john banks

I had higher boost and the same TPS on the TD05 but did not hit maximum load. The difference is the higher airflow voltage. Of course the MAF calibration can be rescaled. I still have a bit of MAF voltage to go. It hit about 4.78V with the APS intake on (which lowers things by about 0.3V).

Only did very brief setting up of it last night with a few maps after the turbo went on. Will have a think about all the helpful points raised. Indeed this thread was opened not as a mine is better than yours because it clearly isn't, but there are issues to resolve with it.

1 bar seems to arrive at about 3200 RPM. Running 1.3 bar is OK. Running 1.4 bar and it surges and hits the load limiter. After 4000 RPM it seems to clear, TD05 16G small (standard was really the same though). Cake and eat it will be nice.

Keep it coming...

Interestingly doing an STi 7 UK today (VF35) with full Miltek and K&N induction, it had light compressor surge at anything over 1.25 bar until 4000 RPM if you were in 6th gear. An induction kit did mean you could hear it more easily, but the MAF voltage was a bit fuzzy in this area.

One issue with bigger capacity - surge is still likely to happen if you bring the boost in too hard because it could be considered a mismatch between turbine and compressor?

So the situation as I see it - can have a bigger exhaust side so it doesn't surge, but then it spools up later, or smaller exhaust side but hold the boost down particularly in higher gears. There has to be a nice compromise.

[Edited by john banks - 3/2/2003 4:07:09 PM]

Andy.F

Mark

The one you posted that is fitted to Christians car. The one which Bob refers to above.Andy

[Edited by Andy.F - 3/2/2003 7:31:25 PM]

R19KET

Edited because I posted by accident !!!!

[Edited by R19KET - 3/3/2003 12:22:45 AM]

Andy.F

Link to 'the' thread. Going to upset people now linking to locked threads (P2 near bottom)
http://www.scoobynet.co.uk/bbs/threa...=173402&Page=1

[Edited by Andy.F - 3/2/2003 11:34:47 PM]

ChristianR

iTrader: (1)

Why not just post the map??



[Edited by ChristianR - 3/2/2003 11:44:28 PM]

john banks

For a turbo capable of 450 to 500 BHP on a 2.0 litre, where in the rev range would you expect it to be able to make and hold in a high gear a surge free:

1.2 bar, 1.5 and 1.8 bar?

300, 350 and (400) lbft?

Just trying to work out what is realistic on something this size so we know what to shoot for. Variables to manipulate include wastegate size, actuator and turbine clip, (as well as ECU control), think we are getting reasonably near... the TD05/06 20G is on my car and making silly airflow numbers, and the poor old JECS is maxxing out the load at 1.4 bar which will require some work

Anyone want to share their opinions or what they have actually got out of a turbo of this sort of size?

Deep Singh

Hi John,sorry to go off at a tangent... Can you tell me about the boost characteristics on the 5/6 hybrid.
1)In gear 2 or 3 where does the turbo come on full song?
2)In gear 4 or 5 where can you get one bar boost?
3)Have the surge issues been sorted?
4)Do you feel the turbo is good for a real 375ish BHP at 1.4bar or less?
5)What do you mean by maxing out the load on the Jecs

Bob Rawle

You can take your choice, there are many turbos and turbo hybrids that could be put together to do that my own offering being one, achieving those figures is not soley dependant on the turbo though. It really depends on your budget, pure Garretts entail custom exhaust and intake piping to suit, mine is a Garrett hybrid using a P20 turbine housing (suitably modified) which allows the use of readily available exhaust components, the compressor map looks good enough for a 2.5 litre let alone a 2 litre (in fact if anyone can dig up the thread where R19KET posted a compressor map of the Garrett that is on Christian R's 2.5 litre its the same map), but I would be a little uncertain of the ability of the P20 to allow enough gas flow through in "real" 2.5 litre use. I'll get around to trying it at some point. My Jecs maxed out (56.5) at 1.2 bar on my setup using this turbo.

I would have thought that the TD0 hybrid "cause" should have been serving for this though?

If you use the "usually quoted" rules of thumb re airflow and boost you can get slightly misled, you will also need to pay attention to intercooler efficiency as well as volumetric efficiency in doing those sums and the turbine map can make a big difference to the end result.

So I would be suprised if you can get any guarentees as the turbo is only a small part of the overall contribution to achieve the end result, the best way to be certain is to follow someone who has already done it but I think that takes away the satisfaction you get when it all comes together.

Andy.F

Bob

From what I observed of your Well Lane power curve when your car made 407 bhp, the torque curve was very steep ie not a lot of midrange. What minimum rpm can you actually get full boost pressure at ?
I noticed that you were boost cutting at at 2.1 bar at high rpm suggesting that it was still increasing boost at 6000+ rpm ?

dowser

I'll be trying the 05/06 hybrid with an ej25 - but it'll just be the 'suck it and see' approach, followed by exhaust trimming if necessary (my engine builder thinks it'll be OK).

I'm not trying to be scientific with it, as I don't know how I'm also making changes to the rest of the inlet path which will see around 13% more flow capability over the ej20, and I've no idea what the builder has planned for the heads except that the OE cams will stay.

All of these things have the potential to affect spool characteristics, and the surge point, no? I will be using a stock 05 during run-in and initial tune to try and give some sort of baseline.

Richard

EMS

John, Bob,

To avoid reaching max load on the JECS, I think it should be possible to use for example a MAF-sensor from a Porsche 996 Turbo. (It´s a Bosch sensor which looks exactly the same as the Subaru one, only bigger diameter housing) Perhaps using the standard MAF, fitted in a custom made housing with larger diameter should give the same result.

This should give a lower output voltage at the same flow, therefore also lower load values. The lower output voltage at a given flow isn´t a bad thing either with big injectors!

This is what I am planning to do in the near future, as I have to map a customer car with BIG mods.

Mark.

Bob Rawle

Andy, the two very deep grooves (necessitating replacement) on my front tyres testified to the wheel spin that I had at Well Lane, that plus the constant spark limiting (as soft boost cut) that the ecu applied as the boost rose in a straight line. If I had left the boost cut off (as I normally do) then I think I would have had a blown engine that day, and all due to wheel spin and rr setup. I should have used cut slicks maybe.

On the road I achieve full set boost (2.1 bar guage indicated) by 4000 rpm with 1 bar at 3600 rpm, since 1 bar is circa 400bhp based on my calcs (and part throttle rolling road runs)at 7000 rpm then the spool rate is somewhat less influencial in the real world than the air flow generated into the engine. I consider I have the spool rate for this particular turbo as aggressive as it can be being mapped to run just to the left of the surge line, since my wastegate is held shut by a 1.5 bar actuator I have no wastegate issues during this period. Since the turbine and turbine housing design needs to allow for enough gas flow to actually allow the compressor to work correctly (without surge etc) then it necessarily follows that a reduction in spool rate is inevitable. That's not to say that your suggested solution won't give the lie to that but then this thread was opened to allow that discussion (I hope)

Rather than discuss my figures I think Johns thread is a good platform for discussion of the differing approaches that could be taken, hopefully without the mines better than yours effect that arose recently, my graph taken at PE was posted clearly on the wall at a recent rr day and several people observed the runs which were carried oput during a normal Friday afternoon so I have a degree of confidence in participating. I would also mention that the first couple of runs at PE also developed wheel spin but with correct tyre pressures and planty of heat in them it settled down and subsequently generated consistent graphs over several runs with and without extra weight in the car. The PE rollers are virtually identical to Well Lane btw but they have recently knurled their rollers to overcome similar wheel spin problems.

John, you stipulate a "held in high gear" range of boost which implies something somewhat differentt that the AndyF post which focused on spool. Is your question related to spool ? If so then there are different observations to be made in addition. The "have cake and eat it" desire in terms of spool and turbo capacity is a perpetual one of course.

Bob

Bob Rawle

Mark, its not maf related for my part, when I max it out I have not hit the maf ceiling, far from it, its the effect of boost and tps (I think) that has caused me this problem. Max maf volts 4.64 .... I'm all done with a maf voltage of 4.26.

cheers

bob

EMS

John,

Sorry, it´s a bit off topic but I think it´s something which is interesting also!

Bob,

AFAIK in the JECS the load is only calculated from the MAF reading. I measured a car with a Superchip module and the ECU only "saw" about 0.1 bar from the MAP-sensor. This didn´t seem to influence the load reading compared to other cars with normal MAP-sensor reading.

At the point where you reach max. load, it´s not necessary you max out the MAF-sensor! It is depending from engine speed. If you have high torque (torque = load) in the midrange, you can easily max out the load range at relatively low MAF reading. If you reduce MAF-sensor voltage, the load definitely will be lower.

Mark.

Andy.F

I have both the 20g and the Garrett compressor efficiency maps, interesting to overlay them
Around the 400 bhp 1.6 bar range the 20g is 2 - 3 % more efficient.
At 450 bhp they are very similar.
The garrett runs on to 520 ish bhp where the 20g chokes around 480.
The 20g needs around 4500 rpm to develop 1.8 bar surge free whilst the garrett requires around 4300 rpm.
The 20g runs to 1.9 bar max at 73% efficient, the garrett runs to 2.05 bar at 60% efficient.

So, similar yet subtly different compressors

Agree with Bob that the turbo is only a small part of the equation, for instance, I'm quoting pressure ratios as boost pressures therefore assuming an identical intake pressure loss and intercooler/system pressure loss in each case.

What would be 'really' interesting would be to see the two power graphs overlaid One day maybe

R19KET

Andy,

Which Garrett map are you comparing the TD06 map too ?

Mark.

Bob Rawle

John, if you have identified the area of surge then could address that by focusing attention on the tps and rpm values wrt duty cycle? you "should" then be able to slow things down unless the dynamics of the exhaust gas flow increasing with ve just makes it not controllable. Basically though you are limited by the surge line and so can only map to it. The compressor/turbine relationship is the key to this, cutting back is one way to go but a more subtle adjustment of the turbine housing might generate a better result possibly.

The other way to do it is to install a more intelligent boost controller (with more mappability) and use the Jecs boost control to run water injection maybe. It would interface with the ERL high speed valve very well I would think.

Bob

Andy.F

We (John) can map around it no problem. The 'problem' is having a turbo that wants to produce more pressure at low rpm and we can't use it !!
See the 'classic turbo' thread for plans to overcome.

Andy.F

Mark

I'll Email you the two compressor flow maps which I have drawn 'my' load lines on. (I used the same maultiplier)
Let me know your opinion.

Andy

Bob Rawle

The thread had a good pic in it, I remember feeling quite chuffed that I had speced 2.5 litre capable turbo as everyone commented on how compatible the map was, difference is the turbine hsg on mine is a modified P20 and on Christians a "pure" Garrett, the core's in both are Garrett though. On Christian's, being 2.5 litre, it was very easy to hit the surge line but its behaviour on his car was very different to mine.

In the interests of science I have decided to install an EVC on my car, hopefully early next week, I can then play with the boost response in islolation to whatever ecu I have in. That will enable some further experimentation to take place in a more detailed way, who knows I may get a bar at 3200 but if I did then I would be on the wrong side of the line. We will see.

Andy.F

Here they are with 'my' load lines. I know they should strictly speaking be curves but I find the apply to my car quite well with the 16G and 20G compressors.

R19KET

Andy,

Oops, my last post was posted by accident, YHM.

Mark.