Engine shaking at idle.

[Madaboutmx5s]

I am after some advice or know how from you guys please on an issue on my red mx5 before I proceed any further.

Background is I bought the car beginning of June as a non runner which would not start due to an electrical issue which got sorted. It had been unused for 2 years previously. Car is low mileage as according to the car history and dealer receipt of sale when it was imported in 2009 total mileage is just over 49,000 miles.


I got it back up and running but since I have had it the idle is quite low, lumpy and the engine physically shakes on idle clearing as soon as you blip the throttle. The car drives fine and starts hot and cold all OK and tbh of all the 50 odd mx5s I have had seems one of the best NA I have driven performance wise for a stand one pulling really well. To begin with given the nice performance I thought it maybe had some uprated cams which on checking today are standard.

To try and cure the problem I have tried the following but with no success.

1. I replaced the plugs and leads.
2. I fitted a new battery.
3. Did a smoke test to check for any air leaks in the system (all OK).
4. Tried a different known good ECU to see if that made any difference.
5. Tried a different known good Cam angle sensor to see if that made any difference.
6. Checked timing was not out.

Possible other issues..

1. The idle screw would turn but made no difference to revs (it was jammed to begin with needed a mallet to shift it to begin with).
2. Trying to do the idle base reset on the diagnostics tool would not work and idle stayed the same so I wondered if it was working. Checked for power and diagnostics box has 12 (something volts) to it.
3. Engine would idle OK and lumpy shaking idle would stop if slightly higher revs if I unplugged the idle control valve ecu connector.

Finally with a friendly mechanic did a compression test and got results of 210, 210, 210 and 180 (which was a stuggle). Also did a leak down test and mechanic felt it was bad from the exhaust.


My friendly mechanic who did the compression test reckons the low compression on one of the cylinders is causing the car to struggle at idle causing the engine shaking, lumpy nature of idle. My question is before I condemn the engine as needing major surgery is there anything else I should try? I had planned a new fuel filter on the car but see no point based on the compression results.


Thanks in advance.

[Slipstreem]

Wow that's really strange. As I was reading this I was thinking injectors until the compression test. At 50k miles it's not even broken in so I highly doubt rings. Does it smoke at all from the exhaust?

[Madaboutmx5s]

Wow that's really strange. As I was reading this I was thinking injectors until the compression test. At 50k miles it's not even broken in so I highly doubt rings. Does it smoke at all from the exhaust?

It is odd. The car pulls fine, given how well it accelerates and the lumpy idle to begin with I though tricked cams till I lifted the cam cover off today. No smoke from the exhaust, etc. I have a spare coil pack and inlet manifold complete with throttle body, injector and idle control valve. I'm tempted to swop over all of this.

what I found strange yesterday was turning the idle screw did nothing to the revs. Also if I unplugged the idle control connector for the ecu or the throttle position sensor for the ecu the revs would increase and the engine shaking would go away.

[NobleRogue]

If I were You I would start by testing my Mas Air Flow Censor and Throttle Position Censor. I would also check to make sure all of my hoses were properly connected to the Air intake Tube as well as to make sure that they were in good shape and fitted to where they should be without leaking. In response to the leak down test I would do a Carbon Cleaning with SeaFoam before anything else exhaust related. www.youtube.com/watch?v=AW5Rf1Wg-Vs . This is a helpful video on how to do this. Buy two cans of SeaFoam. one for the intake and crank and the other to be added to a full tank of fuel. If the SeaFoam does not work you may have some bigger issues with the exhaust valve. Could be a sticky lifter or a bad valve. Also you should have no more than 20 psi difference from one cylinder to any of the others. If you do you will have a rough idle and various other performance issues.

[Madaboutmx5s]

Thanks cylinders are 210, 210, 210 and one that was a struggle to 180. It sounds lumpy on idle and with the engine shake does point do a dodgy valve based on the other items I've changed to no success. I try something similar to seafoam and likewise maybe a different coil pack and inlet manifold possibly to rule out the idle control valve/throttle body. However idle is spot on just like I say lumpy like it's got cams

[Madaboutmx5s]

Ok left the car over the weekend as on daddy day care duties. Decided to go back and check the diagnostics port as when I was with my mechanic friend he could not get the GND and TEN to alter revs for a base idle reset which it had done a couple of weeks ago when I played round with it myself but found the idle screw locked in place like a bastard. I got another bit of copper wire, warmed car up and repeated the base idle reset and revs increase from about 700rpm to about 900rpm and the engine shake stops as does the lumpy idle. Unplug the wire and the revs drop back. I didn't have a flat headed screw driver on me but hoping the idle screw might work now if I have the GND and TEN reset working. I'm guessing I cannot drive round with the wire plugged in to the port? Based on the know how on here am I right in thinking the revs increasing is putting the idle where the ECU thinks it should be when I to the GND and TEN test?

[Satisaii]

Never mind you already tried this.

[Madaboutmx5s]

Never mind you already tried this.

Never done as on Friday with my friendly mechanic the bridging of TEN and GND did nothing which I now know can be done. Digging around Google I've also ascertained that turning the idle screw without it in reset mode does nothing as I found out as ECU is not over ridden.

[NobleRogue]

If I'm not mistaken. Yes The ECU is trying to put the Idle where it should be. This cars Idle speed is supposed to be between 800 and 900 rpm when at operating temp. When messing with the Idle be sure that the engine is at operating temp. Drive it around the block a few times. Here isa video on how to do an Idle reset. www.youtube.com/watch?v=PVi2uaxAwnU. Use only the first 3 minuets of this next video for checking the TPS and making sure it is working and set properly. When it gets to adjusting the Idle he messes up. www.youtube.com/watch?v=nRGmB9Vbct0

[Madaboutmx5s]

Thanks for the info and links. I had a play around today with the idle base reset. When I bridge the GND and TEN revs increase. I've adjusted the idle adjuster screw anti clockwise (not fully) and the revs have increased but I cannot get it to match the ambient idle (but it's nearer). I've noticed doing this the engine now shakes less but still sounds a bit lumpy from the exhaust. As mentioned I've checked and replaced ecu, cas, engine mounts, wires, plugs. I haven't had a chance to drive it since the change yet but may adjust further if I can but found then whilst the idle screw turned after a point it didn't seem to adjust revs higher.

[NobleRogue]

Do you have the 1.6 or the 1.8?

[Madaboutmx5s]

Do you have the 1.6 or the 1.8?

1.6 NobleRogue. Engine is short crank but pulley is not wobbling. Now I can put the ecu in the reset mode my mechanic mate could not do when checked before I am going to check the timing again. Going to do another compression check as results seem high and reading around seen your suppose to remove all the plugs before doing the test, which didn't get done (took them out one by one and replaced all at end) and still thinking of another coilpack.

[NobleRogue]

Ha! yah you are supposed to remove all of the plugs. Also, don't forget to block open the throttle (No air in means no air to compress). Do each cylinder dry. Then do them each with about three small squirts of oil (2oz) from an oil can. http://www.hirespares.com/images/CAN2-1.jpg. The wet test should provide the same results as the dry + maybe 5psi. any more than that means you have a bad piston ring. If the wet test does not show any difference then your faulty compression is coming from a bad valve. Unless you have two cylinders next to each other showing a similar drop in compression. This means you have a bad head gasket. http://www.freeasestudyguides.com/compression-test-injecting-oil.html ... www.youtube.com/watch?v=X_tbksFYhl4 . I'm not sure about the 1.6 but on the 1.8 you compression should be around 190psi. On the 1.6 there is an air valve that sits on the intake manifold above the fuel rail. This valve will sometimes get stuck closed or shut. Stuck open it can cause your idle to be to high. Stuck closed can cause your idle to be to low. Try cleaning this out with carb cleaner, if it is stuck. Be sure to let it dry before putting it back on. I hope all this is helpful. Also you may want or need to check your CPC (Cam Position Censor) and MAF (Mas Air Flow censor). If your Roadster is a "97" and subject to the OBD2 emissions set up you may also want to look at your CAC (Crank Angle Censor). If you wish or need to look at thees let me know. I will send you links on how to do them. All of thees censors can have an impact on your idle and or timing.

[Madaboutmx5s]

Thanks. Given the high compression recorded on 3 of the cylinders and the fact the test wasn't done right I don't trust the results so going to do again. Thanks for the advice on how to do. If it comes back better on Friday when I'm hopefully have a chance to have a go I then try timing and maybe chuck a spare known good coiloack I have on to rule that out for the lumpy idle sound and now slightly less shaking engine since I've got it revving a bit nearer the ambient Rev ranage

[Madaboutmx5s]

Likewise thanks for the advice on the idle control valve.

[NobleRogue]

glad to help where I can.

[Madaboutmx5s]

Well I had a play around with Friday:

1. Did the compression test again the proper way: 180, 180, 180, 150 (same cylinder that was down before)
2. Checked timing and even with base idle reset in place difficult to get a reading due to engine shaking and best guess was it was around 16 degrees.
3. Checked for air leaks again but found nothing.
4. Played around further with base idle reset to adjust revs but still its lumpy on idle.

To add to the fun the clutch slave cylinder seems to have packed up as clutch wouldn't engage and foot went straight to the floor with clutch pedal. I've hopefully done a temporary fix as clutch fluid was low so topped up, pumped system and got it going again but imagine slave cylinder has failed. That adds to the slipping clutch I had.

Likewise spoke to a mx5 specialist I know well and given the short crank engine and the other bits I've replaced they also mentioned it could be a bad loctite fix done before I got it. They have a car running lumpy like mine before but like a bandit down the road (like mine) and this was the route cause.

So plan of attack is new clutch, new slave cylinder and I think a decent replacement engine. I've found a 108000km so around 67000 miles long nose crank JDM 1.6 engine with good even compression around 175 across all cylinders. Likewise will come with the inlet manifold, fuel rail, idle control valve, throttle body, CAS installed and I've heard it running so makes the change easier.

So the red car will now owe me more and seems extreme replacing the engine but given the tests all seems to point back to the engine don't see anymore point in wasting time on it. Whilst I bought as a non runner I was told it ran fine before being parked up but think the issues I've found was probably why the car was laid up for the last 2 or so years.

[NobleRogue]

My advice would be to keep the old engine after the swap. Do a leak down test to make sure the block is not cracked. If it is not cracked. Keep it as a back up. Looking at the compression numbers you are looking at a top or bottom end rebuild maybe even a complete rebuild. Or if you are particularly lucky, it may only need a new head gasket. Also your timing should be at 10 degrees before top dead center at 850 rpm. This to is obviously a cause for your rough idol. Before you go replacing your engine. Adjust your Cam Position Censor and fix your timing. Then do another compression test. You might be surprised at the result. Then again like you said you may have gotten a loctite P.O.S.. If this is the case and the block is not cracked you could use it as a project block. Bore it. Stoke it. do whatever you want with it. make yourself a nice little sleeper Miata. You will be taking down Porsche's in no time. :)

[Madaboutmx5s]

Thanks for the advice. Even with the ecu in reset mode on the diagnostics and adjusting the cam angle sensor it was difficult to get a timing reading as engine shaking at idle. Best guess was 16 degrees when adjusting it to improve the lumpy idle a tiny bit. Whilst it going to wipe me out I think replacing the engine, clutch, slave clylinder will hopefully resolve the issues. Positive is the car is a 2009 Japanese import low mileage 1990 roadster and never welded solid shell so out of all of this I'm putting together a nice (abit more expensive then I planned) restored roadster.

[NobleRogue]

I hope you enjoy it.

[Madaboutmx5s]

I hope you enjoy it.

Thanks.

[alexlula]

Could be the engine mounts ??

[Madaboutmx5s]

Hopefully now resolved the issue as fitted a long nose crank engine which fired up today. This had even compression of around 175 and had done 108,000KM so about 67000 miles. Whilst it went in also replaced all the coolant hoses with silicone versions, replaced the radiator, fitted the missing engine under tray whilst the clutch that was shagged has also been replaced with a new one together with a new clutch slave cylinder. Started fine first time and warmed up to temperature OK and the lumpy idle and vibration has gone away. Still on the ramps at the moment but hopefully will get a drive tomorrow.

[NobleRogue]

Glad to hear it. One thing though. Since you replaced the Clutch slave cylinder you will want to replace the clutch master cylinder as well. If you go to long without replacing it you risk getting stranded. Often times the new slave cylinder will cause an old but good master cylinder to blow out. It's like replacing the wheel cylinders on drum breaks. If you replace one you need to replace the other.