Shaking at idle and low-mid throttle on Closed Loop

Gap Yah

New Member
Authorized Seller
Joined
Apr 26, 2020
Messages
4
Hello! I own a 1989 7MGTE, R154 Supra with around 270k original kms on the engine.

So recently my Supra started behaving weirdly when it has warmed up and, I'm assuming, enters into closed loop operation. On idle, the car shakes like the engine is missing a beat and the exhaust sounds like a Subaru rumble. I can't say the idle is stumbly because it isn't, it stays quite still but the car shakes. On low-mid throttle, the car is accelerating fine but it feels like it is missing a beat, the exhaust sounds like a lawnmower and the car is shaking constantly, but I don't feel like it is a proper misfire.

If I put my foot down and go to WOT, once I hit around 3500 rpm or so in 2nd gear, the car suddenly transitions and is absolutely perfectly smooth again with full boost and no shaking, I'm assuming this is when the car goes into open loop operation. However once I change gears and start to cruise in low-mid throttle again, the shaking comes back after maybe 1 minute or so (the car is actually smooth again in low throttle, but only for a few gear changes, then the car starts to shake again).

The moment I start the car (cold-start), the car is butter smooth, no shaking at all. This is till the car warms up to operating temperature (so around 10 minutes of driving) then it starts to shake. This pretty confirmed to me that it is due to closed loop operation.

Things that I have checked:

o2 sensor: I assumed this was the first issue as it is only operating during closed loop mode.
- I replaced the sensor with new wiring all the way to the plug, no difference (I did this twice as I have 2 new DENSO sensors)
- I checked the continuity from the o2 sensor to the ECU (OX, HT, B+ at the diagnosis box) and it is all good
- I unplugged the o2 sensor, started the car and it turns on the CEL. I plug it back in and the engine light turns off.
- When I drive the car whilst it is shaking, the AFRs are quite stable on low throttle cruising at 14.7 or so (I have a wideband monitoring this). This makes me think the o2 sensor is doing its job in low-mid throttle, but not 100% sure.
- My last thought is that the wiring from the o2 sensor plug to the ECU is faulty, but this problem only recently started and I've been driving the Supra perfectly fine for many months, so it is quite unbelievable to me that the wiring could just deteriorate like that... but correct me if I'm wrong.

ECU: the next problem I thought of was the ECU being faulty, being 30+ years old. I have 2 spare ECUs for my Supra, one is for a M/T and one is for A/T.
- The M/T spare one came with the car when I bought it, so I am not 100% sure that is a good working one (potentially previous owner fried that one and bought a second one), but I plugged it in anyways to try it out. Weirdly it comes with a CEL, Code 21 once I start the car, regardless of the O2 Sensor plugged or unplugged. This is a bit suspicious to me so I have kind of categorised that ECU as faulty.
- The A/T one is a mate's working ECU and I read that you can plug a A/T ECU into a Manual with no issues. I plugged it in and the issue is still there. No CEL unless I unplug the o2 sensor. So now I can rule out the ECU being faulty, no issues there.

AFM: I have a few spare AFMs which I have replaced as well to try it out but it did not make a difference (I don't think it'll change with closed or open loop operation).

Other things to note:
- My Supra is running a Lexus AFM housing with RX7 injectors, boosted to 10psi.
- It has a wideband and Apexi SAFCII, but I have dialed all the settings to 0 to make it as factory as possible.

I want to note that the car is perfectly smooth in open loop (cold operation and WOT) before it enters into closed loop operation, which tells me mechanically the car is fine and there should not be anything physical/mechanical that is making the car shake and sound like a lawnmower i.e. fueling, ignition or timing issues.

Thus I believe it is an electronic/signal problem. I wanted to ask what else the Supra uses as an input for closed loop operation? o2 sensor is the obvious one but how about the coolant temperature sensor? Does that play a part in how the car runs in closed loop operation? I feel like the car is going into some limp mode of some sort once it gets a bad signal from something which is causing the shake. Keeping in mind the AFRs while the car is shaking is still in the acceptable range so I find it hard to believe why it is shaking...

Before you ask if the car is misfiring, I'm not 100% sure. I feel like a misfire would cause the car to not have any power at a very short period of time, so the car pulls back a bit while I am accelerating, but the car still accelerates fine, it does feel sluggish though like it is down on power. Also the AFRs don't suggest a misfire as it would run really rich right?

At some points whilst the car transitions into closed loop operation, I can actually feel the car trying very hard not to go into the shaking lawnmower mode (as I have called it), like it is jumping from being smooth to shaking to smooth to shaking, and then it gives up and starts to shake.

Sorry for the long read but I wanted to be as comprehensive as possible, I will answer every response and further questions people may have.

Thanks in advance!
 

Gap Yah

New Member
Authorized Seller
Joined
Apr 26, 2020
Messages
4
Thanks for your input it is much appreciated!

Although how does compression play a part in this? Wouldn't a compression issue be evident regardless of open loop (cold start/WOT) or closed loop operation? i.e. the engine would run badly overall in every situation right?
 

Koenigturbo

Well-Known Member
Authorized Seller
Joined
Oct 4, 2006
Messages
1,574
Location
Oxnard CA
Here's my wild swing:: a sensor that is run by both temp and vacuum??? or maybe first by temp then vacuum? SVS?
 

Piratetip

Far From Maddening Crowds
Staff member
Super Moderator
Authorized Seller
Joined
Dec 30, 2005
Messages
2,445
Location
MKE, WI
My guess is a bad or stuck/sticking actuator somewhere.
VSV/EGR/IAC/Something
Or an intermittent/cracked intake leak post AFM somewhere.
Or an intake hose not connected or a vacuum leak.
 

Gap Yah

New Member
Authorized Seller
Joined
Apr 26, 2020
Messages
4
Thanks for your input it is much appreciated!

Hmm with that thought a bad VSV could be the cause, as they are temperature dependent. I will check the fuel pressure VSV and EGR VSV next.
In terms of any air leaks, would this also affect the car when it is open loop (i.e. cold starts, WOT)? I find it hard to believe that it could be a hose leak unless it was dictated by temperature e.g. from a VSV. I will come back after those investigations though!
 

3p141592654

90T
Authorized Seller
Contributor
Joined
Oct 11, 2005
Messages
4,261
Location
Thousand Oaks, CA
I suggested the compression check because it rules out anything mechanical. A weak cylinder may not fire at low rpm, but will be fine as the rpms increase and there is less time for the charge to leak out.

The next area to focus on after that is EGR, The car will run badly if the EGR is on at the wrong time or all the time. The EGR vac modulator can be easily checked and is a known failure point after 30+ years. Its a 5 minute check to verify the vac modulator..

 

Gap Yah

New Member
Authorized Seller
Joined
Apr 26, 2020
Messages
4
Hi guys!

An update on my search for solutions. I have checked both the Fuel Pressure VSV and the EGR VSV and both are in good shape in accordance to the TSRM specs. Weirdly though, the first time I checked the fuel VSV it had no continuity nor resistance, but then I put 12v through it to see if it clicks and it did... then I tested it again and there was continuity and resistance up to TSRM spec! How crazy is that! I have no idea... but I tried this on 2 VSVs and both had the same behaviour... the EGR VSV is all good though.

I think checked the whole EGR system, I performed all the tests on TSRM and it seems to all check out quite well. When I put vacuum from the intake manifold to the EGR valve, the car just dies. With the engine stopped, the air I blow into the pipe and it comes out the filter. With the car at 2,500rpm, the air I blow into the pipe comes out the other pipe, so the modulator is working. The car is still shaking whilst I perform these tests as the car is at operating temperature.

As for compression, you make a good point but I find it hard to believe it is genuinely any mechanical issue as the car runs super smooth when it is cold (open loop). There is no stutter or shake and idles fine too (when cold). I can't believe that compression would be an issue only when it is up to operating temperature.

I don't believe it is any air leaks as that would also be present during open loop / when the car is cold? Other than the VSV is there anything vacuum related that would be temperature dependent? I have a boost gauge and I have around 18-20 InHgs at idle. The boost gauge is tapped into the factory boost pressure module thing that comes off the intake manifold so it should be accurate.

It is worthy to note that, when I was performing these tests and I rev the car up to 2,500rpm or so standing still (in neutral), the car is running super lean. The Wideband is going up to 20 AFRs (with the car shaking as I have described), but when I am driving the AFRs go back to a normal range or perhaps a bit leaner than normal. But when I put my foot down and the car goes open loop it is nice and rich as it is supposed to.

The next thing I will replace is the IACV. I have already taken it out and checked all the specs from the TSRM, e.g. whether it clicks after engine stops, grounding the pins to move the valve, it all checks out... so I don't think there is something wrong with the IACV.

I will also replace the coolant temperature sensor for the ECU (the green plug) as that is cheap and quite easy to get so, I will come back with an update after the above is complete. Thanks for your input and looking forward to more inputs if anyone has any other ideas!
 

3p141592654

90T
Authorized Seller
Contributor
Joined
Oct 11, 2005
Messages
4,261
Location
Thousand Oaks, CA
Good, your EGR checks out. if you believe the issue is mixture control, then the next thing to test would be cross counting of the Ox sensor. I know you replaced the whole sensor and wiring, but let's verify it actually is working. When you unplug it the code you are setting is the O2 heater circuit, so that is really not relevant to your problem. Follow these directions with a voltmeter connected to E1 and VF1. An analog display is easier to use, but you can still get results using a digital gauge if it updates fast,