My S6 work in progress with dyno figs and pics

[JP]

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

The ZT-2 unit indeed has a narrowband output signal, this wire would normally connect to the signal wire that goes to the ECU disconnecting the narrowband probe entirelly. I personally didn't opt for this as I didn't feel like upsetting a perfectly working factory system, for third party aftermarket stuff. Likewise I feel this function is aimed more at people with aftermarket management that only supports narrowband for close loop function.

The other thing I've picked up on is that the ZT-2 is more of a datalogging unit as opposed to a realtime monitering unit as the information brought up on the display screen is delayed to the fact that certain values are not displayed in realtime on the display unit. Whereas once plugged into a pc all the values are seen. The principle being that if you can see that at x rpm your throttle is at x % and your fueling is at x AFR vs boost and EGT you know whether to add fuel/remove fuel advance/retard timing at certain points of the map without having to tweak it whilst driving allowing a single individual to carry out the job instead of two.

[Smiley]

The delay is interesting.
Don't know whats the 'reaction time' of the WBO2 probe itself, as there is some control involved that itself takes time..
It would be interesting to measure the time delay between the (nonlinear) probe input and the linear WB output. That would clarify if the probe control or the display itself is the reason for the delay.

[JP]

Thinking about the display a bit more I actually think it is more to do with the sampling rate of the signal being send to the display rather than an actual delay.

For example if I were to do a full throttle pass with someone visually monitering the display they may for example visually record the following peak values

14psi, 12.2:1 afr and 726 EGT.

Whereas if you then use the button on the side of the display to recal peak values it will read back

14psi, 11.7:1 afrand 742 EGT.

I'll get the laptop on it properly once I'm ready to start testing and verify whether these peak values are shown as small spikes in the datalogging graphs which will in turn verify my suspicions on the sampling rate.

[JP]

Well first of Seasons greetings to all following this thread.

To start with I didn't realise how difficult it was to accumulate 500 miles to bed in the new clutch and flywheel combo when mainly doing work from home this month. As normally a couple of onsite jobs are plenty to clock up 500 miles in a week. Once the 500 had been achieved I started looking into using the ZT2 unit properly. From which I can confirm that there is apparently a refresh rate issue with the lcd display screen that's used for incar monitering, when compared side by side with a laptop running a 1.7 celeron processor with 2gb of ram. Although not as much a delay on the lcd screen more as the ommision of results. So although good to keep an eye on things the ZT2 is really more a through and through datalogging unit for which I could not fault it.

My final chosen data feeds for logging are as follows

AFR - Wideband probe 16" back from turbo in downpipe after joining of wastegate.
Boost - Tee'd into the original factory boost line after the watertrap by the ECU in psi
EGT - probe fitted approx 4" away from combustion source on cyl 5. in celsius
RPM - Signal taken from ECU wiring loom near the ECU in 1rpm increments
TPS - Signal taken from ECU wiring loom near the ECU 0-100%
MAF - Signal taken from ECU wiring loom near the ECU 0-5v

Using the standard chipset I proceeded to lay down some base comparison figures for later cross referencing using various testing parameters. These will however until the final results assesment be kept secret as I'm currently not in possession of all the chipsets. Obviously I don't wish to give any unfair advantages to those that have the potential to tweak prior to finalising the chips that are being send to me.

With a series of roadtests under my belt with interesting and informative plots I was ready to begin on the first chipset. Still awaiting for a friend of a friend to solder in the chip sockets into the ECU I got restless on saturday and took matters into my own hands. Taking advantage of the sale in Maplins I managed to pick up a solder station, desolder pump, solder and 2 chip sockets for under £50, bargain!

Having been 15 yrs since my CDT A levels it's been a while since having done PCB's but the whole process wasn't too bad getting a nice clean result in a relaxed 90min for both chip sockets. Making sure all was well I fitted the first set of chips and made sure the car started for piece of mind. I waited till saturday evening for a good friend of mine of this forum to come round for the first stage of analysis on the chips, providing him with the laptop in the passenger seat keeping an eye on all of the afforementioned variables. After around 3-4 miles of gradually building up different throttle position vs gear and rpm combinations after the engine was warmed up we were both happy that the figures we were looking at were close enough to the stock factory chip (except boost ofcourse) to warrant some proper testing.

After a good 20-25mile initial roundtrip we logged some interesting data and have drawn some general conclusions using our butt dyno's. I'll go into greater depth at a later date once all the results have been collected from all the chips but as an interim no names mentioned stage the following points have been noted in the initial miles of the first chipset.

1) The power delivery is incredibly smooth throughout the rev range although it feels like it is running out of steam at around 5000rpm and onwards.

2) Going up a good size gradient the car easily pulls 3figures without feeling that it's straining itself.

3) Part throttle acceleration is what can only be described as vicious in the lower gears whilst in the higher gears means no downshifting was required in areas where a standard chipset benefited from a down shift. In fact down shifting to full throttle acceleration appeared slower than part throttle higher gear acceleration. (Possibly related to the flat feeling over 5000rpm at WOT).

4) After having covered some more mileage today (53miles total now) the ECU is showing the same self learning characteristics with the chip change as it did with cam change when it comes to wastegate control. On initial use a constant 2.2 - 2.3 absolute was shown on the dash whilst it is reading 2.3-2.4 after 45-50miles of driving.

5) On full throttle acceleration on the standard chip boost would drop of as you progressed through the gears on WOT, on the new chipset it maintains full boost throughout the full acceleration run well into the top of 4th gear.

6) Some prelimenary testing with the Gtech is throwing up some temperamental results which is cross referencing to an anomaly on the ZT2 data log. Before drawing final conclusions on the cause I have to see if this occurence is repeated across all chipsets. However more will follow on this at a later date after some more indepth analysis.

7) Taking an average of the gtech runs the car is currently pulling 211hp in 2nd gear and 257ft/lb. Which is only a minor 5hp increase over previously but a whopping 50ft/lb of tq.

Peak hp is achieved between 5000-5100rpm much lower than previous results. This also correlates to the car feeling flat in the upper rev range. Inadvertandly it also cross references with the afforementioned anomaly I'm logging on the ZT2 at 4750rpm onwards.

For the moment I'm concentrating on the cause of the repeated flat spot over 5000rpm at WOT as I feel the fault does not lie with the chipset looking at the results. Therefore before I report any further findings I will strive to find the cause as I feel there is a lot of hp to be gained if the problem can be overcome looking at the tq gains and the power curves as a representation of the potential.

Obviously as more conclusions are drawn from further data collection and analysis I'll post them up as they are achieved.

[JP]

Bit more of an update, after a chip swap and similar results with power delivery and high rpm flaspots although not as noticeable a flat spot as before.

More datalogging and in depth discussions with tuners that are not involved with this experiment, we've collectivelly pinpointed the cause of the flatspots to the MAF. Quite simply it is going out of range over 4750rpm at WOT as the airflow is exceeding what the MAF is capable of reading causing the MAF to flatline at 5v.

It's a similar problem the jap cars suffer from when they crank up the boost. Different ECU's respond differently when this happens. In the case of Jap motors the ECU cuts the fuel to hold the engine back for safety. However looking at the AFR readings on the datalogging it appears that the S6 ECU is achieving safety by cutting timing. So unlike the Jap boys who overcome the problem by fitting something called a fuel cut defender (which basically stops the ECU backing of the fuel system on overboost), the S6 ECU appears to overcome the problem in the map itself. However this is just a theory it'd be nice to confirm this if anyone can shed some light on this?

This would in theory explain why currently one chip is less sensitive to the MAF topping out than the other, one being a better map than the other in this particular scenario. However as I don't personally know how to map or even read a map it is for now only an assumption based on information gathered to date.

As a reference point I've attached two screenshots of the onscreen information of the data collected on the standard chip and one of the uprated chips. Both datalogs have been taken on the same starting stretch of road. The first one (standard) was with an outside ambient temperature of 4 degrees in the mid afternoon on a dry sunny day. The second screenshot (uprated chip) was taken with an outside ambient temp of 3 degrees in the early afternoon on an overcast day. Both pulls where rolling starts in 2nd gear from 30mph.







As you can see on the standard chip there is an occasion where we just clip 5volts on the MAF, whereas once an uprated chip was fitted for a similar pull it pretty much flies straight to the top of the scale. The engine is also running leaner however EGT temperatures are considered to be more than safe and pretty much identical to that of the standard chip so I have no real concerns on the 13:1 ratio at 17psi of boost.
Obviously any independant feedback on the above info is more than welcome. Likewise if anyone wants more detailed info in a more specific testing condition I'd be happy to oblige as ultimatelly knowledge is power (literally in this case ).

Out of curiousity I've also taken the worst offending chip of the two tested so far with regards to the flatspot syndrome and fitted it to a mates S2 coupe with an ABY lump and 3" exhaust and standard head. So very similar to my car except for the fact mine has a stage 1 head on it.
Although the car had not been plumbed for datalogging purposes it is quite apparent from our 'butt' dyno's that his S2 doesn't have the same flatspot and comfortably carries on pulling into the 6000rpm+ zone. Not as torquey or potent as my car but fast none the less.

Which brings me to the final conclusion for the moment which is that the flow claims made for the stage 1 head are apparently sufficient enough to peak the MAF which makes me curious to what the real potential is of the K24 with a correctly sized and scaled MAF and the RS2 or 7A cams instead of standard ones on the head I currently have. So possibly an experiment for later on this thread?

[JP]

Well it's been a while since the last update. Asides from the fact I was still waiting on some chips, holidays, weather being more suitable to drifting opposed to dragging etc....

I'm currently still awaiting 1 chipset to complete the test of 4. For this reason I'm only posting up some brief results to keep the thread alive and make sure people know I am still working at it.

First of all the roadtesting I will report on once I've had the last chipset so no unfair gains can be had from the info I'll be testing. Secondly all the power curves from the Gtech will remain un-named to keep people in suspense for the final results.

The attached power curves reflect 1 x average standard chip one from last year when I started this and 3 chisets fitted one after the next this fine evening on a private stretch of road. Outside temperatures remained constant at 7.5 degrees throughout all 3 chipsets. each chip had a total of ten runs 5 times in each direction of the tarmac and after sifting through the info in the last hour I have chosen the figures that look the least outrageous and are approximatelly in the middle to upper range of the average.

The weight remained a constant at 1720kg which is what the weight is at 1/4 tank full. After fitting each chip the car was driven about for a bit until she idled at a perfect 14.7:1 and the oil temp was getting almost spot on with the first main increment of the gauge after the min mark. A petrol station being round the corner I topped her up with £3 of super to try and maintain a steady weight inbetween chips. The attendant did give me a funny look when i came back for £3 a second time but hey who cares

Anyway the only other thing to note is now with the warmer outside temperatures none of the chips suffered from the horrible flatspot syndrome which I noted earlier in the thread when outside temps where pretty much zero degrees celsius.

Putting all that aside enough of the waffle I'll let the figures speak for themselves. I've got 2 jpegs the second zoomed in on the main power area so you can see the difference in the power delivery clearer with respect to area under the curve which I feel is most important for driveability.

Remember Run 1 is the standard chipset.

[JP]

Almost forgot. I'll be going to gti festival at the end of the month so I can do a back to back drag strip comparison for the final test before doing the dyno test.

[Rusty]

Good work. Be interesting to know the differences in the testing. BTW, I got 258bhp without a remap . 210bhp at the wheels in 2nd is about the same as mine is now on the GPS.

[JP]

What was the WHP? As it's a better reference point as I didn't use the dyno operators Flywheel power I calculated the fly myself and it came in 10-11hp less iirc. WHP was 202 ish.

The other thing I noted as a whole is that the stg1 headwork is giving minimal top end gains instead it is providing a responsive flat power curve driveable car.

More importantly it is pushing the torque through the roof once the boost is cranked up

I also did a max speed test which I've decided to post up whilst I got the info at hand, again corresponding to the same chipsets. I've done it in rpm in 6th gear (tires 235/40 17's) as I'm not a 100% on the accuracy of the speedo at the sort of speeds the car was travelling.

However once again on a nice private straight


Chipset corresponding to run 2 pulled comfortably to 5100rpm and slowed down at that point gradually pulling through to 6100rpm before an imminent corner came up at the end of the straight, the car was still pulling at 6100rpm and there was a strong indication that with enough straight the 7K redline could be met.

Chipset corresponding to run 3 started flatspotting at 4600rpm and then slowly pulled though to a max of 5600rpm before wind resistance became too much for the car and didn't want to go any further.

Chipset corresponding to run 4 pulled straight through to 5800rpm then slowing down through to 6300rpm again there was no indication of running out of steam as I ran out of straight a little too quickly again and had to back of before the 7K redline.

[Rusty]

Don't think I got power at the wheels... My power at the wheels on the GPS is subject to rolling resistance, air resistance and inertia effects which all add up. Not sure what the GTech guesses and what it doesn't. Losses at 7k are around 1/3. When I've got the energy to get the cam in I'll get it on the rollers. Might have a new chipset myself by then too.

[EspenW]

Don't think I got power at the wheels... My power at the wheels on the GPS is subject to rolling resistance, air resistance and inertia effects which all add up. Not sure what the GTech guesses and what it doesn't. Losses at 7k are around 1/3. When I've got the energy to get the cam in I'll get it on the rollers. Might have a new chipset myself by then too.

G-Tech numbers include rolling resistance, air resistance, inertia, etc. Also need to be corrected for temperature, etc.

[Rusty]

Yeah, that's the thing with the GPS, you can control and apply all the correction factors. I need to look at the temp compensation although it's uslally a small % over here.

[S2central.net]

Espen posted a link before with a site that does the SAE conversion to standardise results. It is only a small %age unless the runs were done in extremes of temperature or altitude.

[JP]

I just realised I made a slight error in my wording in the previous report but the forum wont let me go back and edit it. So I thought I'd write it here to save confusion.

Run 1 in the report = Main run on the screenshot
Run 2 in the report = Run 1 on the screenshot
Run 3 in the report = Run 2 on the screenshot
Run 4 in the report = Run 3 on the screenshot

Sorry for that

[JP]

Just thought I'd add this diagram to the thread for an interesting comparison of UK figures vs the US ones from 034 motorsport.

I'm going to see if they are interested in donating a chip for comparison testing as they are claiming what we are experiencing but with lesser figures to what we are recording obviously theirs will be on a factory exhaust and head, however it will be nice to see how it translates across.

lifted from their site



weblink HERE