Honda L15B7 Direct Injection G4+ Force GDI

[John Cousins]

Hi there Link crew.

As far as i'm aware, I'm the first person in Western Australia to dive into an L15B7 swap into an EG civic. I wanted a forever car and i'm having this car properly engineered in the hope that it's a perfect mix of old vs new with quiet, mild road manners but plenty of punch when asked.

The engine I have purchased has come with the factory ECU but obviously being from 2018 i'm expecting headaches with regard to the ECU not seeing multiple CAN nodes and probably de-rating the engine, or worse still, not allowing it to start at all. That's before we get to the whole immobiliser side of things!

I guess what I want to know is, do you think the G4+ FOrce GDI would be an easier solution? There is a Hondata Flash pro plug in for the factory ECU but I think this is specifically designed for a 2018 Civic Installation and I have no Idea how capable this would be in terms of disabling CAN nodes and allowing it to run in an EG.

Just wondering if anyone has played around with a Honda GDI engine and a Link or would I be kind of pioneering my way through this?

Obviously passing an emission check is critical for this build but it's not the full check, I think WA requires just requires the vehicle to pass a 5 gas analysis.

Also keen to know what electronic dash displays play well with link as it's probably just easier to swap to a completely new display than make the existing dash work with the new engine.

Thanks!!

 

[Adamw]

Yes the Force GDI would be an easy option for this, there is a well-set-up base map for the GDI L15 in PC Link.  This combo is quite commonly used in aircraft.  

[John Cousins]

Hi Adam,

Thanks for your reply.

Looking to dive into this swap as soon as possible. I simultaneously sent a message to Technical Support and they are currently trying to steer me toward the G5. (when it is released)

Strange.. as they didn't even know that the G4 had been proven on the L15 or that it was being used on the Viking aircraft engines?

Are you aware of any real benefits that the G5 would offer in my L15B7 installation? As if there aren't any real benefits, id rather just get busy with a known quantity like the G4 GDI

Seems like you are a bit more in touch with what is going on with regard to my engine.

Also, are pinouts available from link to assist with wiring up the L15B7? I have the factory loom that came with the engine and i'd like to start prepping it while I wait for my transmission to arrive. The loom is only a couple of years old and in great shape so i'm more than happy to break it down and re-plug it to suit the Link.

Thanks again

John

[Adamw]

Ok, I just dug a bit deeper, I didnt initially realise the L15B7 was a turbo'd version, the L15B that our base map is from is the common Jazz/Fit engine which is NA.

So that means there may be significant hardware differences compared to our L15B base map.  Metrics such as injector flow rate, GDI pump lobe profile and electronic characteristics etc are likely quite different.  Unfortunately this data is very difficult for an end user to reverse engineer so I no longer feel the Force GDI is a good fit.  The G5 would be even less so. 

You really want an option where the supplier has done all the reverse engineering for you and can supply a fairly complete map or configuration that will allow it to run "out of the box" with no major engineering required.  If it were the NA L15B, then the G4+ Force would have been as good as any, but with the unknowns of the B7 I have concerns it would be a big job. 

It would be worth discussing with Hondata if the flash pro has enough configuration to allow standalone use, otherwise the only other ecu company I have seen run the L15 is Emtron but I have a feeling they may only have experience with the NA version as well.  

[John Cousins]

Good news.

I have found someone who is running a Link Force GDI on an L15B7.

They are racing a Honda CRZ and have performed the complete 2018 L15B7 Civic Turbo engine and transmission swap.

Just in case there are any others interested in this swap.

https://youtu.be/kflFpYA2UH8

[Adamw]

So did they characterise the pump and injectors themselves and they are going to provide that information to you? 

[John Cousins]

I've made contact with the owner of the team who said he will pass my details onto the tuner of the car.

Hopefully I hear something from them soon.

I've actually discovered a whitepaper document by the EPA in the US which were doing a study on the emissions of modern direct injection cars. The car they were using just happens to be the turbocharged L15B7 civic. The report goes into massive detail about the engine. Things like ignition advance angle under different loads, AFRs etc. all captured by sniffing data from the stock engine whilst running under controlled conditions on a dyno. It's very well graphed and probably extremely useful information for setting up a map close to the stock engine. Quite surprised Honda would allow it to be published TBH.

I'm still trawling through all the details to see what I can apply.

[Adamw]

8 hours ago, John Cousins said:

I've actually discovered a whitepaper document by the EPA in the US which were doing a study on the emissions of modern direct injection cars. The car they were using just happens to be the turbocharged L15B7 civic. The report goes into massive detail about the engine. Things like ignition advance angle under different loads, AFRs etc. all captured by sniffing data from the stock engine whilst running under controlled conditions on a dyno. It's very well graphed and probably extremely useful information for setting up a map close to the stock engine. Quite surprised Honda would allow it to be published TBH.

This is all stuff that your tuner would typically optimise for MBT on the dyno in a few hours, its not all that relevant to your application unless efficiency and emissions is the goal.  

The important thing to understand about GDI is that the characterisation of the fuel pump and injectors is very important and must be accurate.  Your high pressure fuel pump is a positive displacement pump and petrol is basically not compressible.  So, to be able to control the fuel pressure in the rail you must pump the exact same volume of fuel into the rail as the injectors have just released from the rail.  The pump volume must be known, the pump cam lobe profile must be accurately measured, the pump lobe centers in relation to the triggers must be known, the pump peak and hold currents and deadtimes must be known, the injector flowrate, deadtimes, P&H currents and boost voltages etc must all be known.  

[John Cousins]

I understand the basic principles but i'm still unsure of the role of the spill valve. From what I understand, the lobe is driven off the end of the camshaft and that is what provides the huge pressure required for injection required to to spray into the pre-compressed combustion chamber. As for the volume and timing of each "press" of the lobe, I will have to delve deep into the web to see if the information is out there. I'm assuming there is no "relief" to the high pressure pump judging by what you have said above. Could you please explain what you mean by the PUMP peak and hold current? What electronic interface is there with the high pressure pump? Currently I only have the injector flow details which i'm assuming would be measured like standard injectors at 100% duty cycle... but don't stress I'm going to approach this methodically and carefully as I'd really like to take this challenge on and do it properly. I've always wanted an EG civic fitted with a latest generation engine to build a form of retro-tech combo. (yes the EG is now considered retro ha!)

I've found some people online who are currently reverse engineering the factory ECU. Hopefully this will provide the information I need.

Wish me luck.

[Adamw]

The spill valve controls the inlet side of the pump, It is normally open.  You will notice there is no check valve on the inlet.  If we dont energise the spill valve, fuel will be drawn into the cylinder as the piston moves down, if we continue to keep it open after the fill event, when the piston comes back up the fuel will just be pushed back out of the inlet (ie "spilled"), no rail pressure will be generated. 

 

If we want to generate rail pressure then we need to close the spill valve.  You will notice in the pic below which shows it starting to pump with the valve closed, that once it is closed you actually have fuel pressure also pushing the valve closed against its seat - what that means is once it is closed you cant open it again by just "turning off" the valve - it wont open again until the piston is past TDC and the piston is travelling down again.  If you pump in too much fuel then your rail pressure runs away, the only way to reduce the pressure is to open the injectors (they do have a relief valve for safety but it is very hard on the lobe and pump to run at the relief pressure).  

The volume of fuel that is delivered is determined by how far up the bore the piston is when you energise the spill valve.  If you energise the spill valve when the piston is at the bottom then you get the full swept volume delivered, if you energise the valve when the piston has already travelled halfway up the bore then you will get half the swept volume delivered.  There is significant complexity however in determining what crank angle to energise the spill valve at to achieve your desired delivery volume.  1. the pump lobe profile means the volume is not linear to lobe angle.  2. the pump lobe is connected to a VVT cam that moves in relation to the crank.  3. the spill valve action is not instantaneous - the length of time it takes to close varies electrical factors such as battery voltage and the peak and hold currents. 

12 hours ago, John Cousins said:

Could you please explain what you mean by the PUMP peak and hold current?

 To get the valve to close as fast as possible you need a lot of current, you want to allow as much current to flow in as the coil can handle (heat/internal melting), however this peak current only needs to be applied for a short time until the magnetic field reaches saturation, once at saturation there is no advantage in still flowing max current - it just creates more heat, you can reduce the current back to a level just enough to "hold" it saturated.  The ECU controls both the peak and hold currents.  Typically the peak & hold currents can be determined by scoping an engine running a factory ECU, the deadtime is more difficult - typically this is done on a specialised test set-up with an accelerometer clamped to the pump so you can "see" how long it takes for the valve to hit the seat after it is energised.  

 

[John Cousins]

Excellent information. Thank you.