[FIXED] 1997 Honda Accord No Start

1997 honda accord EX 4CYL SOHC vTec 2.2L Auto. Stock. F22B1 Engine? 116k miles?

Hi. My car won't start and I'm here hoping for help.

It spins over fine and it tries. Just won't catch. It's Vrooom, then spins, then Vroom-vroom, then spins.etc. {Attaching a short mp3 so you can hear this for yourself. oops, the forum won't let me attatch an mp3. }

History:
-Aprox. three months ago, noticed oil puddle under car. Leak appeared to be coming from under the distributor. Topped off oil. Then cruising home about 50mph, car suddenly bogged down, losing acceleration. Gave it a bit more gas but made no difference. So I floored it and the car dropped into passing gear and zoomed ahead. No misfiring detected. Man, this car is fast! Everytime I backed off the gas it started to die, but I kept it running and made it to a safe spot. Popped hood for visual inspection. Found nothing. So... I fired it up again... it ran great. And I drove about 5 miles to home. No more probs.

-Ran great for the next few days, then acted up again. Alternating between bogging down, then surging, then running good. Didn't ever cut completely off or strand me, Googled symptoms and found that lots of people had this problem due to oil getting inside the distributor. I looked inside, however, distributor looked clean. I then purchased lucas oil stop leak, in hopes that it would stop the leak until I had time to address it. In 2-3 days the oil leak had stopped and the problem didn't resurface, so I fuggedabout it, until....

- day before yesterday: It was a hot afternoon and I was parked, waiting in my car (for about 15-20mins) with the AC running. I cut if off. I come back about 5mins later; it won't start. The battery too weak. I jump it off. It starts fine and I begin the drive home. About ~20mins of driving time later, it died going down the same road and in almost the same spot that it did ~3mos ago. Stuck in heavy traffiic, I tried to restart. It started and ran smooth for about 1/2 mile, then started bogging. Managed to make it to a parking lot: Inspected it. Removed distributor cap. Contacts were dirty but there didn't appear to be oil inside. Cleaned contacts. Fired it up. Wallaha, it was running smooth. This time I made it about 1 mile and barely coasted into a better parking lot where I could take time to investigate.

-checked for gas: loosened the bolt on the fuel rail, and gas leaked out. Turned the key to activate the fuel pump and fuel sprayed out.
-checked for spark. pulled plug wire, inserted old spark plug, Got good spark. Checked two others this way. All good spark that remained all the while I was cranking.
-checked breather. Looks almost new. Anyway, just in case, I pulled the breather tube off the throttle body and tried to start. No luck.
-sprayed starting fluid into the intake through the power brake line - quickly reconnected it to booster (so no vac leak), and it started, ran really rough, backfired, then died after about 3 secs.
-looked for loose or bad electrical connections. Tried jiggling the wires to the cam sensor. Tried jiggling all the wires.. LOL
- Tried jiggling the key, & smacking the steering column, in hopes that it might be the ignition switch. Jiggle jiggle smack jiggle jiggle
No luck.
-Finally, the battery died. I Gave up.

Had it towed home...
-Threw battery on a charger.
-Rented OBD II Scanner. Put charged battery back in and realized I had cleared any codes by removing the battery. (Smacks self in head!) Tried to start a few times hoping to generate a code.
-Scanner test results: P1361 TDC Sensor 1 Noise.
-Ran battery monitor test too: Voltage PID 11.30v. Checked the battery with multimeter, key off- it read 12.65v. Also checked battery with the key in the on position: result 12.5v (I don't know why the OBD scanner reported low voltage). Also tried several times to check for ABS codes but scanner kept saying it couldn't connect. Ran OBD code again. Had no problem connecting for that. Why it could connect for the OBD but not the ABS - I don't know?

Right now I'm thinking that the battery dying is unrelated (unless somehow the alternator is shorting out the system?) to the starting/running problem and that it's either crank or cam sensor OR the distributor. But I'm guessing. Which is why I'm here.

Questions: Also, Could a blown fuse or a relay cause this? If so, which? Will pulling a plug tell me anything? Could the two bottles of fuel cleaner that I used in the last two tanks have clogged up the old fuel filter? (but then it is getting gas... and it wouldn't run right even with starting fluid, hmmmn..) Is there a way to test the suspected position sensors and/or the distributor using my multimeter, or some other affordable tool? If not, can I pull them and take them somewhere for testing? Or best just to replace them like a parts replacer? hehe (yeah, I'm trying to keep a sense of humor about this, but it's really not funny. I need wheels and soon!)

Can you advise me on how to proceed next? {Keep in mind that I'm a bankrupted political refugee with little funds - and I'm not kidding about this..}


Thanks!

Hey sir, welcome to the new forums! Also love the detail you included, great info to work with.

I think the testing you've done already covers a lot of ground. You've got spark at the plug wires and fuel pressure at the rail. Mechanical timing is possible but unlikely given how your symptoms were initially intermittent. For me, that leaves injection pulse, which may tie in to the P1361 code you pulled.

Doing some research, found that the TDC sensor is part of a combination CKP/TDC sensor on this engine, behind the timing cover. It's a somewhat strange setup, and I can post waveforms later, but it's not really important right now. Just know that the sensor lives behind the timing belt, and is a variable reluctance sensor (produces it's own voltage).

Do you have an incandescent test light? If you do, or can get a hold of one, we can pretty quickly figure out if you have injection pulse or not. Check out this ScannerDanner video , it shows perfectly how to do this test, far better than I could explain it. I'm gonna guess that you either have no injection pulse, or it's gonna be erratic/inconsistent.

Another simple check you could do would be to find and disconnect the TDC/CKP sensor connector, probably located somewhere near the timing cover, and ohm the sensor itself. It's not the ideal test, but you can do it easily with a cheap multimeter. The info I can find says the resistance of the TDC sensor should be 1800 to 2450 ohms. The TDC sensor wires are orange/blue and white/blue at the sensor connector.

Let us know if you have more questions, or have any new test results, and we can go from there!

Throwing this up for later reference.

Hello Tyler, and thank you Sir, for taking the time to help me and for the great welcome.

Here's an update: My gut was telling me that this really feels like a fuel problem so I decided to try again to see if spraying something into the brake line (intake side) would work. This time I sprayed Carb Cleaner (last time all I had was starting fluid) and to my surprise the car ran. It ran horrible, bogging down and trying to cut off, but it kept running as long as I kept spraying carb cleaner AND my helper kept pumping the pedal. Anyway, this changed my working theory to the possibility that my fuel filter was clogged or my fuel pump was weak.

So later this afternoon, I removed the old fuel filter and tried to blow through it and didn't get much air flow at all. Then I installed a new fuel filter, AND turned the key to prime the pump (did this 3x for 10sec. each time), then when I turned the key to start - my car started and ran (lousy) for about ten seconds before dying. I couldn't keep it running. However, this is new behavior - so something changed. I did this a few more times, but my battery started to lose power and I decided to call it a night.

Tomorrow or Tues I am going to rent a fuel pressure tester . Not sure how I will hook it up though since there is no schraeder valve.

Also wondering if this could be two problems since I am throwing a code for the cam/crank sensor...

I do have a test light but I didn't know you could test the pulse like that. Cool! Thanks for the link to that awesome video and thanks for helping me work through this.

Sorry for the late reply, sir!

I think your work with the carb clean definitely backs up our thoughts on a fuel delivery issue. However, I'm still not convinced it's in the actual fuel system, since you were initially able to loosen the banjo bolt at the fuel rail and got fuel spraying out when the pump ran. I've never personally seen a fuel filter cause a no start, but I see new stuff all the time.

I'm not saying NOT to check fuel pressure, I just know how tough it can be on this engine. You'll have to get a test kit that has banjo fittings and longer banjo bolts to get the pressure gauge in-line on this design. Have a good look at the kit before you leave the store.

My money is on having two separate problems, the battery going dead and the no start. I could be WAY off by associating the P1361 with the no start, we'll have to see!

Love that Honda PGM-FI main relay video, so much good info in there. If you get a chance, check for injection pulse and let me know what you find.

Did the test. Got no injection pulse to any of the injectors. Also checked to see if I had power to the injectors and I did have power.

Now I'm lost. What to do next?

A little about me that may help in guiding me. I have a great deal of past experience with old carbureted engines. Having done just about every repair imaginable numerous times (except for rebuilding trannys). Thus the SN of OldSchool. As a kid, I was hand-lapping valves at the kitchen table for my grandfather, and by age 17, I was completely rebuilding engines. Throwing them in my old hot-rod and driving across country (after a break-in period). I continued to own & drive a carbureted car until around 1992, when I sold it, and bought a newer car. And from then on I kept a newer car so I never had to do much except for reg. Maintenance, brakes, etc. Thus, explaining my lack of experience & understanding of modern auto electronics. Anyhow, you never know what curves life is going to throw you... But that's a story for your intro section. So... here I am. Anyway. Still Alive & kicking - which is more than I can say for many people that I've known.

And... I really do appreciate your help!

Thanks Man!

Alright, no injection pulse! Sorry, not trying to sound excited, but that explains the no start and why it runs on starting fluid. I'm not sure I'd worry about fuel pressure at this point.

I'm at work, let me get home and I'll let you know about the next step.

I forgot that you had mentioned how to test the sensor. I measured it.

Results : .52 with my meter set to "2k".

I guess this means, it's got 520 ohms of resistance? Which is way out of spec.

Couple Questions: 1. since this is a a combination cam/crank sensor would the range that you gave me still apply?
2. If so, I guess that it's definitive then, the sensor is bad?

Tyler wrote: Alright, no injection pulse! Sorry, not trying to sound excited, but that explains the no start and why it runs on starting fluid. I'm not sure I'd worry about fuel pressure at this point.

I'm at work, let me get home and I'll let you know about the next step.

Without a scope it is going to be tough for him to check these distributor signals. It looks like by your diagram Tyler that these are VR type sensors. If we knew which one was used for injection pulse, maybe we could try the bypass test on it and see if the injectors fire.
I never did this to a Honda yet but it is worth a try.
Thoughts?

To the OP here is a video I did on doing VR bypass testing. In the video, I am creating spark but it is the same concept for injection pulse.
Just make sure the key is on when you try it.

VRS Bypass Test

VRS Bypass Test

Tyler
How do I get the video itself to show up here and not just the link to it?

Wyatt explained it like this:

onyx wrote: how this works is:

Copy and paste the link without any addons. but if you want to control the size follow this:

you have the youtube link here:

Code:

https://www.youtube.com/watch?v=nfnkT0Iki8s

We need this part "nfnkT0Iki8s", its the part right after the v=. That is the video id.

once we have that we can go to the action bar at the top and click on the film strip icon.
In that icon we can choose youtube as the provider in the dropdown and paste that code we saved from earlier. also set the size to be in a hd aspect 500x280c works well. That gives us this:

and the code looks like this:

Code:

[video width=500 height=280 type=youtube]nfnkT0Iki8s[/video]

So...
if you right-click the video player and select "copy video url", you can just paste it into the text of your post.

Like that?

I agree, testing these sensors is gonna be challenging, but I thought a resistance test would be a good place to start given the tools available. I've been trying to figure out how these signals work together for timing purposes, and I *believe* the TDC sensor produces two pulses every 360 degrees of the crank. The CKP appears to be higher resolution, more teeth, not sure if has a sync notch or not. Finding a known good waveform for this setup has been tough.

One thing I can't figure out is why this Honda has spark but no injection pulse? This screams input problem to me, but still not sure about the CKP/TDC/CMP setup.

I love the idea of a bypass test! Though, I'm not sure if the engine would need to be cranking or not. I'd be inclined to try it both ways, with a test light on the injector control wire, to watch for results.

OldSchool, thanks for getting back to me on the TDC sensor measurement! 520 ohms does sound low, but let me recheck that spec when I get back to work tomorrow. Like Paul mentioned, it's not a smoking gun, but it may not bode well for that sensor.

The design of this sensor IS a little confusing, no worries. It's really just two separate sensors packaged together, using one electrical connector, reading different tone wheels on the crankshaft. The individual sensors themselves share no wiring at all, so your ohms measurement is valid as long as you're sure you were on the correct pins.

Hah, Noah got in first, nice!

Anyway OldSchool, let me know if you're interested in trying the bypass test, and we can help you set it up. You'd essentially be creating the TDC sensor signal yourself, using a test light, and seeing if you can make the injectors pulse. It's an amazing thing to watch when it works.

Woah Noah, I saw that graph, then started to read your post - and (for a moment) was thinking "what the heck is he talking about?" LOL

ScannerDannner, thanks for dropping in to help.

Tyler, yes, I'd def. like to do the bypass test - as I'm hoping it's not the sensor because I have no idea where I could do this repair. My apt. complex would kick me out. Plus, I can no longer tolerate being out in the sun all day. Which brings up what is prob. a dumb question. I'm wondering if there is some sort of affordable aftermarket ignition part or system that I could install so it could run without the sensor? If, it turns out to be the sensor.

Please instruct me on how to do the bypass test. Thanks Again!

Cool OldSchool, I'll get on and talk you through it later today. Sorry about my timing, work is stupid busy.

I actually had the chance to work on a '98 Odyssey with a 2.3L engine today. Got my scope on the TDC, CKP and CMP sensor signals, and it's like I thought. One cam pulse per 720 crank degrees, two TDC pulses per 360 crank degrees, and about 15 teeth on the CKP with no sync notch. Short version of all that is the TDC sensor looks like the critical input we've been thinking it is. I'll post up the waveforms later when I get home.

I am truly sorry if we end up at a TDC sensor problem Better to be sure about this call given how much work is involved in replacement.

Alright, got some info for OldSchool (and anyone else following along).

Got some known good waveforms of this setup, mostly for my own information. First up is the TDC sensor in green, and the CMP (in the distributor) in yellow.



Then we have the TDC sensor in green, and the CKP sensor in yellow.



I also wanted to get an idea of how the TDC sensor plays into timing, so I also grabbed a shot of the TDC sensor with an ignition pickup on the coil.



I don't think it's chance that the coil fires every time the TDC pulse occurs. I'm still somewhat confused how your Accord has spark, 'cause I'm not sure how the PCM can figure timing based on the CKP and CMP alone.

Anyway, onto the bypass test! If you haven't already watched it, that GM ignition module test is great stuff. Another video about a Saturn no start (with a bad crank sensor) is also worth watching:



Referring back to the diagram on the first page, you'll want to backprobe the orange/blue wire (sewing T-pins are perfect for this). Turn the key on, leave it there, and hook your test light up to B+. Rapidly tap the test light tip on the backprobe while listening for the injectors to fire, or maybe even spark.

If this doesn't work, don't panic. It may be that the engine computer needs the other two signals at the same time to fire the injectors, so have someone crank the engine over while you're tapping the test light on the backprobe.

Getting the injectors to fire in either case tells us that the computer CAN fire the injectors, and that it was missing the input that you're supplying with the bypass test. Let us know how it goes, or if you have more questions!

Sorry for not getting back sooner. Here's an update: 2 versions. One brief, another a rant. (Please ignore that if you just want the imp. details.)

Brief: Rented noid lights. I do have an injector pulse, but it seems to pulse erratically and with varying strength. Usually weak. Going to pull the main relay today and examine it. Then, to check some grounds - and then on to other things, if necessary.

Long Rant: No need to read on - unless you wanna read a rant!

I'm very frustrated with this vehicle. And I've been searching for answers and found many interesting (sometimes wild) online thoughts about the honda accord. For example: one guy on youtube says that if the car doesn't start right away that the computer senses low oil pressure and turns off the fuel pump and the injectors. I don't believe that, but.... I've seen lots of crazier things in my life. At this point, I'm not surprised by anything people do... Anyhow, if that's true, then it would invalidate my quick test of the fuel pump and possibly the injectors too.

In several other forums, I found others who said that the TDC/CKP sensor (for my mm/yr vehicle) should read between .5 & 1000 ohms - which would put mine in spec. Fortunately, I have a haynes repair manual now and am able to look it up for myself. Haynes agrees with you, Tyler!

I didn't feel confident in the injector pulse test that I performed using my test light. So I tried this on a friends vehicle - and I disconnected an injector (and I heard the engine start to struggle - so I know that it was injecting fuel) - and the test light didn't come on. Ever. I tried this on a couple more. Same results. Could there be something different about my test light that makes it not work with this test? This is a simple test. Don't see what I could've done wrong.

Having said all that, I'm grateful for any help given in this forum.

I'm really burned out on this vehicle and at this point I would take it to an honest (and experienced) mechanic in a heartbeat - if I knew one. I'm in my 6th decade of life (rapidly approaching my 7th), and I've only met 3 or so mechanics that I'm confident were honest. Which is why one reason why I can count the times that I've let someone else (since my late teens) work on my vehicle. Times when I just couldn't get around to doing it. Even then I usually ended up going over-their-work (tightening loose bolts, installing missing bolts, putting wires back like they should be, re-torquing, etc.) Then there's the "parts-replacers". I remember waaay back when I was working 7 long days a week and I told my wife to take her car to the shop because I had no time. The local shop said only a dealer could fix it. She took it to the dealer. It's a long story but they replaced 5 different parts/5 diff times - and never even touched the problem. Disgusted, I went to the dealership and told his mechanic that none of the parts he replaced could cause the problem it was having and said, "now you, stand there and watch as I fix this...." 15mins later, I went into the parts dept, bought a TPS. Another 15mins later - my wife's car was fixed. And to my surprise, on my way out, the mechanic hollered something to the effect of, "but she brought the vehicle in again, and we replaced X (another part - I don't recall what it was) - and you still owe $400 for that?" As I drove off, I hollered back, "sue me, you ______ idiot!" And btw, at the time, I had no idea that my wife had taken the vehicle to them that many times. She spent over $1200 bucks and on five occasions for unnecessary parts and still it wasn't fixed! No way they were getting more money out of me!

Anyhow, I'll stop there. And if I violated any rules by my rant. I apologize.

I'm very frustrated with this vehicle. And I've been searching for answers and found many interesting (sometimes wild) online thoughts about the honda accord. For example: one guy on youtube says that if the car doesn't start right away that the computer senses low oil pressure and turns off the fuel pump and the injectors. I don't believe that, but.... I've seen lots of crazier things in my life. At this point, I'm not surprised by anything people do..

I put VERY little faith in anything posted in online enthusiast forums. Obviously the same doesn't apply on THIS forum, but some of the advice people are giving and taking as gospel out there is scary.

I didn't feel confident in the injector pulse test that I performed using my test light. So I tried this on a friends vehicle - and I disconnected an injector (and I heard the engine start to struggle - so I know that it was injecting fuel) - and the test light didn't come on. Ever. I tried this on a couple more. Same results. Could there be something different about my test light that makes it not work with this test?

Depending on the on the car, some will disable a fuel injector when a misfire is detected to protect the catalytic converter. I could see this happening especially if you connected the noid light with the engine running. The way around this that I know of, is to connect the noid light, then start the car, and check for one or a couple flashes of the light. Generally, the PCM won't disable the injector on initial start up, so you should see some activity at first even if the computer is shutting it down strategically.

Anyhow, I'll stop there. And if I violated any rules by my rant. I apologize.

No apology necessary.

Hey OldSchool, no worries about getting back to us at all! Always happy to see members posting when they can.

You're right to calla BS on the whole low-oil-pressure-no-start thing, total crap. That whole myth comes from older Chevy designs with the oil pressure switch tied into the fuel pump circuit, but for redundancy instead of fail-safe purposes. That .5 to 1000 ohm spec... I just... I don't even know what to say, lol.

About your injection pulse test, I think you had to have gotten SOMETHING right, 'cause you noted that you had power at the injectors previously. Having power at the injectors AND fuel pressure also proves out both sides of the PGM-FI main relay.

Simple test or not, it's still tough to have confidence in the results until you've done the test a few times. Great idea to try it on another vehicle, by the way! There should be nothing wrong with your test light that will change the results, other than a problem with the light itself.

If you feel like revisiting the injection pulse test, here's another video that describes everything in more detail. The problem with the vehicle in the video is different from yours, but the methods are the same.



Don't feel bad about the rant, sir! It's all still true If you do end up taking this to a shop, I completely understand. We're still happy to assist if you want to take the diagnosis further!