithECUs-TechCorner.htm
 

·
Meat Popsicle
91 CRX Si
Joined
·
2,936 Posts
Discussion Starter #16
Thanks for confirming. Believe the dead times are just too long with the EV14 600's.

Looks like I'll be going the RC 550 route.

Thanks everyone
 
1 - 16 of 16 Posts
Top