Failed CO emissions - all controls removed - where to start?

[Dave WM]

On ‎11‎/‎14‎/‎2019 at 10:34 PM, thetwood said:

 

For fun I decided to go check out my connections, Now just recently I noticed my normally super easy to start car has getting a bit iffy on starts, more cranking than I like lower idle once it does (not the AAR checked it). On a hunch and with all this talk about EVAP I decided to check the function of my carbon canister. I hooked up the mini hand vac to the purge side (the line that goes to manifold vacuum, the large one) of the canister and to my surprise it would hold no vacuum what so ever. If I understand this it should only open and allow airflow when the control vacuum is applied to the purge valve (small hose). So I assume my purge is stuck open, creating a pretty good sized vacuum leak. I don't know if this is new or always been this way, but you can bet I am going to plug it up tomorrow and see if it has any effect on the start and idle of the car.

Next up is IF the purge valve is indeed bad, what are my options? I does not seem to be serviceable (could not see how it is attached to the canister). But if it is stuck open I was thinking I could just ad a inline purge valve. Looking into this I find they are crazy expensive new, so the plan is to hit the JY tomorrow for some mid 90's era nissans that seem to have a vacuum operated purge valve.

[Dave WM]

ok found what may work for me and my inline solution a SP CP105, for a GM product mid 80's to mid 90's

I don't know if this is a problem but for 20$ its something to try. maybe OP can test his canister as well, just see if you can suck air thru the purge fitting on the canister. I assume it should not allow air to flow since there will be no vacuum applied to the control line (do this with the motor off).

[Captain Obvious]

10 hours ago, Zed Head said:

I don't think that anyone has actually measured.it.

You want to see movies, or just the summary?

[Zed Head]

Very nice.  But.  I hate picking on people's hard word work, but the lowest RPM shown is 2300 RPM.  Crack of the throttle should be down around 1000, I'd assume.

As far as effect, the biggest might be that the timing advance to full vacuum advance setting.  So timing probably jumps 18 degrees as soon as you crack the throttle.

[Captain Obvious]

Actually the whole point of my study was to determine where "on the engine load curve" does the peak vacuum (and hence, ignition timing) occur? The reason behind the project was I was adapting the throttle body from a Sentra onto my Z and I wanted to make sure I wasn't changing the ignition timing much by moving the vacuum advance characteristics.

So I drove around a bunch with a vacuum gauge teed into the ported vacuum source with the gauge on the interior so I could see it while I drove. I was trying to measure something "portable" between different throttle bodies so I could compare them. In other words... A measurement "standard".

My "standard" went like this:

"While driving flat and level, what speed would the vehicle be going if the pedal position was held constant at the point where the ported vacuum was at it's peak?"

Even though the absolute number does not matter, it gives me an objective way to compare port locations on different throttle bodies.

10 hours ago, Zed Head said:

the lowest RPM shown is 2300 RPM.  Crack of the throttle should be down around 1000, I'd assume.

As far as effect, the biggest might be that the timing advance to full vacuum advance setting.  So timing probably jumps 18 degrees as soon as you crack the throttle.

So, directly to your question... The vacuum peak is not right off idle. It's significantly deeper in the pedal than that. In fact, I found that in my car, I would be cruising flat and level on the highway in 5th gear at 65 mph when the ported vacuum was maxed out.

[thetwood]

Installed the carbon canister today. Two of the four mounting bolts were stripped out completely, so will need to figure that out one day. @Dave WM, I'll try to test the vacuum tomorrow. Everything was running fine after install and hooking everything up, so I'm taking that as a good sign.

I found a good timing light on craigslist and picked up last week. Finally got around to using it today. It has RPM as well, so was able to adjust that too. Timing was at 0 initially. I adjusted to 10 degrees, which bumped the rpms up. With a few adjustments I'm at 10 degrees and rpms are around 800 (ish). Sounds better than it did before.

Tomorrow, I'm going to swap the spark plugs out for the new and take it out to drive a bit (though my temp tag just expired... will need to be a bit careful of where I go...). See how the plugs look after an initial drive. I pulled one of the ones in now, which don't have much driving on them and they're very black.

[Zed Head]

Edit - responding to CO's post....

I can't tell what RPM that would be but I've wondered in the past about how much total advance might be seen under certain conditions.  I had a distributor with "11" weights and a vacuum canister that gave 18, I think.  So, 10 + 22 + 18 = 50 degrees.  If I bumped my initial timing up to 14, I'd be at 54 degrees maximum.  The centrifugal maxes out at about 2400 - 2500 RPM, commonly, so it's up there pretty quick.  The vacuum curves are in the FSM.  With your data you could probably put a map together, like the ones you use for programmable electronic control systems.

Edited November 17, 2019 by Zed Head

[Captain Obvious]

Glad to hear your making progress.

10 hours ago, thetwood said:

my temp tag just expired... will need to be a bit careful of where I go

LOL. Don't get busted. 

[thetwood]

Good news is that I didn't get pulled over. Drove 10+ miles getting gas and just enjoying having it out of the garage.

Car felt like it was running pretty well most of the drive, till the very end. After coming to a stop to pull into our neighborhood, the car didn't respond well finishing into home. As I'd give it a bit of gas it would sputter, then catch when I gave more.

After letting it cool down, here's what the spark plugs look like.

Also, don't think my temp gauge is working correctly, as it said I never went over 120 for the entire trip... Another thing to check. Just put in a brand new temp gauge sensor, but may need to check it more now.

Edited November 18, 2019 by thetwood

[Dave WM]

for ten miles that looks rich to me. you will need to determine what the actual temps are. IF its 120, then running too cool. Get a meat thermometer, remove the rad cap, put that in the rad, start the car, let it warm up. If the thermostat is working properly you should see it fully warmed up in 5-10min of idling. Just keep an eye on it you should see cold for a while then the T stat will open and the rad will suddenly get warm a the top. That would be the T stat opening. I don't think you are way off, just a bit too rich. Hopefully now its just down to fine tuning by correcting small issues. The ideal temp would be between 170-190, Getting the temp gauge to match up with what is going on should be next on the agenda. You really want that working right since its very bad to overheat the engine.

[Captain Obvious]

Actually for ten miles, those plugs look good to me. Maybe a tiny bit rich, but not bad. Remember this isn't a closed loop engine burning all nice and clean and stuff.

You do need to make sure the engine is getting up to temp though. Figure out if you have an instrumentation issue, or are you really running that cold?

 

[superlen]

My first thought on the plugs were that they looked pretty good too, but it's often hard to tell with a picture, and everyone's eye is calibrated a little differently anyway. I could easily be persuaded to slightly rich opinions. They certainly don't look far off.

I'll add that the stock ECU does it's best to maintain proper AFR during warmup, but the design lacks great accuracy in this mode & would tend to run slightly rich during this period, purposely. Given that your actual CLT temp may/may not be at 180...the plug readings are a little suspect.  +1 on what Captain said...ensure you actually are up to temp & verify your gauge isn't lying to you. You can then repeat the test, in theory with the exact same result, but you'll know for sure you were at temp, ? 

BTW, You may know this, but the stock ECU uses it's own sensor for CLT and does NOT share the same as what feeds the temp gauge, so don't get sidetracked thinking that the ECU was being fooled just because the gauge might have been reading too cold. 

Len