harvey

iTrader: (48)

Now this is total bollocks.

First you go into a detailled account of Subaru oil temperatures when we are talking about 12:1 AFRs and EGTs of 915 to 935 not 913. Apparent and actual AFRs This is more bollocks. At 12:1 AFR and 950 C EGT the life of a road going Subaru engine at high power is less than that of a May Fly.

Second. There is something far wrong if you need your head in the cabin scanning guages to know when it is time to lift off when using the power.

Third, the ONLY Subaru "engine failures" I have ever experience were the need to remove a cylinder head on two separate occassions. Once the plug ceramic insulator had dropped off and caused subsequent valve and valve seat damage and the second time there was exhaust valve and seat damage. Both failures occurred during experimentation and testing (on my own car) and were related to high EGTs and may have been avoided with lower EGTs.

I am not sure these class as full blown engine failures.

Fourth, I cannot recall you or anyone else advising me not to run Denso Iridium plugs. On the contrary they were the rage on this board and the HKS I had my failure with were highly recommended. So much for following the masses and perceived wisdom on this and other boards. At the time of my failure, Bob Rawle had a similar failure and I became aware of others.
Like I said, I have no recollection of a warning from you or anybody else and it was only sometime after my failure and Bob's that I think Bob and I were the instigators, not of mutterings but statement of facts. I don't know if there was a bad manufacturing batch of plugs or an inherent design fault and the manufacturer was unable and not really interested in giving any explanation.
I am sure Bob's recollection is exactly the same but if you can find any warnings from yourself to me on diff ratios or plugs I will be most embarrassed, humble and owe you a thousand appologies.

Fifth,I cannot ever recall you advising me that my front and rear diffs were possibly different ratios (courtesy of Pete Croney, Scooby Sport Mk 1). If you did, was it in plain terms or dressed up in allegory like so many of your posts.
I found the problem when I destroyed the centre viscuous diff. The gearbox itself was OK and is still in use today in its original form with a second hand centre diff and the same ratio diffs front and rear.
If you did tell me, the message was not received and understood but in the event you actally did, let me see your warning as I must apologise as outlined above.

Lastly, Some free advice from me. Don't try to run customer road going cars on road fuel at 12:1 AFRs or with EGTs of 950 c. Ultimately it will end in grief and cause lots of customer upset.

I look forward to hearing from Tony when he has had his mapping issues resolved, the AFR values and EGTs he then runs.

swisstonihasher

Not that I want this topic to be such a hot issue (as it wasn't what I wanted when I started the post) I have to add a few comments re Pat's very long script.

Re the EGT's and oil temps...

Firstly, my EGT gauge takes a reading from the up-pipe, so def reading the mix from all cylinders and confirms valves running at these temps. It's an SPA gauge, reads upto 1000 deg.

Next, oil temps...well, funny enough I've got a gauge for that also . Secs monitor with oil temp sensor fitted. Warning alarm set to fire off at 105 deg! It does go off at this point sometimes, but I always lift off by then. The car runs these EGT temps of over 900 deg all the time with foot down (upto max of 935 deg) and I guess the oil temps would creep up too if I didn't monitor them - the way I drive my car is short blasts, I dont let the car get very hot and keep my foot down - being an ex-mechanical engineer, I know not to push engines from cold starts and from hot temps on a turbo car.

Oh, and yes, my car was the two tone motor that had nothing put problems from the start. Firstly it was injectors stuck closed (as they'd been on the shelf for a while - something I didn't know could happen ) - at the time the fuel pump was tested by the gaffers chaps and it produced upto 7 bar - so can't see it not being upto the job!! Next was a samco hose that was leaking boost (SC did a modification to resolve this, although I cant see anything different to when I originally fitted the FMIC myself - not complaining, just saying). This was the point I decided to change the turbo for SC's 450 item.

I'd like this to end now until I report back on the finished product...EGT's, AFR's etc will all be confirmed and yes Kev, I will be running more fuel with more boost

pat

Harvey,

I used the oil temp as a separate example of where the accepted conventional wisdom of "experts" is wrong. We know that it is wrong because if it were right, then many thousands of engines which DO run high oil temps regularly would now be dead, but that is not the case. Just because the "experts" say something is so, does not mean it actually is. The empirical evidence supports that conclusion. In the same way, just because some people think that EGTs over 900 degrees will kill an engine, does not mean that is actually right. Furthermore, the safe limit is not a universal constant, and is dependant on the engine in question.

The empirical evidence would tend to disagree with that assertion. I've done many tens of thousands of miles in my own cars running EGTs up to and beyond 950 degrees and never had a problem. Many customer cars have put in many hundreds of thousands of miles with EGTs over 900 degrees and they have never had a problem. I have yet to lose an engine to high EGTs and I have no intention of starting now. You were unlucky to have suffered a failure, but if you'de thrown a rod through the block at 280lb/ft would you then automatically conclude that anything over 280lb/ft is not safe also? There's hundreds of cars making more and enjoying a long and productive life with more, some very significantly more. But likewise I am sure there will have been instances of broken rods at 280lb/ft. You can't always go by the lowest common denominator.

OK then, to illustrate a point, here's a nice simple test. Take your car to the Autobahn when it's ready and mapped, brim the tank and apply full power until such a point as you run out of fuel. I guarantee you the engine will expire before your fuel supply does. Does this then mean that the map was unsafe ? Or is it simply abuse ? The reality of high power turbocharged cars is that we are getting more power out of a small engine than we really should be, this means there is more heat in a confined space than if you were to do it the old fashioned way, with displacement. You won't kill a 500CID V8 making over 500BHP in the same test because it's not working anywhere near as hard. I'm sure if I asked Andy, or Paul, or Bob, or anyone else to perform the test I described I would expect them to tell me to go away, in a rather less polite way

Subaru introduce safety features into their OEM maps to try to keep the cars together under such conditions; for example they shut the boost down above a certain ground speed. Of course, in the eyes of owners, this is not desirable, they want more, that's why they are having the car mapped. So as a mapper you face a dilemma. You can either tell customers that it is possible to remove the limit but they are then responsible for keeping an eye on things, or you can keep the limit and get complaints that the car shuts the boost down at 230kph. In reality it's not a mapping issue, it's just that the standard equipment cannot cope with that level of abuse, and it is extraordinarily rare to see people spend money to improve safety without gaining power.... you could improve the situation, for example, by using a bigger turbo which will allow lower boost pressure for the same power, and thus be kinder on the engine. But who buys a turbo to get the same power that they already have ? No, the reality is that people want the bigger turbo because they want more power. So just how do you draw the line? Once you accept that you cannot make the thing indestructible or Ming-the-Merciless-proof, you are left with two choices, one to take the responsibility away from the customer by de-tuning the engine, or two, to let the customer be responsible for their own actions.

I am 100% convinced that better plugs would have saved at least one of those failures.

Well, the engine had to be removed, stripped and rebuilt, it's not like you cracked a set of headers that could be replaced with relative ease.

It was common knowledge at the time that they were prone to disintegrating and taking engines with them.

Perhaps on this board, but let's face it, there is more BS and more keyboard heroes on this board than I've had hot dinners. Oven on the "technical" board where the signal to noise ratio is somewhat better and the contributors are better informed, they certainly weren't highly recommended.

But this post-dates the time where the problem was identified. The first engine I peronally saw destroyed by a Denso plug was Danny Fisher's, which expired on the way to Prodrive Live (I don't recall the year specifically, I'de need to look it up). Datalogs from the MoTeC confirmed there was nothing untoward going on at the time, it was just poor plug quality. This was not an engine that I had mapped.

I make "mutterings" so as not to leave myself open to libel action. If Denso had taken issue with me saying "don't use these plugs they will cause your engine to fail" and decided to take me to court, I'de only have "evidence" of a few failures, which they could rightly argue is insufficient to make such a bold statement and I would have been in a heap of trouble. Now the number of failures is so high that Denso really don't have a leg to stand on, it's also all over all of the performance BBSes, so it's much less risky stating that Denso plugs in high power engines will increase the likelihood of an engine failure. At the time I was more cautious, simply saying "we've lost engines, be careful guys!".

Even today the situation is no better, only a couple of weeks ago I saw a dead engine which was down to plugs, seems that people are still using them, and I was not at all surprised to see the Denso logo on the side of the plugs.

Like I said, I alerted you in person to the diff ratios. This would have been around the time of TOTB1. I saw no reason to post that up on the BBS because I'de spoken to you in person. At the time when you had your diffs changed to make the gearing taller I was doing a lot of research into what went with what and it became apparent that if they had simply fitted UK front and rear RWPs without addressing your drop gears, then the thing would blow the centre diff. I suggested that although you had been assured it was right, it would be wise to verify that was indeed true, because I don't think that Pete knew about the drop ratio issue at the time.

Regards plugs, I'll see if I can find something in some old posts of mine, but there are hundreds and to find a single instance predating your failure will take some time, my posts are notoriously long!

Thank you for that, Harvey As already stated, I don't tend to run cars that way as a general rule of thumb, about 11:1 is closer to the norm for a normal road engine, but as explained previously, this is not a normal road engine and therefore its behaviour will not be the same as a normal road engine. When you have your car running with different cams I'm sure you will observe that, for example, cruise EGTs are not the same as they were with standard cams.

Cheers,

Pat.

Tidgy

So what first gave you the idea there was a problem with a/f mix?

you mention that you have gauges all over, but don't want to be looking at them, so just out of interest y have them? is it the gauges that have suggested the mix isn't correct?

swisstonihasher

Lucky enough the gauges give max readings (which can be reset), and its this that made me first think the EGT's are high. Then I had the roller graphs looked at because of the EGTs and didn't like the AFR's. Trust me in that I only get one shot maybe two to look at any of the gauges when driving her fast.

Car goes like stink, but even in semi warm weather, she runs like a dog.

I think that will do on this topic now...I've had enough.

sifly

What did you expect when it was posted in the way it was?

ZEN Performance

Toni,

It you were after a second opinion, I would recommend you do this in private in future. It is obvious that you had heard some people comment on your graph as being lean, and wished to follow it up with further opinion.

I will not comment on the technical aspects of this thread.

swisstonihasher

Agree there but then that's not exactly easy when you dont know anyone else thats going through the same process with a scooby....All I wanted to do was look at a few people's SC roller AFR results and also understand the scale/confirm if SC's readings are accurate/true - which I now know they are. Trouble is, I dont know anyone else with SC's roller results .

The fact still remains I'm not 100% happy with the car, be it EGT's or AFR's and I will be doing something about it when I can.

harvey

iTrader: (48)

Total bollocks Pat and a lot of it is untrue or off the point.
As someone who frequented 22b and S/Net, as far as I know the failure I had was the fist I heard of. I can check my diary but I think that rectification of my diff issues pre dated TOTB. One head was removed the engine was not stripped down etc etc.
If you want to carry this on start another topic and I will deal with this fully.

Hope you get it sorted soon Tony. Let us know.

swisstonihasher

Okay Harvey, will do.

What spec you running now? I remember you being a front runner for power not that long ago!

pat

Harvey,

Perhaps you would care to elaborate on exactly what you think is total bollocks, or what you think is untrue ?

The first failure I knew of was, as far as I can tell, on 09 Feb 2002, Danny Fisher had a Denso plug disintegrate in his engine on the way to Prodrive Live. I had heard stories before that but it was the first time I had actually seen it. I have too much on my plate at the moment prepping pretty much half of Team Subaru for this year's TOTB to waste my time trying to prove a point, so I did a very quick search and found this thread. You'll see that I did NOT list an HKS Iridium plug as a recommended plug for Danny's engine, despite Bob having suggested it above (but did agree with him on his other recommendation). Read into that what you want.

This was not the first failure I had heard about, but it was the first that I had actually seen, and I think it significantly predates your own experience.

I think you're probably right, I wasn't **** enough to look up exact dates. I would imagine that it was some time after Feb 2002. As near as I can tell TOTB1 was on 4 August 2002, and I don't think it took that long for you to identify the problem and rectify it, so perhaps at a chance encounter at one of the dyno meets, perhaps at Well Lane ? Trouble with verbal communication is it doesn't leave much of a trace for you to look up four years after the event!

You're a braver man than I am in that case. There is no way I would simply replace a head on a performance engine that has just suffered a failure of any kind without completely stripping and assessing any collateral damage. What if some of the debris has got caught down the side of the piston ?

There is no need to start another thread. You can continue to make unfounded statements and I'll keep showing why they are at best misleading, at worst simply untrue.

Cheers,

Pat.

The Gaffer

Hi,
this has got a little out of hand,
only a week ago I was getting almost daily updates from Swisstoihasher about how it feels like we have strapped a missile to his car, how he has been power sliding off rounderbouts with his mates in the car, how by the time he has saved to upgrade the ecu and have a boost controller fitted he still wont be used to the power.

Then someone (who ?) tells him its running to weak, we invite him back to check it over and put his mind at rest.
He mentions a tickover issue and a recorded fault code, cylinder missfire, I ask if it has actually missfired, he replied no.
Now those of you who are genuine subaru specialists will know as I do that more often than not " missfire cyl 1 or 2,3,4 " is actualy not a missfire at all its a quirk of the ecu puting up that fault code for the wrong reason, I am not about to reveal our secrets but it is a quick, cheap fix and is very comman on that generation of car, Swisstoni was invited to have this checked out FOC and if proved to be the fault,( both tickover and the cyl missfire codes are affected). we could do it whilst he waited.
So far he has not taken that option nor has he digested any thing that subaru guru Pat has said.
Pat is one of the top mappers in the country, most possibly europe, he knows his stuff, is hands on and a top bloke, he certainly does not tell lies as has been implied, but unfortunatly as the saying goes "you can take a horse to water but you cant make it drink" if swisstonihasher wants to bellieve other "experts" then its up to him, what more can we do.
Cheers
Kev.

Ozzy_B

iTrader: (2)

From first hand experience of this I totally agree with Kev, within 15 mins of my car getting looked at, the problem was diagnosed and resolved. Regarding how they resolved it, I don't know, I just left them to get on with it, then drove away a happy man. Since then I've had no probelms at all regarding tickover & missfire codes and the cars been running spot on, so cheers guys

swisstonihasher - Just take the car to S/C and let the lads have a look at what is actually going on, rather than them trying to piece together/diagnose the problem by reading the posts on this forum, or over the phone. I've no doubt they'll be able to resolve any problems for you

DaveBlueRA

I know where there is a link ecu sat doing nothing


Looks to me like there is an offer to have it looked at and IF there is reason to, be sorted out


I will second that. Done great jobs on my car and im sure will do another great job when he maps it very soon. only 206 miles to go

AlanG

I don't get it.
If the car was mapped on the rollers, then i assume the car was given a final power run to see what the results are and as a check to ensure mapping has been done correctly.
If that is the case then the AFR's seen on the printout will have been the AFR's the car was using at that time and hence the mapper will have been happy with that, so why do we have a difference from appearance and actual?
I can understand if you said AFR would be different out on the road as against what was seen on the rollers but that statement doesn't make sense to me as it infers that what was seen during mapping is not what is being indicated on the graph.

As for AFR's in general. Every tuned engine is different and due to parts used, ecu etc etc, you cannot compare like for like, there are too many variables. Taking my own as an example (a non std 2.5 build), i ran 12.0:1 AFR on a GT30R.
At the time of mapping (on a DD dyno) by Steve Simpson, there was all manner of testing done to see how the engine responded taking into account such things as egt, boost and afr's. We ended up at 12.0:1 on this particular build as this was deemed to be where we felt the engine was "happiest".
I'm not suggesting anyone else to run that AFR, all i'm saying is if you haven't the knowledge to dictate what your engine needs, leave those decisions to those that are being paid to get the best out of the mods you have done on your engine.
I don't have the luxury of egt probes in collectors or manifold runners, all i have is a sensor in the downpipe immediately behind the turbo exhaust housing which as many professional people have told me is a waste of time as you really need to have the sensor in a collector to produce meaningful results. Anyhow, the maximum i seen during mapping which included a number of back to back runs, the highest egt i seen was 630 degrees (or was it 620.. doesn't matter ) Adding approximately 200 degrees as a pessimistic figure, egt's were around, say 850 tops, but what was more important to me at the time was that with this build, i didn't want to have to continually look at gauges as i feel this detracts from the experience. I still have gauges, don't get me wrong, but i only really use them as a sort of confirmation that everything is okay after "playing". So far, it's been good and builds up confidence in the car over time.

FWIW the engine has been in the car for almost 2 years and been running for about 18 months. So far it's been trouble free, though i do wonder when expiry is due, since there's not that many road going cars with 500+ in the UK to dictate if these power levels are sustainable for any length of time. Only a few years ago 350+ was in the same position as we are with 500+ now, but look how many engines are out there these days with 350+...

Jolly Green Monster

iTrader: (4)

Alan,

Actual AFR used in the burn, rather than appearance in the exhaust.

Simon

Jay m A

iTrader: (2)

I think what Pat is saying, is that an AFR seen in the exhaust of 12:1 could be made up of 11.5:1 combustion AFR diluted with non combusted intake charge due to valve overlap.

I'm still confused since the intake charge has fuel in it, so there will be unburnt AFR in the charge going straight down the exhaust. Unless the lambda can only read hot gasses etc etc etc. I need to read up on this!

Jolly Green Monster

iTrader: (4)

Depends if the injection duration has also been extended..

AlanG

Thanks Simon.
I still don't think i get it.
Is it a reference to what an afr sensor was reading in say the downpipe, as opposed to a sensor reading off the tailpipe? (in which case, there may be a disparity, i think)

Why is a comment being made with regard to "appearance" and "actual" AFR's anyway when all you have to do is map it to whatever that particular engine likes best.

AlanG

Thanks Jay m A
I'm slow at typing, so didn't see your post right after Simon's..

AlanG

swisstonihasher
Are you using the same turbo that SC have posted these results on?

pat

Alan,

The procedure is normally to adjust the mapping on the dyno until we are happy with it, which will normally yield a final dyno plot and then the car is tested on the road to identify any disparities between the way the car runs on the dyno and where it will actually get used in real life. Most often a tweak of the boost control system is in order, and it's not uncommon to make further adjustments to the timing; sometimes a car will take more timing on the road, sometimes less. It varies from car to car.

By definition, AFR is Air/Fuel Ratio. It's a mass ratio. In simple terms, if the engine draws in 147 grammes of air per second then you need to inject 10 grammes per second of fuel to achieve an actual AFR of 14.7:1. But there is nothing in the rulebook of physics that compels every single last molecule of oxygen to react with every single last molecule of fuel during the burn. Just because the fuel is going in there does not automatically mean it's going to burn. We "measure" the AFR at the exhaust with lambda sensors. Lambda sensors measure how much oxygen is left in the exhaust gas, to figure out how rich or lean the mixture is. There are a myriad of reasons that there would be more oxygen in the exhaust gas than there "should" be, so there are also a myriad of reasons why the reading may be wrong.

Here's a simple "yawot?" example. If, at 14.7:1 there is exactly the right amount of air present to entirely combust all the fuel present, then there should be no oxygen left at the end of the burn. We could measure the resulting exhaust gas and we could see there was no oxygen left. So far so good. But if lambda sensors measure oxygen, and we've already used it all up running at 14.7:1, how could they ever read mixtures richer than 14.7:1 ? You can't have less than zero oxygen, so how can you ever detect that you have more fuel than is necessary to use up all the air ? The reality is that you don't ever use up all the oxygen, regardless how rich you run it, and so there is always oxygen left in the exhaust gas that we can measure. It doesn't take a genius to figure out that if for some reason there is more oxygen present then it will read lean. A leak in the exhaust pipe drawing in fresh air will make it read lean! Of course the leak in the exhaust pipe doesn't affect the actual amount of fuel and air going through the engine, but it does alter what our gauges see, and therefore what we believe to be true and correct. There are a myriad of reasons why you may have more oxygen in the exhaust gas that would make the lambda sensor "lie"; if you have cams with a lot of overlap fresh air gets blown from the inlet through the cylinder and out the exhaust without ever doing anything useful in the cylinder, and so that air distorts the reading on the guage. But there are also more sublte reasons. It takes time to burn fuel and air. What if it hasn't finished by the time the gas is expelled from the cylinder ? There would be oxygen left in the exhaust gas that hasn't had a chance to react yet and would make the sensor read lean.

In short, then, actual AFR is what actually goes into the engine. It's pure hard fact, there is X number of grammes per second of air, and there is Y number of grammes per second of fuel. We can measure the air with an airflow meter, and we can measure the fuel directly with a flow meter or use injector duty cycle to work out how much or their known total capacity the injectors are actually delivering. Appearance AFR is what we see coming out of the engine's exhaust gas and is not the same. You can quite happily be putting in enough fuel to achieve 10.0:1 AFR yet only read 13.0:1 AFR on your wideband. A four gas analyser would tell you why there is a disparity; if the HC figure was quite high you'de know you've got a poor, incomplete burn. If the HC was low then you'de know you have a good complete burn but there's excess oxygen contaminating the exhaust gas, say from a leak or overlap. A wideband sensor cannot give you that level of detail, so always treat the readings with a pinch of salt. Widebands do NOT tell you what's going in, only what's coming out, and that might be polluted and thus not representative of what's going on inside the engine.

Cheers,

Pat.

pat

Oh my God, you're so lucky it hasn't blown up yet Alan! You really need to take it to another mapper to sort out this dangerous map! Maybe you need to take heed of Harvey's advice, because.....

I cannot believe how lucky you have been!

But, that can't be, it's not possible, according to the guru Harvey, it should have blown up by now! How can this be ???

People are free to believe whoever they want... but perhaps they should occasionally question just how knowledgeable the people that they are believing really are!

Cheers,

Pat.

Tidgy

rafpmsl

Neilo

Looking at this graph, is it norma the for torque to drop off quite this much? seems to be a massive drop from peak?

john banks

^^ that is how you make a **cking quick real world road car. 90% of peak POWER is available over 2300 RPM. Spool up the turbo to a good boost level as early as possible and then let it roll off whilst the revs reduce the need for boost to hold a similar maximum airflow across a wide range - you keep it pegged near peak power for ages. You get massive torque and have a flat wide POWER curve. This makes a generally quicker car than a flat wide TORQUE curve.

AlanG

My interpretation of your reply Pat is that what your saying is similar to what i've commented on?, that the customer should rely on the mappers experience to dictate what the engine spec needs, than what is deemed to be "the norm"?

Interesting power graph and impressive torque. On paper.
The shape of the graph reminds me of mine when i ran the GT30R last year with a big fat torque band, though as Pat kindly pointed out at one point, doubts were cast as to how accurate the figures were, since the dyno hadn't used an ignition pick up to measure the rpm and so even with the peak power figures probably being true (in SC's instance 451bhp), the torque figures (which are measured from rpm and bhp) would be artificially high.
I've seen a number of SC graphs posted on these boards and quite a number don't use the ignition pick up which is a little odd, especially with the higher power cars we are seeing these days and so can give rise to more inaccurate (torque) figures as a result.

Neilo

John im confused....you say 90% of peak power is available from 2300, yet on that graph the power at 2300rpm is about 135bhp?

Im guessing im missing something about the way to read the graph.

Aztec Performance Ltd

iTrader: (234)

'over' not 'from'

Neilo

ahhhhh, doh.