hey!
i have a 2001 prelude and i'm going to obd1 setup
imo i have to get the ecu jumper for the ecu conversion.
ok my Qs is do i need a resistor box for this kind of conversion sense obd1
has one on it

 

As long as your resistors match your injectors you are fine. Your 2001 has saturated injectors with the resistor on it. The ECU injector driver is the same.

You only need a resistor box if you change to peak and hold injectors.

 

i don't have a resistor box on my 2001 prelude

 

Exactly. This is because 97-01 Preludes have saturated injectors (i.e. no resistor box, they are on each injector instead).

As long as you match resistors and injectors you are fine. All Honda ECUs drive injectors the same way.

What do you mean by ECU jumper? A converted wiring harness?


[Modified by satan_srv, 10:37 AM 7/22/2002]

 

o im talking about the ecu conversion for obd2 to obd1 but still do i need to change my injector and plug in a resistor box inorder for the ecu to work fine.

 

I know. Ok I'll answer your question for the 3rd time. THe P13 (Prelude), P72 (GS-R), P28, P30...all of them drive/send signals the EXACT SAME WAY. Thus resistors must match only injectors.

Saturated Injectors: Have the resistor ON the injector, that's why there is no resistor box in your lude.

Peak-and-Hold Injectors: Not true pnh injectors. Have a separate resistor box, like on my 4th gen prelude.

You can keep your injectors as is. The other ECUs will drive them just fine. However if you change your injectors for some reason, you will need a resistor box IF they are peak and hold injectors. But I doubt you will be changing your injectors unless you need to, so don't worry about the resistors, they are fine.

 

ok fine
i have a friend that owns a hoopup shop and he has one coversion harness
i'll try it today with a p13 ecu
i'll let you know the out come

 

I'm going to disagree with you there. The way that peak and hold injectors trigger VS saturated injectors is completely different. The current requirements and signal durations are different. And the waveforms of the signals are completely different.

To the original question... On a OBD2 Prelude, in order to avoid rewiring injectors, you will want to use a P72 OBD1 ECU reprogrammed with H22a maps. You will need a OBD2->OBD1 convertor harness. And in my experience, you will get a crap-load of CEL codes unless you switch to a 92-95 Prelude distributor and re-wire it in.

Andrew

 

thank you! ANDREW

 

Yeah I've seen you post about this resistor/injector thing and I'm not sure you are correct either. Honda injectors are not true peak and hold injectors, there is not two seperate leads for the two different current levels. How can the ECU vary the current in the same wire with the resistors outside of the ECU? I know someone that can explain it better, hopefully he can post about it.

https://honda-tech.com/zerothread?id=187600

I have other links on it as well. Tell me how the ECU signal is different.


[Modified by satan_srv, 1:17 AM 7/23/2002]

 

Changing distributors are not required either.

Here's a discussion about it and some links to another PO.com thread about it

http://www.hondaprelude.to/forums/sh...&threadid=7358
http://www.preludeonline.com/showthr...threadid=43523

Actually it's funny aklucsarits you posted in the PO.com thread as well stating this stuff, unless you need to change the base timing, you don't need to change distributors. Like Dirtylude said, if you are going Hondata it doesn't matter.


[Modified by satan_srv, 1:21 AM 7/23/2002]

 

Trust me. I have tried this myself with a friend of mine. We got all kinds of error codes with a GS-R ECU and a P13 Ecu on two different OBD2 Preludes. One was a 1997 Prelude, the other was a 2000 Prelude. The results were the same. With the P72 and the P13 ECU, they gave us 3 or 4 CEL codes. I don't remeber all of them. But one was CYP sensor code. (The CYP sensor is near the crank on OBD2 cars, but in the distrubutor in 92-95 Preludes). The others I can remember were misfire codes.

My friend I was working with on this test/project with ended up working with DRT in NYC, and they could not figure out the CEL issue either. He also talked to Kenji quite a bit about it and he could not explain it either. The end fix was to get a 92-95 distrib and rewire the distrib harness to use the sensors in the 92-95 distrib. Some people, Dirtylude and Sun among them, say this works fine with the OBD2 distributor. On paper, it looks like it should work since the sensor signals and ranges should all be the same for OBD1 vs OBD2. But in my experience in our real-world tests, it did not work without the distrubtor replacement.

If you don't believe me, give it a try for yourself with the OBD2 distributor. Maybe you will have better luck with it than we did. If you get CEL codes with the new ECU, replacing the distrib is probably neccessary. You can use any 92-95 H22A or H23 distributor assembly complete with all sensors.

Andrew

 

It's not a question of belief, I'm sure you saw it happen like that. I just know Mark (Dirtylude) and trust what he says. Either way I will know within a week or two, once the harness is done. I'll post back in this thread and let you know.

 

OK, now lets talk about injectors and the difference between the way peak and hold and saturated injectors trigger...

This is a trace on an ocilloscope of the signal that triggers a saturated injector:

A saturated injector typically has an impedance around 14ohms. The voltage on the plot is 14v. Ohm's law = I=V/r, I=14v/14ohms = 1amp of current needed to drive this saturated injector.
So now lets look at the signal as the ECU triggers the saturated injector.... At point B, the ECU connects the circuit to ground, turning "on" the injector. The ECU maintains the grounding for the entire pulsewidth of the injector firing (the area labeled C). In this plot, the pulsewidth is about 4.2 ms. At point D, the ECU has disconnected the ground. An inductive kick occurrs as the voltage spikes as the magnetic field in the injector collapses. The saturated injector is now closed.


This is a trace on an ocilloscope of the signal that triggers a peak and hold injector:

A peak and hold injector typically has an impedance around 2ohms. The voltage on this injector plot is 14v, the same voltage as the saturated injector we just looked at. So I=V/r, I=14v/2ohms = 7amps of current needed to drive this peak and hold injector. 7 amps is a very large amount of current to be drawing from a low voltage source like the ECU. The injector driver cannot push 7 amps for a very long amount of time, so the signal for the injector pulsewidth looks very different on a peak and hold injector.
So now lets look at the signal as the ECU triggers the peak and hold injector.... At point A, the ECU again connects the circuit to ground, turning "on" the injector. This is the "peak" time, and it is independant of pulsewidth. This time, the ECU maintains the grounding for only 1ms - in this case only half of the entire pulsewidth of the injector firing. In this plot, the total pulsewidth is about 2.1 ms. At Point B, the ECU has reached the limit of current it can deliver. Current drops off dramatically and an inductive voltage spike occurrs, however, due to the low impedance of the injector, it remains open. Between B and C is the "Hold" period. Its length is determined by pulsewidth and controlled by the injector driver. If we wanted to have a 4.2ms pulsewidth like in the saturated plot above, we would have a 1ms "peak" and a 3.2ms "hold" time. After point C, the ECU has disconnected the ground. The voltage spikes as the magnetic field in the injector collapses. The peak and hold injector is now closed.

So knowing that, what would happen if we drive a peak and hold injector with a satutated driver? Well, the current draw on the Saturated injector driver will be at least 7 times what the driver was designed for. The driver may be able to deliver that kind of current for part of the injector pulsewidth (it may be able to "peak"), but it will not be able to keep the injector open for the entire pulsewidth (the injector will close instead of "holding"). So what you get is the injector closing "early". How early? 1/4 pulsewidth? 1/2 pulsewidth? Who knows? You clould run a trace to figure it out. But you are still getting unpredictable fuel delivery with a bias towards running very lean - which is not desirable at alll.

What about if we have a Peak and hold driver, with saturated injectors? Some ECUs with Peak and Hold injector drivers are capable of driving saurated injectors as well. This works to an extent but sometimes not very well. What can happen in the "peak and hold driver with saturated injectors" scenario is that the saturated injector may only stay open for a fraction of the injector pulse width. It opens on the "peak" signal, but then due to it high internal impedance closes early during the "hold" transition when the ECU thinks the injector is still open. This can cause unpredictable fuel delivery - again, running lean is usually the symptom.

I am by no means an expert at fuel injection systems, but I feel I have a good grasp of basic EE concepts and I have done quite a bit of my own research on this subject for my own projects. All comments are welcome. But please back up your assertions with facts or mathematical evidence, not "My cousin's neighbor has been running peak and hold injectors with a GS-R ECU for 6 months and has not had any problems yet..." I'm very interested in learning more about this subject...

Andrew




[Modified by aklucsarits, 10:43 AM 7/23/2002]

 

I've seen these graphs and posts by you before, so I'm familiar with what you're trying to say.

I agree with the Sat. Injectors requiring 1-2 amps in a constant pulse, and I agree that "true" peak and hold injectors require 6-7 amps, and then drops off to a lower current.

However, how can the ECU change the amount of current given that in a Honda engine, all the resistors are either on the injector (Saturated) or in a separate resistor box (Peak and Hold)?

You are using ohms law in conjunction with the impedance of the resistor box to calculate current, which is correct. So, in the case of the peak and hold injectors: I=V/r, I=14v/2ohms = 7amps. In the case of the saturated injector, the resistor on the injector is rated at 14ohms or whatever so: I=V/r, I=14v/14ohms = 1amp. So would you not agree the the sole determinant of the current supplied to the injector is determined by the resistors? I think that is a fair assumption.

However, this begs the question, given that current is determined by the resistor box, how can the ECU reduce the current on the injectors for the "hold" period? You would need another set of resistors with an impedance of 14ohms or so to actually produce that current wouldn't you? Actually a true peak and hold injector would require a second lead with a different set of resistors, and the ECU would divert the signal to that lead for the hold period.

Since the ECU does not have resistors within the ECU itself, it seems the only logical conclusion is that the ECU sends one signal, at 14 volts for the entire pulse duration. The current is determined by the resistors, which are external to the ECU. Also, you are saying the ECU could not take running at 7 amps, how is this so? The amount of current the ECU could handle would be determined by the quality of wire/circuitboard, which I would believe to be able to support 7 amps of current. 7 amps isn't THAT much current, but I don't know the specs of the ECU for sure in terms of ability to handle current.

Anyways, this is why I'm thinking the ECU signal is always the same, and thus the ECUs are entirely swappable, as long as the injector/resistor combo is appropriate.

I would like to see an oscilloscope for a Honda Peak and Hold Injector, rather than the GM ones you are posting. I think they are quite different. I welcome your comments on the issue as well


[Modified by satan_srv, 11:15 AM 7/23/2002]

 

ok here the test result on my 2001 prelude
it gave me code 9 and some other staff the idle went crazy (1000-1500)
i've tried adjusting my idle screw but it didn't work.
i think i have to rewire my dist. and get a peak hold injector to compensate with the ecu or something

 

Interesting, code 9 is the CYP sensor... http://www.hondaprelude.to/main_checkengine.html

This should not have anything to do with your injectors.

How did you wire the conversion? Hondata? Skunk2?


[Modified by satan_srv, 2:51 PM 7/23/2002]


[Modified by satan_srv, 4:10 PM 7/23/2002]

 

Ah thank you! 9 is one of the codes we got. I'm not that crazy after all...
It has to do with the position of the CYP sensor. On the 97-01 H22a, This sensor is on the block down by the crank. On 93-95 H22As, the CYP sensor is in the distrubtor... That's why switching distributors and rewiring the appropriate sensors works...

Andrew

 

On paper it works and in practice it works. There not just one other person using Hondata and OBD II distributors on the Prelude. What ECU where you using? Very old and some older JDM OBD I ECU's have a totaly seperate trigger configuration with with a 16 tooth CKP rather than a 24 tooth wheel.

The CYP sensor is in fact the only sensor that has not moved to the crank. If you were getting a code 9 it was because of some other issue. It is in the same location as it always was and works exactly the same.

Here's a picture of the OBD II distributor:
http://www.higginstribe.com/gallery/...e=DSC00053.JPG
That single tooth wheel in there is the CYP sensor. It fires at TDC cylinder #3. This has not changed.

This is a picture of an OBD I distributor. You can see in order the CYP (1 tooth), CKP (24 tooth), and TDC (4 tooth) trigger.
http://www.higginstribe.com/gallery/rebuild/index.php?imageld-dissy1.jpg

The CKP and TDC triggers have moved to the crank because OBD II vehicles need far more accurate signals. They can't be affected by valve lash because the ECU detects misfires by looking for an variance in the trigger timing.

This is a picture of the TDC wheel and sensor on the crank. You can see it only has two teeth. This is because the crank turns twice for every cam rotation and any sensor on the crank is going to have half the number of teeth.
http://www.hondaprelude.to/images/tdc-1.jpg

This is the CKP wheel with it's 12 teeth. Half the number it would have on the cam.
http://www.hondaprelude.to/images/ckp-1.jpg

...and here it is on the crank with it's sensor.
http://www.hondaprelude.to/images/ckp-2.jpg

In short the triggers, even though there in seperate locations give exactly the same signals and are transparant to the ECU.

If you are doing this for a Hondata install I would talk to them directly about how to fix the Code 9 issue or what the issue might be.


[Modified by DirtyLude, 4:26 PM 7/23/2002]

 

...


[Modified by satan_srv, 4:28 PM 7/23/2002]

 

...and there's no OBD I misfire codes. Idle will likely be off. The P13 ECU is still used to 330cc injectors while the 97-01 injectors are 290. This shouldn't affect the whole fuel map alot, but idle is very sensitive.


[Modified by DirtyLude, 4:35 PM 7/23/2002]

 

i should get an peak and hold injector and obd1 dist. to get it running
thanks for the input

 

That's not what he's saying...the idle may be off a little but the code is from something else. Don't toss money away, distributors are $$$

What ECU are you using?


[Modified by satan_srv, 12:44 AM 7/25/2002]

 

Ah, sorry for the misinformation. Thank you for correcting me. I think I got it confused with the CKP... There goes my theory...

I don't beleive he's doing Hondata. Based on past posts, I think he has just a regular OBD1 P72... Can you confirm what ECU(s) you are testing with, WICKED?

Satan_SRV: Please keep us updated on your project. I'm very interested to hear the results of your testing...

Andrew



[Modified by aklucsarits, 10:33 AM 7/26/2002]

 

p72 hondata stage2