AFR shooting up and down

[zetecspit]

I have a recent issue with the AFR gauge going from full rich to full weak. It could be the sensor, but with the lambda control off I can feel the difference as the gauge modulates.

Nick was witness to this behaviourat Castle combe, and the issue is MUCH worse at idle and low revs/load. 

But as I am sitting dying from manflu my little brain has been whirring away. I am wondering if teh fuel PRV is teh issue. I reckon I have the bits to check the pressure tomorrow (assuming I do a lazarus), It would explain a lot. 

Any other ideas? I intend to fit another oxygen sensor that I have in stock. 

[rogerguzzi]

3 hours ago, zetecspit said:

Any other ideas?

That sounds promising but what about the pump or filter? or a rubber fuel line that has gone soggy!

Roger

[zetecspit]

5 minutes ago, rogerguzzi said:

That sounds promising but what about the pump or filter? or a rubber fuel line that has gone soggy!

Roger

Thanks, that all has potential. Under high load the AFR was nicely buried around 12, and seemed sensibly steady (if that makes sense, it never went weak) 

Tomorrow I will rig up a pressure gauge, see what I get. Sadly Gill will be at work so I won't have an assistant. But I am sure I will manage.

 

[Nick Jones]

Hmm, no harm in checking fuel pressure etc, but the sorts of extreme end numbers we were seeing I don’t believe could be produced by a running engine, much less than one apparently running happily. If your wide-band controller has a reset and/or re-calibrate function, I might try that first and also check that there are no big air leaks around the sensor location.

It is true that it could just be going dead rich at idle and that dead rich can cause misfires that can show as lean on wideband….. but it would be running like crap and stinking of fuel and I wasn’t noticing misfires…..

[zetecspit]

I am an idiot.

And impulsive. So I popped to the garage and changed the fuel filter. Err, that was a few years old, I can't remember changing it before.

Sarted teh engine, seemed reasonable but cold so a bit misearble as usual. I then fitted a ghetto fuel pressure gauge made from T and a spare pressure gauge for my compressor. Started, and sat bang on 4bar. The ford regulator is 3.8, but importantly it wasn't wafting around at all. 

I have come inside to have a cuppa and declag myself, then will take the car for a spin. If it is fixed by the filter I will owe Roger a bottle of something. 

If not, change of sensor is next, but that requires delving under a car, and I really don't feel up to that today.

[rogerguzzi]

49 minutes ago, zetecspit said:

Started, and sat bang on 4bar. The ford regulator is 3.8, but importantly it wasn't wafting around at all. 

Hello All

             I have leant something new as Spitty runs at 3bar(40psi) and drips a bit with manifold vacuum!

I foolisly thought all modern EFI was the same on pressures?

Roger

[Nick Jones]

Assuming the pressure regulator has a vacuum connection to the manifold it should be maintaining its rated pressure relative to the manifold pressure. This is to maintain a constant differential pressure across the injectors. 
 

More recent stuff doesn’t bother with this and just has a fixed pressure, relying on the ECU mapping to maintain control.

[JohnD]

We will await the test drive, but is the AFR meter connected to an EMU that control fuelling?   Because it's normal for them to 'hunt', as this allows the EMU to get to a near perfect Stoichiometric mixture, faster.

https://www.researchgate.net/figure/Lambda-sensor-signal-for-40-km-hour-and-3-rd-gear-110-gram-silica-formed-The-red-lines_fig6_321759631

John

[SpitfireBGT]

Hi all,

How was this resolved? I'm having the same issue. The AFR is all over the place from rich to lean. The ECU isn't adding or taking away fuel according to the PWM gauge. I have no exhaust or vacuum leaks. 

Also I have a hunting idle that I can't tune out. But if I turn EGO control off by turning the authority to 0% it gets better.

I have a 10 year old AEM controller that I suspect to be the problem.

Thanks

[Nick Jones]

Based on my recent experiences, my first question would be - how good is your ITB balance?

Especially if things seem to stabilise at large throttle openings……

[SpitfireBGT]

I got that dialed in really good. I started with the ball bearing method to get me close. To fine tune it I used a MAP sensor, volt meter, and a 5V power source to get them all the same value. I pulled the vac line off of a TB and plugged it. Then put the MAP sensor on it to get a voltage reading. This was repeated on the other three then all of the butterflies were adjusted to get the same voltage in each one. That should have worked........ I think?????

I've got the car running pretty well with a lot of help from Roger. The Accell Enrichment has been giving me fits though. I get lean spikes as the car accelerates. 

I have never had nice smooth AFR curves in the data logs. Tonight I'm going to check how the controller is grounded. There isn't a separate ground for the electronics and sensor heater. I'm going to try grounding to the Microsquirt sensor ground instead of to the block. Does the heater draw enough current to fry something in the ECU?

[spitfire6]

On 5/9/2023 at 8:47 PM, JohnD said:

We will await the test drive, but is the AFR meter connected to an EMU that control fuelling?   Because it's normal for them to 'hunt', as this allows the EMU to get to a near perfect Stoichiometric mixture, faster.

https://www.researchgate.net/figure/Lambda-sensor-signal-for-40-km-hour-and-3-rd-gear-110-gram-silica-formed-The-red-lines_fig6_321759631

John

Hi John,

The trace is for a narrow-band sensor. Most aftermarket is Wide-band now. The hunting is a requirement of NB Sensors & is quite normal.
Cheers,
Iain.

[spitfire6]

5 hours ago, SpitfireBGT said:

I got that dialed in really good. I started with the ball bearing method to get me close. To fine tune it I used a MAP sensor, volt meter, and a 5V power source to get them all the same value. I pulled the vac line off of a TB and plugged it. Then put the MAP sensor on it to get a voltage reading. This was repeated on the other three then all of the butterflies were adjusted to get the same voltage in each one. That should have worked........ I think?????

I've got the car running pretty well with a lot of help from Roger. The Accell Enrichment has been giving me fits though. I get lean spikes as the car accelerates. 

I have never had nice smooth AFR curves in the data logs. Tonight I'm going to check how the controller is grounded. There isn't a separate ground for the electronics and sensor heater. I'm going to try grounding to the Microsquirt sensor ground instead of to the block. Does the heater draw enough current to fry something in the ECU?

Hi,
When I had sensor grounding issues, my CLT and TPS had loads AC. I unplugged the main connector, and then, after replugging it for testing, the fault disappeared. All power and low-level grounds are connected correctly. 

I think the heater is 2.5 amps.
Cheers,
Iain.

[SpitfireBGT]

Unplugging it then plugging it back in sounds like witchcraft but I'll try anything.

The Inovative controllers have a separate ground for the electronics and heater. My AEM shares a common ground. Do you think the MicroSquirt can handle 2.5 amps going into it's filtered ground?

[zetecspit]

On 9/4/2024 at 1:55 AM, SpitfireBGT said:

Hi all,

How was this resolved? I'm having the same issue. The AFR is all over the place from rich to lean. The ECU isn't adding or taking away fuel according to the PWM gauge. I have no exhaust or vacuum leaks. 

Also I have a hunting idle that I can't tune out. But if I turn EGO control off by turning the authority to 0% it gets better.

I have a 10 year old AEM controller that I suspect to be the problem.

Thanks

Having prepped the car for the 10CR I noted the ITB balance was out.Again. 30 mins later with my Morgan Carbtune (brilliant bit of kit, and fascinating to see how adjusting one intake affects the other 3) I am resigned the linkages/setup are far from perfect especuially at idle.

On the run we were still getting cyclical readings. Turning the wideband control off fixed all that, getting very steady AFRs (at 4k it was twitching 14.6-14.8 but quite slowly, with wideband on 13.2-15.6!) Quite possible I need to learn how the wideband control actually works, and making adjustments may "damp" things down. 

However, I am turning my thoughts to a single throttle body/plenum setup. I have a suitable throttle body, and looking into plenum design. I will start a thread.