swan

'ello all,

I've now got a few logs and thought I was understanding them pretty well until I did a search through the archives and saw the comments made on other logs. Would someone(s) be prepared to review the log(s) and advise on any problems? Which logged fields would be relevant if not all and what kind of driving? Around town? Highway? For what duration? Most of the logs I've got ATM are around 15-20 minutes but obviously it would be easy to trim them down. TIA

swan

Ok, if anyone would like to blind me with their brilliant translation skills, a log of a run partly on private property and then onto public roads (see the low speed section )can be found here. It's all of 22Kb zipped.

Feedback welcome here or via email, thanks.

SiHethers

Swan

I've had a look at your log. I take it you have an MY01 WRX looking at the logs. You really need to 1)log fewer sensors: speed;rpm;front o2 sensor;knock correction;manifold relative pressure;mass air flow voltage;fuel injector duty cycle; throttle position would be a start, 2)log a full throttle power run say in third gear from 1000rpm to redline, 3)have a question that needs answering: for example fault diagnosis.

I'm not an expert on ECU tuning, and have only looked at logs for MY99 UK cars, but if you do the above I don't mind giving an opinion

Simon

swan

Thanks, will do so and mail you directly if that's ok.

torsteinvh

I'd be happy if you posted your logs here as well.

I am quite new myself, but I have logged a few 3rd gear runs I can compare to your logs if you want. I have a MY00, btw.

-Torstein

swan

Ta, will do so then.

john banks

Yes, collect some shorter logs, use the defog switch to trigger and do a few runs from say 2000-7000 RPM in different gears, each as a separate short log with fewer sensors to avoid the clutter and improve capture rate.

swan

Ok, a few runs, 1st into 2nd, 3rd, 3rd into 4th. Thanks all!

SiHethers

The capture rate is still very low. Were you using the defogger switch to start ant stop logging as John recommended, or are these edited from a longer log? I suppose if your laptop isn't very fast this will affect capture rate too. Come to think of it, have you selected a capture of say 200ms from the options bit?

From these, however, it just shows what a difference climate makes. You're only holding 11psi of boost in 3rd. I also note that at approx 6500rpm the ECU briefly flicks into "safe mode", ie cuts boost to around 7psi which is the default wastegate actuator pressure. Just before this the ECU pulls 5.5degrees of timing out, but your AFR remains around 11:1. To me this would tend to indicate you may be getting some sustained det at this rpm which is causing the ECU to pull boost back. As the AFR's are still relatively rich this would most likely be due to a rise in charge temperature. Although I thought the newer ECU's stopped listening for det above 6000rpm. I stand to be corrected by John Banks

My question would be what RON fuel are you using. For safety's sake I would only try and use 97-98 RON + octane booster, but I don't know what's available in SA. That would allow the ECU to run more boost and reduce the chance of det, thereby giving a little more oomph with lower risk of bang

Oh the pleasures of living in cooler climes.

All purely my opinion and probably wrong

Simon

swan

>switch to start ant stop logging as John recommended, or are >these edited from a longer log? I suppose if your laptop isn't

They were edited from a larger log, does the rate vary depending on whether you use defogger or not?

>very fast this will affect capture rate too. Come to think of >it, have you selected a capture of say 200ms from the options

It's a 1Ghz P3 so I doubt that's the problem. I didn't realise you could change the timing intervals, can't see it under options either though.

>From these, however, it just shows what a difference climate >makes. You're only holding 11psi of boost in 3rd. I also note

And this was done fairly early in the morning, was probably 18 or 19 degreees celcius.

> that at approx 6500rpm the ECU briefly flicks into "safe mode", >ie cuts boost to around 7psi which is the default wastegate

Thanks for the comments, they help me understand the figures better.

>My question would be what RON fuel are you using. For safety's >sake I would only try and use 97-98 RON + octane booster, but I >don't know what's available in SA. That would allow the ECU to >run more boost and reduce the chance of det, thereby giving a >little more oomph with lower risk of bang

Pump fuel in Johannesburg is only 93. Coastal areas get 95 and 97. I had added 5l of 102 and a bottle of NF one shot to this tank.

>Oh the pleasures of living in cooler climes.

Only reason I enjoy winter is because the car goes better.

swan

>The capture rate option is found in the live data capture >window. I think 200ms may be the default rate though. I'm not

AFAIK the number of items you're logging will affect the intervals, will check interval settings on my next live log.

>have around 96-97 RON. Shame you can't get it any higher.

Filling up with racing fuel will do it but it's a touch inconvenient (and more expensive) when the car is your daily drive.

>Charge temps rise very quickly if the cars been stationery for >any length of time, was that contributory? If not I'm a bit

For the 1st gear runs it could have been but not so with the 3rd gear runs.

>planning on running around 15.5psi of boost, plus as I'm using a >mechanical boost controller, the ECU can't cut the boost to

I had my EBC removed - I'd rather give the ECU the ability to pull timing/boost, etc. when it needs to.

>7psi. I am however going to run optimax + booster for this >period which should see me close to 100 RON.

Should help a lot.

>Edited to say - maybe I should just get a remap and be done with

Ah, joys of having a knowledgable and wide range of after market tuners...

torsteinvh

Here is a log from a 3-rd gear run in my MY00 with a TeK1 map. Temperature was around 6 degrees C, fuel was 98 RON.

I'm very new at all this, but from what knowloedge I have gathered on the internet, I can't find anything obviously wrong from my log.

What puzzles me though, is the total lack of knock correction. Actually, I have never seen any knock correction in any of my logs.


-Torstein

SiHethers

Torstein

With a peak boost of 13.3psi, an ambient temp of 6 degrees and 98 RON fuel you will be very unlikely to see any knock, especially with a Tek1 which has already had its fuel and ignition maps optimised. Your AFRs are nice and rich on boost. What is disappointing is that the weak spot around 2000rpm during spool up is still there, and I would have thought that would have been addressed with a Tek1. Doing an ECU reset will erase any learned knock correction(e.g. if you had a bad batch of fuel) and you may see some following that. I didn't see any KC until I raise the boost over 15.5psi with a standard ae800 ECU and 98 RON fuel.

Out of interest, have you (and swan) downloaded the beta version of delta dash with the road dyno, as it would be interesting to see some wheel torque/power figures. I have some deltalogs and graphs from road dyno but alas am not able to post them via a link but could email them if you're interested.

Cheers

Simon

swan

On the day I had my car dyno'd I couldn't get DD to pick up the dongle - needless to say it picked up the dongle fine that evening but then it was a bit late...

torsteinvh

Simon, thanks for your reply.

I have seen higher boost peaks in 4th and 5th gear, so maybe I will see some knock when summer arrives and the temperature (hopefully) rises.

I don't think a ECU reset will make any difference, I bought a spare ECU with the TeK1 map, so it was without power for several weeks. I would think that would have erased any learning?

As for the DD Road Dyno, here is a 3rd gear run. Nice cool evening, 4-5 degrees C ambient.

-Torstein

SiHethers

Torstein

I'd be much happier if I didn't see any knock

On the road dyno, how did you calculate your car weight? I can see you have a Magnex BB, do you have any other mods?

Simon

torsteinvh

Simon,

Regarding the knock I agree, but I would like to see it just once, maybe on a very hot day, just so I know the knock correction works

The car weight is a 'guesstimate'. I have a wagon, which according to my registration documents weighs 1280kg. Adding drivers weight and various tools, i get 1410 kgs.

I have no mods other than the TeK1 remap. The Magnex was in the boot, the run was made on my way home from picking it up

I haven't had a chance to fit it yet, so I am still running with the original backbox

-Torstein

SiHethers

Well that's quite remarkable, that dyno figure would indicate between 265-280bhp at the flywheel(depending which multiplier you use), which cannot be possible from a wagon only with a tek1 running standard boost. At first I thought the weight may be overestimated, but being a wagon, that's only 110kg extra for you + fuel etc, so that's not likely.

Your in gear acceleration times would fit with the quoted power, but I wonder was it a downhill stretch of road you tested on, thereby increasing acceleration and fooling the dyno that you have more power. The cool ambient temp will release more power, but wouldn't make that much difference.

To put it into context, my MY99 puts out barely 10bhp more but that has K&N filter, samco hoses, full 3inch sport cat exhaust and Dawes Device with 16.5/16psi peak/held and 14psi at 6000rpm.

What a puzzle

Who knows what'll happen when you put the Magnex on

Simon

torsteinvh

I thought the figure was a bit high myself, and there is a possibility the stretch descended slightly, but not by much. It appeared quite flat to me.

Here are a few more runs, maybe the can shed som more light on this:

This is the standard AE802 ECU: AE802.DDF

This is a run with the TeK1 the day after i fitted it:
Tek1-1st-run.DDF. I had to abort at about 6000 RPM, so it is probably not accurate.

This is another run from the same evening I did the run I posted earlier: Tek1-week-later.DDF. As the filename suggests this is one week after fitting the TeK1 ECU. A different stretch than the other log, but again it appeared quite flat.

I will try to make a few runs both ways on the same stretch before and after fitting the Magnex, so I can what difference that makes.

The car was previously owned by a company, and as far as I know it was completely original when I bought it.

-Torstein

swan

Possibly a really foolish question here but what impact on the accuracy of the dyno run figures will it have if you do the run on a private road/track versus an actual dyno?

SiHethers

Andre

The road dyno uses acceleration, weight of vehicle and drag to determine power/torque. Acceleration is calculated from change in engine revs with gear you're in, wheel/tyre size and gear ratios factored into the equation. Therefore there is multiple factors in which slight variations may affect the result. For example: tyre pressure, tread depth, wind, not being on a completely level road etc. Not knowing the exact weight of the car will make a relatively big contribution to the power/torque figures, but if you're comparing before and after mods using the same guestimates it is likely to be representative.

There's been a lot of discussion on here about whether rolling roads are true reflections of road conditions, especially when it comes to ecu remapping. Theoretically, if you had a completely level road with no wind and knew the exact weight/drag of your car +had accurately measured the rolling circumference of your tyres the road dyno would be a more accurate reflection of power/torque in various gears because you are achieving realworld load/cooling effects. But the conditions would never be perfect in the real world. All IMHO

I use the road dyno for before/after modification comparisons, and in gear acceleration readings. So I know that I've gained 40bhp/50lbft at the flywheel with my current mods compared to standard and knocked 0.5-0.8 seconds off each 20mph increment in third

Torstein, I've yet to look at the other logs, but an each way comparison down the road you test on would allow you to calculate the average power, so long as there isn't a strong head/tailwind.

Simon

Edited to say - a may of missed the point of the original question. You couldn't use the DeltaDash Road Dyno on a real dyno because the dyno operator controls how fast the wheels accelerate by varying the load on the rollers. It has to be done on the open road.

[Edited by SiHethers - 4/8/2003 9:29:49 AM]

[Edited by SiHethers - 4/8/2003 10:16:13 AM]

SiHethers

Torstein

Check your mail. You dyno logs show very impressive low down torque, even as standard, and what looks like overboost. This could be due to either very low ambient temp, or there is some non standard exhaust components, either DP or centre section. I hope it's the latter

I think the decline on the last log must be more significant than appreciated. Best thing would still be a run in both directions. Tek1 made a significant difference in the mid to upper range, did it feel quicker?

Simon

torsteinvh

Simon

How do you see the overboost from the dyno logs? Is it by looking at the shape of the torque curve at low revs?

The log I posted on the first page of this thread is from a 1500-7000 RPM run in 3rd gear, and I cannot see anything abnormal with the boost curve. But then again, I'm not exactly an export on this :|

I compared my logs to yours and some other examples, and the shape of the torque curve change quite a bit if you switch between medium and heavy smoothing. If I use medium (or even light) smoothing, the shape of my curve looks more like yours.

Anyway, I would think any overboost is due to low ambient temperatures, in all the runs the temperatur was only a few degrees above 0.

The car was bought from a Subaru dealer, with no mention of any mods. I have also had the car checked recently to see if anything needed fixing before the warranty expired, and no mention of mods then either.

The car definitely feels quicker after the TeK1. It feels like it pulls harder over a wider rev band, and I think that is what EcuTek said it would do

-Torstein

SiHethers

Torstein

I see what you mean. I think I was talking about comparing my standard curve to your ae802 - as boost comes in, there is a relative peak in torque between 3000-4500. This would fit with overboost, although I agree it is most probably due to the cold ambient, and not necessarily a bad thing. In fact many EBC's allow you to deliberately produce overboost for a few seconds to aid overtaking. I think it was a standard feature of the Escort Cosworth.

I've now compared your deltalog of 3rd gear to one of my standard logs and the interesting thing is you spool up 300rpm before me, despite identical fueling and my ecu showing more advance in the 1700-3000rpm range, but then I hold 0.5-1.0psi more boost to 6500rpm. Injector duty cycles and ignition timing are almost identical from 3000-6500rpm, with mine showing the occasional 1 degree more of advance. Both our logs were snap WOT's from 1700rpm so are directly comparable. Your better spool up may indicate better VE due to lower ambient temp. Still doesn't explain how you've gained 10bhp with each road dyno run since fitting the tek1

Anyway, get that Magnex fitted and dyno some more.

Simon

torsteinvh

Simon

Do you mind mailing me your 3rd gear DD log, so I can compare with mine?

Maybe the 10 bhp gains are caused by the ECU learning fature?

I will notify you when I have the Magnex in place. It may be a while though, I have some work commitments coming up

-Torstein

swan

Thanks Simon, I totally lost the plot as far as the road dyno function was concerned.

SiHethers

No prob Andre

Torstein, the learning feature is only supported in the new-age impreza's. Our less sophisticated ECU has a standard map which, if it senses knock, can retard the ignition. If it sees knock frequently at a certain rpm/load combo it will learn this, i.e. automatically retard at that point rather than wait for the knock to occur. This then becomes the base map. If there is no knock consistently for a prolonged period it will then advance the ignition again slowly, but can not advance further than the programmed map allows. This slow process of relearning can be overcome by the ECU reset, say for example if you had to use 95 RON fuel for one tank, then back to 98RON with the next.

New age impreza's, as far as I understand, use a base map and advance/retard limits, so as soon as you start driving it will advance the ECU to just below the point of knock and hold it there actively across all loads/rpms by advancing and retarding as necessary. This process will optimise the ignition curve over the space of a hundred miles or so, hence the learning aspect/gains in power. I suppose it means the base map doesn't need to be so finely tuned to get the max performance, and makes the ECU reset obsolete.

I'll mail that standard log, and one with current mods.

Cheers

Simon

john banks

Swan YHM.

Torstein, don't think your Tek 1 is lean during spool up. After more than a tiny bit of boost the narrowband is over 850 mV. The drop to zero is the vertical response of the narrowband - the IDC rises smoothly through this zone. If you logged the pseudowideband through this zone it would give you more info.

swan

Thanks for the input John.

torsteinvh

John, appreciate your input
Can you tell me what you mean by the 'pseudowideband'?

I guess the 'narrowband' is what DD calls 'Front O2 Sensor' and IDC is Injection Duty Cycle?

-Torstein