Mazda Project Idle Surge

[TheArete]

Alright so it’s been about a month working on this idle surge problem to little avail so I’m admitting I’m missing something that’s probably right in front of me.

Topic

[TheArete]

I can do a second round of testing, or maybe the gauge is miscalibrated? possible. It’s close enough to specs that I’m not worried considering this was a cheap gauge and the Dry tests are pretty close to 170. Misreading gauge (marked in 5psi increments) compounded with slight miscalibration could easily explain that.

In any case “betsy” most definitely does not have aftermarket anything other than a coat of rust. Lol. But I love her nonetheless. Don’t think that’s the problem though, too much compression just means I’d be getting maybe more backpressure but through the O2 removal we know that isn’t what’s causing this idle surge and if the piston were traveling to far we’d see spark plug damage wouldn’t we or other nefarious issues.

I’ll go grab another gauge at some point to make sure that one is accurate but from what I’ve tested so far… removing the intake to the throttle body should have no effect on the surge (which it does) if the issue was compression so it’s unlikely that’s the cause.

Thanks for the factory specs man, much appreciated!

[dreamer2355]

Did you do your compression test at WOT?

I would definitely look into your engine timing seeing as you installed a new timing belt.

Do you have a scan tool that can read PID data so you can see exactly what is going on while the engine is running and experiencing this issue?

I’m not sure either how effective testing a MAF with a DVOM is but i would prefer to use a scope.

Awesome job thus far in your approach, testing and information. C8-)

[TheArete]

Compression test was done with ignition disabled and Fuel Injectors turned off, I don’t have scan tool right now with Live PIDs to see what the compression is at WOT, so I guess the next step is to really grab a scan tool.

A MAF can be tested relatively well with a DMM as can the TPS, you just have extend the testing as the refresh rate isn’t fast enough to capture everything that is happening on a multimeter, so basically you’re doing it for 5 minutes trying to catch the peak and lows or you’re actuating the sensor slowly to try and catch all the readings. As effective as a scope… negative, but it’s what I got to work with unfortunately at the moment. Am I confident in my MAF readings… relatively but not well enough to put it out of the picture. I’m thinking I’ll bow to the pressure and snag that Launch Tech and get some live data graphing which will capture at a much faster refresh rate and I can pull those stored codes.

New development, the Distributor is most definitely leaking oil, after my cleaning frenzy I can be 100% sure that oil is new and coming from a busted seal. Question is, can I just replace that seal or given that my rotor and cap already have some wear, buy a new OEM distributor for 200 pop?

Alrighty, timing up next given that we are having misfires and the t-belt is new. Let’s see where this goes.

[dreamer2355]

A DVOM will never be as good as a scope for testing MAF sensors as the DVOM’s do not update fast enough and are not valued in MAF testing when you are looking for frequency glitches.

Check out this video –

http://www.youtube.com/watch?v=p2QLxxst … r6qfmOsm56

As for the compression test, you need to manually depress the accelerator pedal while cranking. The throttle plate has to be wide open.

[TheArete]

update #1204

New Fuel pressure Tests–
1st test- 33 psi on idle, up to around 38 psi when WOT, and goes down to about 31 psi after the car starts bogging down. Remains at 31 psi even 10 minutes after car was turned off.

2nd Test- 33 psi on Idle, up to around 38 psi when WOT and went down to 34 psi after car starting bogging down. Remained at 31 psi even 45 minutes after car was turned off. Both Test are w/in specs (should be no less than 30 psi and no more than 38 psi)

*Edit* With regulator vacuum hoses disconnected we get a rise of psi to 42, w/in spec. Almost certainly not a fuel issue from here to the tank.

14.57 volts on battery while running and no significant drop when it begins surging.

Removed 02 sensor. Idles high still and then when it begins bogging down you can distinctly hear a “brrp brrpp” a second before it bogs down…. then it goes back up and then immediately preceding any more bogging down by about a second you hear what I presume is a misfire/s. Any other explanation for hearing this? It’s not extremely loud or high pitched, more of a low sort of cough or sputter.

*Edit* removed Intake up to throttle body including IAT, MAF, and the main vacuum hose from the valve cover to the intake and a smaller line from the intake to the EVAP system, with the O2 downstream sensor still removed. Throttle body looked clean on the intake side and really wasn’t bad (a lil carbon buildup) on the other side of the butterfly plate. Cleaned Plate and body.

Removing Intake sensors/ hoses obviously caused a rough idle BUT… no more surging, and the biggest fluctuation I saw was about 200 rpms. No more high idle, started right up around 1000 rpms (which is were it was before this whole thing began). The sound out of the O2 hole greatly diminished. Thoughts? We’ve checked the MAF but maybe there is still something in that system beyond the checks I did. Maybe in the harness (the conductor) between that and the PCM?

[djdevon3]

The distributor does have an O-ring and it should be oiled before installation. It’s a cheap part, try a dealer or auto parts store.

Yeah I’m pretty impressed with the amount of diagnostics you’ve done so far. Certainly more than I’ve ever done. This is all highly enjoyable to follow. Keep up the excellent work. I’m sure you’ll get it soon. You’re certainly ruling out a lot of things. You’re doing it the right way but I’ll agree it’s better to have an analog voltmeter or oscilloscope to pull MAF and O2 sensor values. Gotta use what you have to work with though. Not everyone has that kind of equipment.

By the way we’re dealing with an MTX not ATX correct? The MTX vs ATX MAF’s output different readings which is why I want to make sure.

[TheArete]

yeah Totally forgot, shes’ a manual transmission.

Thanks DjDevon, yeah it’s just the O-ring that’s gone, I did another cleaning frenzy on the valve cover and engine as well as the disty and the Disty is looking in used… but decent condition. She reads a mitsubishi, so I’m guessing she’s aftermarket, and since the plug wires are not OEM here’s my thought. Bare with me.

This engine does have a lot of wear, rust and broken/missing parts (I just easy-outed two siezed valve cover bolts). The Engine lift mounts however.,.. are bent… as if the engine has been pulled out. This coupled with the compression, the non-OEM wires, plugs, and possibly the Disty as non OEM, and some strange bolting… I’m thinking this sucka was overhauled at some point in time and she was rebored, new rings and maybe new pistons before I bought her. What do you think? Anyway to tell for sure?*Update*
IAT- The resistance was higher than expected but I’m not sure if that was because it’s about 50 out here and not the room temp 69

[dreamer2355]

You should look at the free Chilton manual link we have in the ‘How To’ forum to see if if has any information on timing belt replacement. You need to make sure you are at TDC. There should be timing marks visible so you can get the engine to TDC. Again if that Chilton’s link has timing belt replacement information, it should tell you where those marks are.

When you are looking at your SGN GRN as stated above, what sensor are you probing? Some sensors work on 5vREF and some do not.

Here is some information regarding compression testing –

http://www.dietersmotorsports.com/Compr … wntest.htm

[djdevon3]

Hate to burst your theory on the rebuild. The 626 did come with a lot of Mitsubishi parts including the distributor, throttle body, and IAC. You’ve got to look real close to find those little symbols but they are all over the car.

The coolant pipe behind the header is a bypass system. That is the only part of the system that is used until the car warms up (the thermostat opens). I have an explanation of that in one of my videos but I have to re-shoot the entire video because I got so many things wrong. I’ve learned quite a lot since then.

Have you tried resetting the ECU? There are 2 things that happen before the car is warmed up. The ECU uses default values for the O2 sensor and the coolant doesn’t fully circulate through the engine. This is so that the car warms up faster. Once it reaches a certain temperature the O2 sensor kicks in and starts feeding live data and the thermostat opens up to allow coolant through the entire engine.

As for the timing it’s exactly like any FS engine. There is an “I” (intake) and “E” (exhaust) stamp on the cam gears. Those marks must be pointing at each other exactly when the engine is at TDC. Being 1 tooth off will cause issues like loping idle. It shouldn’t cause surging though (to my knowledge). Just so happens that a YouTube channel that many of us like (including Eric) called RealFixesRealFast has a video that explains most of it and you’re really in luck because it’s for a 1997. How about them apples? 🙂 http://www.youtube.com/watch?v=HBhNITlXZ6c

You can download the EU (European) version of the factory manual from pmx626.info but there will be differences to look out for. The EU version came with a 1.8L, 2.0L, and 2.5L. The US only has 2.0L and 2.5L. The 1994 manual on that same site would probably be more useful for the engine specs and timing stuff. It’s mostly the same.

[EricTheCarGuy]

As the idle problem occurs try putting your finger over the inlet hole for the IAC in the throttle body (before the throttle plate), if it starts to run normally or stalls then your problem is most likely mixture related and my #1 suspect is that the IAC is not functioning properly. It could still be a vacuum leak however but considering that car sat for a while it’s possible that the IAC got corroded and is not functioning properly. It’s also possible that someone either made the throttle cable too tight or adjusted the mixture screw on the throttle body and the computer is not able to correct for the extra air.

[TheArete]

Dreamer, I have a Haynes and some of the FSM so I’m set there. As far as the MAF goes, it’s only a 3pin so it does not have a 5 volt reference.
It has a B+ (power) MAF (Signal voltage) and GRD (Ground). Ohh borrowed a PID scanner so I can give you guys the full stats after we reassemble Betsy for the Umptenth time.
Interesting link dreamer, it sounds like they’re suggesting its more accurate because your engine has full access to air (leading to higher numbers off the bat), which means equal access to air intake across all cylinders, lending itself to really highlighting the cylinder/s which are lower. So if you’re not extending the crank time or really cranking the EXACT same amount each time, you might not get full compression on a cylinder. Do you guys have a preference on Compression gauges? I’m thinking about shelling out and snagging a more accurate one for this exact reason.

DjDevon, Ahhh man way to shoot down my theory. lol. Thanks for the info on that hose, I was curious as to why it was attached to the thermostat housing and it’s purpose. Mits part on a Mazda engine put together by Ford…the more you learn everyday. Never would of Guessed. As for the video that’s an okay one, I think there’s a better extended on on youtube that I used last year when I was looking for help on the timing last time. Nice thing about those guys is they take the time to explain some of the theory though NOTHING beats MSDignition videos on Distributors and Timing and MossMotors.
A lil heavy handed on the Theory but FANTASTIC nonetheless. Check it out http://www.youtube.com/watch?v=mPdSCWZbHEU

Eric, Alrighty I’ll pop off the IAC and manually open and close it with my thumb after we get the baby put back together. Though I should mention that she did sit awhile BEFORE I bought her. The problem manifested itself during the summer (which I had driven the car the whole time with mainly just overheating as an issue (bad thermostat).

I didn’t fully explain the timing shenanigans. I’ve done the timing before and 2 different shops checked the timing. Each time the timging has been done or checked, or adjusted… the problem has worsened or gotten slightly better. Here’s my concern… these idiots at the unameable shop I’ve vented on before, took off the cam sprockets to help remove the timing belt and did NOT put the sprocket back on correctly (maybe 180 off), that’s if my cams don’t have just one key and gateway like the crank does (I’m thinking two). If this is the case… then even though they are right now TDC on the crank while the cam marks are aligned perfectly horizontal as well… she’s not in time. The exhaust or intake valves would be slightly open and therefore off because the sprocket was put on incorrectly.

My plan today… Check the crank and make sure the balancer is TDC and manually check the TDC not with the mark but by piston stop. That’ll let me double check the marking on the balancer and engine housing.

2nd. Rotate Sprockets till the valves on cylinder number one are at TDC (both closed) and make sure the Timing mark on the cam sprocket is correct.

3rd Realign the distributor housing and make sure the rotor is pointing exactly at cylinder number 1 (cap) while the Crank, and Cams are set at 0 degrees or TDC. Then once all this is verified I can adjust the distributor to get it to 12 to 15 degrees BTDC.

I have a timing light, a old timer next door neighbor mechanic whose got an ear for Timing and some beer….We’re set. Also while I’m at it I’ll probably check and clean all the grounds while the Intake system is off and everything else is out of the way. Thanks for sticking with me, I know this whole thing is pretty heavy in info… just trying to provide the utmost information, details and reasoning.

[TheArete]

The Plot thickens. I have the end of this book coming to a close now and it’s strange… I mean what are the freakin’ chances /probability… I mean I’m looking at the issue to all this right now (after we spent the day tearing her down) and I gotta say… I’m not sure if anyone would of caught this issue. Anyone… Unless they saw what I saw… two little white marks. I see now why several Techs/Mechs/ me and my bro and everyone else whose looked at it… couldn’t figure out the issue. Very heavily obscured.

Problem fixed? I’ll let you know tomorrow but we have isolated it. I’m being intentionally obscure because I just can’t type this out… I’ll have to show you a video of it. Lol. It’ll be good… trust me and something you probably have not seen that often if ever.

[dreamer2355]

Most 3 wire sensors do use a reference voltage so i am pretty confident yours will too S:). Check to see what voltage you have on the SGN wire KOEO.

And where are these white marks you are talking about? I am very interested to see what you have found.

[djdevon3]

Eric, neat tip on plugging the IAC air inlet. I’ve never thought of that. Really interested to give that a try but my engine is still in pieces.

TheArete, what white marks? Guess I’ll wait to see the video. I love watching 626 videos. 😉

[dreamer2355]

Thanks for the update.

I think the timing was mentioned on page 2 of your thread for checking TDC W;)

Don’t forget to change the traffic light to green if the issue is resolved so it will update our database and help someone else out who is using the search function of our website C8-)

[Author]

Replies