KA24de won't idle - leans out until dies

[JohnnyBlu]

Pic for attention

 

Hi Guys,

 

I'm working on my KA swap and am having trouble getting it to run with no load once it comes up to temp.this is an S13 KA24DE, EGR deleted, ECU is #24 for a S13 5-speed

 

All things run good until the coolant temp comes up to about 170 ish. Then I see AFR start to climb until the engine stumbles and dies. Reving the throttle can get it to stay alive but once back to idle it will just lean out and die. It takes about 5 seconds from taking your foot off the pedal at 14:1 AFR for it to climb up to a point of stalling.

When driving, it feels great once you're going; engine speed is up and there is load on the engine. Pulling away from a stop is not so good, as you have to fight this low speed lean issue.

 

Injectors are refurbished

I had a new cheap knock sensor but was getting code 34, so i jumped it out with a 560 ohm resistor. Still get code 34.

Unplug the ECU temp sensor, still same issue but with ECU code 12

unplug the MAF adn it will run but with an AFR of 9 or 10:1

I tested the TPS and found a bad wire so i replaced it. calibrated the new one and still have the same problem.

Have also fiddled with all the IACV stuff on the back of the intake manifold.

 

Any thoughts or help is much appreciated.

[datzenmike]

TPS tells the ECU that the throttle is closed. If that doesn't get through it may think you're driving along at low speed?

[JohnnyBlu]

Hi Mike,

I verified that the TPS was getting 5V supplied. Also checked the output voltage back to the ECU while moving the throttle plate. Ranges from ~1v to ~ 4V if i remember correctly. TPS also checks out with 2 to 10 ohms across the range of closed to wide open throttle. Could this mean bad ECU if the TPS passes all the troubleshooting tests in the book?

[datzenmike]

No but you did find a problem with it and maybe not resolved.

[Stoffregen Motorsports]

Sounds like a temp switch issue. You could get a volume knob at Fry's and wire it into the temp switch, play around with the resistance, and see if you can make it run. If you get it to run well with the volume knob, check the resistance and check it against the factory service manual temp switch resistance, then check your temp switch and see if it is the same.

 

Sounds weird, but I've actually done this.

 

The TPS, MAF and temp switch work in conjunction with each other to set timing and make air/fuel adjustments so it could be a fault with any one of these.

 

Questions? Did it run well before the swap? Did you buy or build your harness? Did you pressure wash the motor before install? Sometimes water can get into the TPS and screw things up.

 

I know there are a few things I'm forgetting to ask, so if they come to me, I'll post them up.

[JohnnyBlu]

I like your idea of the volume knob testing! After some poking poking around I see this range for resistance.

68 degrees....... 2.10 - 2.90 k ohms

122 degrees..... 0.68 - 1.00 k ohms

176 degrees..... 0.30 - 0.33 k ohms

 

Seeing as I just replaced the TPS with new, and the MAF has expected effect on engine running when unplugged - the temp switch seems like the next place to go. The issue does not occur until the car comes up to temp hints towards that as well.

 

other install notes:

Engine was a Craigslist special that a friend of mine picked up for me when I was out of town. It showed up with a bad head gasket and most all accessories in boxes. I should find a pic of the piles of crap he delivered! All the bolts were loose in one box. That was fun to sort out. I tore it down and put in new rings, bearings, gaskets, etc.

 

I did the harness myself. I unwrapped the wiring harness and pulled all the not needed stuff, and connected to the car for power and relay switching, start signal, and so on. My connections were similar to the CAN-AM setup, but without buying the box. I just used relays. I worked that from tracing the S13 wiring harness diagram. I think the harness that I got with the engine was originally from an automatic transmission car.

 

I am running a wideband o2 probe that is doing double duty, sending the 0-1 volt signal to the ecu and wideband signal to a gauge in the dash.

[Stoffregen Motorsports]

Another part that can have heat related issues is the cam sensor or crank angle sensor. This problem usually occurs after heat soak, so it doesn't sound like your problem.

 

Also, I assume you've verified your fuel pressure and line routing...?

[JohnnyBlu]

Well, I did the pot of hot water test with the temp probe and it passed. matched up pretty well with what the book said it should do. 250 ohms at 200F, 450 ohms at 170F, 1k ohms at 125F and 2.2k ohms at room temp.

 

I also verified that the harness is giving 5V to the sender. I can still try running the car and varying resistance with a volume knob in place of the temp sender and see what happens.

 

Fuel: yes fuel pressure is good (38-43psi). That was one of the first places I went. I now have a fuel pressure gauge installed in the supply line where the filter is seen in the picture above. I moved the filter back near the pump. I am running an inline pump and added a bung at the bottom of the tank.

 

Do you have any more insight on this distributor heat soak idea? I haven't heard about that.

 

Lastly, the only code I get from the ECU is 34, which is the knock sensor. This seems odd since I have a 560 ohm resistor in place of a knock sensor.

[Stoffregen Motorsports]

You are talking about the temp switch, not the temp sender, right?

 

Heat soak can affect cam and crank sensors, or an ignition module. A very common issue, usually fixed by simply replacing the sensor or module. On a vehicle that I built a few years ago (engine swapped 4Runner with a 1GR-FE 4.0L Toyota V6), we had constant problems with the cam sensors, which were hall effect sensors. We replaced them with magnetic pickup sensors and the problem went away. You'd have to ask my electronics guy about the how and why, I just remember the problem.

 

You need to remember that codes can just be a symptom and are often not the problem. I would try installing the actual knock sensor and see if the code goes away.

[JohnnyBlu]

The sender on the left as you look at the front of the engine; Red with 2 probes. My understanding is that the single probe one on the right is for the 240sx temp gauge in the OEM instrument cluster.

 

In this drawing it is F22.  This is the drawing I used as a reference when i made my harness up.

 

I tried dropping a 1.1 k phm resistor in place of the temp sender, and it had an effect. at idle I was now at 15.5 AFR but stable there. Something is going on there. I'll try to get an adjustable knob as suggested to play with it and see what happens.

 

 

 

It looks like there are a few more troubleshooting tests in the book that I can run on the distributor for the crank angle sensor.

 

I was getting code 34 with the stock knock sender, which is why I put in the resister. With continued codes i checked the harness side of it but things checked out. looking for correct supply voltage and continuity to ground on their respective pins.

[Stoffregen Motorsports]

I don't actually know where on the engine the two sensors are. I just wanted to be sure you knew there was a difference between the temp switch and the temp sender. Sound like you got it.

[JohnnyBlu]

Still no real progress on this, but further observations.

 

When the engine is cold, it will idle, but lean. Like 15 or 16:1 AFR. plugging in a potentiometer as a variable resistor in place of the temp sensor didn't tell me much. changing the resistance changed idle speed, but not AFR. odd.

 

what I'm seeing is that after about 5 seconds at a constant rpm, things start going more lean and it stumbles.Letting off the throttle and it will recover, with a slight increase in rpm before coming back down to idle. When I say idle, it's more like 1200 rpm. All of this is no load on the engine, car in the garage. On the 2 short drives I had on it, it would do well while accelerating, after you were going. Terrible to get going from a stop.

 

I have been trying to think of all the things that have an affect on the ECU controlling fuel and injectors.

 

-Crank angle sensor tells engine speed and piston location

-MAF measure air flow rate - more air = more fuel

-TPS sees throttle plate position to allow ECU to stay ahead of Air flow measurement. also cruise vs. hard accel.

-Temp switch - has some effect of fuel - should cause engine to go rich if engine is hot. resistance of sender drops as temp increases

-O2 sensor - maintain AFR. should fail rich

 

MAF tested out good. I checked voltage with car running and all things matched the manual.

TPS is new and calibrated per the book

Temp switch - I pulled and tested by pot of hot water with thermometer and volt meter in the kitchen. Resistance values matched the book, sensor passed.

o2 sensor - will try a stock narrowband 02 sensor tomorrow. right now I have an innovate wideband performing double duty.

 

All the IACV stuff works by pulsing open/closed and letting more air into the engine to maintain idle. What I am seeing appears to be the ECU reducing the amount of fuel being delivered.

 

other ideas? I'm banging my head against the wall.  I'm having trouble believing it could be the crank angle sensor sense it runs at all.

[JohnnyBlu]

Swapped in a stock narrow band one-wire o2 sensor. Problem seems to be gone. Always the little stuff. If i read the manuals correct, the wide band should be capable of sending the ECU the same 0.1-1.1v signal as a standard narrow band. Oh well. works now! Feels great to be driving it again.