KPower Industries K24 swap

[Irace86.2.0]

Originally Posted by geraldjust

sorry for anyone that got confused about my video but let me shed some light into this.

The fuel level system is a hybrid system. The cluster reads the raw value (analog resistive sensor) level sensor. This lets the cluster know the rough estimate of fuel. To combat variations in the level sensing, (high G turns, going up hill, down hill, sloshing etc... ) the cluster also uses the fuel economy to determine fuel usage. this is a combination of a special engine parameter being transmitted. This goes along with the ABS to determine the fuel economy that then the cluster uses that with the combination of the analog sensor.

I already fixed this issue with my own products (translator modules), and helped "wiringspecialties" with this too. I will get in contact with both haltech and kindustries so see how i can help them fix this.

Gerald R.

Kpower seems to think Haltech can fix this with software only. Maybe they believe Haltech can modify the CAN integration software. Seems like this shouldn't require extra hardware, so I tend to agree with them. Regardless, I'm sure I can speak for everyone here that we appreciate the assistance.

It is still odd to me that the system seems to be accurate until 1/4 tank and then it jumps down to empty. Does the hybrid system take over after the tank reaches a certain level because, as you say, the level can be affected by external forces in ways that matter much more? Does it operate purely analog and then switches to a hybrid system where it uses the analog data and then calculates the output, or is it always hybrid?

[Irace86.2.0]

Originally Posted by geraldjust

you referring to the fuel pump controller?

Originally Posted by brzlegend

x2

Sorry if that was confusing. I meant it is from the Kpower installation writeup that gets bypassed. Because they messed with it, I had thought maybe there was a feedback system in some way to the main ECU for some type of signal, whether used computationally or something that effected the fuel gauge, but I see now from the wiring diagram that it isn't really an issue.

And yes, it is labeled the Fuel Pump Control ECU from the wiring diagram. I omitted Control because I thought it was superfluous.

[brzlegend]

Originally Posted by geraldjust

you referring to the fuel pump controller?

Irace is just repeating what Kpower calls the fuel pump controller since they call it a fuel pump ecu.
I was lost, but I guess Kpower was the one lost.

I guess I am still lost. lol

Weird they call it an ecu.
Electronic control unit?

[geraldjust]

Originally Posted by Irace86.2.0

Kpower seems to think Haltech can fix this with software only. Maybe they believe Haltech can modify the CAN integration software. Seems like this shouldn't require extra hardware, so I tend to agree with them. Regardless, I'm sure I can speak for everyone here that we appreciate the assistance.

It is still odd to me that the system seems to be accurate until 1/4 tank and then it jumps down to empty. Does the hybrid system take over after the tank reaches a certain level because, as you say, the level can be affected by external forces in ways that matter much more? Does it operate purely analog and then switches to a hybrid system where it uses the analog data and then calculates the output, or is it always hybrid?

Yeah i already send a email to kpower and il send one to haltech soon. Well depends it the part of the code is hidden then only they will be able to add the support as you said. But by the initial thought it might be that haltech will not know exactly what to add so il give them some help. same issue happened with ecumater. But they just blew me off saying they did not need any help... lol anyways..

So no. When you initially fill up the cluster will set that as the new max point and then decreases from there. the hybrid system is always there... it only resets with a power disconnect or a fillup event. Think of it as a compound problem. initial value is exact for the tank. but as your driving, you get better accuracy knowing how much fuel your actually burning. When looking at the code, you would notice that the when doing a full fillup doing some math.. the... "average number - total number (full fillup) = 75% " so it leaves 25% to be calculated via ECU usage. Its a complicated system but its necessary as the real world can make the sensors fluctuate like crazy haha.

[Irace86.2.0]

Originally Posted by brzlegend

Irace is just repeating what Kpower calls the fuel pump controller since they call it a fuel pump ecu.
I was lost, but I guess Kpower was the one lost.

I misspoke on a previous page, which could have been confusing. I corrected a statement and bolded the correction for transparency and clarity.

ECU is much more broadly used, if that is the confusing part too. Many controllers or modules are called ECUs. More ECUs and a ECM... By convention, ECU is electrical instead of engine because there are so many difference units controlling everything in the powertrain from the transmission to ABS to steering to batteries and motors in hybrids and EVs.

[brzlegend]

Neat, I have dual fuel pump ecu s

[Irace86.2.0]

Originally Posted by geraldjust

Yeah i already send a email to kpower and il send one to haltech soon. Well depends it the part of the code is hidden then only they will be able to add the support as you said. But by the initial thought it might be that haltech will not know exactly what to add so il give them some help. same issue happened with ecumater. But they just blew me off saying they did not need any help... lol anyways..

So no. When you initially fill up the cluster will set that as the new max point and then decreases from there. the hybrid system is always there... it only resets with a power disconnect or a fillup event. Think of it as a compound problem. initial value is exact for the tank. but as your driving, you get better accuracy knowing how much fuel your actually burning. When looking at the code, you would notice that the when doing a full fillup doing some math.. the... "average number - total number (full fillup) = 75% " so it leaves 25% to be calculated via ECU usage. Its a complicated system but its necessary as the real world can make the sensors fluctuate like crazy haha.

Hopefully you can help, and they can compensate

I'm still confused. If the gauge is entirely analog without CANBUS input, as we have it now, right, then it would just be working from full to empty, but it would be twitchy or variable as the car pitched and accelerated like you mentioned. I don't know if that is what these guys are experiencing, a twitchy fuel gauge. I feel like I have seen this in classic cars before. It seems like they would have mentioned this if it was happening. If it is not happening then what is smoothing out the signal if it is not calculated?

I don't know if the car isn't progressing past 1/4 tank only when the car is running, but if they turned it off and back on then it would show 1/8 tank or some true value before maybe failing to progress. I'm a little confused why 25% is left to be calculated. Is the accuracy of the sensor bad past 1/4 tank or what? Once the sensor reads zero, is the reason why the gauge suddenly drops to zero because any calculation multiplied by zero is zero, and maybe the fail safe up until then is to output 1/4 tank. Trying to wrap my head around it.

[Ebush]

Originally Posted by Irace86.2.0

Hopefully you can help, and they can compensate

I'm still confused. If the gauge is entirely analog without CANBUS input, as we have it now, right, then it would just be working from full to empty, but it would be twitchy or variable as the car pitched and accelerated like you mentioned. I don't know if that is what these guys are experiencing, a twitchy fuel gauge. I feel like I have seen this in classic cars before. It seems like they would have mentioned this if it was happening. If it is not happening then what is smoothing out the signal if it is not calculated?

I don't know if the car isn't progressing past 1/4 tank only when the car is running, but if they turned it off and back on then it would show 1/8 tank or some true value before maybe failing to progress. I'm a little confused why 25% is left to be calculated. Is the accuracy of the sensor bad past 1/4 tank or what? Once the sensor reads zero, is the reason why the gauge suddenly drops to zero because any calculation multiplied by zero is zero, and maybe the fail safe up until then is to output 1/4 tank. Trying to wrap my head around it.

Gauge isn’t twitchy at all. I fill up and it goes to full. Seems to progress it’s way down to 1/4 tank like normal and then just stayed there at a little above the 1/4 mark for a while. What sucked for me, was I pulled up to the post office and parked the car. Came out and it just wouldn’t start so I tore the car apart for over an hour looking for a wiring issue. I guess atleast it didn’t die on the highway or on Main Street.

[geraldjust]

Originally Posted by Irace86.2.0

Hopefully you can help, and they can compensate

I'm still confused. If the gauge is entirely analog without CANBUS input, as we have it now, right, then it would just be working from full to empty, but it would be twitchy or variable as the car pitched and accelerated like you mentioned. I don't know if that is what these guys are experiencing, a twitchy fuel gauge. I feel like I have seen this in classic cars before. It seems like they would have mentioned this if it was happening. If it is not happening then what is smoothing out the signal if it is not calculated?

I don't know if the car isn't progressing past 1/4 tank only when the car is running, but if they turned it off and back on then it would show 1/8 tank or some true value before maybe failing to progress. I'm a little confused why 25% is left to be calculated. Is the accuracy of the sensor bad past 1/4 tank or what? Once the sensor reads zero, is the reason why the gauge suddenly drops to zero because any calculation multiplied by zero is zero, and maybe the fail safe up until then is to output 1/4 tank. Trying to wrap my head around it.

haha not looking for compensation mainly. hey at least they listen compared to other ecu makers lol.

well 25% is just the number if you assume the "rough" variable (analog) analog value is X ... and the "fine tuning " (digital) variable is 0%. then added together it would be 75%.. Its just code basically that helps these things be like that. is complicated to explain ahha. But the symptom is basically running out of fuel at 1/4 tank (displayed). or the gauge going from 1/4th to 0/4 after you turn it off and back on. Its all due to the same reason. Older cars that were lets say.. 90s used to have a needle that has a very viscous lubricant that would slow down the movement of the needle. Thus very slow to react to stuff. Then early 00s cars would be analog to digital.. but they did not need to be so precise. just enough for visual reference. But as part of emission system got complex and more exact. the need to have a extremely precise amount of fuel is what is needed to comply with modern standards. you could always hookup the analog sensor to the haltech. But then you use the oem display :/

[Irace86.2.0]

Originally Posted by geraldjust

haha not looking for compensation mainly. hey at least they listen compared to other ecu makers lol.

well 25% is just the number if you assume the "rough" variable (analog) analog value is X ... and the "fine tuning " (digital) variable is 0%. then added together it would be 75%.. Its just code basically that helps these things be like that. is complicated to explain ahha. But the symptom is basically running out of fuel at 1/4 tank (displayed). or the gauge going from 1/4th to 0/4 after you turn it off and back on. Its all due to the same reason. Older cars that were lets say.. 90s used to have a needle that has a very viscous lubricant that would slow down the movement of the needle. Thus very slow to react to stuff. Then early 00s cars would be analog to digital.. but they did not need to be so precise. just enough for visual reference. But as part of emission system got complex and more exact. the need to have a extremely precise amount of fuel is what is needed to comply with modern standards. you could always hookup the analog sensor to the haltech. But then you use the oem display :/

My first car in 1998 was a 1961 Plymouth Valiant. The needle was bouncy…pretty sure.

Seems odd that 1/4 would lead to such a change. Weird. Im any way, hopefully they get it figured out.

[Irace86.2.0]

What is this for?

[captain awesome]

I have no clue either. I have been meaning to ask but figure the car hasn't exploded yet so can't be too important.

[Ebush]

If I remember correctly it was in the wiring box right? Maybe a new ECU mount? I have no clue. I forgot all about it and I’m now at just over 500 miles with the swap. For the most part I’ve forgotten I even swapped the car. It starts up like stock and drives like a stock car. But that’s exactly what I wanted. It does feel faster in the corners, but in a straight line it’s still slow. I imagine swapping an LS into the car makes it feel like a totally different car (I’ve had an ls swapped s2000, s14, and a few rx7’s and they kinda lost their original feeling). So I’m very happy that the 86 maintained it’s character.

[Irace86.2.0]

Originally Posted by captain awesome

I have no clue either. I have been meaning to ask but figure the car hasn't exploded yet so can't be too important.

They said it is a wiring harness bracket for the driver’s side of the engine. It mounts near the bellhousing.

The edges are sharp, and the wiring harness was easily zip tied at that spot, so I think I’m good without it. It would be too much work to try to install it at this point.