Bouncing RPM, is this mechanical or tuning related?

Ever since I have switched onto a standalone ECU, I've experience various degree of jolting when accelerating/moving at a low speed. The car would feel like it is hitting some on/off type limiter where it would hesitate and jolt forward in rapid succession with rpm jumping up and down.

This usually happens when I am moving at a very slow speed with light throttle opening in a low gear.

I've since switched from distributor to individual COP setup and have also rebuilt the engine but the problem still persists. I've came up with a theories including the ISC valve cycling on and off so I have tried plugging it and creating a fixed air source for idle without the influence of an ISC to rule it out but the problem still persisted. The strange thing is, the problem could be very mild one day, and very severe the next.

Attached is a log file of it having a bad day and amplified by me driving up a narrow multi story car park where I was constantly in 1st gear with very light throttle on a slight incline (see last 6 min of the log) where it was the easiest way to trigger the problem. The bouncing in RPM is very apparent in that log.

Is there any insight as to what is causing this? I've went through the settings in my tune over and over again and could not find any setting that might have related to this.

If you can put that into an Excel compatible format it would be easier to import. Can you post a screenshot of your timing map(s) ?

Do you have a full mod list? It sounds like unstable combustion causing torque fluctuations. You may need more spark advance. Try adding 2-4 degrees in the affected area (it might actually need a lot more than that). When you are at low speed and low load you can get very slow combustion requires a lot of spark to run right. From what I can see you are at 15 to 20 degrees in some of these areas. It's not uncommon for some engines to need 40+ degrees at low load to run stable.

I'd say your ignition table at low load at 1500 to 2000 rpm needs a bit of attention. Over those 500rpms you're advancing by 6.5 degrees.

My guess is, that even some 3 degrees of ignition advance I'm seeing in that bouncy RPM trace, will produce that much torque, that your car is bouncing back and forth and your foot is applying +/- 1 to 2 % of throttle in a frequency manner during that.

So that must have been a shaky drive in first gear.

Thank you for the help.

Yes, this make sense. A couple days ago I was adjusting the overall base timing and ended up advancing everything by 2 degrees. After that, it did improve things quite a bit and was much less shaky but I never thought it was related to the timing. I just thought the engine was having one of its better days again since this problem had always been inconsistent.

Attached is a screenshot and excel of my ignition map for ease of access.

Thank you!

Made some changes to the main ignition map and will give it a test drive once I find time off work.

Small changes were also made to the high RPM area just to keep the graph smooth.

Oh and some my major spec are as follows:

Gen 3 3S-GTE engine

Precision 5558 turbo 0.48AR

Kelford 262 high lift cams

~8.8:1 CR

You have a few things going on...

1. Idle ignition control is bouncing on and off, this is due to two reasons, RPM lockout is too high, Speed lockout is also set too high (more typical would be 2kmh) and you dont have a non-driven wheel speed source assigned so the speed lockout doesn't work at all anyway. So change those lockouts and assign a non-driven wheel under the chassis and body menu.

2. Overrun fuel cut is bouncing on and off. This is applying an extra -6 degree ignition trim on top of the already unstable ignition. You need to increase the de-activation RPM so there is a good margin above normal idle speed and probably increase the TP lockout a little too. It may even be a good idea to turn off over-run fuel cut for now and get the other items above working right first.

Your bigger cams are going to have a higher overlap and a lower effective compression ratio due to the later intake valve closing timing. Directionally speaking you can expect the engine to burn slower and require more spark advance from this hardware change.

Your second ignition table is going in the right direction. I would experiment adding in 2 degree increments all the 60 kPa and below cells. Maybe add 2 degrees at 60 kPa and 1 degree at 80 kPa.

I've tried various settings today and here is what I ended up with.

The problem still persists despite ignition had been advanced by 5 to 9 degrees at the low rpm/load area and effects from idle ignition and overrun fuel cut had been minimized.

I cannot say whether things have improved yet given how inconsistent this problem had always been but at least it had not gotten worse. I might try and add in even more igniton advance and disable overrun fuel cut completely but given the amount of advance I have already added, shouldn't I have felt some notable improvement if ignition was indeed causing my problem?

Thank you for all the suggestions and comments given so far!

This still comes down to the cams and the heads. Did you port the heads? That slows the burn down even more because you lose any tumble or swirl effect that speeds up combustion. You've got two options here. More spark or richer mixture (richer target Lambda). I'm not convinced you've given it enough spark. Look at stock spark maps on port injected engines with variable valve timing (running overlap during cruising). Attached is one from a 2005 Subaru STi, which has intake variable cam timing (similar to early vvt-i like my mom's 1ZZ Corolla had). The load axis is based on the mass airflow sensor, but vacuuum/low boost area is under about 1.40 load. You can see at lot places having 40+ degrees of spark, not to mention the knock control system adding even more.

After tinkering on the ignition map further and added more advance which still made no difference, I simply went ahead and did a blanket change to ignition timing between 800 to 2500 rpm to 40 degrees from -60kpa and below and 35 degrees for -40kpa. Result were still the same even when my logs showed a stable ignition timing at 40 degrees across the board.

I did not port my heads but for the cams, my power output had always been lower than expected with torque falling off much earlier than expected compared to others with a similar setup. Most people attributed it to the degreeing of the cams after looking at the dyno plot. I was pretty certain the degreeing was done correctly but have planned to try adjusting the cams again while on the dyno. However, I wanted the engine to run right before heading to the dyno. Given your comments, seems like the cams timing may be playing a part in my problem and I may need to head to the dyno first.

Ok, you probably spent enough time on ignition timing. Try richening the mixture up: target lambda no leaner than 0.90

Just looked at my older logs from the track. Seems like there are similar oscillations in RPM at higher load and rev range during partial throttle looking at the logs although I couldn't feel it. Could this mean something else is wrong?

Hmm. It's starting to look like electrical noise. Are all the grounds clean? Shielding and harness good?

Yes, the hall sensors are separately connected on its own shielded raychem 55A wires and the shielding is grounded to the sensor ground on the ECU side.

If it was actual interference rather than the RPM itself fluctuating, I shouldn't be able to feel it after I have made the blanket change to 40 degrees for ignition right? As the false RPM fluctuation no longer results in a change in igniton timing which means no change in torque due to ignition? Or does that mean it could result in errors in real timing and therefore the bumpy ride?

Try logging the Engine Speed ROC to see if it's trigger related. Also, isn't the 3S-GTE usually equipped with a 36-2 wheel on the crank and a reluctor sensor?

Thank you, I'll try logging ROC next.

The stock 3S-GTE is actually equipped with a 24 teeth (don't remember whether it had any missing teeth) wheel on the distributor which reads off the cam along with an additional wheel for sync and does use reluctor sensors. But I have since switched it to a 12 teeth wheel on the crank for more accurate timing.

Here are some logs on the stock trigger system when my car was first tuned by a "professional" tuner. Note the signal had more noise at high RPMs on the stock tigger stsem and also the the bumpy acceleration under low loads still existed (along with 10million other problems which I have since fixed).

The original tuner did a very poor job at tuning my car and had some very weird settings that made the car drive horribly but most of those problems had been corrected now. In stages, engine had switched from a distributor system to COP, triggers upgraded to crank vs distributor and a full rebuild but this problem still persisted over each stage. Since my current map is still built on top of his original map, I really cannot help to think there might be some other whack settings in the original tune that is hidden somewhere that is leading to this problem.

In your latest log I dont see any obvious signs of trigger problems, but there are not really enough parameters recorded to say that confidently. For the next log can you please do a PC Log instead of an ECU log so we get all parameters. In your latest log at least it appears you have improved a lot of the earlier problems. Fuel and Ign are quite stable and both with realistic numbers in the areas where your RPM is bouncing so I'm now not so convinced it is a tune problem.

Sorry this took awhile but finally here is the PC log. It's recorded in the same car park I was at for the logs in the initial post. Weirdly, the problem will always be most severe when I am at the specific car park.

One thing just came to my mind though. I have had miss-firing problem at high boost some time ago when I first switched to 1ZZ coils for COP coversion due to too large of a spark plug gap. I've since gapped it way down to 0.6mm (also swapped for another set of 1ZZ coil just in case). Could the small gap be causing my problem? Anyway, I will be switching to LS2 coils shortly (yes, this is way overkill) and will be rewiring the ignition system and if my problem is ignition related, that will solve the problem.

I've uploaded the log and map to google drive below due to trouble uploading large files to this forum on a slow connection:

https://drive.google.com/open?id=1rJCwpJp0JK47hV2YFnsMxRMLYrGWjgbT (MAP)

https://drive.google.com/open?id=1kY42BTrK138TaXegOhIzuzzRWWwhINyl (log)

There are no symptoms to suggest you have a trigger problem in your log. One thing that looks a bit odd to me is battery voltage, it only shows a max of 12.7V in your log and is quite unstable. In the "bouncy areas" it is 10.9-11.5V. This could possibly mean dead times could be having quite a significant impact. I notice also your idle area is targeting 1.0 lambda, have you tried it richer in this area to see if it has any influence? I would try something like 0.93 just as a quick test to rule that out (just change the target table, rather than the VE table).