Low rev misfire

[Jhonno]

ChimpOnGas said:

Jhonno said:

ChimpOnGas said:

Good luck with it Jhonno, misfires can be a biatch but once you've eliminated them throttle response will feel like it's come alive

I've had glimpses of it amongst it all, as it does run well intermittently.. It is a beast hiding behind this issue.

The truth is most TVRs are beasts hiding behind a series of issues

Releasing the beast just takes:

1. A good understanding of four cycle spark ignition internal combustion engine fundamentals

2. A systematic approach using good diagnostic equipment

3. Patience!

The golden rule with TVRs is most idle and drivability issues can be traced to poor earthing and low voltage conditions, however never discount other critical elements such as fuel pressure, if you are not recording a spike in fuel pressure when you stab the throttle suspect issues with your FPR.

In your case you absolutely must analyse and log your coil voltage and injector behavior during misfire or you'll be going nowhere fast with a blindfold on, and the only way to do that is with an oscilloscope! This was all very well understood some 40 years ago which is why we had Krypton and Sun tuner machines, these days any guy off the street can access the same diagnostic functionality for just £85, for geeks like me running an alternative fuel and anyone like you with a misfire issue this a wonderful thing.

I have no idea what control features you have with your ECU but I can offer you food for thought in the fact I can literately dial in and dial out tip in misfire on LPG just by altering my coil dwell time and or timing numbers in certain rpm x load cells where hard to strike propane becomes highly intolerant of combustion initiation.

Somewhere in your ECU software there will be a dwell vs voltage compensation strategy, don't mess with it but do take time to study why the ECU is calibrated to increase your coil dwell as battery voltage drops. Understanding this simple strategy will help you to appreciate why you need to capture coil voltage at the split millisecond misfire occurs.

At this point it's important to understand all spark ignition internal combustion engines misfire to some degree, even the most modern and engines using the best ignition systems have an accepted misfire rate which the sophisticated OEM ECU is closely monitoring. The OEMs know misfires are inevitable so they set a threshold before the ECU throws a fault code, this will be long after the engine has started to misfire which is considered normal but long before the misfire can be felt by the driver.

A misfire that can be felt by the driver is a very severe misfire indeed!

Study and understand the fundamental first principles of internal combustion engine ignition systems, then use the correct diagnostic equipment to understand what's really going on when you experience the misfire. This is the way to fix your issue because armed with the right knowledge and kit..... the fault will be sure to jump out and slap you square in the face

That's got to be way better than the 'Parts Darts' approach, right

Read, acknowledged, taken on board and appreciated btw Dave..

[Jhonno]

Thom said:

Sounds good. Will you replace the remaining 6 plugs too?

No, I will leave them. They are Iridium and only done 2k odd.

[Jhonno]

Is it possible to have a lazy/intermittent lambda?

Noticed my Lambda 1 keeps dropping to about 0.01-0.05 and staying stationary for a bit, coinciding with the low rev stutter, when it starts flicking the values jump about a lot more and it pulls cleanly from low revs. It also seems to be inactive when cold/for a period after lambda 2 comes in also.. With my symptoms being worse when cold..

[Jhonno]

Thom said:

Could the fouled plug(s) have let too much unburnt fuel into the exhaust and damage the corresponding lambda sensor?

I would also replace the other plugs, they may have done only 2k miles but then the failed ones also did...

I am swapping them all out for NGKs.. I am having to a do some parts department diagnostics as I don't have time to order diagnostics and then order parts. Plus these were a bargain @£18 for a full set.

I have heard of other AJP's not liking Iridium plugs, copper ones being more powerful. The iridium plugs are 1.1mm gap, NGK's are 0.7mm, so it might not like the larger gaps/weaker spark.

I have also ordered a spare lambda, in case. No harm in having any of these bits as spares, but I only have this week to solve it.

[Jhonno]

Thom said:

If 0.7mm is the plug gap originally specified for an AJP running on the standard ignition set up then I would not try another value for the plug gap, regardless of the type of plug used.
If your engine has a slightly higher compression ratio than standard then it might even benefit from a gap tighter than 0.7mm.
A good thing about setting the gap down is that it's free

Yep, agree. It says they come pre-gapped, and not to change it, but could be worth a gentle go. Or just wait for the NGK's to arrive.

[Jhonno]

FarmyardPants said:

Is it possible to revive a borderline o2 sensor by blowtorching it, or is this an urban myth?

I have read that too.. I don't know how true it is either!

[Jhonno]

Lambda from the bank that keeps dropping out..

https://photos.app.goo.gl/DudTUHnp5bkrG74h8

No idea what it should look like.

[Jhonno]

Thom said:

The symptoms you initially described with the car missing below 3k rpm and fouling the plugs suggested an excessively rich running condition which would have had the lambda sensor(s) look positively black, at least for real world road driving where going full load all the time does not happen.
If the engine is now running better with the replaced plugs and after setting the throttles again then the lambda sensor(s) may not be sooted anymore, however this picture suggest an excessively lean running condition.
How does the other lambda sensor look like and can you swap them to see how or if the adaptives and lambda voltages change?

So I went back to the car and it was refusing to idle again, well it will idle at 1300rpm+.. The missing is better, but still there. I noticed that the Lambda 1 value kept dropping to 0.01-0.04 (remaining stationary, whilst Lambda 2 flicked) and it would miss when you tried to accelerate. When it "came to life" again it would pull cleanly from any revs.

I was thinking the lambda condition was showing lean, but wasn't sure, I had a couple laying about which looked darker/more sooty. The plugs from that bank looked OK.

Edited by Jhonno on Tuesday 21st May 09:12

Edited by Jhonno on Tuesday 21st May 09:52

[Jhonno]

Lambda traces..

https://photos.app.goo.gl/WGoNnk9VAur3wFYV7

[Jhonno]

Another..

Having to hold about 1.3k due to lack of idle..

https://photos.app.goo.gl/nhTJSVHrooqYTdqY7

[Jhonno]

spitfire4v8 said:

In a galaxy far far away about 10 days ago ...

spitfire4v8 said:

We could do with seeing a trace of the live graphing screen showing the lambdas spiking




The adaptives trim is based on the lambda feedback .. some lambda outputs look ok on the main diags screen, but when you view it in the realtime logging they don't spike sharply like they should do, the traces might be gentle meanderings which is not correct.

I was just checking that is what you meant.. Still learning some of the more in-depth bits. Never seen a lambda trace before.

[Jhonno]

For reference..

Lambda trace with 1 new lambda and one heated with a blow torch and reused.

https://photos.app.goo.gl/rKi3M9AXxd3vbxwf8

[Jhonno]

spitfire4v8 said:

That looks better, but with the lambda traces both being the same colour its impossible to tell the two apart. Does the tvr diags not work for you?

They are different shades of green..? The issue is they follow each other quite a bit and so are over laid.

I have the rewritten version of the software on my laptop..

[Jhonno]

Here is an interesting one..

NGK's fitted.

Started straight up and settled into a nice 1k idle.

Adaptives about +3

Comparison of iridium vs. copper plug. I wonder if the spark position/shrouding is why some Cerbs don't like Iridiums.. Mine used to, but now maybe it doesn't?

https://photos.app.goo.gl/HcCiqQSZ3Ugaa2dU7

[Jhonno]

FarmyardPants said:

So how does it drive now?

So much better.. It still has a miss when pulling away, and brief one at 3k, but it is pretty transformed. Pulls cleanly in fifth, much harder than before. I haven't even reset the adaptives yet. Forgot to plug my phone in to do it.

[Jhonno]

Thom said:

Would be interesting to gap the Iridium plugs down to 0.7mm. I would be really surprised if the nature of the plugs caused the issue rather than the gap.

In any case, congrats for the progress!

I still have the plugs, I may experiment when it is run in..

So looks like it was a combination of unbalanced cylinders, lambda, plug lead and plugs, and setup to get it to 90%..

[Jhonno]

I notice the odd bank lambda still spends a bit of time at 0.01, is this lean or rich?

[Jhonno]

Drove her in to work this morning..

Still loads better.. Idle slightly high, and there is a pull away stutter still present, and an occasional 3k miss, but the improvement is night and day.

[Jhonno]

I think I have worked out what the remaining stutter is.. Play in the even throttle arm. Which I think can be taken up with a spring.

[Jhonno]

CerbWill said:

The arm connecting the rod linking the 2 banks to the spindle running through the throttle bodies? There's a grub screw holding the arm on and, on the 4.2 arrangement at least, it's on the underside. Maybe check that's not vibrated loose?

I didn't realise that.. Good shout ta. Nice and awkward to get to I am sure!