Performing Honda K24a3/AST5 6spd Conversion

Hussein, you are so close to the answer it will hit you in the face when I tell you the solution. Abarth swap has the same issue, solution, install the X1/9 idle up AC air valve. The ECU will see the change in MAP value and adjust the injectors accordingly. Feed the solenoid valve from the clutch wire of the AC compressor. On the Abarth it works so fast that you really have to pay attention to a very small hiccup in the idle.

TonyK.

Grimsby Ontario Canada.

Thanks Tony - I was hoping to do it without adding an auxiliary device, but I think I must follow your advice. I can't live with it as is.

Now I have to find it - I removed it years ago when I did the Volvo 240 Turbo CIS (constant idle speed) conversion

EDIT - I think I have some Volvo EVAP valves, and MAC valves used for waste gate control. One of those should do it also...
 
Last edited:
Well you could make an emulator that clones the CAN packet and put that on the CAN bus when the AC switch is pushed to allow the ECU to bring the idle up for the AC load, but connecting 2 wires of a DC solenoid and a bit of hose routing seemed to be a simpler solution from my way of thinking. Now if you like messing around, then you could find the climate control module from the Honda and use it to put the CAN Packet on the bus as well.


TonyK.

Grimsby Ontario Canada
 
Well you could make an emulator that clones the CAN packet and put that on the CAN bus when the AC switch is pushed to allow the ECU to bring the idle up for the AC load, but connecting 2 wires of a DC solenoid and a bit of hose routing seemed to be a simpler solution from my way of thinking. Now if you like messing around, then you could find the climate control module from the Honda and use it to put the CAN Packet on the bus as well.
TonyK.
Grimsby Ontario Canada

Hey Tony - yeah that would be much work. I have found another circuit for the power steering that bumps the idle through a pressure switch. Looking at the wiring diagrams I would say it supplies ground for the ECU E16 connection.

The ECU I/O chart says 0V on pin E16 at idle with wheels ahead, and Vbat with wheel turned to full lock at idle. PSP switch is wired from ECU E16 to ECU A24 (goes to signal ground on engine)

I assume that means open circuit at idle/wheels straight, then grounded via the PSP switch at idle/wheels at lock.

Does that make sense? I'm wondering if I can ground it through a relay coil, I assume a hard ground is not OK, or maybe it is since the ground circuit is through the ECU?

Screen Shot 2020-03-20 at 9.23.33 AM.png
Screen Shot 2020-03-20 at 9.15.12 AM.png
 
Well the way the switch is drawn it is normally closed which in electrical drawings means de-energized state or at rest, wheels straight. When turned it opens. ECU's normally put out + voltage and change in state applies a ground to bring the status to low. That being said, you now need an interposing relay to accomplish this.

TonyK

Grimsby Ontario Canada.
 
Well the way the switch is drawn it is normally closed which in electrical drawings means de-energized state or at rest, wheels straight. When turned it opens. ECU's normally put out + voltage and change in state applies a ground to bring the status to low. That being said, you now need an interposing relay to accomplish this.

TonyK

Grimsby Ontario Canada.

Just wrapping my head around it so I have to restate it - the E16 wire is live with the engine running, the pressure switch is closed, grounding the wire through a resistance (I assume).
Turn to full lock, the switch opens, E16 is no longer grounded, that causes idle to elevate. If so, that would mean I've had this condition from the start - I never connected that wire to anything. I would need to ground it through a resistance no? Otherwise it would be a dead short, no?

So far, I have gone through 2 new IACV and checking the diagnostic agian today, I fail the test at step 15, no Vbat at pin 3 - I only get 1.4v

I don't know where to look for another new IACV - Honda sells it for stupid money. Aftermarket, seems like the $100 + ones are no different than the $30 ones, just better warranty.

Screen-Shot-2020-03-20-at-8-58-19-PM.png
 
Last edited:
Well the way the switch is drawn it is normally closed which in electrical drawings means de-energized state or at rest, wheels straight. When turned it opens. ECU's normally put out + voltage and change in state applies a ground to bring the status to low. That being said, you now need an interposing relay to accomplish this.

TonyK

Grimsby Ontario Canada.

Tony, based on your knowledge, does this internal connection imply the E16 output is safe to run to direct ground through a switch(relay)? I don't have access to a PSP switch to test for a resistance value - but since one of the test procedures is to bridge the connector, I'm assuming that would work - I just don't want to fry my ECU

15848086756659164282813448380833.jpg


15848087371436345999807710121751.jpg
 
Looking at the diagram E16 on the left is showing a blocking diode and a resistor. I think if I am correct those componets are in the ECU. So applying a direct ground is all the switch does. But let's give this a bit of thought. I say that the input at rest must be grounded and when the switch goes open it provides a higher idle. You say you have a high idle without reason, possibly you have stumbled on the reason that the high idle is being produced because E16 is not grounded. I say ground E16 and see if idle goes to normal that would explain a lot of the issues you are having with the idle control valve. Possibly the idle control valve is doing it's job, the ECU is telling it that the power steering pump requires a higher idle because the circuit is open and not grounded.


TonyK.

Grimsby Ontario Canada.
 
Looking at the diagram E16 on the left is showing a blocking diode and a resistor. I think if I am correct those componets are in the ECU. So applying a direct ground is all the switch does. But let's give this a bit of thought. I say that the input at rest must be grounded and when the switch goes open it provides a higher idle. You say you have a high idle without reason, possibly you have stumbled on the reason that the high idle is being produced because E16 is not grounded. I say ground E16 and see if idle goes to normal that would explain a lot of the issues you are having with the idle control valve. Possibly the idle control valve is doing it's job, the ECU is telling it that the power steering pump requires a higher idle because the circuit is open and not grounded.


TonyK.

Grimsby Ontario Canada.

Many thanks for outlining it for me Tony :). I have a new Honda IACV on order, I'll ground that wire when I install the IACV.

I have found another issue - the connector I bought for the IACV is numbered backwards - meaning the ground and signal wires have been reversed all along. I reversed them. Went through the test procedure again, and still don't have the values I should have at step 15, and now the Check Eng light is on all the time with the 0511 code for IACV Circuit Malfunction.
Beyond that, resetting the ECU to eliminate the P0511 code caused another problem, now the motor won't rev, off idle it's going super lean. Contacted K-Tuner, not sure where to got with that problem. If it's not one thing, it's another. I'd really like to get beyond this phase :D
 
Last edited:
Engine management by ECU can be daunting. At least you have support. In the swap I did, because it is not a Fiat car, the dealer really doesn't want to get into it and I did have to spend some time with the dealer and their diagnostic tools that are for dealer eyes only. Engine management now in a word is complicated, but what it can do is beyond Weber DCNF's, camshafts and pushing the timing beyond spec. Today's engines can be just about bullet proof it is just understanding the parameters placed in the ECU on what it does, what it limits and then there is the unknown. An example of this was when I was probing some terminals on the Abarth, Key off and the fuel pump started running. What ever that ground placed on a wire did, for some reason started the fuel pump and kept it running. Probing again turned it off. You have to follow the OBD2 codes and eliminate what you can, pull the battery terminal to hard reset the ECU at times as well and keep working to resolve the DTC's. Being an electrician helps, I was trained as a trouble shooter so that helps as well, yet there are times I would just rather cut a section of metal and weld it, then stand back and look at it. Always keep an open mind, if you change something and things go south change it back, but realize that it too may be hiding another problem. In my car I extended some wires with a multi conductor cable and reversed at one end two wires. One being brown the other brown with a black tracer. There was never an issue until I reflashed the ECU and could not get the car to come off of idle. That took weeks to find, once the wires were swapped the ECU saw what it was looking for and everything was fine, there were no DTC's apparent so my scan tool did not show the issue. The dealer found the issue and said correct that and you see where it takes you, that was after 4 hours in their shop running through diagnostic trees. I trailered the car home and found my error and had it fixed in 45 minutes and all was well until the ECU hit a limit and would not allow more than 12 PSI boost, the engine would just cut out. The resolution to that one took months to sort out and it was all by trial and error, no DTC's posted, I understand now how to beat it. I suppose and I do not know your platform, you could first install a stock ECU and make sure everything works correctly and no DTC's are posted that are fatal. Then swap the ECU knowing that there are no issues that are under lying. Just one other point when working in this theater, only change 1 thing at a time because several changes can have you going down a rabbit hole chasing your tail.

TonyK.

Grimsby Ontario Canada.
 
Unfortunately I can't install a stock ECU - they all have an immobilizer feature that locks it to the chassis. I can try re-installing the Accord ECU with the K-Tuner board installed - of course if the problem is the K-Tuner user board the problem will just transfer over. I may have no other option though. It can't be from switching the IACV signal & ground wires. I also unplugged the IACV, so those wires are not even in an active circuit, and the symptom persists.

I've disconnected the battery, restarted the laptop & K-Tuner app, and reflashed the ECU without any change. I ended by leaving the batteyr disconnected & I'll see what's up tomorrow or Monday.
 
Unfortunately I can't install a stock ECU - they all have an immobilizer feature that locks it to the chassis. I can try re-installing the Accord ECU with the K-Tuner board installed - of course if the problem is the K-Tuner user board the problem will just transfer over. I may have no other option though. It can't be from switching the IACV signal & ground wires. I also unplugged the IACV, so those wires are not even in an active circuit, and the symptom persists.

I've disconnected the battery, restarted the laptop & K-Tuner app, and reflashed the ECU without any change. I ended by leaving the batteyr disconnected & I'll see what's up tomorrow or Monday.

Hi Hussein,

I preface my comments by pointing out that I have absolutely no knowledge of the idle control system in a Honda, so my comments may not be relevant at all ;). With that said, all the idle control systems I have ever seen are not simply "on / off" systems where the ECU supplies a binary signal but rather PWM systems where the amount that the valve opens is proportional to the duty cycle of the PWM signal from the ECU. The valves I've used in the past are the 2 wire variants similar to these (BMW part):

volvo_940_1995_idle_control_valve_oem_0_280_140_516_280140516.jpg

There are also 3 wire valves. There is a small vane in the valve that will open proportionally to the input signal (or more specifically the duty cycle of the input signal). You could use an oscilloscope to check the signal that is coming out of the ECU to validate this and also validate the operation of the ECU concerning idle control. If the idle valve is indeed PWM controlled than you could (as a troubleshooting method) use something similar to this to drive the idle control valve independently of the ECU.

In any case, having an oscilloscope record the ECU output (for the idle control) and how it responds under various conditions would certainly provide you with some valuable additional information. You can get portable oscilloscopes pretty cheaply these days (e.g. SainSmart DSO211).

Cheers,
Dom.

PS: This is what you would expect the PWM signal to look like (where the width of the square wave will vary):

main-qimg-e2d124568f8486e50c681eab5e6ed69a.webp
 
Hi Hussein,

I preface my comments by pointing out that I have absolutely no knowledge of the idle control system in a Honda, so my comments may not be relevant at all ;). With that said, all the idle control systems I have ever seen are not simply "on / off" systems where the ECU supplies a binary signal but rather PWM systems where the amount that the valve opens is proportional to the duty cycle of the PWM signal from the ECU. The valves I've used in the past are the 2 wire variants similar to these (BMW part):

volvo_940_1995_idle_control_valve_oem_0_280_140_516_280140516.jpg

There are also 3 wire valves. There is a small vane in the valve that will open proportionally to the input signal (or more specifically the duty cycle of the input signal). You could use an oscilloscope to check the signal that is coming out of the ECU to validate this and also validate the operation of the ECU concerning idle control. If the idle valve is indeed PWM controlled than you could (as a troubleshooting method) use something similar to this to drive the idle control valve independently of the ECU.

In any case, having an oscilloscope record the ECU output (for the idle control) and how it responds under various conditions would certainly provide you with some valuable additional information. You can get portable oscilloscopes pretty cheaply these days (e.g. SainSmart DSO211).

Cheers,
Dom.

PS: This is what you would expect the PWM signal to look like (where the width of the square wave will vary):

main-qimg-e2d124568f8486e50c681eab5e6ed69a.webp

e.g. this is from a Honda forum:

upload_2020-3-22_11-11-53.png
 
Personally when I have had these problems, I have started tracing wiring pin for pin, wire for wire. In the process I found this, if you have 1 wire rolled to a different terminal you know there is another. In my case 3 wires were rolled and the symptoms will not resemble anything in the car's DTC's trouble shooting guide. Remember that the manufacture's frame of mind is that they are starting with a correctly wired stock car and the issues that could present themselves will be shorted wire, open wire, high resistance connection. A miss wired terminal would have been proved at the factory before the wiring harness was installed in the car. So that is where I would start first, prove the wiring is correct.

TonyK.

Grimsby Ontario Canada.
 
Thanks guys. The Honda IACV is pwm, as noted in the schematic you posted.

Dom, the Honda version is neat in the sense that being integrated into the TB, it has no extra tubes/hoses required to attach it, and takes up very little space in the bay compared to the style you pictured (which is the same as older Volvo also).

What is also interesting in the article you linked to me is this: This would explain the jerky conditon when cruising with steady speed & almost closed throttle, fuel cut would make sense to explain that symptom. I will have to look at how those parameters can be manipulated.

Screen Shot 2020-03-22 at 9.48.16 AM.png


I had reversed the ground and pwm signal wire at the IACV plug, based on the connector numbering (not a factory connector), the ECU pinout I had checked pin by pin as I added wires, that was only for the I/O connector E. For connector A, where the IACV I/O resides, I only had to move two pins for the IACV and TPS due to the change from drive-by-wire TB to drive-by-cable TB. I checked & rechecked those two, they are correct. I will recheck the pinouts based on the '05 RSX schematic, vs. the '05 Accord I was using.


I don't know if sending the IACV battery voltage to ground at A12 signal input (through the coil) would fry the ECU, I don't see it though, as the engine ran fine (besides the idle/tip issue) under load until yesterday. I will need to revisit the A connector & make sure I didn't somehow dislodge any wires for the front O2, etc, however no error codes were set that would indicate a problem with anything besides the IACV :(

EDIT: found a thread on K20a.org with similar 'flat on face' symptom - his was caused by TPS - I'll have to recheck the signal & TPS response (% can be viewed in K-Tuner) & make sure mine is operating as intended.

IMG-20200321-125954.jpg


Screen-Shot-2020-03-22-at-5-00-42-PM.png


Manual confirms wire order shown above

Screen-Shot-2020-03-22-at-6-12-05-PM.png


I gutted one of the dead IACV & added a sealed shaft connection that allows me to manually set the air bypass. I used this, however it requires setting in one position for start & warmup, then closing for driving. After fixing the AC, it requires a further change in position. That is a PITA and not acceptable for daily use obviously.

IMG-20200307-182242.jpg
 
Last edited:
Always keep an open mind, if you change something and things go south change it back, but realize that it too may be hiding another problem.
TonyK.
Grimsby Ontario Canada.

Rechecked my wiring again. Everything is as it should be, but it doesn't register throttle position - I can see it in the live data. So, I followed the above advice, and with the signal/ground back the way I had them, the engine starts & runs as it did before, in other words normal as it was. I don't now what to do next, as this is not as it should be.

I grounded the E16 wire, and I think the idle is more stable, but I didn't drive it much. Still having the bucking off idle, so I need to investigate the Fuel Cut settings.

This is perplexing, as according to the wiring I definitely have the IACV wired backwards in terms of ground and signal I/O. I don't want to install the new Honda IACV when it arrives with the wiring backwards.
 
Last edited:
Rechecked my wiring again. Everything is as it should be, but it doesn't register throttle position - I can see it in the live data. So, I followed the above advice, and with the signal/ground back the way I had them, the engine starts & runs as it did before, in other words normal as it was. I don't now what to do next, as this is not as it should be.

I grounded the E16 wire, and I think the idle is more stabile, but I didn't drive it much. Still having the bucking off idle, so I need to investigate the Fuel Cut settings.

This is perplexing, as according to the wiring I definitely have the IACV wired backwards in terms of ground and signal I/O. I don't want to install the new Honda IACV when it arrives with the wiring backwards.
If the ECU does not "see" the throttle position, that could cause problems both with idle and with the fuel injector cut-off. Do you know how the throttle position sensor works? I know simple pots (variable resistors) have been used in the past, but would assume something more sophisticated is used now (something that does not wear like a pot does).
 
If the ECU does not "see" the throttle position, that could cause problems both with idle and with the fuel injector cut-off. Do you know how the throttle position sensor works? I know simple pots (variable resistors) have been used in the past, but would assume something more sophisticated is used now (something that does not wear like a pot does).

What's wierd is that I can check and set the TPS calibration in the software. With it properly adjusted it shows 0%. If I press the throttle to the floor it maxxes out at 65%. Don't know if that is an issue, I'll have to research it.

So, with IACV wired correctly:

With the key on, engine not running, the throttle % will sweep. With engine running, % doesn't not change, and engine runs like crap. No TPS codes are set

With IACV wired in what is the incorrect order according to the diagrams, throttle % sweeps the same with key on/engine off and accelerates properly when running. I can't se any point where the TPS & IACV wiring connects, it doesn't. Uses different signal and power supplies and the TPS grounds through the ECU unlike the IACV that uses a 'hard' ground.

I will have to go through the TPS fault tracing with the IACV wires as they are now, and then with the IACV wires as they should be.

Idle-Drop-IACVcovered-Frank-Tune.png


some of the tests ask for resistance values, so it must be a pot of some sort

Screen Shot 2020-03-22 at 7.51.46 PM.png
 
So, I looked again at my wiring diagrams, and of course I wired the TPS 5v and signal ground backwards also - I labelled them 1 2 3 in my sketch of the connector, but that's from the terminal side, not from the wire side. So, Both TPS & IACV are backwards.

IMG-20200323-094846.jpg


I will reverse both sets of terminals ( 1 & 3 on TPS) (1 & 3 on IACV) & see if the TPS symptom I created by reversing the IACV wires is corrected. If so, then I know I can install the Honda IACV with the corrected wiring when it arrives later in the week

ediT - corrected numbering
IMG-20200323-210100.jpg
 
Last edited:
So, I looked again at my wiring diagrams, and of course I wired the TPS 5v and signal ground backwards also - I labelled them 1 2 3 in my sketch of the connector, but that's from the terminal side, not from the wire side. So, Both TPS & IACV are backwards.

IMG-20200323-094846.jpg


I will reverse both sets of terminals ( 1 & 3 on TPS) (1 & 3 on IACV) & see if the TPS symptom I created by reversing the IACV wires is corrected. If so, then I know I can install the Honda IACV with the corrected wiring when it arrives later in the week

View attachment 30108

IMG-20200323-094906.jpg


your Sketches look better than my full-take-all-day drawings.

Odie
 
Back
Top