4K Efi Query

[Banjo]

Hi Guys & Gals,

Currently doing a 7KE EFI setup on my 4K test rig engine, which if it works well, will go onto a fresh 5K engine, yet to be completed.

 

Nearly got everything finished, and the Haltech E6K I'm using, is all wired up.

 

Got a brand new Haltech O2 sensor & mounting boss today, and am not sure where the best place to fit it ?

 

I'm running a 4 into 1 extractor, and thought possibly the easiest & best place would be where I've marked with a "white" circle.

 

 

 

However, it occurred to me this is a fairly turbulent point, as all four exhaust pipes meet here. Maybe a better place would be a little way further down the exhaust pipe ?

 

Most manufacturers of modern engines put the O2 sensor up on the exhaust manifold somewhere, relatively close to the engine. Is this for convenience, or would a better spot be further down the pipe.

 

Appreciate if anyone who has done this before, could give me their thoughts.

 

Here's a couple of pics of the 7KE EFI intake on the 4K.

 

 

 

 

 

 

 

No distributor

 

Pajero acelerator cable (thank you Nesta)

 

No mechanical fuel pump

 

Cheers Banjo

[Taz_Rx]

A little further down the pipe should be ok....where you've marked it I would be more worried about clearance on the under side of the floor!

[Banjo]

Thanks Taz,

My research has shown that if the O2 sensors are too close to the engine they can over-heat & result in a shortened life. Too far away from the engine they can suffer from condensation problems, which is why they suggest strongly that O2 sensors be fitted at 12:00 o'clock, or at least between 9:00 & 3:00. Apparently condensation can kill them pretty quickly.

 

However, as the Haltech unit I am using is a "heated type", I agree, I don't think it will be too much of an issue if I move it a bit further downstream.

 

I will check floor clearance in the area I put it in.

 

Thanks again !

 

Cheers Banjo

Edited October 27, 2014 by Banjo

[Taz_Rx]

Oh btw what are you using for the ecu trigger?

[altezzaclub]

I put my pair about chassis rail level, before the collector. getting them to fit was 'interesting'.. Even though they are heated it takes a minute of driving before they settle down and give consistant readings, they seem to need a warm-up period and need to be hot to work.

[Banjo]

Hi Taz,

This has been an ongoing project. The initial idea was to experiment with various trigger systems to get the ignition side perfect, before doing the EFI side of the conversion.

 

Initially this 4K on the test stand had a carby fitted.

 

 

I wasn't too keen on the "after-market" add-on crank shaft pully sensors, with brackets & adjustments, all being sources of potential problems.

 

I was keen on a flywheel sensor, using magnets & Hall Effect sensor.

 

Initially I drilled two holes into the flywheel, 180 deg apart, with 25mm x 10mm rare earth magnets fitted.

 

The face of the magnet was accessible via a slot in the bell housing.

 

 

 

 

This worked well, but I needed a CAS to determine whether the trigger was 1 & 4, or 2 & 3, as I was running a waste spark system.

 

This I did by gutting a distributor, and again fitting some magnets & Hall Effect sensors. This worked well, using the Haltech E6K ECU.

 

 

I even experimented with a Nissan slotted disk CAS fitted to a 4K dizzy, but was not satisfied with the accuracy & stability of the result. It was all getting a bit too complicated for the simplicity & reliability I wanted from this system.

 

I then discovered a Hall Effect sensor which could tell the difference betweem a "N" & "S" pole of a magnet.

 

I then revisited the flywheel sensing, but this time fitting two rare earth magnets through the flywheel, rather than radially. One has a "N" pole facing out, & the other a "S" pole facing out.

 

 

That's the magnets sitting proud at 10 deg BTDC & 170 deg on the other side of the flywheel.

 

The sensor is on the little green PCB, for the experiment.

 

The system worked perfectly & the stability and accuracy of the strobe was unblieveable, compared to the other experiments I had tried.

 

Ultimately, the magnets were moved to 80 deg BTDC, so that all computations could be done in the E6K before the firing point, on the same revolution. The sensor is finally located on the "flex" plate between the engine block & the bell housing, down under the oil filter mounting block. It's accessible & proteceted, and doesn't require slotting the bell housing. The Hall Effect sensors position is fixed; not requiring any adjustment as the initial advance setting (10 deg BTDC) can be set in the Haltech. It achieved the simple accurate system I was after, with direct fire, and "no distributor".

 

Got a high pressure fuel pump arriving today, so shouldn't be long before I fire it up, & start playing with the ECU maps to see what it can do.The E6K ECU is being fed with all the normal signals, like TPS, Coolant & Intake air temps, a MAP sensor & the O2 sensor.

 

Cheers Banjo

[Banjo]

Altezzaclub,

When you say a pair, I gather you have a narrow band O2 for control, & a separate wide band O2 for display ?

 

My research has shown that the ideal operating temperature for O2 sensors is about 320 deg C.

 

I suppose I could drill a little hole in my daily drive exhaust, at the point where I intend to fit the O2 sensor, and stick a thermocouple in there and measure the actual temperature ?

 

Appreciate these comments & feedback.

 

Cheers Banjo

[Banjo]

The sensors only work effectively when heated to approximately 316 °C (600 °F), so most newer lambda probes have heating elements encased in the ceramic that bring the ceramic tip up to temperature quickly. Older probes, without heating elements, would eventually be heated by the exhaust, but there is a time lag between when the engine is started and when the components in the exhaust system come to a thermal equilibrium. The length of time required for the exhaust gases to bring the probe to temperature depends on the temperature of the ambient air and the geometry of the exhaust system. Without a heater, the process may take several minutes. There are pollution problems that are attributed to this slow start-up process, including a similar problem with the working temperature of a catalytic converter.

The probe typically has four wires attached to it: two for the lambda output, and two for the heater power, although some automakers use the metal case as ground for the sensor element signal, resulting in three wires. Earlier non-electrically-heated sensors had one or two wires.

 

Found this on Wikipedia.

 

Seems to confirm what Altezzaclub notes about delay after start up, (in a cold place in winter like Orange) and my reading that ideal operating temp is about 320 deg C

 

Cheers Banjo

[Taz_Rx]

So with your new flywheel sensor setup you're still obviously restricted to a waste spark setup and can't run quad coils?

I'm just about to test and fit my new hybrid k/sr20 cas so that I have a dual pulse Cam sensor. When you say your Nissan setup was a bit temperamental were you just using the 4 slots or a combination of the 4 and 360?

[Banjo]

Hi Taz,

Correct, the ignition as it stands can only run direct fire waste spark. To give sequential firing, it would need a simple CAS sensor, to direct signal to ignitor for 1 or 4, and 2 or 3. Not very hard. That's next step.

 

It will be interesting to see if there is any difference between the performance of the two types of system. I'm currently running platinum plugs, so that should remove any issues in relation to spark plug tip wear.

 

The 360 deg Nissan CAS disk, has apparently had some issues with some afternmarket ECUs reading the pulses coming that fast.

 

In my experiment, I settled for a 24 tooth one from Yoshifab, with a synchronising pulse.

 

 

What I found was that some of these disks are water laser cut. The error between edges of the holes in the disk, can give "jitter" in the strobe. However a CAS disk, only used to synchoise the flywheel pulses, is not critical at all, as it just has to overlap the whole advance/retard operating area for each cylinder. The pulse edge from the flywheel sensor determines the trigger point, not the CAS.

 

Most of the high end after market ECUs allow for individual trimming of the trigger points, if the "mechanicals" for a 4 cylinder engine, are not exactly 180 degree, or 90 degree, (if a CAS signal is used). My dear old Haltech E6K ECU doesn't have that luxury.

 

My issue has always been with using the camshaft as a trigger point for ignition, when in any old K series engine , invarably has some "slop" in the timing chain and oil pump/distributor gear. With the trigger taken off the crankshaft, whether pulley or flywheel, this issue is illiminated altogether.

 

 

 

Cheers Banjo

[Banjo]

Hi Taz,

Is this the one you are using in your hybrid ?

 

 

How have you incoporated it into the K series engine ?

 

I tried a CA18DET CAS, built into a 4K distributor. It was a very tight fit, but we got there. Then I discovered the Toyota 7K dizzy was much large in diameter, but had issues in other areas, as it is not interchangable with the 4K & 5K distributors.

 

Cheers Banjo

[Taz_Rx]

Yeah that's the type of disc I'm using. I read on some Nissan forums that some rb & vg engine owners could suffer top in misses caused by timing belt slop through the 360 slot reference trigger. I am only going to use the 4 slot sensor with an extra hole drilled after 1 to give a cylinder 1 dual pulse. I have another custom cas made from a k lower dizzy half welded to the top of a stanza cas. I wired it to my ecu and tested it, worked well but I then discovered the welded housing was miss aligned. Hence I've started again and retro fitted the sr20 sensor into a k housing. I'll be uploading a whole heap of pics of it in the next couple of weeks.

[Banjo]

Hi Taz,

Looks like we have travelled down a similar path. I'd love to hear what your results are, and see the pics. Here's some pics of the one I made, in a K series distributor body.

 

 

 

 

 

 

Cheers Banjo

Edited October 28, 2014 by Banjo

[altezzaclub]

I then revisited the flywheel sensing, but this time fitting two rare earth magnets through the flywheel, rather than radially. One has a "N" pole facing out, & the other a "S" pole facing out.

 

Damm that is clever! All those 4AGE conversions where you have the dizzy on the back of the motor, poking into the firewall... You've just solved all those problems!

 

I'm running two narrow-band oxy sensors, one in #1 and one in #3 for reading each of the twin carbs. I have a switch on the display to flick from one to the other and make sure they are the same.

[Banjo]

Hi Altezzaclub,

Now I understand why it was an interesting exercise getting your two (2) off O2 sensors fitted. They had to be in the #1 & #3 pipes of the extractor, but as far away from the head as possible, making it just behind where all four pipes converge on the extractor.

 

The technique for installing the magnets is not hard. The magnets are enclosed in some heatshrink with the adhesive already on the inside of the heatshrink tube. The heatshrink prevents s/c magnetic path through the flywheel, forcing the field to shape well around the poles. The two (2) holes in the flywheel are drilled to the diameter of the rare earth magnet, plus heat shrink. The holes are then lined with fresh "Araldite", & the rare earth magnets fitted with their faces flush with the backside of the flywheel. The Hall Effect sensor to magnet gap, is not critical, with the rare earth magnets. I'm running a gap of 3-4mm and the signal is very stong over a wide temperature range.

 

When I had them radially fitted on the edge of the flywheel with my early experiment, I was always concerned that they would fly out of the flywheel at high revs, despite the Araldite. I therefore sunk them slightly below the surface of the flywheel, and "peened" the holes around the edge. Fitting the magnets across the flywheel overcomes that concern all together.

 

Cheers Banjo