1998 Mazda 626 uneven idling

Reply

  #1  
Old 03-31-04, 02:48 AM
ayoshak
Visiting Guest
Posts: n/a
1998 Mazda 626 uneven idling

Hi,
I owned this Mazda since 2000. It is a high mileage car, 150000m.

Only a few previous known problems:
1. Cam belt tensioner?, repaired under recall by a garage,
2. Start problem, the key not recognised, bad contact at sensor coil, fixed by me,
3. Engine would stutter bellow 2000rpm when accelerating, bad spark plug wires, replaced sparkplugs, wires and repaired damage to ignition coil block (part of plastic shield eaten away by spark, fixed with epoxy), fixed by me,

New problem:
Engine idles unevenly, when around 650 rpm as if not all four cylinders are engaged, and then it would jump to 800 and be smooth for a while, and then back to 500-600 and rough. I had this problem for more than 6 months but was waiting for a regular service to have it looked into. It drives OK. No signal indicators on the dashboard. It feels more like a fuel problem than a spark problem.

However, of late the problem is getting worse. The engine will stay at 1500rpm after foot removed from accelerator, and then after a pause of 30-60s come suddenly down to idling speed (and run rough as described above). Occasionally it will not settle to idling but swing between 500 and 1500rpm and sometime stall. I couldnít influence that by turning A/C on-off like in one of previous threads. Still no warning lights, the engine runs smoothly when driven. Coolant OK.

Any help or even info of how and why is welcome. I donít trust my nearest Mazda service too much and would like to be fully informed when speaking to them.

Thanks
ayoshak
 
Sponsored Links
  #2  
Old 03-31-04, 03:49 AM
Desi501's Avatar
Member
Join Date: Oct 2003
Location: Boynton Beach Florida
Posts: 2,207
I'd look for a decell valve problem. Sometimes they come apart internally causing a large air bleed.
 
  #3  
Old 03-31-04, 04:22 AM
carguyinva's Avatar
Member
Join Date: Nov 2003
Location: Shenandoah Valley
Posts: 1,147
probably also a good time to...

...inspect the entire idle control system and clean the throttle body as well...when was the last time it was tuned up?
 
  #4  
Old 03-31-04, 05:11 AM
ayoshak
Visiting Guest
Posts: n/a
Hi,
Thanks for replies.

It was serviced by Mazda garage in June 2003.

Yes, it looks like I will have to take it appart.

Just tested it again from cold:

Start no problems,
Idle and driving when cold no problems,
Idle and driving when warm (normal idle speed but before ventilators starts to work, temperature half way between normal and cold) no problems,
Idle when hot, problems as described.
Driving when hot no problems.

All connectors treated with Stabilant (done while searching for previous problem). All visible hoses OK, no leaks.

From another thread, another group, another vehicle, it was suggested O2 sensors start measuring only when engine hot, might not be related to my problem.

Thanks
ayoshak
 
  #5  
Old 03-31-04, 05:58 AM
carguyinva's Avatar
Member
Join Date: Nov 2003
Location: Shenandoah Valley
Posts: 1,147
o2 sensors...

...have little effect on idle speed unless there is a bad air leak into the engine.

you might check the "snorkle" tube which connects the air filter assembly to the throttle body...sometimes they crack along the bottom due to engine movement...this allows unmetered air to enter the engine
 
  #6  
Old 03-31-04, 09:11 AM
ayoshak
Visiting Guest
Posts: n/a
I dived in with my multimeter and checked some sensors. My reference book is Haynes Mazda 1993 to 1998.

TPS (Throttle Position Sensor) is OK,

MAF (Mass Airflow Sensor) is OK,

IAT (Intake Air Temperature Sensor) is OK,

ETC (Engine Coolant Temperature Sensor) measures 14kOhm when cold and 4kOhm when engine hot. This differs from Haynes values (2700 Ohm cold to 300 Ohm hot). Could that be the problem, if PCM registers the engine as cold when actually hot and then tries to correct emissions fiddling with IAC Valve?

CKP ( Crankshaft Position Sensor) not tested,

CMP (Camshaft Position Sensor) is OK,

PSP (Power Steering Pressure Sensor) is OK,

O2S Front (Oxygen sensor Front) is OK, when cold it measures below 0.05V, hot at idle at 750 rpm 0.09V , when throttle opened it jumps to max 0.8V. When idle and it surges to 1500rpm it registers jump to >0.5V. It seems to work OK. Didn't test rear one.

Knock Sensor, didn't test,

IAC Valve (Idle Air Control) resistance checks OK. There is a difference in engine operation depending on if the valve is connected or not. If not connected the engine is idling OK at 750rpm, until something turns on (like A/C or ventilators) when it drops to 500rpm. When connected the engine would idle between 600 and 900rpm trying to correct for power users. But then it would suddenly surge to 1500rpm or drop to below 500 and stall.

It still looks like the computer is getting some inputs wrong and trying to compensate causing uneven idling. But what?
 
  #7  
Old 03-31-04, 09:26 AM
carguyinva's Avatar
Member
Join Date: Nov 2003
Location: Shenandoah Valley
Posts: 1,147
you'll have to...

...reconcile that coolant temp situation before you do anything else. you need something to accurately measure the engine temp...an infrared temp gun works well...aim it right at where the t/stat goes. you also need a chart that correlates resistance of the sensor to the actual temp...bad inputs = incorrect outputs.
 
  #8  
Old 03-31-04, 10:03 AM
ayoshak
Visiting Guest
Posts: n/a
Hi Desi 501,
Didn't find any reference to decell valve in Haynes nor in Chilton.
Could you elaborate?
Thanks
ayoshak
 
  #9  
Old 04-01-04, 03:11 AM
ayoshak
Visiting Guest
Posts: n/a
Just ordered the sensor, £20. Long waiting expected as none in stock anywhere.
 
  #10  
Old 04-01-04, 04:28 AM
Desi501's Avatar
Member
Join Date: Oct 2003
Location: Boynton Beach Florida
Posts: 2,207
Originally posted by ayoshak
Hi Desi 501,
Didn't find any reference to decell valve in Haynes nor in Chilton.
Could you elaborate?
Thanks
ayoshak
Couldn't tell you for sure if yours is equipped with one but most Asian engines use a "deceration valve". It creats a metered vacuum leak as the RPMs are coming down as to bring it down slowly. They are adjustable for the duration that the RPMs are held before releasing to idle. I had one that the internal parts vibrated loose and couldn't seal off the vacuum and caused a symptom exactly like your describing with a "rolling" idle. It's usually mounted to the side of the plenum and sometimes has small coolant hoses passing through it.
 
  #11  
Old 04-02-04, 05:48 AM
ayoshak
Visiting Guest
Posts: n/a
Hi,
Thanks Desi501, I will look into it next in depth.
In the meantime

ECT sensor arrived today, next day delivery, aren't we getting spoiled these days by these efficiency improvements?

Anyway I just measured it (the new one).
According to parts guy this is a standard sensor that is used for a lot of models and a wide span of years.

Deg Celsius.........................Ohm as measured
0..........................................100k.......(icy water)
36.........................................24k
60..........................................9k
80..........................................4k
100........................................2k......(boiling water)

This is more or less as the old one still in my car (14k warm 4k hot).
And there are Haynes and Chilton manuals which state figures which are 10 times smaller. This is not the first error I found in these books. For example wiring colours are all wrong for my model.

To definitely confirm that I just tried to simulate ECT by connecting 600 and 300 Ohm. The engine went crazy, raising to 3000 rpm and turning on all ventilators (engine was warm not hot) both times. Obviously it registered serious overheating and reacted.

My conclusion is my ECT sensor is OK.

Any ideas what should I try next?

Thanks
ayoshak
 

Last edited by ayoshak; 04-02-04 at 05:59 AM.
Reply

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Thread Tools
Search this Thread
Display Modes