1UZ UCF20 runs horribly rich

PARTS FOR SALE: Limited availability on EGR Delete Kits, Torque Converters and Supercharger Brackets. Email for info: [email protected]

smokeymx5

New Member
Messages
7
Hi all, I created an account just to try and solve this problem on my project car swap. I'd really like your thoughts on the issue, but I guess even if noone has any ideas it helps my head to write it all in one place.
I've had this issue for probably 2-3 years, and since I'm finishing up everything else and trying to get it roadworthy I really need to sort this, but I'm just struggling to work out what to do.

Basically, I have a later non-vvti 1UZ swapped into a miata. TL;DR, it runs so rich and pours smoke out of the exhaust. Long version of the problem is this:

ECU power on, key on, gives code error 31 for MAF.
Engine turned on, it starts perfectly and settles into a good idle. 14.7afr, spot on. Then... after about 5 seconds, the afr drops slowly. After about 10 seconds, the afr is on the minimum reading of 10afr, and smoke starts to come out the exhaust. The smell of fuel is horrendous. However, it runs and idles just fine. Revs up nicely. Seems to have decent power, I dont think its in limp. When codes are checked after engine on, I get the following:

21- o2 sensor (AEM output wired into ECU, 0v-1v)
27- sub o2 sensor (it doesnt have one)
28- o2 sensor (AEM output wired into ECU, 0v-1v)
29- sub o2 sensor (it doesnt have one)
31- AFM error
71- EGR (deleted, blanked)
78- fuel pump (Uses aftermarket bosch 044 wired to a panel switch)

It is running stock ECU and wiring harness it came with. Manualised auto box. Plastic maf. AEM wideband with narrowband 0-1v output into stock ecu.

Okay so lets list the diagnostics I've done so far:

Coolant temp sensor test at sensor, resistance within spec for ambient temps.
Opened ECU, checked for faulty caps. Seems fine, no evidence of leaks or damage.
I was using thin oil, swapped it for thicker stuff in case it was leaking past stem seals or something.
Wired in o2 sensors. I was originally using no o2 sensors, so wired in the AEM gauge narrowband output of 0-1v to the two main o2 sensor wires to the ECU.
Replaced MAF
Replaced MAF plug (wire was split)
Checked maf wiring. THA and E2 have correct resistance across them, 2.6kohm. E2, THA, VG all have continuity to ECU. I cant figure out what pin B and E2G connect to? I cant get continuity of B to anything. E2G has a connection to ground with negligible resistance. E2 has connection to ground with negligble resistance. THA has 5v.
When MAF is disconnected, there is no change at all.
And no, I havent put any cleaning solution through the MAF!

To me this seems to point to the MAF. The 31 code, unplugging and it does nothing... OR injectors stuck open. OR the classic faulty ECU. In that order.

So next I think to diagnose on the MAF is finding continuity between B and something. I dont have a frequency thingy in order to test the output frequency, I assume is the VG wire. I might have to invest in one. I might even buy a THIRD MAF just to make sure. But I suspect wiring is the culprit.

Lambda sensors - its just the narrowband output from AEM wideband running stright to OXL1 and OXR1 together. Voltage has been verified to be correct. Is it really as easy as this? I assume the errors are for things like the heater circuit having an open reading as the correct sensor is not present?

As for the injectors - when I disconnect one or two the smoking is reduced and the afr rises to about 10.2, but I dont really think im proving anything here because it must be all of them together to be smoking this bad. It doesnt sound like its just one cylinder, it sounds pretty smooth for its condition, no misfires. If they were stuck open, why would the car run so well in the first 5 seconds and drop down the afr range so smoothly? The same for fuel pressure. I had suspected the pump I was using was too good and is overwhelming the regulator and causing rail overpressure, but if this was the case then it would be rich right from the engine start.

Really any thoughts are appreciated. Especially on MAF and lambda sensor error fixing. I'm fairly confident the error is somewhere in the wiring. If I can get rid of these main codes I might be able to solve the issue. I've probably spent upwards of 100 hours just trying to figure out why its smoking so much. So sorry if this is a bit rambly, I think I'm losing my mind.
 

gloverman

Well-Known Member
Messages
3,031
Location
Hamilton. New Zealand
I would be sorting the list of codes , a couple of which are wiring faults. Search CartuneNZ on YouTube. I may have helped out there otherwise I sell instructions.
 

smokeymx5

New Member
Messages
7
I would be sorting the list of codes , a couple of which are wiring faults. Search CartuneNZ on YouTube. I may have helped out there otherwise I sell instructions.
I'll check out the channel for more info. I've seen a few vids but there is so much info on there!
Yeah thats what I'm working down, but I'm a bit stuck on my MAF diagnosis as to where pin 1 and 5 connect to in the ECU for a continuity test. I think I'm getting confused over earlier wiring pinouts for a 92 as opposed to the 96. Which ties in to:

Just a minor point: Is the AFM the correct one for the ECU you have?
I got given the MAF, harness and ECU along with the engine and trans when I bought it. But since you mention it I shall check. My ECU is 89661-50251, which I believe is UCF20, the later non-vvti, and the later plastic MAF. So I think its the correct one.

This makes me think about differences in early and old MAF and MAF wiring... I need to go back and double check what I'm testing is correct against the correct year as opposed to old pinouts... I know its going to be something really simple such as that lol. Thanks guys, I'll have a look at it tonight
 
Last edited:

smokeymx5

New Member
Messages
7
As for the lambda sensors, is it acceptable to just tie the same narrowband output to both the OXL1 and OXR1? Or should I go for all four, front and back? Not sure why its still giving front sensor errors, I guess its not seeing the heater resistance?
 

smokeymx5

New Member
Messages
7
So I watched some more of that channel and found something interesting. I have a further theory regarding the AFM, but I am struggling to find/read diagrams to prove or disprove it. The white plug that is on the engine harness that doesnt plug into the ECU, (Chassis plug maybe?) could be a culprit. As shown in this video at around the 2:00min mark:


I dont have that many wires going into my white plug. I have two 12v leading in (one is starter), and I think from memory I have a ground running in. He says one of his wires is power to the afm. Perhaps my AFM needs a 12v from somewhere in this plug? This would explain why the air temp sensor works (no codes thrown) because its a seperate circuit... And would make sense why I cant find continuity from the AFM to something to supply it power other than the 5v to the IAT sensor.

Anyway, I suspect if I can find a pinout for that plug, I might be able to fix a possible power to MAF problem.
 

smokeymx5

New Member
Messages
7
Funny enough I have the pinout to that plug in my wiring instructions.
I'd absolutely love that pinout if you have it please!! How much do you sell your instructions for? Or even just that pinout, because I have the rest.

Also, I managed to fix the smoke issue. Whilst I didnt have a pinout, I did have my multimeter so I started probing. I probed a red/black wire in between the large ignition and starter wires I have on that plug. Turns out it leads to B red/black wire on the MAF. Maybe leads to more things too. I put 12v to it and hey presto, maf code dissapeared. Turned the car on and no more smoke on idle! Idled for about 20 seconds at 12~ish afr before spluttering and cutting out because I ran out of fuel in the tank. Doh. But that is that code and the majority of the issue fixed... Now onto the o2 sensor codes...

It makes me wonder what else is on that pinout that I need. I think there must be power to igniters on there, I had to cut into the ignitor wire to give it power a year ago because I couldnt figure out what was supplying it! In the video it says power to injectors but it seems to be happy enough without it so not sure on that one.

Then my roof started leaking so I couldnt play with wires anymore until I've fixed that, sigh.
 

smokeymx5

New Member
Messages
7
Still having issues with this, turns out it wasn't running out of fuel, it just runs super lean and wants to die unless I rev it a bit. Then it injects fuel as normal which rules out a lot of stuff. Any return to idle will eventually lean out and cause a stall after about 10 seconds. I guess this is because the lambda sensor codes mean whatever signal I'm putting in to the ECU isn't being used for closed loop?

So, trying to find a way to get rid of error 21 and 28 whilst only having the single wideband with gauge and narrowband simulator... Thinking about wiring in an O2 sensor or two to the stock ECU purely for the heater circuit. Although a better solution might be to put a resistor across them to simulate one, if I can find the correct resistance.

So... @gloverman ... How much for these wiring instructions? Either that or I go custom ecu but I'd really rather save the money for now.

edit: also starting to suspect all the throttle/idle junk is not set correctly or stuck. Makes me want to set the thing on fire to be honest. So gotta go down that route of checking and adjusting, if I can find documentation on how to set everything up on the throttle body from scratch.
 
Last edited:

smokeymx5

New Member
Messages
7
My wiring instructions for ucf20 non vvti are $175nzd at present. They also include 20 minutes of my time for personalized questions. Email is [email protected]
Sounds reasonable. I think if I had a better guide than the 1uz doc by Nigel Wade and various forums I could fix a lot of issues much quicker. Or if I went custom but really don't need that at the moment.

I'll shoot you an email
 
Top