My S6 work in progress with dyno figs and pics

[S2central.net]

@RWD - For your testing purposes I would avoid any worries about guesstimating power at the flywheel... Power at the roadwheels as measured by the GTech is a perfectly good reference point. Those should compare well with accurate dyno/rollers - at the wheels. You can do the drag strip runs in two ways with G-Tech...

1 - Full beans launch and maximum attack for best 0-60, quarter time and ET - a normal drag run if you like
2 - Gentle start, but keep forwards momentum, and change gear so you get into 4th at about 2000rpm - then go WOT to the redline

Method 2 provides a more accurate / repeatable way of getting nett horsepower at the wheels. Method 1 gets the best times.

I'd suggest two runs in each direction of this private test strip in close succession to get four runs which provide a nice averaged result.

[EspenW]

NOTE: 1 litre of petrol weighs 0.74kg, so the difference between a full and empty tank on your S6 is approx. 59kg.
(Full tank is for extended trips only, no need to haul around 30kg of dead weight for daily fun. Weight is performance enemy #1.)

At peak power in 2nd gear, aero losses on the S6 should be in the 8-12hp range.
2nd gear is fine for comparative purposes, but you will see lower numbers due to higher inertia losses vs. 3rd and 4th and less spool.

Weather correction: Any good dyno will automatically correct to either 25degC (SAE standard) or 20decC (DIN standard), so ambient on dyno day should not matter (however, heatsoak can affect results)
Testing at 5.5 C, you need to apply a correction factor (SAE) of approx. 0.96 (assuming you were testing at sea level, etc).
Use this calculator to compare runs made at different atmospheric conditions: http://www.csgnetwork.com/relhumhpcalc.html

For the comparisons, use the raw G-tech data, corrected for different temp, etc. Trying to calculate back to flywheel hp only increases the margin of error.
Make sure that you use the same stretch of road, and start from the same exact point every time.
Try to start in 2nd gear too, if that is not possible, just get rolling in 1st, then shift into 2nd as soon as possible.

It seems that your average corrected (atmo and drag) whp from the runs is approx 207whp, and that indicates quite a bit more hp than stock. I've measured approx 250whp in 2nd gear on a stock RS2 (boosting 1,5bar) and approx 250hp in 2nd on a tuned S6+ (V8 with 326hp stock, tuner claims close to 400hp ) I have posted the graphs earlier in this thread.

[S2central.net]

Good point about ambient corrections on the G-Tech results - I forgot that one...

[JP]

Excellent info from both S2Central and Espen.

Never actually sat down and weighed petrol or admittendly in utter laziness bothered to look it up on the net, it was an assumption made from years ago from my schoold days about weight of h2o obviously all fluids have different densities so I should have considered that before assuming.
I will also use the supplied calculator to factor in atmospheric changes into the results to try and keep everything as scientific/fair as possible.

I did try several 4th gear pulls earlier today although had to back off on all occasions before I got past 6000rpm due to not enough straight.
I have also noticed that hp results, although approx the same coming in at the low to mid 190's since the pitch and previous weight correction (not current revelation from Espen) the peak was coming in again around 5600rpm instead of the previous 5900rpm I achieved in 2nd gear after changing the pitch setting - which is also the rpm the dyno was peaking at.

Other things I've noticed is in 2nd gear with the Gtech I'm now getting the same boost spike as seen on the dyno (3rd gear pull) whereas this is not experienced in 4th with the Gtech instead I'm getting 230+tq results at around 2300-2600rpm peaks.
I can only assume that the ECU is not trying to catch the boost climbing too quick resulting in a more consistant spool throughout the rev range and more boost earlier on giving the shift in results? Obviously any input on this theory would be good as it is only a theory as of yet.

NB The above theory I've based on the fact that according to flow figures of the head on the car there is an average of around a 10-12% increase in flow from the exhaust port (depending on which section of the lift range you look at). This in turn means there is a larger volume of air available on each combustion stroke to aid spool up. As we've already become aware, there is a certain degree of self learning with the wastegate control from cam changes this may explain why the power was coming in late on the dyno as the turbo had hit it's peak too quickly resulting in an excess quantity of exhaust gasses being bled off by the wastegate as a safeguard, the result being the exact opposite of what's desired as the turbo has to start spooling all over again from a higher rpm start point. This in itself will be interesting to moniter once the Zeitronix unit has been hooked up along with the results of spool up on the chips. As the turbo is not going to be held back at 2.1 absolute like it is on a standard chip.

I will also try a 3rd gear pull to see if I can balance the results a bit better as the straight in question is obviously not long enough for 4th. Ultimatelly irrespective of the end hp/tq results of the Gtech I am seeking a method/system that is going to give the best repeated results time and time again. As this is the most important as a back to back comparison. instead of saying look how many ponies we've made. We'll leave that to the dyno results. I will also see if we can do the dyno runs as a 4th gear pull from now on to minimise transmission loss and the potential of boost spikes.

[knobstar]

Any further updates?

[JP]

Sorry for delayed reply. A combination of internet being down due to service provider change and an overwhelming amount of work has meant I haven't had too much time to sort everything out.

However at the moment the progress is as follows

Further test runs have been made with the Gtech (still standard chip) in second and 3rd gear along the same stretch twice in both directions when outside temps we're 2.5degrees the other morning. I haven't had a chance to put them on the comp yet and properly analyse them.
In the interim though it appears that both are as consistant as each other with obviously the third gear pulls reading slightly lower due to the added wind resistance. Until further analysis next week when I got a moment, I am leaning towards 3rd gear pulls to ensure the turbo is hitting max boost from the extra load.

On the Chip front I am currently in possesion of the chips from urquattro-island which arrived safely packaged up last tue/wed (can't remeber which now) and the ones from Quattronics arrived this morning. S2 Central I haven't spoken to but originally he mentioned December so still plenty of time there. The others I should hopefully have my hands on aswell once I get send of some blank chips next week.

Clutch wise I've decided to go single mass, so in order to maintain as much torque as possible I've opted for the 7A flywheel over the 3B. This is currently in my possesion in the boot of the car freshly refaced ready for fitting next week with a brand new LUK clutch kit, rear crank seal and clutch slave cylinder. As the exhaust has to come of to do the clutch I will use the opportunity to also weld in the extra lambda boss for the wideband and an EGT boss.

That's it at the moment all being well some time early to mid December I'll start fitting the chips and get some results down on paper/screen for comparison.

[urquattro-island]

great

to hear the chips have arrived...

[JP]

Little update in the form of advice needed.

I'm currently in the process of wiring up the Zeitronix data logging unit.

So far I've plumbed in the following

Power up on ignition
Boost sensor
Wideband Lambda probe
EGT sensor (placed b4 turbo)

I would like to plumb in the rpm signal aswell which it advises to be the squarewave signal that goes to the clocks. So does anyone know which wire I need to tap into on the S6 or is this going to be done with a power probe and haynes manual?

I also have the option of adding an extra wire into the harness for a TPS signal to add to datalogging completeness so I'm hoping to do this some time over the weekend. Again does anyone know which pin out it is at the ECU end to save pulling extra wires into the bay? Again just trying to save on some time wityh having to test for continuity.

Lastly I have a user variable input of 0-5v, so are there any other sensors which operate on this range that could be of interest to the datalogging process. I did think intake air temp would be nice but I believe this is based on resistance? Obviously correct me if wrong on this one I'm no ECU/Electrical expert.

[twoqu]

This is usually a good site for reliable info:

http://www.sjmautotechnik.com/trouble_shooting/20vpin.html

Obviously wise to check yourself!

[JP]

^^^^^ Good link thankyou. Like you say wise to check myself.

Just wanted a second oppinion before checking in real life from anyone in the know.

If I want the variable TPS signal i.e. not the earth or the +ve 5v feed can anyone confirm I'm looking at pin 53. Throttle Position Potentiometer on that link.

Next for the rpm signal should I be looking at pin 40. Tachometer on that link.

Lastly looking at the pinouts there is a ground for the air temp sensor and a signal wire and no power feed (obviously as it's a 2 pin sensor) which indicates that it is not a 0-5v range but instead based on a resistance to ground which I thought. So I figured it may be interesting to get a MAF reading instead? in which case would I be looking at using pin 7. Mass Air Flow Sensor (G70)?

Obviously I'll use a multimeter to verify the variables and a power probe but just wanted to double check I'm not completely barking up the wrong tree before starting to probe wires.

[twoqu]

There is some detail of sensors etc on that site:

http://www.sjmautotechnik.com/trouble_shooting/ecu20v.html

[JP]

Bit more progress over the weekend.

Started by going round to my mates workshop near Hitchin to do the flywheel/clutch and plumb in the wideband and EGT.

Pretty much plain sailing as we took the exhaust and box of.

Took the downpipe away with me back to my place (20-25min away) to weld the lambda bung into the downpipe. This turned into a bit of a mission as the ideal placement is 16" back of the turbo and located between 10 and 2 o'clock on the pipework. Which unfortunatelly at 16" back the downpipe was at a vertical drop. We settled at coming in at an angle on the gearbox side and hoped for clearance upon our return. The obscure angle and lack of length of the bung did mean we ended up with a thread cleaning mission as no one had an M18x1.5 tap on a sat afternoon, but that's another story. (Pictures of lambda location will follow when I get round to charging the battery on the camera).

Next we changed the rear crank seal as a precautionary as added boost = more blow by = potential of rear crank seal leak on a 172k engine. As we where there anyway it wasn't a lot of effort for th £8 or so in parts from Elring.

Then problem no.2 arose the spigot bearing of the 7A flywheel sits a lot further back than that of the AAN flywheel when compared from the crank mounting face. This means less physical support and the potential risk of clutch slippage under high shock/load conditions (quarter mile launch). So a couple hours on the laithe ensured an interference bush that was pressed into the centre of the flywheel after heating the flywheel centre up. So that the standard AAN spigot bearing design could be fitted in to the correct offset.

With the gearbox all back on and a new slave cylinder fitted and bled all that was left was to put the EGT probe on. This proved a bit more interesting than anticipated as there was no real place to put it to get a collective reading of all the cylinders before the turbo. So instead knowing that no.5 cylinder is the one that runs hot I decided to locate the EGT probe as close as is physically possible to cyl.5 exhaust port. Working on the theory that if no.5 starts to read too high then the others are getting lean aswell. (Again will post some pics when camera charged).

Later on today I'm going to have a look at the tps and rev wiring, I'll also have a look at seeing the possibility of logging the MAF. With hopefully tonight getting round to another mate to have the chip bases put into the ECU.

Unfortunatelly I then have to play the waiting game as I wish to put 500miles on the new clutch before loading her up and fitting the chipsets for testing. It sort of feels like being handed the keys to your dream car and not being allowed to drive it for 2 weeks.

[S2central.net]

Weird place to fit EGT I think... Why not take the easy option of the factory fitted gas probe point in the exhaust that resides just after the turbo ?

And why not fit a WBO2 probe in the existing lambda location - a probe that can feed Zeitronix in WB mode and clone the narrowband signal into the ECU is available AFAIK.

Seems like you went to a lot of effort on this when it could have been done more simply - what am I missing ?


Paul

[Smiley]

It is mentioned to fit the WBO2 probe at least 16" past the outlet, to be safe on high temperatures. So the factory location might be too close/hot.
Anyone using this location for a WB probe?
The Zeitronix ZT-2 has a non-linear NB output to feed the ecu. It is not clear to me if that can be done in parallel with the WB output on another line

[JP]

^^^^ yep, also on the wideband front the factory lambda bung is before where the wastegate joins the system. In an ideal world I want to have a reading of both as a composite especially when the car is on part throttle acceleration and the wastegate is open for a large percentage of the time as you'll find that the readings will show leaner by mounting the probe slightly past the junction with the wastegate. The narrow band isn't affected by the location as much as the wideband as it is a relativelly primitive measuring device compared to that of the wideband probe.

As for the EGT I would never place a probe after a turbo for two reasons, there is a considerable heatloss in exhaust gas temps through the transition of the turbo which is a difficult to calculate variable as it is as it varies from turbo to turbo.
Secondly this loss means that if I decide to ever use it as a live mapping reference I can be busy melting the exhaust ports of the head whilst the temp reading is saying everything is spot on. In an ideal world a probe would be placed into each individual runner into the centre of the gassflow 3-4" away from the combustion cycle just after the flange to the head. This is where the most accurate representation of EGT is.

Although using what's available on the shelf to me without financial investment where not necessary I had to make the choice do I take a cumulative EGT reading after the turbo knowing it's going to read low.
Do I take a cumulative reading before the turbo in the collector knowing I'm going to get an average of the hottest and coolest. Also being a logistical nightmare to mount.
Or do I take a reading from as close as possible to the optimum position on what is widely recognized to be the hottest running cylinder?

Weighing the pro's and con's out of the above three I decided with the third as at the end of the day if you're measuring the hottest then the other four can only be safer.