Xmt - Sarty's Xsara GTI6 Turbo

[welshpug 1,621]

700 would have been ballpark for a manifold mike already has a jig for, on an n-a engine, like a tu5j4 for example.

 

hell of a craftsman Mike is, look forward to seeing his work on this.

[Type-R 3]

On 25/10/2017 at 2:09 PM, Andy said:

Right. You have had more advice , opinion and ideas than you can shake a stick at. So here is my experience , albeit of the Fiat/Lancia 16v turbo as found in the Integrale ( had three over twelve years) . Anyway, in my last one, I did some interesting development work with Guy Croft and a company called RP labs who specialised in ecu work. In particular, they produced some hardware and software upgrades for the Marelli P6 ecu that was fitted to all Integrales after 1988.The are a serious piece of kit and with the RP labs upgrades, extremely serious. Only snag is that only about three places in the uk can access the bespoke software and map. So I had a long trip to Blackpool to Motorsports Developments who are an RP labs agent.

Where is this all going you ask, whilst trying to stay awake. Guy and I wanted to see what the standard Integrale engine (210bhp) would produce with just head and management mods. So, bottom end was completely as God and Lancia intended, complete with cast pistons, the wonderful F/L nitrided crank ( yes, standard, but utterly impossible to regrind) and standard 8.5:1 compression ratio,. Guy did a job on the head with all the usual stuff so that it flowed as well as it could still using standard valves . In the meantime, I worked on the loom, engine management , fuel supply( surge pot, dual pumps) oil supply and cooling, inter cooling etc. In other words, spent more on the installation than I did on the engine. Engine management from the mighty P6 included full wastegate control, over boost management etc. Turbo, manifolds standard but the exhaust system was 75mm stainless . . . So what did it produce. ? 274bhp at 6000 rpm. The mapper as MSD was ultra keen to spin it up to 6500rpm but I had imposed a6000rpm rev limit ( standard bottom end) . He was confident it would easily have made 300 bhp had I allowed him to go on. At the point he stopped, the Power was still climbing and the torque was flatter than the Cambridgeshire fens. Boost began at 2400rpm ( and remember that this turbo was a standard out of the arc Garret T3 , and not very big to boot) .

To drive , it was an absolute dream and stone reliable. The new owner of this car ( sob) who has now had it for nearly 4 years uses it hard and has not put a spanner to it in all that time, except to change the wonderful Recarro seats for some nasty Evo 1 ones( he has an ‘ample fundament’

And the point of all this? The F/l 2.0litre 16v engine is remarkably similar to the Peugeot Xu10j4( with or without the RS) . You are going for the best flowing head you can lay your hands on, a modern turbo( do not go mad on compressor size. You want it to spool up from low rpm) . Standard bits will cope, whatever bottom end you use as long as you are not power greedy( and you are not). But, and this is, in my humble opinion, very important, engine management and power unit installation are critical both to reliability and output, not to mention whether you end up with a car which is nice to drive because you are not endlessly worrying about fuel pressure, or oil temperature etc. . Be prepared to spend more on all this than you do on the engine . And that includes good instrumentation , oil, air and water cooling, fuel managemt ( I would go for a swirl pot every day of the week with the in- tank pump used as a lift pump and a Bosch motorsport pump taking care of the injection from the swirl pot.

Sermon over. ‘This is merely my experience . I am not a motor engineer ( o.k at quantum electrodynamics and high energy particle Physics but bugger all else) , so please take ( or leave ) my views with however much salt you prefer.

Yours,

Andy

If I'm following this, could a Fiat/Lancia P8 ECU therefore be modified to work on any 4cyl turbo engine?

 

BTW glad to see this coming along Doc.

[petert 549]

I think it's bollocks on the 8 injectors idea. A modern ECU and four new injectors can do your power range easily and still idle like a kitten. 

 

Have I used the word "bollocks" correctly? It's not in the Straylan vocabulary.

[welshpug 1,621]

Seems an apt use of it yes, however it is Richards money so we'll let him have a play and learn

 

Theory is that shifting this volume of air with the aid of boost the position of the injectors will make less difference than it would in a Naturally aspirated engine.

Edited December 13, 2017 by welshpug

[Andy 105 1 Cars]

Yep. In principle the P8 ecu could be used to run any 4 cylinder engine provide that you bought with it the RP labs software and hardware upgrades. The only difficulty is the paucity of people who are happy to map the bad boy, as they need access to the ( restricted ) software. ( or they have managed to set up an emulator ) .

When buying the ecu and upgrades from RP labs, as a customer one is given the relevant software and mapping tools , but you still need a rolling road and someone to drive the laptop who is familiar with the RP labs software .

[Andy 105 1 Cars]

 

Peter,

 I am slightly surprised to,learn that the term’ bollocks’ is not in the Straylan vocabulary. It is derived from the Middle English and is often used colloquially to mean ‘ useless’ . Given the origin of the word, I am astonished that it did not travel to the Southern Hemisphere on one of the many journeys made by so many from the UK several hundred years ago .

 Sorry for the hijack Rich, but I have just had a couple of glasses of wine .

p.s It is possible to summarise this post in one word. I think Peter has mastered it already 

Andy 

[DrSarty 90 1 Cars]

A long time coming, but here's my reply to the last posts (querying my sanity) and a general update. Pics to follow shortly after this.

 

I'm not being a numpty nor a dreamer re the 'staged' injectors. It was simply a question of economics.

 

I had an opportunity to acquire 2 brand new sets of injectors which combined, were cheaper than a set of suitably high flow-rate injectors; so I took it.

 

In my research - and perhaps stating the obvious here - it's all a question of chemical properties/potential energy exchange. One needs X amount of fuel and Y amount of air (at the appropriate circa 13-14:1 ratio to achieve optimal burn) to generate Z amount of explosive energy suitable for driving a piston down (or around if you're thinking rotary). There are plenty of on-line calculators relating to N/A and F/I engines, which I'm sure are backed up by sensible science. The Forced Induction (F/I) needs are far higher than the Normally Aspirated (N/A) ones, and understandably so, because with F/I there are more air molecules per volumetric unit meaning more fuel is required. If you get this right, bigger bang for the displacement and Bob's your uncle.

 

From memory, 300bhp ish in an in-line 4 cylinder (with certain assumptions) requires something like 520cc/min.

 

So whilst - as PeterT and others are alluding - there are injectors out there that can (easily) support that (and then some, as I've heard of 2000cc/min plus on my travels), there are 2 considerations: how much do the injectors cost, and are they any good under 'normal' driving conditions?

 

With an apparent 80% ideal duty cycle, I suppose I was always looking at running 600cc+ injectors. But I'm guessing these cost money, and in my project explorations I found 2 sets of brand new 320cc injectors for under £100 giving me 640cc/min flow, i.e. some headroom. I've always believed headroom/flex is a positive thing, so despite the imbuggerence of running 2 fuel rails (in series and power/signalling them appropriately), as I've done it before it's not really that difficult. AND...who knows...despite your scepticism it MAY make a difference.

 

While I'm open to criticism for many things, including over-complicating sh1t, at least I'm prepared to try things. Most differences in stuff these days are negligible rather than radical, and small steps often get us there.

 

I MAY get better idle, better fuel economy and (as often happens) better throttle response/performance for minor design nuances like this; but ultimately this was just an ease and economics of choice solution, and something to try for either my loss or everyone's benefit. In a nutshell: wind your necks in. Someone's gotta try, plus I may have saved a few quid.

 

If everyone just wanted an 'ultimate', easy to obtain, all-round, dependable road and track weapon, we'd probably all just buy a WRX, Civic Type R, a Mitsu Evo (lower spec), a Golf GTI or Golf R, or maybe an Audi S3/RS3 (the RS1's too small). Other candidates include the BMW M3 or Z4M etc, a Mazda MPS 3 or 6 (I have a MPS6 BTW) or perhaps a Focus ST or RS. Either way, for budget money (thrills vs bucks), I think the choice is quite limited, hence why people like us try and do something different.

 

In closing....what is the f*cking point of doing what loads of others do? You are undoubtedly correct - all who've commented - that I could just run 1 fuel rail with 550cc/min+ injectors, but I would have to buy the injectors. If it's just a matter of buying sh1t, then where's the fun, and where's the experimentation?

 

Rant over and generally speaking:

I've been to see Mike and the car in Wisbech. He has quite an impressive workshop and apparent clientelle! A gorgeous orange mkII Escort, another lovely Ford, and a 6-figure Mitsubishi Evo...and then my £290 Xsara!

 

The exhaust studs starting snapping in the (old) head, so the guys had to remove the engine. I wish they'd told me, as they put it back in and reconnected the driveshafts. They could've left those out, as the engine in there is just a dummy really; it's only for turbo placement and manifold clearance reasons currently.

 

The fabrication looks as genius and quality as expected, and looking forward to collecting it in a few weeks. £1,200 well-spent one hopes........

[wicked 103 3 Cars]

Did you notice that the K6 ecu doesn't have 8 injector drivers?

Denso 630cc injectors work perfectly fine under normal driving conditions (if your ecu gives proper edges on the switching sigals) and can be found for 180 quid/set. 

[petert 549]

This is how I think it is, maybe I'm wrong, someone please correct me if so. Great discussion regardless! Good brain food.

 

Fifteen plus years ago, it was fashionable to run staged injectors for a numbers of reasons. Firstly, ECU and injector technology didn't allow small enough injection times to control the mixtures at idle, when a single injector was used. Primarily this was because tuners favoured multipoint/batch firing rather than sequential injection. The latter halves the injector dead time. With multipoint and early injectors, it was difficult to get the idle injection time under 2.5ms. Running staged injectors gave the best of both worlds.

 

Secondly, on a normally aspirated engine, mounting the injector at the bell increases the efficiency. Don't quote me on the numbers, but let's say efficiency goes from 95 to 105% percent. I doubt there can be any increase in efficiency on a forced induction engine, when you're already pushing air in at say 200%.

 

Haltech for example, dropped staged injection at least ten ago, as the combination of modern injectors with better spray patterns and less dead time, coupled with the trend to sequential injection, meant that control of idle mixtures was easy, even with 900cc injectors.

 

Injectors in my mind have never been cheaper. Running an additional four pairs of wiring and an additional fuel rail is unnecessarily complicating a modern N/A installation. On forced induction, I just can't see it.

 

If you've already got all the parts, then sure.

 

 

 

 

 

 

[wicked 103 3 Cars]

I agree with PeterT; I would not bother with staged injectors (yet), given that you have plenty of challenges already. I would first make it work, maybe with 440cc injectors, and if the engine is running properly, scale up and make things more complicated. Even with 4 injectors, it will not be the  'ultimate', easy to obtain, all-round, dependable road and track weapon. 

 

As said; great thread and looking forward to the fabrication!

(pls read all the comments as constructive criticism; not intended to flame you, but to help to make till the end) 

 

 

[B1ack_Mi16 67]

Just for the sake of it, my T16 (still just roughly mapped), idle quality on 1000cc/min injectors fully sequential.

 

Almost just after first start-up. Injectordynamics ID1000 ones, which apparently are both flow- and deadtime matched.

20161215_213308_zpsj4iw9gy1.MP4

20161216_220653_zpsw0spv6qe.MP4

[DrSarty 90 1 Cars]

Nice bolt balancing trick!

 

As Pete said, I have the parts. Is it worth it technically speaking? No.

 

Watch this space. I now have the swirl pot too, and the engine can go away to be built. Pics of the parts and the turbo manifold to follow soon.

 

Thanks for contributing.

[Andy 105 1 Cars]

Not directly commenting on the above thoughts, the Emerald I assume you will wire semi- sequentially. That is , Injector drivers. 1 and 2 will run the inner 4 ( 1+4 and 2+3) and drivers 3 and 4 will perform same trick on the outer set. Complication of two fuel rails? Guess so but then you are having to fiddle around with an inlet manifold anyway, so I guess an extra four injector bosses will not add to the cost significantly . Given what you already have and what you have to do, it has to be worth the experiment.

[DrSarty 90 1 Cars]

Basically yes. I didn't want to get into the detail, but the K6 has 4 inj drivers which means batched inner and outer is possible; I should know, I've done it twice before.

 

Ironically - and this is a sign of the times, technology and opinions changing, and even on this forum - many people used to say that sequential injection was a waste of time!

 

I will just have 4 injector seating and 2 rail mounting bosses cut and welded into the GTI6 inlet manifold, probably with the outer rail being underneath the runners. People just have to accept that I'm doing it this way as I have the bits (and mindset) already.

 

I suspect the alternative is different/bigger (more expensive) injectors running sequentially. I'm just committed to 2 rails running batch; I've already done the wiring and have all 8 injectors (brand new, 640cc/min total flow capacity) all ready to go for £100. Not really an issue for me.

 

For it to be 'an experiment' (read: test), we'd have to have something suitably equivalent to compare it with. This isn't practical, and also boring. It'll be what it will be. We could be talking 0.2% power, torque or fuel efficiency difference, and if we're into that level of detail we're not having fun are we? My only submission is that the mapping may be slightly tougher and I have more components to go wrong. Ho hum...tough titty...I'm doing it my way!

 

@Wicked: Anyone can 'flame', criticise or otherwise all they like. This forum must know me by now; I go 'fishing' for responses and ideas and take it ALL on board. I then make my 'own' decisions, rightly or wrongly, but each time it's part of a unique journey.

And to clarify: I was never proposing my car would be the ultimate anything. I was just talking about the concept of the 'all rounder car', e.g. the MkVII Golf GTI type of thing. Why would we (idiots like me) try and build something that competes with a car like the Golf or Civic as I mentioned? Those cars will always have the upper hand, and we all know why.

But when we build a car we have to have a goal of some sort, and even if it misses/falls short of matching (or beating) any of the 'off the shelf' all-rounder cars out there, the point is we tried and did something different. A Xsara 1.9RS twin-scroll turbo doesn't exist anywhere but in my plans as far as I'm aware, and I'm going to make it happen. It'll be different if not optimal; but optimal is either limitless funds, or something off the shelf, and I desire neither of these.

Edited February 6, 2018 by DrSarty

[wicked 103 3 Cars]

I did not intend to say that your car would be the all rounder car if you use only 4 injectors; I meant that building a  Xsara 1.9RS twin-scroll turbo is all ready so non standard, that you'll have enough issues to solve. I think we've seen project fail on this forum because people wanted just to much custom/special on their car. 

So respecting your decision, I'll add some advise:

 

K6 manual:

Quote

Using a multi-meter measure the resistance present across the 2 pins of the fuel injectors - be sure to measure the resistance of the fuel injector and not the plug on the wiring harness! A reading of between 12 to 16 ohms categorises the injector as “High impedance” and is suitable for use with the K6 ECU. A reading of between 2 and 4 ohms means the injectors are “Low impedance” and that they can not be used with the K6 ECU unless suitable Ballast resistors are fitted.

 

If you put 2 injectors on 1 output in parallel, it will result in an impedance of ~8 ohm. You risk frying the end stages in the ECU or get crappy switching characteristics on your injectors (=crappy idle).

So you better add an extra driver stage before the extra set of injectors to keep the impedance high enough.

 

[jackherer 543]

http://www.emeraldm3d.com/faq

 

Quote

The K6 ECU has six injector drivers and six ignition drivers therefore on a four cylinder engine you can run four sequential injectors and four batched semi-sequential.

 

They're overload protected outputs too so you wont damage them.

[Andy 105 1 Cars]

Couple of K6 related points here. ( Nod off young Rich as you have a plan that will work) . If one needed to, one could batch fire two injectors in parallel. That would produce a sinking current of  around 1.5A. Given that an ignition driver can be configured to drive a COP which will sink 7 or 8 amps, the drivers are fine up to 10 A  and, as stated above, overload protected. The snag with using all the available drivers as ignition or injector drivers is that it prevents them for doing other jobs. For example, pin 4 is potentially ignition driver 3. It can also be usefully employed as the  main ecu power supply switch on . Apply 12v from the ignition switch to pin 11 and pin 4 will ground, which is a handy way of triggering a relay to supply power to pin 28 ( main power supply ) The real advantage of configuring the ecu this way is that, in the event that, say , the cooling fan is running when the engine is switched off, the ecu will remain powered, keeping the fan running until the cutoff temp is reached . If 11 and 28 are wired together , then this function is lost and one needs to wire ,( via a second delay and the switch on the rad) the fan to stay on if the engine is hot and the ignition is off .

Having said this, I agree, lots of options to use the available drivers in the K6 in lots of ways . What I have discovered is that Emerald ( ie Dave Walker) is not overkeen on using all four ignition drivers to run COP. The K6 will do it , but when I discussed it with him, he was very strongly in favour of using just two ignition drivers on a wasted spark coilpack ( or indeed  the splendid PeterT solution of four LS1 coils running in wasted spark format but each coil dedicated to just one cylinder )

And with that , you may all reasonably conclude that I really do need to get out more .

[DrSarty 90 1 Cars]

Xsara comes back to me next weekend. Mike's work is done. More pics with the old engine out soon.

 

4 matched exhaust runners in pairs (1+4 / 2+3) to feed the twin-scroll, which has a mounting bracket off the gearbox between the turbo and downpipe. Turbo 'clocked' so oil drain is at bottom and to aid plumbing. Lambda boss pictured with a bung on downpipe which currently ends just short of the car's underside, so rest of system (to follow in 6-8 weeks) will start there with a flange and flexi.

Edited March 24, 2018 by DrSarty
Spelling

[DrSarty 90 1 Cars]

Before I lift the engine out, I intend to mock up the intercooler & pipework; that means I need an intercooler. Any recommendations on a donor, at the very least in terms of size? I've heard too big is bad.

 

Quite naturally inlet & outlet placement on the IC should be optimised to pipe routing, but I'd like to get an idea of size. I intend to change to a pull fan & use the space freed up on front of the rad for the IC.

 

Battery box needed for boot (plus wiring) now snail is in the way. Swirl pot can be fitted too with pipework, then the engine can come out, be split, & sent off with the Mi rods & low comp Mi pistons, GTI6 crank & head, Kent (Piper?) cam pulleys & turn into a hybrid 1.9 something.

 

Probably another £2,000 in costs to look forward to which means a forecast of £4k overall. Considering my Mazda 6 MPS is now valued a about £3.5k, it's not a silly question to wonder why we do this!! 

[Andy 105 1 Cars]

Looking good Rich. Why do we do it? Not everything in life is about value for money , and if you manage to do it on £4K , you will have done very well in my view.

Andy

[petert 549]

That exhaust manifold looks the goods! Love it. Well done.

[allanallen 528 2 Cars]

Very nice. What’s the spacer between the manifold and turbo for? 

[welshpug 1,621]

the the support bracket.

[allanallen 528 2 Cars]

17 minutes ago, welshpug said:

the the support bracket.

What’s the reason for doing it like that rather than directly to the manifold? 

[DrSarty 90 1 Cars]

Quite a lot of weight for the runners to support.

 

It also means the turbo, downpipe or manifold can be removed independently. Everything stays where it is.

 

Plus it all moves together, which overall means it should last longer.

 

Apology: Support bracket - the black bit - is between the turbo & manifold. (Not the downpipe)

Edited March 25, 2018 by DrSarty