Yet another XU9/10 engine build

[Andy 105 1 Cars]

Update, but on ancillaries . My Satchell inlet arrived recently. It was derived from a GTi6 inlet, so one eBay purchase later and I became the proud owner of a gti6 fuel rail. The reasoning is that the Satchell inlet mounts the bodies upside down, making the injectors slightly more difficult to get at , so I plan to used the Satchell sourced blanking plugs on the bodies and put the injectors in the o.e position on the manifold. However, all my fuel hosing is in -6jic. I was going to weld on some adaptors to the rail but then, on reading a thread in the 2010 archive, came across a reference to some -6 adaptors from Torque u.k which clamp on to the 8mm return and 10mm feed pipes via an olive. I do have some weld on adaptors too, but my welding skills are marginal when it comes to a delicate weld on rather thin 8mm tube, so I will try these olive located adaptors and see .

The fuel rail responded well to a good soak in Nitromors , stiff wire brush, then an equally good soak in Bilthamber deox ‘c’ ( brilliant stuff) . Paint to follow and I should have one fairly neat fuel rail installation .

Oh. Inlets. The photograph shows both the Satchell and the Jenvey version ( yes, I seem to have both at the moment ) . As I have a BBM radiator , I am hoping that the angle of the Satchell will point the filter in the nice big gap between the top of the radiator and the underside of the slam panel, where there is nothing but fresh cool air ( he said hopefully ) 

[jackherer 543]

Very nice!

 

This is going onto an XU9J4 head isn't it? If so have you checked the alignment dowel holes on the GTI6 inlet? I had to add some new ones as they are slightly different on the GTI6.

 

What are the Satchell injector blanking plugs like? I'm using some of the Jenvey ones that they don't recommend for permanent use so I should probably change them sometime...

[petert 586]

I'd still find someone to silver solder the 'weld on' fittings on.

[Tom Fenton 1,541]

Andy, I'd consider leaving the injectors in the bodies, theory being better homogenous charge through more time for fuel and air to mix gives better torque.

[Andy 105 1 Cars]

1 hour ago, jackherer said:

Very nice!

 

This is going onto an XU9J4 head isn't it? If so have you checked the alignment dowel holes on the GTI6 inlet? I had to add some new ones as they are slightly different on the GTI6.

 

What are the Satchell injector blanking plugs like? I'm using some of the Jenvey ones that they don't recommend for permanent use so I should probably change them sometime...

 

1 hour ago, jackherer said:

Very nice!

 

This is going onto an XU9J4 head isn't it? If so have you checked the alignment dowel holes on the GTI6 inlet? I had to add some new ones as they are slightly different on the GTI6.

 

What are the Satchell injector blanking plugs like? I'm using some of the Jenvey ones that they don't recommend for permanent use so I should probably change them sometime...

 Photograph of blanking plugs below. Looks like they are designed to take an ‘o’ ring. I will oblige before fitting 

[Andy 105 1 Cars]

Thanks for the comments. The injector location really is a topic for debate .There are advantages both ways, as I understand it . Modern injectors with multi hole nozzles running at 45psi should give a good spray  pattern and good atomisation without the need for the aerodynamic aid of the 150mm of extra pipe flow. Having said that, 

point taken about helping  produce a homogeneous mixture prior to cylinder entry . I must admit Tom that the neatness of installation with the built in FPR, rather then mounting my FPR remotely , as I did last time via a ( very good) Satchell adaptor which I would need to do if I retain the Jenvey fuel rail for body injector location does feel attractive to me. In the end, open to experimentation as I now have both types of rail and all the other bits required to put the injectors in either location. 

Peter, as for the pipe fittings, I will start with a pressure test. I take your point about being Silver soldered ( or brazed) into position . If I can find someone local to do that little job, I will whip off the screw on fittings and get some -6 ones ( which I already have) fitted instead 

[jackherer 543]

I wanted to try both injector locations when I mapped mine but I ran out of time unfortunately.

 

The Jenvey plugs are similar but in two pieces with a bolt to pull them together. https://store.jenvey.co.uk/injector-pocket-plug-ipp02

[Andy 105 1 Cars]

56 minutes ago, jackherer said:

I wanted to try both injector locations when I mapped mine but I ran out of time unfortunately.

 

The Jenvey plugs are similar but in two pieces with a bolt to pull them together. https://store.jenvey.co.uk/injector-pocket-plug-ipp02

When I finally get to mapping , it is one of the things I would like to try . The other will be any difference between semi sequential and fully sequential injection. I have my injectors wired independently and each with its dedicated driver via the ECU and a cam sensor , so changing between the two is not much more than a couple  of mouse clicks . Swapping injector locations it a bit more tricky but if I have a set of hoses all ready attached to the Jenvey rail, it should not take too long 

[welshpug 1,657]

i'm using the standard rail and reg with two sets of injectors, not sequential as I only have an S40 so not enough drivers!

[Andy 105 1 Cars]

That’s the problem with two sets of injectors . The Emerald has the same limitation. However, for the engine I am building,where maximum power is not the main build criterion, one set of injectors will, I am sure , work just fine .

[welshpug 1,657]

Twin injectors isnt about peak numbers, its about cleaner running and better torque.

 

mine isn't built for peak numbers at all, its a standard head!

[petert 586]

Twin injectors is a bit old school now. Most ecu's don't offer staging anymore, as a single modern injector, mounted out, will do the same. Modern ECU's have the ability to minimise injector dead time, which was the problem with using one large injector. It's not uncommon now for a turbo 4 cyl to idle at 800 rpm and make 500+hp.

 

Below is a stand off injector setup, but you can only run that if you have full sequential injection, eliminating the chance of fuel reversion.

 

Over here, silver soldering and brazing are two different procedures. Silver solder, with an Ag content of approx. 45% will melt at approx. 600ºC, compared to a brazing rod (more like brass) which melts at approx. 900ºC. Thus silver solder is a much cleaner process and easier to clean up, as the steel won't oxidize, leaving the inside of the pipes clean.

Edited August 18, 2018 by petert

[Andy 105 1 Cars]

Morning. Yep. Brazing much more aggressive as the temp is much higher. Silver solder on the hose fittings is the best solution. As for staged injection, it clearly has advantages for some uses as folk on here report the benefits, and I am pretty sure that Richard (Sarty) is going down that route with his Vts turbo. I do not plan to add the complication, especially as I have used the drivers in my ecu for other things and do not plan building another loom , having only completed one last early Spring 

[allanallen 528 2 Cars]

We normally tig braze fuel fittings and fuel rails. We’re down in Matlock Andy if you have no joy with the compression fittings and want them brazing on. 

[Andy 105 1 Cars]

49 minutes ago, allanallen said:

We normally tig braze fuel fittings and fuel rails. We’re down in Matlock Andy if you have no joy with the compression fittings and want them brazing on. 

HI. I think I had better drop round next week. The compression fittings went on, but a permanent tig welded joint seems far better. 

[DrSarty 90 1 Cars]

Great stuff Andy. Keep it up.

[petert 586]

21 hours ago, Andy said:

 The other will be any difference between semi sequential and fully sequential injection. I have my injectors wired independently and each with its dedicated driver via the ECU and a cam sensor , so changing between the two is not much more than a couple  of mouse clicks . Swapping injector locations it a bit more tricky but if I have a set of hoses all ready attached to the Jenvey rail, it should not take too long 

There's a bit more to it than that. Some ECU's use the cam sync sensor to determine either the start or the end of the injection period. Others calculate both via software. Thus if considering it, you need to contact your ECU manufacturer to determine if there are any hardware considerations before you press on the cam sensor tang. I've sent you the attached tool which visually lets you see where it's going to fire, based on cam timing and injector duty. The two attachments are for 2000 rpm cruise and 7000 max load (with the injectors maxed out at 85%). You really need to find the optimum closing point for each load range. Also attached is how that's done in my ECU. It takes considerable dyno time to fine tune each point. Here's how it's done:

 

"The Injector firing angle is the angle at which the injector will SHUT after the injection even has occured (relitive to a full camshaft revolution). The injection angle depends on things such as camshaft profile, intake manifold design, fuel type, injector location as well as other external factors.
Injection angle is also used in certain applications to cool in the inlet valves.
When tuning the end of injection angle (with the exception of above) you want to make sure all injected fuel is taken in by the motor and not "blown" back up the inlet manifold. (Look on Youtube for old F1 engine dyno tests to see the fuel spraying all over the place)

In practice the easiest way to do this is by tuning your engine to a known AFR. Next, adjust the injector firing angle until the engine becomes as rich as possible, now we know that the engine is burning the maximum amount of injected fuel. Do this for each RPM site you require.
You will notice that the engine will be slightly more responsive at lower RPM with the correct injection angle.
Please note, if the engine is not tuned to accpetable mixtures (rich and lean spots all over the place) that injection angle will not make much of a diffrence and you should spend more time on the tune." Scott from Haltech

 

From the Injector Angle app., you can see why Sequential Inj becomes somewhat irrelevant at full load (85%). ie the injector needs to open for the NEXT cycle before the inlet valve is shut. The benefit comes at low-med revs., the stuff that pulls you out of corners.  Thus it's worth spending some time from 2000-4000 optimising the injector period.

[Andy 105 1 Cars]

Wow. Thanks for this Peter. I now need to look at the section in the 63 page Emerald guide . I think that I will need a chat with Dave Walker , who developed the Emerald, as I am sure that the guide does not go into this much detail, but merely offers the option of sequential injection . 

[Andy 105 1 Cars]

That reminds me. I read a paper some time ago that discussed some engine dyno results on a Vauxhall XE 16 valve engine . It was done at an engineering department in a University in Ireland . Anyway , the conclusion concurs almost precisely with your post, but until I had read your post, I  did not understand why. That is, at high rpm/ full throttle , the difference in engine output between fully and semi sequential injection was within the measurement error, ie negligible . Different story at low to medium rpm, as you say above 

[Duroc 2 1 Cars]

I've used those compression fittings on a gti6 rail, and they've been fine for the last couple of years 

[Andy 105 1 Cars]

I put mine on, and then  undid them. The olive on the feed pipe pulled off almost immediately. The return one seems to have held. Off to Bridgecraft tomorrow to get a permanent job . When I did them up, the return tightened progressively, so I think that one is secure, but the feed did not behave similarly.I think the olive was too looose a fit on the pipe prior to tightening, so it did not clamp sufficiently 

[Andy 105 1 Cars]

Afternoon. I have been doing some reading to update my understanding of the sequential injection protocols in the Emerald. The set up menu does not require input of cam timing data. Further, it does not suggest that the cam signal position is critical, but rather invites using a default value of 0 degrees. I will talk with Emeral tomorrow, but all the above leads me to suspect the the ecu uses the cam trigger signal to initiate the start of the injector  firing cycle .At high rpm, when the valve is open for less time ( in ms) than the injector needs to remain open in order to fuel correctly, this is not important. However, at low rpm when the injector will need to fire for much lesstime than the inlet valve will be open, it would be nice if the injector firing and the valve opening coincided, as per the screenshots that Peter posted above. 

If this is indeed the case, then I think I can ‘fix’ it by fitting the tang on my inlet cam so that it is aligned with the hall sensor at exactly 30 degrees btdc . That will coincide with the inlet valve beginning to open.

This is a bit of speculation , give that I have not talked with Dave Walker yet, but I will report back . I imagine that others on here will have used the K6 with sequential injection , so will have travelled this path before 

[Tom Fenton 1,541]

Andy, I'm sure theres a table in the Emerald software somewhere for injection angle?

[Andy 105 1 Cars]

Cheers Tom. I will power up the old laptop on which the software resides and have a look.

[Biggles 65]

8 hours ago, Andy said:

I have been doing some reading to update my understanding of the sequential injection protocols in the Emerald. The set up menu does not require input of cam timing data. Further, it does not suggest that the cam signal position is critical

No offence intended but I would suggest some more research is called for.  As Petert posted above, the injection window relative to valve opening is significant (which requires cam position to be known).  Having spent many an hour on the dyno being paid by a major OEM to look at such things, take it from me that even at WOT the effects are measureable and at part load the opportunity for getting it badly wrong is significant.