Astra Coupe Turbo 273bhp, but 25bhp less power with WI.
 

Hi. I have a Mk4 Astra Coupe Turbo running 273bhp and 297lb.ft torque.

When I was on the rollers a few weeks back, we tried the car with and without WI (ERL 1s kit) and I lost around 25 to 30bhp at the top end of the revs. Lower revs seemed less effected, but as the revs rose, the losses increased.

I am running a single 0.5mm jet after the IC and I was running off standard washer fluid. The car over-boosts to 19.5psi and holds around 16psi.

If anyone can give me some advice on getting the most from the WI it would be much appreciated. Please note the car has not been mapped for WI and it is unlikely the I will do this in the future.

Here is the current power graph without WI. Cheers.
http://www.migweb.co.uk/gallery/data/516/medium/CS4.jpg

Hello Mike,

Thanks for posting your set up and hope that ywe can go forward from here and taking your car beyond 300 brakes.

I am not familiar with the Astra but as you are not too far away, I am willing to learn soemthineg from it and hope to be able help you tuning it.

I always hope a few Astra owners herecan chime in and help. Loosing 30 horses is no fun.

Richard

Thanks Richard. By the way, today I was visiting some Southern Water sites in Hampshire and I saw a couple of your ERL power supplier in some of the telemetry boxes. :)

I should be down lunchtime on Friday. If you could be around it would be great. I can't stay much more than half an hour though.

Thisis interesting, I didn;t realise we supply Power Supply Unit to Southern water authority.

Tell me the ETA on Friday so I make sure I am at the office to meet you.

Richard

Hi Richard. I'll be there for 12:15pm on Friday.

I took this photo of the power supply, it says ERL on it, so I assume it must be one of yours.
http://www.migweb.co.uk/temp/erl.jpg

Yes it is - Not using the un-interrupible function??? I will sue the company if my water is not pure during power cut periods.
:lol: :lol: :lol:

Anyone?

Hi Mike

I suspect that you might be injecting too much, too early.

You don't say at what boost pressure your WI is activated.
Also what do you inject? is it straight water?
Have you got a picture of the nozzle setup?

If your car is running rich while the WI operates, then it's normal to lose power, in my experience.

No photo, but it is mounted right after the IC and before a long straight section. The spary comes in at about 12 psi. I will be trying a 0.3mm jet soon.

One Friday, I see what I can do in 30 minutes to get to know the car a bit more.

I think timing and fuel adjustment is the key to your WI success. Just need to find a way to modify them without a special tool.

Possibly delve ino ther coolant sens0r and air temperture sensor and MAF sensor, the management is only as intelligent as the sensors.



Richard

I'd try the 0.4mm jet activated at 16psi. See how that goes.

Also if the AFRs at 16psi are way richer than 12:1 it will be hurting power.
You need to monitor that as well. 10-20% methanol helps reduce the surface tension of the water. Make sure you only use distilled water and pure methanol in the mix. Nothing else.

Thanks John.

Richard - I'll be there at the time above, but should be able to stay much longer now if required.

I will watch this which much intresst as i still need to fit mine. Is you jet before or after the inlet temp sensor ?

Before - I have the Courtenay IC which has a long alloy pipe from the IC outlet. Also, I have bought the IC to plenum pipe from an Astra VXR which relocates this sensor further up tpwards the plenum - this is supposed to give better readings.

Can you let me know of what is your concluesion is after you have had a look on friday. cheers.

Had a good time at ERL today - many thanks Richard.

Richard plans to make an interface to go on the air flow meter signal wire which will be set to lean the mixture slightly when the WI is running. Should be interesting to see if it works. I confirmed with Courtenay Sport this evening that the map I have does go rich at the top end for extra cooling and knock resistence - so leaning it off and using WI may give good results.

When do you plan to do this ?

I will be the responsible to making things happen by designing the interface electronics to modified the MAF sensor. It will be a time well worth spending if the car can repond to the modification.

I have look at the car careful on Friday when Mike came. I am thinking of two other possibilities to trimming ignition and fuel, but MAF sensor seemed to be the obvious, if anyone have any suggestion, please chime in to help.

1) MAF sensor. MY ASSUMPTIOS: if the upper portion of air flow signal (on boost) can be scaled down so the fuel will be trimmed by the ECU. Since the engine is seeing lighter load, it might automatically put some timing advance. Plausible?

2) Air temperature sensor MY ASSUMPTIONS: if engine is seeing hotter air, it will trim fuel but it might also be retarding timing (not good). Suggestion?

3) Engine Coolant Sensor MY ASSUMPTIONS: when engine sees higher than normal coolant temperature, it should reduce fuel, but will it retard timing also?


Mike, I have a question. Have you ECU been reprogrammed? If so, what modifications was done? It might affect my work on your car. If a data looging facility is available, it will make life a great deal simplier.

WORK DONE TO DATE:
I have spliced in a from the MAF sensor to the cockpit so Mike can read the maximum MAF sensor voltage.

At idle the sensor reads 1.2-1.4V and increase to nearly 3V when throttle is blipped. Thas is what I have time for on Friday. Will update when more work is being done.

Richard

Richard, I have yet to test the voltage range. The ECU has been reprogramed and I confirmed with Courtenay Sport that the engine does run richer at the top end - but not overly so.

Today I re-disconnected the WI as the higher revs felt very flat and lifeless. I'll leave it off until will have a chance to play with it again.

I'm just trying to work out if it is better to put the jet before or after the air inlet temp sensor. If it is after, the inlet temp would be cooler than the sensor thinks. Would this not also lean off the mixture? Could be worth a try. Also, I heard WI kills these sensors, so could be worth doing just to save damaging it (about ?70 a throw).

It makes sense that fooling the ECU into thinking there is less air flow would reduce fuel and advance ignition - although I will try to confirm the latter.

Anyone know what a target lambda should be at the top end when running WI? Courtenay aim for around 0.78, but this is running rich. The WI will take the place of the extra fuel, so what should we aim for? 0.8?

Since the air temperature sensor is placed after the water spray, the cooling effect from the additional methanol will make the ECU think the air is much colder than this it. I would tend to re-position the sensor before the jet and perhaps you will notice a difference.

By working through one thing at a time, we can then have more consistant results. I think repositioning the ait temp/MAP sensor is the next thing to alter before anything else such as the MAF.

We normally aim for a lambda valve of 0.86 or 12.5:1 afr. Some engine can run lean and some reich, we shall see on your car. At prresent ypor afr is about 11.3 (0.78L) according to courtenays.

RECAP:
We have reduce the jet size form 0.5 to 0.4mm and added 25% methanol to water.
CONCLUSION:
The engine seemed to run flat and lifeless, it could be the extra cooling effect of methanol at lower air tamperature range as the methanol freezes at -90 degC and boil at 70 degC.

A BMW increased the methanol concentration from 25% to 50% and experience misfire - could be the reason for loss of performance.


Richard

I forgot, if the air temperature sensor is seeing lower temperatures, it will add more fuel?

Richard

I know gelf has his jets in the same positoion as mike. Did he have any problems ? Mind you he is running a 2d setup !

If the water is injected before the temp sensor surely the engine is getting cooler air and leans the mixture out and advances the timing ?

What mixture of water / methnol is being run in mikes car ?

Hi,
I`m new in the forum. Searched for "Temp.Sensor" and found this thread.
We did an installation of the 2D system at a Porsche 933 twin-turbo engine with KKK24/26 Turbochargers and 540 BHB w/o WI system.
After we installed the system we some problems at the point of activation of the water-jet. Something like stucking accures.
We installed the water nozzle directly opposite of the temp-sensor inside of the intercooler approx. 15 cm in front of the butterfly. We think, that the spray on the temp-sensor will be the reason for this behavior. The A/F mixture is to rich for the reason of the "goiing more rich and ign. advance effect" of the ECU.
For better investigation of this problem we will install a complete Datalogging/AFR Displaying unit in next days. Also we want to install a second temp-sensor in the intake manifold after the throttle-body. So we can compare the readings of the original sensor (which affect the ECU injection signal ! ) with the real intake temperature of the air.

I think, that we need to measure the real intake temperature of the air at the relevant input signal sensor. Richard, you are right ! The closer the waternozzle to the temp sensor the more the readings will go down. The sensor is directly placed in the water/air stream and it could be possible that due to a condensation effect of the water the readings are not realistic. Fact is, that the injection time is going up and that means that the ignition behavior of the A/F mixture is dropping so that some sort of stucking is possible.
Next week we will do the same installtion at a Porsche 996 twin-turbo.
Will keep you informed !


rgds
Thomas

Thomas,

Thanks for the input.

I am suspect this might be the case, the Astra was loosing so much power that water along can not be held responible expecially the water jet has been reduce to 0.4mm from 0.5mm. A 40cc/min flow reduction.

Mike is experience the same power drop effect.

Even the air temperature is quite low in England (8 to 9C), the methanol is still evaporating and if it hiots the temperature sensorm it will supercool the sensor and make the car runs too rich.

I will wait for Mike's comment , if he is willing to relocate the sensor before the water spray so it won't get sprayed.

I think the engine will do the opposite, if the air is colder, it should increase fuel and not leaning it.

Richard. I relocate the jet after the sensor. I'll try to do this soon. I think I have a blank somewhere for the old jet position.

Be cautious about triming the MAF signal.

Some ECU might 'work around' the changed signal & keep the a/f constant by the O2 sensor feedback. I suspect, it does so even in open loop mode...

I've tried triming the MAF signal by more than 20% of the voltage, but the ECU eventually adapted to it & kept the engine running in the original condition. All I've done to the MAF signal seems in vain.

Later I tried to trim the O2 sensor signal, by very small steps, then it worked. So I guess in this case, the O2 sensor signal is in a higher priority than the MAF in fueling control. Interceptor has its own limitation in such cases.

This was my own experience with Bosch Motronic ECU. It might be varied a lot in different cars. For your reference.

Thanks for the input, I would only trim the region where water injecting is activated.

I am not sure if the Astra uses a wideband or not, but I don;t think I will go anywhere near the closed-loop region. That was my plan anyway.

I will trim the fuel's duty cycle as the last option, seemed more complicated way but can be done within our ability.

I have been told previously that the lamba doesn't have a clue what is going on at higher revs as it isn't quick enough to react.

hmm....

As I see in my homebrewed O2 signal display, in the mid-high rpm range, the up-down voltage sweeping is pretty quick. At 4500rpm or slightly above, if throttle input is small, the engine is still in close loop. At this stage, the voltage is bumping up & down quickly. If the padel is depressed, the signal will go high & stay. I think the reaction is quick enough.

If the exhaut gas monitoring-feedback system is too slow, how can it catch up with changing load & keep the a/f right?

I remembered you making a bar graph unit on your dash, nicely made unit. I wonder if you have solved the LM3914 overloading issue by reduing the drive current to the LEDS.

I think it will be useful for Mike to have one of those to monitor his lambda values as we start to modified his MAF signals. Care to share the circuit?

Richard

Circuit diagrams:

http://www.taunus-biker.de/~mdvp/Lambda/Lambda.html

http://www.fabrio.it/puntogt/tecnica/lambda.htm

They are not in English, but supposedly the diagrams should be easily understood.


LM3914's datasheet:

http://cache.national.com/ds/LM/LM3914.pdf

About the LED current, haha, I fried some chips. I like them bright :twisted: Right now I attach a heatsink on it. Hope it helps...

Guys, these are all displays of the narrowband signal.
Quite useless at the sort of revs/load we're talking here.

Richard, of the three signals you mention I'd only fudge the AFM a bit (only at revs/loads where it runs open loop ofcourse!)
In fact I'm doing this on the supra, gaining an extra 5% power (at a minimum)

I'd avoid fudging the CTS signal because of the risk that the ECU decides to swap maps altogether. Not good.

narrow band can be used for reference - but better to run on the rich side of 14.5 afr than leaner. It is quite possible to tune to 14.5afr and than richen it up a little.

Although the curve is non-linear, we can see voltage variation between 12.5 to 14.5. I agreed, is not ideal.

Richard

Mike

How is your jet held in the air pipe? Mine was in an in pipe "gromit" and because it did not seal I had less BHP but no change in LBft due to the air loss.

I'm about to use my WI to spray the cooler (I now have the same one as you and the same turbo and map). I know ho effective spraying the cooler is as I have seen it on Courtenay's RR ( reduction of about 20 deg C at 6k rpm).

Hi mate. Now you are running pretty much the same spec as me, do you also have the dip in power like me?

Re. the jet gitting, I had a section of alloy welded onto the pipe (thanks dad), then drilled and tapped that - it does seal, no worries there. The loss in power can be seen by turning the system on and off, so it is nothing to do with bad installation.

I have the bits to plumb in a second jet, which I will do once these existing issues are rectified. I will most likely spray onto the IC too.

As above, I'll be moving the primary jet to a point after the inlet temp sensor and Richard will be working on "fudging" the AFM signal.

Mike

Only fitted the FMIC Thursday so not run the car on the RR yet.

Speaking of the RR, I know that spraying the I/C reduced my inlet temp from 65 to 37 deg C with my SMIC, this is why I am taking the alternate route.

I would not recommend altering the AMM reading, but if Richard thinks it will help then go for it.

Richard, sorry about the delay, but I found my mulitmeter and I can now report of the AFM voltages.

At idle the car sits at an AFM volage of 1.37. When booting the car, the upper limit I saw was 4.93v, so very close to the maximum 5v.

I hope this helps.

Oh, any chance I could pop up at some point please to reposition the jet as I don't have the tools available to do this? Basically all that needs doing is blocking the current location with a blank, and drilling a tapping a plastic pipe the other side of the temp sensor.

Has there been anyone with a sucessful 1s kit on an astra turbo or any z20let ? I who of a fair few people who have WI on there car but no reals gains ! Richard from a turbo car running around 250 - 270bhp Should you see much improvement in bhp / torque with an 1s kit ?

The MAF sensor voltage is goodto be played with.

No problem, which day do you have in mind?
I cannot do Wednesday. Thursday or Friday is OK.

Richard