I have a 91 CRX DX. Since the last time it passed I did a DPFI-MPFI swap. All the parts minus the IM and TB were from and 89 Si. The IM came from a Z6 and TB from a B16. I also replaced the cheap non OEM cat back with the OEM one from the Si, installed a header and intake. All the wiring is correct and the car runs good.


The ECU shows no codes (I had reset after replacing the O2 sensor). Afer failing my first test I put a new NGK O2 sensor and Magnaflow cat (standard, not high flow). My old cat was wasted lacking even remnants of the ceramic material. We have an old Heathkit emissions tester that showed a drastic difference (positive) from these changes, reading far below the readings that have always proved to be passing Washington State emissions standards. This was in the evening, the following day, having changed nothing (still the same 87 octane), I drove the car about 20 miles prior to going in for testing and I failed worse then the first time through.


First test results:

HC (pass required)

Cruise: Limit 180, tested 249
Idle: Limit 220, tested 495

CO (pass required)

Cruise: Limit 1.3, tested 2.8
Idle: Limit 1.2, tested .86 (only pass)

CO+CO2 (not required)

Cruise: Limit 6, tested 14.2
Idle: Limit 6, tested 10.26

Second Test (new cat, new O2 sensor)

HC (pass required)

Cruise: Limit 180, tested 283
Idle: Limit 220, tested 393

CO (pass required)

Cruise: Limit 1.3, tested 3.82
Idle: Limit 1.2, tested .12 (only pass)

CO+CO2 (not required)

Cruise: Limit 6, tested 14.92
Idle: Limit 6, tested 11.02

Today we retested with the home tester and it is showing results like it did before I put in the new O2 sensor and cat. Replaced plugs..no change.


Any help will be greatly appreciated as this is my daily and the tags have already expired!!!!


Currently car does this: Start and drive around the block, park and test with Heathkit home tester. Results are good, on a scale of 0-8 (0 good, 8 bad, 2 or below allways passes WA DEQ) it will register almost 0 at idle and will peg on 0
at 2k rpm initially. Holding the rpm at 2k for more then say a minute or 2 the readings shoot up to 4+. From this point on, whether idle or at 2k rpm the readings are higher then before, say between 1 and 2 at idle and 3+ at 2k.


I have thus far gone through and tested, cleaned, adjusted or replaced every part I added/changed during the swap and a bunch of others aswell....very frustrating. The car has now been expired for 23 days.

Stuff already covered so you don't have to read the whole thing:

ECU: Good
FITV: There isn't one
EACV: Removed and cleaned, tested electronically
MAP: Seems to be working fine
FPR: Seems to be working fine, Idle 28-32 with vac, 38-41 without vac
ETC: Replaced
O2: Replaced
CAT: Replaced but misamatch inlet tube OEM, header has 2.5" collector. Result slight leak.
TDC: Good
TPS: Good, tested to spec
CAP: Replaced
ROTOR: Good
PLUGS: Good
WIRES: Good
AIR FILTER: Good
TIMING: Set to spec

No leaks on vaccuum hoses, gaskets or injectors.




Modified by cornflakes at 5:43 PM 9/23/2007

 

u sure ur timing is to spec? seems like most other stuff is in check, but that miss needs to be checked out. check out your timing.

 

Thanks for the reply. The timing is about 10 degrees advanced, that is with the dizzy at full retard, full advance it is about 10-12 degrees advanced. Checked the timing belt, it is not off by a tooth or anything.

I also just started checking sensors. EACV = good, TPS = adjusted to spec, FITV no present, PCV vented to atmosphere with plug in IM.



Modified by cornflakes at 2:26 PM 9/23/2007

 

There seems to be a bad switch, though I can't figure out which. With the home tester the car will show passing numbers, then without changing a thing it will roll over the the failing side of the tester. Often, the fan will kick on and the numbers will go to crap. Though there have been times the number looked good while the fan is running.

This is driving me nuts. I have checked about every sensor I can think of. Could it be a bad ECU?

Someone has got to have some more input on this, please help. Any suggestions are appreciated.


Modified by cornflakes at 2:00 PM 9/23/2007

 

On the distributor, if you take off the cap, their is a plate on the distributor, which the tdc sensor is mounted on. If your timing is off, but your out of adjustment, you can adjust that plate some. Their are three screws which hold it on, with slots.

I'd get the timing set correctly, then I'd recheck your readings. If their still south, then

 

I ran into these problems too. I failed 6 TIMES! I finally passed when:

I put an OEM cat on it and plugged up the IACV. I also got some " guaranteed to pass" junk from autozone. This isnt the best way to pass but hell, i passed. Also, I drove the car a good 30 miles before entering emissions and had it sitting at 3K sitting in line so the cat was nice and hot. Did the trick.

 

Thanks and thanks! I will try all things mentioned and appreciate the responses, keep em coming


and update: I have since last, tried 2 other known good ECUs, a PG7 and PM6 because of the intermittent issue I am having and neither made any difference.

 

ur running rich at idle..that means u have too much fuel not enough air....is ur air filter clean,fpr could be stuck on high pressure...there is a diaphram and a spring that regulate the fuel pressure under load and idle.

 

are sure it doesnt go idle then cruise?

 

If you have the dizzy retard/turned all the way and you are still 4 deg. adv. I would say you are off a tooth. Are you running the stock dizzy?

My book says your fuel pressure should be 35-37 with the vac hose connected and 35-41 with it disconnected. You are running the stock FPR correct?

If your fuel pressure is low you will be running lean and if you go too lean then you start to missfire. A missfire will raise the HC because it is not burning the gas.
Now if you are having missfires...CO is from poor combustion or partial combustion.

All this said I believe it is your timing.

 

co=anything over 1.0 or higher indicates a rich mixture... high hc (ppm) indicates a missfire.

 

Thanks for the replys, much appreciated.

On the test sheet it may go idle then cruise but the results are correctly corresponding in the post. It seems to have a rich condition during both idle and cruise.

I am running the stock FPR (stock to the 89 Si I got the rail from atleast). I have only gone as far as to remove the vac hose from the IM, no further testing of the fpr has been done (suggestions?)

Regarding pressure: At idle with hose connected 28-32, disconnected 38-41. These readings are from a cheap B&M gauge permanently mounted at the end of the rail.

Regarding the timing, though being a tooth off would make sense, checking shows that it is not. Also the car was perfectly timed prior to the DPFI-MPFI swap. The dizzy is not stock to my 91 DX but was to the 89 Si I got the majority of the MPFI parts from.

You can refer to this post: https://honda-tech.com/zerothread?id=1671903, showing the list of parts and items I had tested to address that issue.


Two other things I have just discoverd and am addressing are that the EACV's filter was very dirty. There is a minor exhaust leak around the header to cat connection due to ghetto rig of OEM diameter cat to the larger diameter collector. (any suggestions for cheap solution to the header/cat thing?)

Thanks again!!!!!!





Modified by cornflakes at 11:29 AM 9/22/2007

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by cornflakes &raquo;</TD></TR><TR><TD CLASS="quote">
Regarding the timing, though being a tooth off would make sense, checking shows that it is not. Also the car was perfectly timed prior to the DPFI-MPFI swap. The dizzy is not stock to my 91 DX but was to the 89 Si I got the majority of the MPFI parts from.
</TD></TR></TABLE>

So you have the 91 dx dizzy on an 89 si motor? Could you go back to the 89 dizzy?

Reason I ask is the Si motor's cam gear is leaned forward one tooth. It is different that the DX I believe and that would throw off the dizzy. On the Si motor's cam gear the word top is of course up, But the two lines at 9 and 3 o'clock should not line up with the head surface. It is tilted counterclockwise one tooth.

I went through this on my motor an 88 Si.

 

Thanks for the reply, that is very interesting.

The opposite actually. I have a 89 Si Dizzy on a 91 DX motor. So then would the condition remain the same? If so, by traditional means of checking it would appear I am not off a tooth when infact I am? If that is the case, I need to rotate the cam gear one tooth counterclockwise?

If true, that sure would have been a nice bit of info to include in all the DPFI-MPFI swap write-ups. But then, though the Si ECU and Si distributor with TDC would agree with each other wouldn't my cam/valve timing be off since mechanically it's still a DX motor?


Modified by cornflakes at 11:53 AM 9/22/2007

 

OK, so it's the si dizzy on the dx motor.

The cam and the dizzy on the si motor would be positioned different than the DX's because of the position that it relates to the crank's TDC. (Damn, did I say that?)

So, Can you use the dizzy from the dx? And the dizzy has nothing to do with a MPFI swap. That is why it is not included in the write-ups I believe.

I personally have not done any swaps yet. But I know the mechanics/ theory very good.

Can some one back me up on my theory?

 

yeah, dude. You need the proper dizzy and set the timing correctly.

 

Yes, that is correct. Si dizzy on a DX motor, it is required as part of the DPFI-MPFI swap.

The Si dizzy is included in the writeups, I was refering to the timing portion being excluded. The dizzy is needed as there are 2 extra wires (dx has 5, si has 7) which need to be run to the Si ECU, if I remember correctly for the TDC sensor. I was not able to find any ways around this (I looked everywhere trying to forgo the expense of the dizzy) so no, I cannot go back to the DX dizzy as far as I know.


Modified by cornflakes at 1:00 PM 9/22/2007

 

OK, so the SI dizzy is needed. I wasn't sure of that, but I did think about it.

This was stated before:

On the distributor, if you take off the cap, their is a plate on the distributor, which the tdc sensor is mounted on. If your timing is off, but your out of adjustment, you can adjust that plate some. Their are three screws which hold it on, with slots.
I'd get the timing set correctly, then I'd recheck your readings. If their still south, then

I don't see how this would change your timing of spark. It would change the TDC sensors position is all. Which would change your injectors timing. (Pause for deep thought)

OK, ouch but I think I got it. So in theory here we go on the Si motor the the dizzy is aligned up for the extra tooth turn, which the DX does not have. So to get the dizzy to time correctly you would have to change the rotor's position and the TDC sensor. Which is not possible. So to get the car timed correctly you have to grind out the mounting holes/ear's on the Si dizzy.

Can you compare the two dizzys? check where the rotor aligns up with the end that goes to the cam. I am thinking that they are off a little. If it is we found the problem.

Turn me onto the link so I can read a little more about it.

 

I want to see him pass.

Took me 6 times!

 

thesmogman: Ok, though the theory is sound, I run into a couple problems with that, the first is I unfortuantely misplaced the DX dizzy zince the swap (was almost 2 years ago) so nothing to compare it to. The second is the ears on the Si Dizzy only have just over 1/8" of material on either side of the current hole. The currrent hole alows for about 1/4-3/8" of adjustment already, which only accounts for a few degrees, there is simply not enough material to grind away to allow adjustment into spec.

Which link did you want to look at?

Thanks again!


Ponda Baba: Thanks for the bump!

 

Ok, with a bit of creativity to make what thesmogman was saying work, the timing is now spot on.

Thanks much!

I have updated the top of the post with the current status aswell as the list of things already addressed.

Thanks to all that have helped thus far, it is almost there, I am pretty sure it will pass the idle part now, just need the cruise part.


Modified by cornflakes at 5:43 PM 9/23/2007

 

Done and done. I stayed home from work this morning to get it tested and it passed. I will post results if anyone cares.

Unfortunately I could not afford to have it tested after each change as I would have liked to, to make this thread more helpful to others in the future but I will elaborate on everything I did and how it changed on the home tester when I get home from work.

Thanks to all that helped!!!!!!!!

 

Ohhhhhh Yeaaaaaaaaah!

Excellent.

And I didn't look at any link it was just theory. But I did hear of that being done when switching dizzy's out.


Modified by thesmogman at 11:18 AM 9/24/2007