There are too many variables to make it any use IMHO. Save your time, go tune your engine well and keep some spare cash for when it blows.

No one can tell you that 100000 miles is reliable at 380 BHP on a standard 2.0 as it hasn't been tested to OEM standards, any other claimed reliability testing is not worth having. You can assume that the JDM EJ207 or US EJ257 are reliable at 276-300 BHP to OEM standards and that is it IMHO.

 

John , I hear what your're saying and I agree - there are no claims that can be upheld-

(Typed and deleted a load of other stuff here, but what the hell).....

I guess that I'm just trying to compile a list of 'what;s possible and been done', 'cos it seems that there are quite a few people reading many threads, all trying to piece the same information together....

Its a simple question, but with no 'real' answer, because the question is not finite...

Does a steel bottom-ended 2.0 compare well to a 'standard bottomed' EJ257 in a fast road car, where the ultimate aim is 400x400 ?

but the question 'how many people have 2.5s making 400x400 are out there, with 'some' miles on them - and what did they do to make them?' seems a bit more reasonable..... (err - which I think that this thread was supposed to be for, before it wandered off a bit - sorry )

Mark

On the subject of the crawford link - anyone any experience of this package ? Seems to be a good one on the face of it - 1500 quid to get a short engine - Personally, I'd need to aquire a second 'donor' 98 engine, so that I can continue to run mine until I had a fully built replacement - (Though at this point I don't have a list of parts even !)

 

Think this http://bbs.scoobynet.co.uk/showthread.php?t=381073 is worth a read for the 2.5 upgraders...

John, would you still advise the same ?

i.e. 257 block, standard pistons/rods ?

If so, what heads would you advise for a 98 UK car (i.e. phase 1.5 ?)

Looking at the crawford set-up (which is obviously with pistons/rods), is it the 257 block (I'm not familiar with the USDM Sti blocks - ?)

Thanks,

Mark

 

I think a steel 2.0 would be stronger than a standard EJ257, but less driveable for a road car making about 400 BHP.

I am intending to push again to 450-500 BHP range but very carefully using what I've learned so far, and see if the standard EJ257 will take it, if it does no information will be held back, it will all be on here, and yes I still think 400 BHP is plenty for the road but I'm daft. If it fails I will either get a built motor, probably 2.33 (unless the 2.65 does well meantime) or go back to standard EJ257 with a 20G, or the GT30R at low boost. Built motors are not the panacea imagined by many by any stretch of the imagination, they can easily introduce more problems than they solve. Crawford stuff though seems a no-brainer purchase though, but they can still have significant piston noise.

But if you only want 400/400, I honestly think you are better with the cheap, easily available and replaceable EJ257. Far more thought needs to be put into the rest of the car before most Imprezas will find the limit of their EJ257s IMHO.

 

The Crawford S1 is to my knowledge, an EJ257 with CP pistons.

It is not unreasonable to use your original heads and enjoy the torque delivery if your gearbox is up to it. You'll need to work out combustion chamber volume, headgasket area, deck height etc to work out your compression ratio.

I have a few temperature issues to look closely at on the 2.5. Due partly to laziness I don't have coolant temp in the car, it would have been interesting to see what it went up to at Knockhill when the oil reached 110C... a 2.0 was about 14 degrees lower, main difference apart from a little less power was no A/C rad. If I get the oil, water and EGTs in check, limit boost to 1.4 to 1.5 bar, carefully use octane booster, map the ignition very carefully, keep it in the low 11s:1 AFR, and use a 7500 RPM limit, I will be very interested to see how the GT30R setup performs and how it lasts. I'm doing it because I have a hope if very carefully setup I might find a nice path to tread as an all rounder, but it may well all blow up...

Had the setup that I took to Knockhill been on the edge the engine I am pretty sure would have been dead by now. As it was, the coolant level hardly moved, which is what you expect, maybe a tiny drop from slight overflow, we are talking that it normally sits just at the cap level, and was literally no more than 5mm below it. I gave it enough of a trial there to know that the headgaskets lived with it. That setup was:

My standard road map for Optimax (very quiet on the knocklink, only flickering bottom green at high RPM, not even 2nd green). Ion P450, 1.25 bar, APS FMIC, headers, 740 injectors etc. Half a bottle NF added for safety during later sessions, but it was fine without it as well. Estimated 400 BHP, 380 lbft.

 

Thanks for your response John - By all accounts I think that my sensible (!) route is:

257 (either standard internals or crawford - depending on cost )
Build this up to match against the current spec of my car:
550s, Hybrid, TD05FE (16), PFC, ported headers, AVC - running low boost to give around the same as current - Until I can get the 'new' (not even fitted yet) UK box replaced with 'something' (Probably 6mt - just 'cos I want one...)

It sounds like this will give me plenty more than I can usefully use on road, and I think I'll be following up with some 6pot APs to replace my 335 Godspeeds on Subaru 4pots....

Thanks again,

Mark

 

Something interesting that came up when my tuner was maping my car was that they didn't want to go over 1.3bar as they have had an ej257 rod bend on an engine they tuned at that level. It did bend when at a track day. They even rang up mrt and talked to them about it and they said they have seen a few ej257 rods die at that level.
seems strange as not many poeple on these forums have bad things to say about the rods?

 

We've all found that the rods are better than the originals, however the Hypereutectic pistons do seem to like cracking ringlands, and from the factory are almost running an interferance fit with the bore (I had 0.2 thou clearance, and I know of another set with 0.1 thou ) Isn't it just as likely that a piston failed first and that lead to the rod failure rather than the other way round?

 

Those are Link timings, so add a static (approx) 10 degress on top. It looks pretty much the same as my 2.0 MY98 at 1.4bar, I can get away with less of a hole a peak VE but that's with better fuel and headers.

Interesting that you've got a 35(+10) advance limit but 36 in some off load areas?

Andrew...

 

The advance limit at 35 must of been a small over sight from the tuner.

Im finding im getting a bit of knock around 4500rpm that the knocklinks seeing but the link cant see it. The link says it only gets up to 3 . I put a new knocksensor on but it made no differance so i might run a new wire to it from the ecu. The knocklink is using its own sensor in an M10 bolt hole under were the aircon compressor was. My tuner said the knocklink was saying the same things as his listing stuff.

 

I thought this 10 degree offset between the Link and OEM was a fallacy? If that row 6 is really 10 degrees more than the OEM ECU running the same numbers I think it wold be detonating like a bomb. 28 degrees on the OEM ECU at 1.2 bar at 7200 RPM on a 2.5 with UK heads sounds a bit suicidal to me. That is methanol/NF/STi heads territory on an EJ257 IMHO.

Comparing 2.0 setups, I think from maps I've seen for 99/00 Links are running very similar timing numbers to OEM - typically in the teens for 1.2 bar type setups on UK heads without lots of octane.

 

Just checked my map, I am running (assuming no difference between OEM and Link) just 3 degrees more at similar boost at 1.2 bar, which I put down to the STi heads. I don't think I'm running 7 degrees less.... Anyway, what I am running is what my 2.5 has needed, it won't take 7 degrees more, not even remotely dreaming of it.

 

I'm unshaw on that john. This is streight from the link book.
"The ignition values displayed are the value + the static value. i.e. ZONE IGN 27=27deg + 10deg statec = 37deg"
When i first put the link in i set the advanced limit to 0deg and checked the timing with a timing light and it was 10deg btdc

 

I do think your flowed p1 (ver5-6) heads will be better than my ver3 sti heads. i think i've got a lower compression ratio to you as i used 1.6mm gaskets. but my heads have a bit smaller volume.

 

Think the OEM ECU must have the same offset then, otherwise none of the Link maps I have on my laptop make any sense at all compared to OEM ECU maps on similar specs that perform similarly, yet the advance numbers appear very similar in idle, cruise and on full load.

Even when I had lower compression and used methanol, I think I was just at about 28 degrees at the top. I don't think this is the same as 18 on a Link, I think it is the same as 28 on a Link, but happy to be corrected.

IIRC, this was one of the misconceptions we addressed ages back, that the OEM and Link timing numbers are very similar.

 

god only knows. all i know is i need to sort this knock sensor crap out so i can do some datalogs to see were im getting a bit of knock and take a small amout of timing out.
did anybody ever come to the conclution if lift off det can be damaging as links seem bad for this. i also use to get it with the factory ecu as well. i.e. a big red when you do a fast gear change.? i remeber pat saying its from the ecu putting timing back in too fast as the map decreases?

 

john, are you comparing base maps or maps with correction added?

Paul

 

With correction added, which gives very similar to what Delta Dash reports. Timing = base + IAM/16*correction + fine learning.

If the Link really is running +10 compared with OEM then I just don't see how the cars can run that level of timing, it would be crazy? Some are running 44 in cruise zones (don't think I would put 54 in cruise zones on OEM), similar to OEM ECU, a lot of bigger turbo 2.0s run in the teens. JDMs at low boost with octane may well run 10 more than that, not a typical UK on Optimax though I reckon.

MoTeC, Apexi, OEM all appear to be the same to me, and I think the Link is the same as well.

 

John, that snapshot of PCLink that Megrac posted is definitely timing MINUS the static offset. As you said the row 1&2 timing looks "quite retarded" unless you add 10 degrees.

So in row 6 he is actually running from 17 to 27/28 degrees at the rev limit. Which is about normal for a 2.0. I can't comment on the later OEM ECUs but the early ones (pre-ECUTEK) had absolute timing no.s (ie no offset) in the maps.

Andrew...

 

I don't think in the zones he would actually hit in row 6 he would ever be less than 11(+10) = 21 degrees though. On the OEM ECU at 1.2 bar in the midrange, 21 is wild though... I wouldn't say that isn't taking timing. Maybe there was a change on the OEM ECU between Phase I and II, but something has to explain this.

On a TD05 setup on a 99/00 UK on OEM ECU I can't run 21 in the midrange on Optimax at 1.2 bar, yet this was with wideband fuelling, made excellent performance on the rollers, didn't blow up, pulled like a train.

At the end of the day, the actual numbers matter less, give the engine what it needs, but I find unexplained discrepancies like this disturbing!

 

need a monkey with a timing light in the engine bay.

 

I have put a timing light on my car with the Link advance limit set to 0 and it shows just under 10 degrees.

I have seen quite a few Link maps set for 1.3-1.4bar with between 10 and 14 degrees in the 4500-5000 zones although I agree more than 10 (20) does require a bit of help in the fuel department.

Andrew...

 

From my 2.0l



Perhaps thats why it went like SOAS

 

Looks like that had a bit of octane/water help?

Anyway, the 2.5 is taking similar timing, which I find extraordinary given that the boost is only a bit lower? That is assuming that 2.5 map is actually running 1.2 bar.

 

That map was just before it died, it had water/methanol injection, but the map wasnt adjusted for it in that range, looking back at the original map, it appeared to take that level of timing from day one.

 

Can i ask again how to and who will insure a 2.0 to 2.5 car?????


please