Hondata Fuel System

[kefi]

Close loop operation with O2 sensor is known. You'll notice when you're idle air fuel will search, which is in control loop narrowing down on control. The important part of that control driving under load is is the fuel-learn (aka basic calibration) can be done. If it cannot, it will search under load as well.

I understand the overall operation of how closed loop works, but not how it determines 12 AFR vs 11 AFR without a table. I know the mechanisms various ECUs use to do it, I just don't know what this one in particular does, outside of the cat-protect operation.

Is that graphic for our ECU or just a generic graphic?

 

[kefi]

Also, to correct myself, the factory ECU doesn't use the cat protect at all. It was a feature in the ECU Hondata enabled specifically to richen it up more when you're spooling. It's 999C on the factory cal.

 

[ayau]

calm down satan, don't make us look that weak


The Bosch ECU has it's own way of determining the A/F target that isn't entirely understood. From factory it will target stoichiometric (14.7:1) near constantly until the catalytic converter reaches a certain temperature, at which point it starts using some static tables for fueling.

Hondata gives us direct control over the A/F by setting the catalytic protection temperature to 400F so these tables are used constantly once the engine has warmed up a little. That's why the tables have 'cat protect' in their name.

However, Hondata's basemaps all use the factory cat protect tables as-is.

The newer Hondata maps run a little richer up top. See 93 Hondata Fuel System map.

 

[Centripetal]

Ok, but what's your point? the sensor reads cooler after the end of a 4th-5th pull because I have it super rich

The topic at hand is the A/F ratio going all over the place and sometimes running lean because of it. Some people have stated they have to return to stock and disconnect the battery to get rid of it, which sounds like fuel-learn (Basic Calibration) is not being done.

What I already said:

I seem to recall reading somewhere that the fuel re-learn procedure will not complete if EGT is too high, which makes sense because the O2 sensor and catalyst are not at their ideal temperature range.​

I will be more verbose. The cat protection temperature is lowered in Hondata tunes, to enable some manipulation of the A/F. I am speculating that lowering the cat protection temperature results in the fuel-learn not being able to be complete. Some swing is going to happen while running close loop because the system has error. An A/F of 17:1 seems a little lean to me but would get you good gas milage.

I understand the overall operation of how closed loop works, but not how it determines 12 AFR vs 11 AFR without a table. I know the mechanisms various ECUs use to do it, I just don't know what this one in particular does, outside of the cat-protect operation.

Is that graphic for our ECU or just a generic graphic?

I think to explain this I would have to explain how a control system works. Generally speaking, you have a command (target A/F - what the table is for), a feedback sensor (O2) and the control system. If its real world, you have error (tolerances of the real world: manufacturing, variation gasoline, etc). A control system will take time "learn" how to compensate for these variations. For example, the fuel-learn procedure takes approximately 220+15 seconds. The values are then stored so the ECU doesn't have to take the 235 seconds (~4 minutes) to "learn" it every time the car starts.

Also, to correct myself, the factory ECU doesn't use the cat protect at all. It was a feature in the ECU Hondata enabled specifically to richen it up more when you're spooling. It's 999C on the factory cal.

Yes, that is already what I have stated.

Consequently, the factory ECU does have a "cat protect". As you notice in the diagram I previously posted, there is a "Lean Lambda limit", which increases rich as temperature increases.

 

[kefi]

The newer Hondata maps run a little richer up top. See 93 Hondata Fuel System map.

Minus that one. I would hope the fuel system upgrade map changes fueling

 

[kefi]

I think to explain this I would have to explain how a control system works. Generally speaking, you have a command (target A/F - what the table is for), a feedback sensor (O2) and the control system.

That's the thing, though. Other than these cat protect tables, we don't have a 'target A/F' table, at least not one that's available to us or seemingly even Hondata. I understand why and how it learns, as well as the general concepts of controlling systems over time with closed loop feedback. It's simply the lack of any fuel tables other than these that confuses me.

 

[AdamD19DFK8]

Here's a log where I just made a basemap with the calibration wizard where the only thing I really changed was ignition.

I'll def have to change the throttle table as it targets 510nm 0-4000rpm lol but fuel looks fine and the car drove normal with the egt disable mod checked. Though IAT gets higher than I'd like running it, didn't need to reflash to stock or anything. Dunno what the issue was before, I didn't try to replicate it just get a good tune going as times of the essences and a bunch of re flashing kills time.

 

[Centripetal]

That's the thing, though. Other than these cat protect tables, we don't have a 'target A/F' table, at least not one that's available to us or seemingly even Hondata. I understand why and how it learns, as well as the general concepts of controlling systems over time with closed loop feedback. It's simply the lack of any fuel tables other than these that confuses me.

There are three basic components that make up the A/F management.

Calibrated: All variations are accounted for and calibrated out by the control loop (aka fuel-learn complete). A/F target stoichiometric (14.7:1) under all operating conditions (Air Charge, RPM, etc). Just think of it of a table filled with 14.7 .

Lambda Instantiation(s): Based on an operating condition, a different instantiation of the A/F table can be used. For example, a warm-up operating condition would be temperature of the coolant and possibly catalyst. Warm up is generally rich because of poor atomization of fuel . Some ECUs just run rich when cold but the Bosch will operate in closed loop but target values in the table.

Lambda Limit(s):
Based on the engines operating point, different lean and rich limits are chosen. For example, during warm up you don't want to be too lean because it will go bang. Also, too much could cause a fire. Therefore, if a Lambda Instantiation's table(s) is being used and values are outside the limits, the targets will be limited based on the operating point.

If you have a good control system with a good mathematical model, it will still have error but you don't need at table. However, there are undesirable condition(s) that could occur, which can be avoided such as engine damage and pollution. Consequently, there are also conditions that could be desirable for example good fuel economy, where you might run bring in the efficiency of the ignition and lambda into the calculation and run the engine lean.

Here's a log where I just made a basemap with the calibration wizard where the only thing I really changed was ignition.

I'll def have to change the throttle table as it targets 510nm 0-4000rpm lol but fuel looks fine and the car drove normal with the egt disable mod checked. Though IAT gets higher than I'd like running it, didn't need to reflash to stock or anything. Dunno what the issue was before, I didn't try to replicate it just get a good tune going as times of the essences and a bunch of re flashing kills time.

I'm looking at my logs ...

and your logs ...

then I realized you might be tapping your foot to the beat of music. lol. It's totally you modulating the gas that makes it look like its all over the place. At least in the latest log that is.

 

[AdamD19DFK8]

Here's a log where I just made a basemap with the calibration wizard where the only thing I really changed was ignition.

I'll def have to change the throttle table as it targets 510nm 0-4000rpm lol but fuel looks fine and the car drove normal with the egt disable mod checked. Though IAT gets higher than I'd like running it, didn't need to reflash to stock or anything. Dunno what the issue was before, I didn't try to replicate it just get a good tune going as times of the essences and a bunch of re flashing kills time.

There are three basic components that make up the A/F management.

Calibrated: All variations are accounted for and calibrated out by the control loop (aka fuel-learn complete). A/F target stoichiometric (14.7:1) under all operating conditions (Air Charge, RPM, etc). Just think of it of a table filled with 14.7 .

Lambda Instantiation(s): Based on an operating condition, a different instantiation of the A/F table can be used. For example, a warm-up operating condition would be temperature of the coolant and possibly catalyst. Warm up is generally rich because of poor atomization of fuel . Some ECUs just run rich when cold but the Bosch will operate in closed loop but target values in the table.

Lambda Limit(s):
Based on the engines operating point, different lean and rich limits are chosen. For example, during warm up you don't want to be too lean because it will go bang. Also, too much could cause a fire. Therefore, if a Lambda Instantiation's table(s) is being used and values are outside the limits, the targets will be limited based on the operating point.

If you have a good control system with a good mathematical model, it will still have error but you don't need at table. However, there are undesirable condition(s) that could occur, which can be avoided such as engine damage and pollution. Consequently, there are also conditions that could be desirable for example good fuel economy, where you might run bring in the efficiency of the ignition and lambda into the calculation and run the engine lean.


I'm looking at my logs ...

and your logs ...

then I realized you might be tapping your foot to the beat of music. lol. It's totally you modulating the gas that makes it look like its all over the place. At least in the latest log that is.

Lol it's because I'm trying to replicate that coasting very light throttle lean condition where it sounds like the car has a burble tune (well gurgle with this exhaust) and feels like it's engine braking.

 

[AdamD19DFK8]

To those running this kit, have you been going all the way to the top with the 7200rpm limiter mod? Before I used it to be able to easily shift past or exactly at 7k.
Since it's enabled by default I've been taking advantage of it. 85mph in 3rd, 115 in 4th, 144 in 5th. I can officially say that this car has significantly more power and certainly torque everywhere. Explosive flat powerband.

It's really a shame that you can't data log the G sensors and individual wheel speed sensors on the 2020. I have logR installed but I hate Android auto. I'll be sitting in a parking lot and it won't let me text because I'm driving lol. But I am going to try it

 

[MoTeC R]

Mine is set for 7200.

 

[AdamD19DFK8]

Yeah, that's enabled by default, I'm just curious if anyone else is taking advantage of it and revving out past 7k. 4th gear will hit 116 lol

So during the installation of the fuel system in either set a rag, or had fuel spray out and into my intake. Enough that if I stuck my nose down there at the intake box I could smell fuel. Last night I pulled out the filter and cleaned it in a bucket of water and dish soap. The smell was gone from there but it at the car in the intake hose you can still smell gas, plus the rag I had stuffed in there while I had the intake out. I should probably pull the hose off too and clean it. I don't think there's to much of a negative effect of running the car with a hint of fuel smell since there isn't actually fuel in there. I figured it would have evaporated away by now.

I can say without a doubt this car is way faster than my 19 was. The other night I did a really good pull, in 6th gear and look down at the speedo and see 168. I debate on to keep pushing as far as I can go, but in the middle of the night on a pitch black stretch it gets really hard to see going that fast. B it's like warp speed. There was a corner coming up eventually so I just braked early for it. I had v so many great pulls along with this, car was running incredible, aaaaannd of course my data log corrupted. I still have the video of it as least. Usually on my dashcam things usually don't look as fast as I'm actually going. Not this instance, it looks like I'm flying

 

[AdamD19DFK8]

Even after flashing to car back to stock and then trying my uploads, the cars idle/coast behavior is still off. While the car sounds really good cracking/popping/gurgling, it's still annoying with the throttle because at low throttle input it will start crackling and I'll lose response in the pedal where it stutters and doesn't maintain whatever throttle I'm at.

Also sometimes, like this morning, when the car isn't warmed up it will throw a
P2422 code which is evap system vent shut valve close malfunction. Unsure if the 2 are directly related. At WOT and more than half throttle everything behaves normally and fueling is on point. It's just if I'm cruising/lightly building speed that my AFR will deviate all around 14.7

Also sometimes when I start up the car it sounds a little rough like it's struggling to maintain idle start speed. If I tap the gas right when I start it up itll idle fine. Or if I don't do anything at all it'll take about 5 seconds for the car to start idling regularly.

Last night doing some testing it didn't matter if I had egt reduction enabled or disabled like before how I was saying it was doing what I described with the mod checked.

Tune is a basemap from the calibration wizard where I only changed ignition tables, throttle tables, enriched high load/rpm lambda tables, and increased some values on the PA Temp table. It would do it even with the basemap with no changes.

I'm baffled at what it could be. Since the car behaves and performs like it should under load, I'm not super worried about it, but I still would like to get this resolved.

Any idea @Hondata ? You can see what I'm talking about in the logs I've posted previously. Here's another 2 logs from last night with it's calibration, thanks!





Calibration:


Don't remember if this calibration is from the first or second log, but the only difference between the 2 is the EGT disable mod.


Lolol since I installed this upgrade, I've had to fill up on fuel every single day this week. Big ol 16mpg

 

[kefi]

Even after flashing to car back to stock and then trying my uploads, the cars idle/coast behavior is still off. While the car sounds really good cracking/popping/gurgling, it's still annoying with the throttle because at low throttle input it will start crackling and I'll lose response in the pedal where it stutters and doesn't maintain whatever throttle I'm at.

I looked at your datalogs and it seems like your fuel pressure isn't responding normally for the first few minutes after you flash, in particular 0:35 to 1:35 on the second log. Even at 17% throttle you're not getting any more fuel pressure or it's delayed. This would definitely cause stuttering.

Compared to another FK8 that fully pressurizes practically every time they touch the throttle:

Another weird thing - your fuel pump duty swings rapidly. I don't think I've seen this before either.

Unfortunately I can't tell you why this is happening, but the ECU is seemingly commanding it. @Hondata will have to chime in from here.

Your trims are also pretty wild in both logs, especially the first.

Since the car behaves and performs like it should under load, I'm not super worried about it, but I still would like to get this resolved.

While most of your pulls are fine, time to time you've got swings most likely caused by MAF calibration. Note how the AFM voltage isn't swinging, TPedal is at 100%, but everything else is swinging. It causes you to rapidly go rich and lean multiple times per second.

This issue is very common in FK8s - people didn't realize how sensitive the ECU was to MAF calibration until recently.

 

[AdamD19DFK8]

I looked at your datalogs and it seems like your fuel pressure isn't responding normally for the first few minutes after you flash, in particular 0:35 to 1:35 on the second log. Even at 17% throttle you're not getting any more fuel pressure or it's delayed. This would definitely cause stuttering.

Compared to another FK8 that fully pressurizes practically every time they touch the throttle:

Another weird thing - your fuel pump duty swings rapidly. I don't think I've seen this before either.

Unfortunately I can't tell you why this is happening, but the ECU is seemingly commanding it. @Hondata will have to chime in from here.

Your trims are also pretty wild in both logs, especially the first.


While most of your pulls are fine, time to time you've got swings most likely caused by MAF calibration. Note how the AFM voltage isn't swinging, TPedal is at 100%, but everything else is swinging. It causes you to rapidly go rich and lean multiple times per second.

This issue is very common in FK8s - people didn't realize how sensitive the ECU was to MAF calibration until recently.

I used the calibration wizard to start out, so I selected all my parts like the HVI and titanium inlet. I don't know if it changed MAF scaling or not.

Any thought on sometimes getting a p2422 early in the morning on the cars first start up of the day? I've been bringing my flashpro to work this week just in case if I have to clear it. I've also gotten a p0560 which is system voltage malfunction.

I really appreciate you taking the time to have a look and hopefully @Hondata can weigh in on this