Jump to content

EFI issue. K3.


Recommended Posts

Hi,

I have and had abysmal MPG since I went EFI in 2010 ish.

Preamble;

I have increased the fuel pressure in the software so that in reality the fuelling is decreased across all sites.

I know it's a dangerous game lowering the fuel pressure everywhere. My MPG has now increased to mid 20's from late teens. Additionally, I rarely run WOT.

I do have some flat spots now, but can live with this for the sake of increased MPG. My LC2 is also running better. ie less error 8's. The K3 is running open loop, and I might set it to adaptive to re-tune.

 

Anyhow, I now get a lovely head turning burbling on the over-run when I close the throttles. I have the ECU set to cut all the fuel above 2250 RPM and throttle closed. Do I need to reduce he RPM to say, 1500 to prevent the burbling? If the burbling suggests too lean, where is the fuel coming from? The overlap? incorrect valve clearance? leaking injector(s)?

Cheers,Iain.

 

Link to comment
Share on other sites

Ian,

I forget what ECU you have but it would be surely worth you while to have it properly mapped on a rolling road and with the mpg figures you quote, should pay back quite swiftly too.  Do you have the software to tune it? 

Presumably your burbling is occurring below 2250 rpm? I would try reducing the cut-off point to 1500 rpm initially but dependent on your natural idle speed you might be able to go lower.  I used 1,100 rpm on the PI but with a standard cam that had a natural idle of 800 rpm so still a reasonable gap.

Cheers

Nick

Link to comment
Share on other sites

Hi Nick,

 It's the Emerald K3. It was tuned by Mr. Walker himself. If I put his map on the MPG is poor compared to when the engine was PI and I regularly got 30 MPG + on the motorway & Autobahn.

That engine died due to Lucas pump not up to job on Autobahn when driving at 130+ with a Escort + Squadie heading to the Netherlands where I was living at the time. Melted earth electrode on #3 and I saw the rings or what was left a year later when the engine was rebuilt.

Now the snow has gone, inshallah, I will bring the fuel cutoff point down to 1500 and see if I can get readings out of the LC2 on a spare 0-5VDC gauge I have (AFR). 

 

Question Nick;

I do not run sequential as I only have 4 ignition drivers. I do have 6 injection drivers though.

I have a matrix for end of injection; TPS x RPM sites, all are set at 318 degrees. Should these not be set different even on my non sequential engine?

Cheers,

Iain.

 

Link to comment
Share on other sites

If you have 6 injector drivers then you could (potentially, subject to other factors like cam position sensor) have sequential injection.  I would imagine that you run 3 of the 4 ignitions drivers to trigger a wasted spark coil.  The injection point (and whether sequential or not) doesn't actually have all that much effect except at (or close to) idle where it helps emissions and can improve smooth running on wilder cams by making sure the fuel actually goes to the cylinder it's intended for.

A Triumph 2.5 delivering sub 20mpg is not correctly mapped.  I don't care who did it.....  My cheapo PI saloon conversion with scrapyard hardware and MS2 had nearly 150bhp and could manage a 30mpg average with 34 - 35 on a run.

Nick

Link to comment
Share on other sites

Hi,

 I could easy fit an M8 prox sensor in the dizzy plate and a mod on the cam gear deal to trigger. 5V & PNP limits my options on EBay! I would not buy an inductive one as have to run near the coil pack. LOL. That will not happen as it will get put at the back of the queue to be honest.

Mini Sport is down the road from me but have no experience or willing to touch the K3.

Adaptive mode is getting engaged again. Easy said and difficult to implement as the P I settings are an unknown. Anybody here with a K3/K6 want to give me some settings?

TB's were new in 2004 and are not the best now, but balanced good. Fucked up a bit by a garage fitting return springs that made the clutch feel soft! Smooth now though.

I drive 40 miles to and from work and 20 MPG is taking the piss.

Thanks for your time Nick. Appreciated.

Cheers,

Iain.

 

Link to comment
Share on other sites

  • 2 weeks later...

Hi Nick and all,

 I am now sure my crap MPG on my road car is because i utilise a TPS as load indicator. I have no manifold pressure sensor & as i'm using the TPS to set spark timing; it is a total waste of ****ing time and therefore useless!

I believe I need a correctly functioning ignition firing point, to have any chance of fuel economy. No wonder the car popped and banged on deceleration. Noise gone now as I have Naughty Rigged the fuelling to cutoff injection if my foot is off the accelerator and RPM above 1500. Being able to set the fuelling to cut off above a certain -ve pressure would be better but impossible on my setup.

First, I need to mount a mechanical vacuum gauge and check i have a sufficient vacuum available. Problem is; normal ports are blocked by fuel rail! My cam timing is unknown, but even if its a 25/65; The PI worked fine for PWR and MPG. I need to fit a mechanical vacuum gauge and go from there.

Lambda/LC2 is not working correctly so this needs pulling and dunking in some petrol with the heater powered to clean it!

cHEERS,

iAIN.

cAPS oFF.

 

Link to comment
Share on other sites

TPS shouldn't be far off a MAP sensor.

I think you need to go through the basics, check TPS calibration, and get the AFR working accurately. Without that it is all a bit of guesswork.

Another thought, what is the throttle linkage like? it really wants to be nice and progressive, first half of pedal movement giving maybe 20% throttle opening. That required some work onmy car as I am using bike ITBs, and solved by changing the attachment point on the pedal. That was not an exact science either.....

Link to comment
Share on other sites

Hi,

 I have been driving around these last few days with the Tough-book. Hooking up the LC-2 and the K3 to check everything.

LC-2 readings are OK and then readings go very very lean before error 8. Sometimes. Will try to clean sensor as not happy about having to buy a third one. Have installed latest firmware and all connections lubed with dielectric grease. Full calibration preformed in exhaust after engine off for 16 Hrs. 

I have downloaded the original Emerald/Dave Walker map to the ECU. This is the first & original map. Power is back after leaning everywhere. Spark advanced at closed throttle everywhere. i.e. from 5 to 25. I'm guessing this is my MPG problem.. 

I believe that the TPS signal is a set-point and not a true load signal. Maybe OK for fuelling, but no good for spark timing. If I lived in the Netherlands, TPS would be more like a load signal. I do not unfortunately.

  • TPS shows the ECU where you have put your foot, that's all.MAP tells the ECU the load on the engine!
  • If you go up a slight incline vs down a slight incline at the same TPS, the actual MAP vs TPS will be different and this in turn will mean the engine goes rich/lean.

I have a prestige setup on my CP ITB's I have renewed all the rod-ends and adjusted the linkage to give the best throttle to pedal response. I have adjusted linkages to give the same airflow across the three inlet manifolds.

I need to fit a vacuum gauge. I need to see what vacuum the engine is pulling. If I am getting a healthy vacuum at idle, then remap to MAP instead of TPS for fuelling and spark. If MAP is not good then I will look at adding an extra table to adjust spark angle relative to absolute pressure to control. Not sure I can run two absolute sensors, so that I can run barometric pressure MAP for fuelling/Spark.

I am having a closer look at the fuel and spark maps that Marcus has sent me, as Marcus achieves power and MPG with a TPS only.

Cheers,

Iain.

Link to comment
Share on other sites

If Marcus wouldn't mind could you send me the timing map with TPS.   Whilst I'm leaving fueling up to the wonders of CD, I am pondering which route to go for secondary input on top of RPM,  I have the TPS version of MJ, as 1) I wasn't 100% convinced that MAP will give me exactly what I wan't, and as the TPS version can be used with an external MAP sensor (which I have), but the MAP version can't be used with a TPS.

My conundrum at the moment is that currently I run fixed mechanical advance (no vacuum advance) and have been fairly happy,  with both TPS and MAP I can see advantages and disadvantages, with TPS you have a sort of "Clear this is what I'm asking for next, but fuzzier now " picture, whereas with MAP you have a "Clear this is where you are now, but fuzzier idea of what's coming next" 

I have a lot to learn!

Alan

Link to comment
Share on other sites

Hi Alan,

 I'm sure that Marcus would not mind if I share his software with you, as I believe he helps anyone he can.

I'm off to Renault in France in the morning. I'm popping into work to drop my Spit off and pickup some spares to take with me. All going to plan, I should have some time to email you the data that Mr M sent to me. If you can PM me an email address or I will try PM at around 6.30 GMT.

Cheers,

Iain.

 

Link to comment
Share on other sites

don't know the Emerald system but on Megasquirt you can use a combined MAP and TPS for the load control. 

This was introduced as MAP can be a problem with an engine with ITB's. I have just set it upon my system and it seems to work well. You need to collect some running data to see what the MAP pressure is when the car is being used then set cross over point in the software. There is a very good  Utube video on doing this.

I am not sure if you can do something similar on the Emerald system. BTW there are people on the TR reg forum who have gone the Emerald EFI route on TR6's.

regards

mike

Link to comment
Share on other sites

Mike is right, Megasquirt has several blending possibilities.

I used MAP-mode first, but the car went much better since I swapped to the  "ITB-mode" on . (I'm using the original PI inlet manifolds)

It uses the MAP signal for lower loads and TPS for full load. In between there is a blend of both, which needs to be set specifically for your engine.

The good thing is, once you've defined this blend you have one and only one "Load", which includes MAP and TPS and you have only one fuel table as f(load, rpm)

This winter a third injector driver was added to the board, so I can run semi-sequential injection. Together with an adaptation of the dead time I'm having a rather smooth idle now.

But Iain's fuel consumption is way too high.

Before Megasquirt I had 3 Spanish DCOE's and Megajolt with TPS as load input for ignition advance. Was much more fuel efficient than your setup.

There must be sth. very wrong with your setup.

Cheers

Patrick

 

 

Link to comment
Share on other sites

Hello Iain

                 I also use ITB mode on my Microsquirt as I did not seem to get enough MAP signal!

Its still not right but I have covered 5,000+ miles and get 35 to 40 mpg depending how it is driven!

I have not bothered to much late last year as I am building a new engine with a 280 deg cam(Newman) and 9.7 to 1 compression ratio, ported head with bronze guides and balanced crank etc

Roger

DSC03580.JPG

Link to comment
Share on other sites

  • 2 weeks later...

Hi all first post so bear with me on forum etiquette etc please :)

You don't need MAP sensor to get the engine running properly, TPS is an excellent way to map the car and far quicker and easier to get a very good base map on TPS than using a MAP sensor.  You do need a good throttle spindle to TPS sensor mechanism though, any play in the system will result in airflow changes that aren't picked up as a throttle signal change at the ECU. This is a particular issue at small throttle positions where rate of change of airflow is at its greatest.

I have a dislike for innovate widebands, too unreliable. My preference is for AEM .. 180 pounds off ebay and so reliable  actually map using one on my dyno.

On the subject of load, whenever you visit any combination of throttle angle v revs the airflow is the same as the last time you visited it (atmospheric conditions excepted) , whether you're going uphill or  downhill is irrelevant .

Hello to all :) 

 

 

Link to comment
Share on other sites

Hi Obe & welcome!

 I have just ordered a "New AEM Digital Wideband AFR UEGO Controller w/ 4.9 LSU Sensor # 30-4110." £170.00 off EBay.

The #30-4110 has a suitable anal output so will work with my K3. The gauge is a bonus and can blank off one off the two spare 2" holes I have. I only have a 270 degree fuel gauge at the moment.

The "never needs free air calibration", sealed the deal.

Thanks for the heads up.

I will soon have a LC1 & LC2 for sale & very suspect!

 

Now, the load equalling TPS is not correct IMHO. I have owned cars with absolute vacuum gauges fitted; throttle position and absolute pressure was not the same, most of the time.

I believe I need to know absolute pressure to set the spark advance to gain economy. I cant see TPS being able to do this?

 

Roger; A very nice looking setup. Thanks for sharing.

Cheers,

Iain.

Link to comment
Share on other sites

7 hours ago, obe said:

Hi all first post so bear with me on forum etiquette etc please :)

On the subject of load, whenever you visit any combination of throttle angle v revs the airflow is the same as the last time you visited it (atmospheric conditions excepted) , whether you're going uphill or  downhill is irrelevant .

 

Welcome obe, I think this forum's etiquette is post what you know or believe to be true, avoid bull shit, and a dash of humour helps!

Your comment on load is true, but I have discovered one significant difference, event duration, on the flat or downhill revs can be rising very quickly and a particular point on the curve can be through in 10th's of a second, whereas on a hill you can remain in a rev band throttle position for seconds (even though at times it feels like hours) this often has the effect of highlighting a problem in timing or fueling that is over in a flash on the flat and not noticed. I now do all my data logging test/tuning runs on the hills now to try and pick things up, its also useful to be able to hold wide throttle settings for longer.

Alan

Link to comment
Share on other sites

Point taken about time spent at any one load site .. especially important if you hold a load site for a length of time and the resulting chamber temperature increase brings about pinking which you'd previously missed if you were travelling through the load sites quickly. Fair point well made.

On the dyno of course it's very easy to hold a load site and dial in the fuel and ignition required .. one point I often see with regards to timing (especially distributors) is to hold the car uphill and swing the dizzy until it pinks, then retard slightly. This is likely to put you too far advanced even if you retard it slightly from the point of pinking. By the time the engine is pinking you're already well past the minimum best timing to make the power. In the absence of any test equipment though it's likely the best method for many people and has been used for decades for setting up timing of course :)

 

Link to comment
Share on other sites

Hi,

The AEM has arrived.

I will fit on Saturday or Friday night after work.

I have received a prompt reply from Dave at Emerald concerning the K3 feedback settings so that I can run in adaptive mode to tune n drive.

I am going to set around idle @ 14.7 AFR. Adaptive everywhere else at ??.?. 

Cheers,

Iain.

IMG_0820.JPG

Link to comment
Share on other sites

Presume you have a target AFR table? Where you want max power you want 12.5. And I have used 15.2 at light load cruise areas. And some interpolation.

And I have had trouble with idle, and got mine set at 14.2 I suspect the coolant temp calibration is a bit off....or the compensation. 

Link to comment
Share on other sites

15.2 will be too lean for a 2 valve Triumph engine - especially a 2.5 with more limited squish.  You'll likely need to be richer than 14.7:1 to get a decent idle.  More like 13 - 13.5 typically - but experiment - it wants whatever it wants.

elsewhere, this is the target table I use.  The 600/900 rpm areas are too lean but I only allow it to tune from 1200 rpm upwards anyway.

New target table.JPG

Link to comment
Share on other sites

Hi,

 The K3 has provision for Open-Loop, Closed-loop & Adaptive.

I will set 0 to 1K RPM at 0 load site for closed loop @ 13.7 & everywhere else as adaptive. 15 AFR at zero load and 12.5 @ WOT interpolated between.

Hopefully a few hours driving will populate the "correct" fuelling numbers in the cells.

Ignition timing is another matter!

******************

Just received an update on this thread.

Will post this first!

************

Cheers,

Iain.

Link to comment
Share on other sites

Hi,

I will use Nicks AFR's and let it tune. I have 32 points and therefore much higher resolution than Nicks, should help the tune better? 200 RPM sites at my end should make self tuning easier..

Fitted the gauge/controller tonight and at lunch-time Friday, paid a garage a tenner to swap the lambda's.

With no sensor fitted the gauge shows 14.8 AFR. Sensor fitted and hanging in free air shows +20 AFR..  Surprised no sensor connector gauge showed 14.7.. Never checked the output to ECU, but hope this was 20+ AFR?

 

Cheers,

Iain.

 

 

Link to comment
Share on other sites

Hi,

 AEM gauge n sensor fitted. Cheers Obe.

Most interesting readings.

Idle 15 AFR +/- 1.0. @1K1 RPM.

cruise around 13 AFR @ 2K7 RPM.

WOT around 14 AFR @ 2K.

Obviously different TPS & RPM's show different. Never saw anything lower than 13 AFR.

Now even more convinced that TPS is not correct choice for engine load! Why? Because, at different engine loads with the same TPS gives same fuel.. Anyhow, going to setup closed loop around idle and adaptive every where else.

Looks like more power available, even though I want more MPG. LOL.

Cheers,

Iain.

 

Link to comment
Share on other sites

On 19/04/2018 at 8:54 PM, Nick Jones said:

15.2 will be too lean for a 2 valve Triumph engine - especially a 2.5 with more limited squish.  You'll likely need to be richer than 14.7:1 to get a decent idle.  More like 13 - 13.5 typically - but experiment - it wants whatever it wants.

elsewhere, this is the target table I use.  The 600/900 rpm areas are too lean but I only allow it to tune from 1200 rpm upwards anyway.

New target table.JPG

Nick, I am a tad confused. You say 15.2 too weak, but your target table goes to similar figures, or even weaker? (though not sure where cruise would be with the MAP figures, but assume approx 50ish kPA?) Or have I missed something? 

Maybe I could go leaner.......And I think a new exhaust collector setup to stop the air leaks may help with economy too.

Link to comment
Share on other sites

That target table is used for the Vitesse (2.0L) which will run leaner than the 2.5 due to having more squish.  As said, the 600/900 columns are definitely too lean, but are not used (by me) as they are outside the area of authority I have set.

The other key point, which you have already picked up, is what MAP you actually have in different parts of the driving cycle.  60 - 70 mph cruise for the Vitesse is likely to be in the 60 - 70 KPa area, but this will vary from engine to engine/car to car.  Below this is the over-run area (or at least minimal load) most of the time.

Your 16V Zetec should be able to run leaner everywhere except full throttle where it's needs are likely to be similar, but should especially be able to do better at cruise.  The seat of the pants method is to keep backing off the fuelling in the relevant areas until you start to get "lean hitch" where it starts to hitch and hesitate, then richen up a bit until that goes away.  At that point it'll probably still be a bit too lean and feel very flat when gently increasing throttle to maintain speed on an upslope or just to increase speed a bit so richening a little more is a good idea to improve response/driveability and will probably give better economy too.

Air air leaks in the exhaust, even slightly downstream of the O2 sensor risk messing up the O2 readings which will obviously influence readings, tuning and, if set to self adjust, affect running as well.

Nick

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...