I have a F20B in my 97 Accord with the JDM PCB ECU. The car runs like crap when it's cold for a while (like more than a couple hours) and I've also noticed that when I throw codes, if I turn the car off and turn it back on the CEL goes away. I am thining the ECU is somehow resetting itself when the ignition is off. The radio doesn't lose presets when it's off though.

This is among other problems with the swap...
Any thoughts? Wiring issues? Any pointers for where to start my search?

 

Check the backup fuse for the radio/clock/ecu.

 

checked it. the fuse is fine. any other ideas? or is it time to just start searching for random shorts?

anyone know if this might be a problem specific to pcb ecu's?

 

so I found out that it holds the trouble codes at least until I start the car again - I can short the diagnostic terminal and turn the key to ON to get them if I turn it off with a CEL. Enough time to diagnose anyways. Any suggestions from anyone on what might be causing an ECU to wipe trouble codes on startup?

 

Your loosing a constant 12v. Its loosing and not holding its memory.

 

Check the codes with the car on & the cel being thrown.. Simple...

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by pgpbubble &raquo;</TD></TR><TR><TD CLASS="quote">Your loosing a constant 12v. Its loosing and not holding its memory.</TD></TR></TABLE>

Are you saying I have a loose wire somewhere? A bad ground? How would I check for something like this?

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by mtber &raquo;</TD></TR><TR><TD CLASS="quote">Check the codes with the car on & the cel being thrown.. Simple...</TD></TR></TABLE>

Yeah. That works too. The only problem is I feel like the car drives crappy when it's cold or it's been off for a while because it seems like the ECU has to relearn the proper parameters... that's another main reason why I want to fix it.

 

Yes; Check the voltage to the corresponding constant 12V pins on the ecu harness.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by pgpbubble &raquo;</TD></TR><TR><TD CLASS="quote">Yes; Check the voltage to the corresponding constant 12V pins on the ecu harness.</TD></TR></TABLE>

Got any idea how to identify the pinouts on a JDM PCB ECU?

 

It would be the same as your stock ecu.

 

ok thanks!

 

I think I might be missing something here. Am I just testing for an open circuit at the 1,5 and 13 holes (yellow/black wires)? I have never done this before and the Haynes manual isn't that clear on how to proceed. If you could help out with a little more detailed instructions that would be awesome.

 

just found out today that I have some screwy wiring going on. the guy who did the swap told me that they routed the CKP pin back to the distributor to fool the ECU into thinking there is a CKP sensor when there is not one on the engine. This doesn't make any sense because the ECU and engine are both obd2b as far as I know, so this trick would only have been needed to put the obd2b ecu in and obd2a accord, i.e. with my old pre-swap f22b2 engine.

anyhow, any chance this wiring is the cause of my ecu problems?

 

Sorry haven't been around to check this thread out; I have my ASE electrical test this fall so I'm trying to help you out as best as possible without being there.

I don't see how doing the CKP fix would cause the ECU not to hold a constant 12v; but you never know with other peoples wiring; and thats the problem. Do you have any other problem when you shut the car off ex; does the clock hold memory? your radio presets?

Also; Just re-read all your posts; have you tried checking the resistance of the map;ect;iat sensors? they all factor into why your car could be running bad while its cold untill its warmed up. Ill get the specs tomorrow for you while im at the shop; and see if your resistance is where it should be; I'll try to find the constant 12v power wires to the ECU for memory also.

 

OBDII H22a 96+

Layout
A1-&gt;A11----Blank--C1-&gt;C10---D1-&gt;D5
A12-&gt;A24--Blank--C11-&gt;C22--D6-&gt;D12
A25-&gt;A32--Blank--C23-&gt;C31--D13-&gt;D16
Pin---Color ---------Desc.
A1----yel------------Injector #4
A2----blu------------Injector #3
A3----red------------Injector #2
A4----brn------------Injector #1
A5----orn/blu-------Secondary O2 heater
A6----blk/wht-------HO2S heater
A7----orn------------EGR Control Solenoid Valve
A8----grn/yel-------VTEC Solenoid Valve
A9----brn/blk-------GND (dig)
A10---blk------------GND (power)
A11---yel/blk-------PGM-FI Power
A12---blk/blu-------Idle Air Control Valve
A13
A14
A15---red/yel-------Evap Purge Control Solenoid Valve
A16---grn/orn-------Fuel Pump Relay
A17---pnk/blu-------A/C Clutch Relay
A18---gry/red-------MIL
A19---wht/grn-------ALT
A20---yel/grn--------ICM
A21
A22---brn/blk--------GND (dig)
A23---blk------------GND (power)
A24---yel/blk--------PGM-FI Power
A25---wht-----------Intake Control Solenoid Valve
A26---red/blu-------IAB Control Solenoid Valve
A27---grn-----------Radiator Fan Relay
A28---grn/wht------EVAP Bypass Solenoid Valve
A29---orn/grn------Evap control cansiter vent shut valve
A30
A31-
A32-
C1---lt grn/blk------Ref. Voltage for ATTS or TCM (not used)
C2---blu------------CKP Detect
C3---grn------------TDCP Detect
C4---yel------------CKP Detect
C5---blu/orn-------A/C Switch Signal Input
C6---blu/red-------Starter Switch Input
C7---red/wht-------SCS
C8---lt grn ---------K Line (scan tool)
C9-- -
C10---wht/yel------ECM Backup Power
C11-- -
C12---wht----------CKP GND
C13---red----------TDC M Side
C14---blk----------CYP M Side
C15---blu/blk------VTEC Pressure Switch
C16---grn----------P/S Pressure Switch
C17---wht/grn-----ALT Detect signal
C18---blu/wht------VSS Signal
C19 -
C20 - -
C21 - -
C22---brn/yel------IMO Immobilizer Code Signal
C23- -
C24 - -
C25 - -
C26 - -
C27 - -
C28 - -
C29 - -
C30 - -
C31 - -
D1----red/blk----------TPS Detect
D2----red/wht---------ECT Engine Coolant Temp. Detect
D3----red/grn---------MAP Detect
D4----yel/red----------Vcc1 (Power for MAP)
D5----wht/blk----------BKSW Brake Switch Detect
D6----red/blu----------KS (Knock Sensor Detect)
D7----wht--------------HO2S Detect
D8----red/yel----------IAT Intake Air Temp Detect
D9----wht/blk----------EGR Valve Lift Sensor
D10---yel/blu----------Vcc2 (Provides Sensor Voltage)
D11---grn/blk----------SG2 (Sensor Ground)
D12---grn/wht---------SG1 (Sensor Ground)
D13- -
D14---wht/red---------SHO2S (Secondary O2 Sensor Detect)
D15---wht/blu---------PTANK (Detects fuel tank pressure signal)
D16---grn/red---------EL (Detects ELD Signal)
Note: some of the pins listed for OBDII don't apply to te 96 and were only added in 97 but the layout is the same.


IAT/ECT RESISTANCE CHART
5 K ohms at 32 deg F
3 K ohms at 68 deg F
1 K ohms at 104 deg F
0.5 K ohms at 176 deg F
0.2 K ohms at 212 deg F

 

thanks man! that pinout diagram really helps a lot. I will try and check into it a little more thoroughly. I have checked MAP and IAT to be within spec. ECT had to be replaced and that is now also within spec. All this is in my other thread https://honda-tech.com/zerothread?id=2071809 - if you want to check it out. I am thinking some of my problems are coming from a faulty EGR valve. And possibly a vac leak from the IACV.

No problems with radio presets or anything. And it doesn't wipe the trouble codes from memory. It just turns off the CEL and runs a little better if I turn off power and turn it back on. Wierd right? Like maybe it goes in then out of limp mode or something? Possible I got a bad ECU? Not sure what USDM ECU I could get my hands on to really check it out... maybe a prelude one? Just wondering if the fuel maps would screw it up because of the smaller displacement (2.0L instead of 2.2).

I just got a obd2 scanner yesterday - and one thing I noticed on the saved freeze frame data is that the error happened with short term fuel trim adjusted - though it wasn't adjusted after turning the power on a second time. I am gonna do a driven test too and see where that gets me.

 

Try and source an OBD2 1997-2001 Prelude ecu; 97-01's use an OBD2A ecu just like the 96-97's on civics;integra's. Maybe accords?

 

I already have the obd2a one from my accord... I checked that one out and I don't think there is anything wrong with the jdm ecu... it just turns off cel's automatically and doesn't run any emissions checks. i think i have to get the car more mechanically sound before i can really say anything conclusive about the ecu.

 

What other mechanical problems are you having besides the IACV/idle problem? If you have any wiring questions/problems let me know.

 

it's only the vac leak... but it has been a real pain. the car actually drives fine otherwise. as per my post in the other thread - replacing the im gasket is next.

re: wiring. I appreciate the help - I may have to hit you up for that. I ultimately need to pass the car through CA emissions. I have been told by the guy who did the swap (he was full of **** and not reliable at all) that I could use the japanese ecu for CA emissions. now that I have a scanner, it seems very unlikely because the JDM ECU, while being wired OBD2, has no OBD support and doesn't run any diagnostic self checks for emissions. Given that, I am eventually going to have to get this working obd2 with a usdm ecu. Since I already have the accord ecu (97 accord, obd2a) i may just use that and do without vtec for emissions purposes. then, once the car is refereed, i can either go obd1 or aem ems and get my vtec back and start modding and tuning for real. i think that plan makes more sense than buying the prelude edu just to have it running legal and with vtec - since the prelude obd2a is just as untunable as the accord one, and it might be expensive.

anyhow - this comes back to me asking you for wiring help because I need to find a way to make my obd2a usdm ecu (which expects a 2-wire iacv) to somehow recognize and operate the jdm 3-wire iacv. if i can't do that, then i am preparing by getting all the parts to replace IM and all sensors with prelude stuff - but then I have to swap out the engine harness or do some other wiring changes to switch it back to a 2-wire iacv anyway. so either way, I will hit you up.

sorry for the novel of a post. thanks for the help!

 

Now for the OBD2 JDM IACV; its a 3 wire sensor compared to the USDM obd2 which is a 2 wire correct?

 

correct. i have a feeling that the 3rd wire is just a ground, but not sure. don't really have much in the way of auto electrical education.

What I am dreading is having to change both engine wiring harness and main (?) wiring harness, like where the pins go into the ECU and which pins go where. It seems I am stuck with this with almost any change I do.

Current setup: JDM ECU + 3-wire IACV
Possibilities:
1. USDM ECU + 3-wire IACV - engine harness stays the same, main harness has to change... though maybe engine harness can change and I could run the 3-wire IACV off of 2 wires?
2. JDM ECU + 2-wire IACV - engine harness has to change, but if I just swap it out for a stock engine harness, then there may be some problems about how the wiring mates up to the main harness, which is still configured for the ECU to run the 3-wire IACV.
3. USDM ECU + 2-wire IACV - engine harness can be swapped out for stock, and main harness has to be restored to stock with regards to the IACV.

It turns out I cannot actually just run a stock main wiring harness from the USDM ECU, I could do that to regain the proper wiring for the IACV, but I would still have to modify it to run the F20B engine. Apparently the JDM motors do not have a CKF (crankshaft fluctuation) sensor, and so you have to splice some wires (specifically CKP and CKF) to fool a USDM ECU into thinking it is getting that signal so it will run ok. (as a side note, if you have any ideas about installing a CKF sensor into the F20B engine, then I could effectively restore the wiring completely to stock and run just fine).

The guy who did the swap confirmed for me that this process was done according to rywire specs and he double and triple checked it... It would make sense if he was setting up the car to run on the USDM ECU, but since we were setting it up originally for the JDM ECU, that may be one thing that is affecting performance when running the JDM ECU, though I am not sure how. It doesn't seem to me that you would need to route signals to fool the ECU when the JDM ECU and engine come from the same car. I asked him about that and he stopped responding to my emails.

if you need info about what color wires are going where in order to help, I would be happy to provide it - and pics too if you like. I am going to Canada this weekend, though... so I won't have access to the car again to really examine it until probably tues or weds of next week. BTW, the JDM ECU has obd2b pin setup, so it is currently connected using a standard obd2a-obd2b conversion harness from rywire.


Modified by 97LXF20B at 12:16 PM 10/11/2007

 

Well most people don't really have the know how or the will power to sit and diagnose electrical problems; but its always been something I was interested in and I have ASE A6 (electrical) testing this fall; so hopefully all will go well with that.

But beyond that; There would be no reason to swap the main/cabin harness for running a jdm engine/ecu. Most JDM ecu's as stated do not have provisions for self emission checks; and most likely none for the evap system. As far as the CKF; I know on OBD2; the CKF is mounted on a gear thats mounted to the oil pump/timing gear and uses a Hall-Effect type sensor to read it. You said that the way that its wired up tho; is using the CFK bypass trick by fooling the ecu using the CKP; as shown by KATMAN. http://technet.ff-squad.com/ckftrick.htm

Honestly; I think the best bet for being emissions compliant and everything; would be to use the USDM 97-99 Prelude ecu. 00-01 Having the immobilizer; 99 might also. So try to source a 97-98 Base model prelude so you do not have to deal with ATTS DTC's. And swapping over to a USDM 97-01 Base engine harness; You could probably get away with using that harness if you just rewire the IACV to the 2 wire and run them to the ECU. Now if that continues to not work with using the above description/process with the CKF fix; you would then need to swap out the F20B oil pump/timing gear for one from a USDM 96-00 H22A1/4 using the CKF sensor and wiring it up that way.

 

This might also help; but this pertains to B-Series; but its the same concept with the F20B and the H22.

https://honda-tech.com/zerothread?id=859796

 

hey man... other people on HT are telling me that all 97-01 preludes have immobilizers. do you know any way for me to verify this one way or the other?