Performing Honda K24a3/AST5 6spd Conversion

Discussion in 'Workshop Forum' started by lookforjoe, Aug 11, 2019.

  1. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    When you are reading 65% TPS is the plate only partly open? We set ours with the Megasquirt, closed is 0% (+/-), open is around 100%. Linkage condition has some effect on all this, but if you are reading 65% at full throttle that's a big difference.
     
  2. Hey Hussein,

    is the diagram above the new (corrected) one or the old one? For the TPS it would seem to be correct, as it corresponds to this picture from the Internet:


    [​IMG]

    https://honda-tech.com/forums/honda...-honda-tps-wiring-calibration-thread-2830131/

    The yellow wire (+5v) will be left most from the terminal side if you were to orient the connector towards you (as per your diagram). Also found this which may (or may not) be relevant:

    https://www.clubrsx.com/threads/k24-swapped-rsx-idle-throttle-issues.821130/

    Also, I suppose you have seen this one already?

    https://www.crzforum.com/threads/ab...ld-guide-with-complete-wiring-diagrams.91819/

    Cheers,
    Dom.
     
  3. lookforjoe

    lookforjoe True Classic

    Hello Gene - thanks for the input. I don't know the answer to that - I have to get a helper to look at the throttle plate whilst I depress the pedal, or vise-versa :D Indeed, if I'm showing 65% at WOT, that is not good, and if it's not opening fully, that's also a problem. I'm assuming the pedal lever is too short & not giving me the throw I need.

    Modified vs stock - I may need to go back to the stock lever, depending on where the throttle plate is sitting at pedal fully depressed state.

    [​IMG]

    I spent several hours today messing with the throttle switch and wiring. The TB has to be removed for any of this, as the TPS is hard wired to the harness, and it cannot be adjusted without removal.

    I rechecked that the TPS wiring at the ECU is correct - A20 has 5v, A23 is the signal ground, and A29 is the signal. Continuity of all wiring is correct.
    IACV wiring is also correct - 1 = int. ground, 2= vBat 3 = A12 signal. Continuity of all wiring is correct.

    I rewired the TPS so that the power was #1 and signal ground was #3

    With that, I got P0122 error codes (throttle input signal low) just turning the key on, and the TP% in the logging window is reversed - % drops as throttle angle opens, instead of other way around. I checked the TPS signal both ways - and the pot works in reverse - which makes sense since in the fault tracing I posted it shows checking the reverse pins for K20Z1 vs. K20A3

    [​IMG]

    So, I removed it all again & rewired the TPS back the way I had it. So, back to square one, really.

    One thing I found going through K-Tuner looking for throttle related settings, is this page:

    [​IMG]

    I could have used the built -in AC and Fan controls by changing what the input signal is to bypass the CAN-bus signals. I can use the PSP switch to trigger the AC related internal calculations, idel adjut, etc.
     
    Last edited: Mar 24, 2020
  4. lookforjoe

    lookforjoe True Classic

    Hey Dom. Many thanks for the research!

    That diagram is the old one. I have reverted to that setup.

    The first pic & linked site seems to be for older setups - not "K" series.

    I'll re-read the 2nd one - the poster is not very clear in his remarks :D Does sound like what I was describing in terms of the alternate TPS wiring options, not symptoms, unless he's mischaracterizing the condition....

    The last one is more of a hybrid - I'll have to look & see if anything is relevant. My setup is straight K24A3 '05 TSX to '05 RSX ECU. The only wiring changes are to the TPS & IACV due to switch from drive by wire to drive by cable.
     
    Last edited: Mar 24, 2020
  5. lookforjoe

    lookforjoe True Classic

    So, with the pedal mashed to the floor, I have 80% throttle opening - shows 79% here

    [​IMG]

    with that, the throttle is actually wide open, it's a fraction off the hard stop

    [​IMG]

    So, the linkage is fine, just have to do whatever calibrations are necessary in the software for the fuel/ign etc., based on the MAP values, presumably, since I'm using a TB that is significantly larger than stock.

    I made a alternate AC clutch cycling circuit using the old Volvo/GM switch (1259519) from the 80/90's intended for that purpose

    [​IMG]

    It would replace this:

    [​IMG]

    It would simplify the install, as I could make one flex line from the frame to the compressor high sideand get rid off all this

    [​IMG]

    And added a relay for the PSP (E16) circuit that I'm repurposing for AC-on idle control. Relay removes the ground when AC is activated

    3 = grounding 4 (E16). 2 = coil ground. 1 = + from AC relay 87 , energizes relay coil, opening the 3-4 connection.

    [​IMG]

    Don't have relay sockets for this style relay. I saved the junction terminals from the Acura fusebox, and soldered wires to them, so it can be serviced if need be
    [​IMG]

    Mounted the relay on the backside of the EMS relay panel support

    [​IMG]

    also figured out a way to make the door stops fit snug in the rear strut tops - put strips of closed -cell foam under to take up the gap

    [​IMG]
     
    Last edited: Mar 26, 2020
  6. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    It may be worth resetting the WOT to 100% before you change anything else. It's possible that the lost 20% is at the fully open end and not really having an effect on idle. If it's spread evenly (most likely) then your idle settings could be confused by the throttle opening quicker than the computer thinks it is. Ex: computer sees 1% open, throttle is open 3%. Idle is fairly sensitive to small changes in plate opening.

    Edit/add:
    A larger throttle body (72?) makes the situation even worse. Even a small change in plate position may double the air entry opening. Our 72 on the Ecotec ran great above 4000, but even with a loose converter it had problems off and just off the line. It turned our that we had a lean condition -- 16:1-- for about 2 seconds after leaving the line. We fixed that but eventually went back to the 58.
     
    Last edited: Mar 25, 2020
  7. lookforjoe

    lookforjoe True Classic

    I don't see any way to do that - there is a TPS calibration, but it's only for closed throttle to establish 0%. I'll have to ask on the K-Tuner forum.
     
  8. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    I edited my last post added some info.
    Being able to set 0% and 100% is a basic necessity, at least for what we've seen on our 2.2 Ecotec.
    Can you remove your TPS and manually run it from closed to open to see if it runs 0 to 100%?
     
  9. lookforjoe

    lookforjoe True Classic

    Thanks again for the input Gene. The TB I have is 80mm - but the opening in the spacers and plenum are about 76mm, if I recall correctly. Still more than 10mm oversize.

    That however is a PITA - the TPS shaft key is wound up when installed, so with it removed there is no practical way to determine exactly where 0% starts. If I remove the TPS with the system connected, it sets an error code & locks the TP% in KTuner, which requires a reset to clear, then I'm back in the same boat. when I had the TB off & was rotating the TPS on the shaft to check range, I don't recall it getting close to 100% before it runs out of range. The manual says to check for 4.8v @ WOT, .5v @ idle. I can check that value at the connector later.

    I did realize that my KTuned TB has a removable bobbin wheel, so I can rotate it so that the plate will open fully & hit the stop without having to completely mash the pedal.
    Screen Shot 2020-03-25 at 8.38.18 AM.png

    That will only result in just over 80% reading though. I'm pretty sure the addtional air flow across the range is corrected for in the fueling and ignition maps, since the MAP is what governs those anyway. I will have to ask the tuner for confirmation on that though. They sell many of these TB's for K20/K24 setups, so there has to be an acounting for the increased volume relative to throttle angle.
     
    Last edited: Mar 26, 2020
  10. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    Can you datalog? That's a big help in determining what's actually going on.

    I would worry if at full throttle and 7000+ that your ECU is fueling at an 80% throttle reading.

    It's a bit confusing how you can set 0% without knowing if you are at 0.5v. Or is that part set? On ours the sensor fits on the end of the shaft, rotates with the shaft. At 0% throttle whatever value is indicated we tell the ECU and it accepts that as 0%. Same with full throttle.
     
  11. lookforjoe

    lookforjoe True Classic

    Thanks again for keeping me in my toes

    I checked voltages at idle and WOT - I have .6v at idle

    IMG_20200325_121220.jpg

    and 4.70v at WOT, so the values are good even if the % at WOT seems off. Honda spec is .5v idle, 4.8v WOT

    IMG_20200325_121246.jpg
    The system in KTuner won't set TPS 0% if it is out of valid range, so I can only presume it sets WOT based on TPS sweep.

    I found a loose ground - thought I might have found the problem with the IACV / TPS interaction - but no. Somewhere the IACV circuit must be messing with the TPS but for the life of me I can't find a common denominator that would indicate a possible backfeed situation such as a bad ground could do.. :(

    I also spent about an hour repositioning the bobbin and also dismantled and rebuilt the IACV, neither one of which made a damn bit of difference to either issue.

    Yesterday I did a fairly hard pull just cos I was annoyed at it. I haven't done any pedal to the metal pulls, this was as close as I got - in 2nd & 3rd. Didn't redline

    2nd
    [​IMG]

    3rd
    [​IMG]

    Looks like it's on the rich side, since LT trim is -2.54?
     
    Last edited: Mar 26, 2020
  12. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    Is there an interaction between the IAC and the TPS or are you looking for one that shouldn't be there?
    On ours apparently the ECU sees a fluctuation in the idle, adds air below the throttle through the IAC and adds fuel at the injectors. All of this has no effect on throttle position or TPS.
     
  13. lookforjoe

    lookforjoe True Classic

    I'm looking for one that shouldn't be there - there is no connection I can find in any diagram. I thought perhaps it was related to where I grounded the IACV on the intake, as I used the same location as the factory ECU ground - but rechecking the diagrams they ground logic/sensor and power grounds altogether at the same eyelet connector on the intake, so that can't be a potential backfeed concern.

    No matter what I have tried, unless I wire the IACV (-) and signal in reverse, the system goes loopy with any movement of the throttle. It sets a P0511 code (IACV circuit malfunction) within seconds of starting the engine, also. I may be going nuts for no reason, as perhaps this is all due to the faulty aftermarket IACV internals. The OEM Honda-packaged IACV will be here late today, so I will check values with that BEFORE I start & run it, and make sure the IACV is wired as it should be at that point. If the problem persists with a factory IACV & wiring as intended, I guess I'll just set fire to the damn thing.

    I had to add a intermediate harness since the TPS is hardwired in my install, no room for a connector at the TPS itself. I was wondering if adding that extra junction was causing voltage drop / connectivity issues. I've checked it (back-probed) with it running though, and all values have been consistent.

    [​IMG]

    Honda uses HX040 terminals for their connector - very small termninals - I used Deutsch for my intermediate connector, which are equally small. There really is no load to speak of with these components, so that should be OK. I'm thinking I will separate the IACV and TPS into their own 3-pole junction connectors & use 090 series connector for the IACV anyway.

    [​IMG]
     
    Last edited: Mar 26, 2020

  14. Hmm, that doesn't sound good. Given the number of potential points for crossed wiring, I'd recommend checking the continuity of the cable directly from the ECU connector to the sensor connectors to ensure everything is as you expect it to be. Ideally (if it were not too expensive) you could even consider buying a second hand factory wiring loom and comparing (by validating the continuity between the connectors for the sensors and the ECU connector). Also, if you were able to use an oscilloscope to measure the signals on each pin you will also get some more insight. I don't want to be negative, but I don't think the factory IACV will solve this. You really need a reference that you can compare with.

    Cheers,
    Dom.
     
  15. lookforjoe

    lookforjoe True Classic

    Thanks Dom.

    I did continuity check all the wires in question, All others I added I had checked at the time of installation.

    The factory IACV arrievd late this afternoon, so I plugged it in & ran the diagnostic checks again - this time everything passed - the main one was that with the key on, I didn't have vBat between A12 (signal wire at IACV) and ground. With the new one, I have the correct value.

    I installed the valve and started the engine. Idle was elevated as it should be for cold motor, and gradually dropped as it warmed up. Motor revved just fine without any funky business. YAY!!!!

    [​IMG]

    So, the bad valve was indeed the cause of all my woes.

    Just have to tidy up my wiring again

    [​IMG]

    Time to move on to other things...

    Modified the stock crank splash guard to fit - reshaped it out & away from the crank pulley.

    [​IMG]

    Made a protector plate for the fuel lines & filter, rad hoses and other bits from a Volvo splash guard..

    [​IMG]

    bolts to stock fuel pump splash guard plate, and to the hole in the other tank strap. Plastic-welded a separate piece back on the end for that.

    [​IMG]

    [​IMG]

    Had a couple drops of AF on the tarp under the forward engine area - so went looking for the source. Found that the Oetiker on the heater hose didn't grab properly on the backside, I could feel it 'bumped' out

    [​IMG]

    looking at the clamp - you can see it's not riding up over the hose at the back
    [​IMG]

    Removed the clamp & found a narrow Norma-style clamp, seems to be good now

    [​IMG]

    No bump out now. I've had trouble getting all the air out of the system, this could certainly have been why

    [​IMG]

    Also found that the vacuum line to the heater valve had pulled off at some point in my back & forth in the bay. I repositioned the fuel return line that had moved out & reattached the vacuum to the valve

    [​IMG]

    Took it out & drove it pretty hard for a short run



     
  16. Glad you finally found the problem. Congratulations.
     
    lookforjoe likes this.
  17. DanielForest

    DanielForest True Classic

    Location:
    Montreal
    Congratulations for your perseverance. I would have switched to carbs a long time ago... I'm not good with electronics...
     
    lookforjoe likes this.
  18. Excellent! I'm very happy to be wrong about the new IACV not fixing the issue :D
     
    lookforjoe likes this.
  19. NEG

    NEG True Classic

    Location:
    UK
    Well done that man!
     
  20. gene cooley

    gene cooley Autocrosser

    Location:
    Henrico, Va
    Glad you found a solution to the erratic idle.
    Sorry if my search for the TPS problem might have caused the vacuum hose problem.

    However, you know that I'm not giving up on the TPS. Having all of this info available means that there is ALWAYS something that you can make better or just fix.
    Looking at your data from above I see that the "65%" is still in the ECU. What's confusing is that in both the 2nd and 3rd gear pulls the throttle % is right at 65% (64 and 66%). The RPM is basically the same at that point in both gears. The AFR is 14.4 in 2nd and 10.4 in 3rd. I can't see what happens before or after these instants, but it would be interesting to see why there is that big a difference in the AFR's under what looks like the same conditions.

    Anyway, congrats on finding the idle problem.
     
    lookforjoe likes this.

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice