LGT AVCS Tuning Discussion

[LittleBlueGT]

I did some more runs on the way home from work tonight. I don't feel like graphing it out, but I basically dropped 5 degrees across the board, didn't see any notable difference.

 

It is interesting that with my "normal" UELH I could run up to 40 degrees tapering down to 0 at about 6400 rpms for best power.

 

Now with the WBR ELH I make the best power at about 30 degrees tapering down to 0 at 6000 rpms.

 

This is not in any way conclusive, but it makes sense. The WBR header only has 1 X-over tube, just like OEM manifold, whereas most headers (UEL or ELH) have two. I would guess that the bigger the hotside of the turbo, and the larger the volume between the cylinder head and turbo the more AVCS you can run.

[SeeeeeYa]

I did some more runs on the way home from work tonight. I don't feel like graphing it out, but I basically dropped 5 degrees across the board, didn't see any notable difference.

 

It is interesting that with my "normal" UELH I could run up to 40 degrees tapering down to 0 at about 6400 rpms for best power.

 

Now with the WBR ELH I make the best power at about 30 degrees tapering down to 0 at 6000 rpms.

 

This is not in any way conclusive, but it makes sense. The WBR header only has 1 X-over tube, just like OEM manifold, whereas most headers (UEL or ELH) have two. I would guess that the bigger the hotside of the turbo, and the larger the volume between the cylinder head and turbo the more AVCS you can run.

 

C.T. says over and over, don't add cam advance without retuning timing and AFR... although AFR is not quite as important. You didn't do that, something he, more or less, berates people for doing.

 

With my APS ELH I have defined only one thing close to certain with the AVCS, advance over 6000 rpm loses power.... just as you have found. Therefore, given that my 5EAT produces the 'quickest' 0-60s with a lot of advance early that extends close to 4000 rpm, the AVCS map's mystery begins to disappear.

 

You are the only person I know who is working on this in a methodical manner and sharing, bravo.

[caramall2]

Any idea what the magical combo is? E.g. Increase or decrease timing with increase AVCS? Increase or decrease AFR with increase AVCS? I'm sure that's what we all want to know ;-)

[Infamous1]

Any idea what the magical combo is? E.g. Increase or decrease timing with increase AVCS? Increase or decrease AFR with increase AVCS? I'm sure that's what we all want to know ;-)

You would in theory decrease ignition timing with more AVCS advance. As for the AFR's they should go leaner with more AVCS advance. This is just the result, you want your AFR's to be the same if it was making the best power where it previously was. In that case just tweak the MAF table to get you back to where you were.

[Infamous1]

For thosse of you that haven't read the big thread on AVCS on Nasioc,

here is an excellent way (without dyno) to see how the AVCS changes effect your car.

 

Here's some test data I collected...

 

What I did was set timing to 12 degrees fixed and then set the AVCS advance to 0 degrees advance, 10, 20, 30 and 40 degrees and captured injector time and AFR (via wideband AFR).

 

This allowed me to calculate cylinder load at each given data capture point (the total mass of air in all cylinders at the point of capture), which is simply based upon the injector time at each data capture point and the measured AFR at the data capture point. I have assumed my injectors flow 565cc/min and that fuel has a mass of 0.76g/cc. This calculated engine load is directly proportional to mass air flow.

 

Here are the results (note that I stopped the 30 degree and 40 degree runs early as I know that much advance isn't very good at high RPMs):

 

Mass Air Flow

http://img141.imageshack.us/img141/5964/avcsloadsi9.jpg

 

Measured AFRs

http://img291.imageshack.us/img291/4166/avcstestru2.jpg

 

The differences are interesting. In theory, the mass air flow graph should be what you work from, but the 20 degree advance result is somewhat misleading at high RPMs. The mass air flow graph indicates 20 degrees advance is better at high RPMs than 0 degrees advance - this is definitely not the case, based upon seat of pants and the fact the car is running about 2PSI more boost with 20 degrees advance at high RPM, indicating the engine is more restricted.

 

Based upon the mass air flow graph, these would be the optimal AVCS settings (assuming your map uses 250RPM increments:

 

4000RPM: 40 degrees advance

4250RPM: 30 degrees advance

5000RPM: 20 degrees advance

5750RPM: 10 degrees advance

 

Based upon the AFR graph, these would be the optimal AVCS settings (assuming your map uses 250RPM increments:

 

4000RPM: 40 degrees advance

4750RPM: 30 degrees advance

5500RPM: 20 degrees advance

6500RPM: 10 degrees advance

7500RPM: 0 degrees advance

 

Next time I'm on a dyno I'll repeat this exercise and see which of the above (calculated mass air flow or AFR) is more aligned with power output...

[LittleBlueGT]

C.T. says over and over, don't add cam advance without retuning timing and AFR... although AFR is not quite as important. You didn't do that, something he, more or less, berates people for doing.

 

 

I agree with him.

 

I also agree with what I said a few posts above, timing is a crutch! I run conservative timing (not like I have E85 and risk going over MBT:lol:) and test AVCS, same for fueling.

 

Once I find the best I then proceed to tweak my timing around it.

 

Fueling I will discuss in my next post.

[LittleBlueGT]

You would in theory decrease ignition timing with more AVCS advance. As for the AFR's they should go leaner with more AVCS advance. This is just the result, you want your AFR's to be the same if it was making the best power where it previously was. In that case just tweak the MAF table to get you back to where you were.

 

I still need to do more testing, but this whole tweaking the MAF table is not always the best answer.

 

First off, why should AFR change as cam timing changes? In my tests it does not, certainly not near as much as the NASIOC graphs do. X amount of air going in = X amount of fuel, it all gets burnt (some may argue with too much AVCS it will get burned out the exhaust).

 

I actually think that the reason why so many people see different AFR readings is becasue the MAF scale is NOT CORRECT at higher voltages, then when the cars flows more air (or less) it is on a different load/rpm cel on the primary fuel table (which is now slightly skewed).

 

Why do I think that?

 

Well, I scaled the high end of my MAF using many many multiple top gear runs, sure there are other ways to do it, but I highly doubt as accurate as mine. I also mixxed things up, I didn't just go WOT in 5th from low rpm (as that would always have the same flow at the same AVCS settings) but did 30% throttle runs, 50% throttle runs, etc...

 

I also noticed no change in my high voltage MAF scaling with my TGV deletes and inlet, and also now my WBR ELH. Bothchanged airflow somewhat, but seemingly not AFR. Note: I have not done this as thoroughly yet, but will this week.

 

What happens if you do tweak the MAF table for this supposed change in AFR? (assuming that these mods did change AFRs) Say you are hitting 160 g/s at 3000 rpm (high AVCS position) but find it is off, well what happens when you are partial throttle at 5500 rpm (hardly any AVCS) and hitting 160 g/s? Same MAF V? It all of a sudden changed at 3000 rpm, but it is not going to change at 5500 rpm?

 

That is my opinion. But you have to admit, I have some very very valid points.

[Infamous1]

Well I know for a fact my MAF scaling is on point more than most . "I" have not found AVCS to effect my MAF scaling at all, however I haven't gone from one extreme to another either (high vs low timing). But others have posted results which have shown otherwise. MAF scaling is essentially the basics of tuning so I can only assume one would have this competely dial in before they ventured into AVCS tuning.

 

With more AVCS there will be unspent fuel in the exhaust, this is a fact and can be seen in gas mileage and EGT's. As far as other mods changing the MAF table that is a completely different topic which has nothing to do with AVCS tuning, so I leave that alone.

[BlackHole]

Any idea what the magical combo is? E.g. Increase or decrease timing with increase AVCS? Increase or decrease AFR with increase AVCS? I'm sure that's what we all want to know ;-)

 

The magical combo will vary somewhat with setup at low load / rpm, but can vary quite a bit a high load / rpm depending on turbo / IC / etc. But those tuners who do know aren't posting, which makes sense since it's their livelihood.

 

Side note - zeroing out the AVCS table reduced Volumetric Efficiency by about 4% on average. Before you ask - I zeroed the AVCS as part of my tuning DOE. I'll post the results when I'm done in a few weeks.

[Infamous1]

The magical combo will vary somewhat with setup at low load / rpm, but can vary quite a bit a high load / rpm depending on turbo / IC / etc. But those tuners who do know aren't posting, which makes sense since it's their livelihood.

 

There really is nothing magical about it, if you just try a few different settings and plot it out you will see what works best. If you really want to see the results just rent some dyno time.

 

 

Just an FYI: My AVCS table looks nothing like the one from the first post.

[LittleBlueGT]

Well I know for a fact my MAF scaling is on point more than most . "I" have not found AVCS to effect my MAF scaling at all, however I haven't gone from one extreme to another either (high vs low timing). But others have posted results which have shown otherwise. MAF scaling is essentially the basics of tuning so I can only assume one would have this competely dial in before they ventured into AVCS tuning.

 

With more AVCS there will be unspent fuel in the exhaust, this is a fact and can be seen in gas mileage and EGT's. As far as other mods changing the MAF table that is a completely different topic which has nothing to do with AVCS tuning, so I leave that alone.

 

I agree, unburnt fuel in the exhaust still gets burnt up pretty darn quick, more like flames still coming out the exhaust valve IMO.

 

I hope to fine tune my MAF today, so we shall see, but with AVCS experimenting with my old header it didn't make any difference at all with AFRs. I just find the chart you posted funny. 20% different in AFRs from the AVCS changes alone? Makes you wonder about the validity of the rest of his findings.

[Infamous1]

20% different in AFRs from the AVCS changes alone? Makes you wonder about the validity of the rest of his findings.

Most likely was on an unaltered MAF scale on the stock box which almost all folks leave untouched.

[caramall2]

Good info and discussion. Infamous, can you post your current table? Just curious.

 

Just an FYI: My AVCS table looks nothing like the one from the first post.

[LittleBlueGT]

Most likely was on an unaltered MAF scale on the stock box which almost all folks leave untouched.

 

Could be, mine OEM scaling were not the greatest.

 

I redid the laborious task of scaling the upper end of my MAF today. Multiple 5th gear runs.

 

I found that it was off by about 2% from my previous set-up. It was fairly consistent from about 1% off at 3 volts to a max of 2.5% off at 4.7 volts.

[NSFW]

First off, why should AFR change as cam timing changes? In my tests it does not, certainly not near as much as the NASIOC graphs do. X amount of air going in = X amount of fuel, it all gets burnt (some may argue with too much AVCS it will get burned out the exhaust).

 

I had always assumed that Clark's point is that with different AVCS settings, the optimum AFR changes. Not the observed AFR. If the observed AFR changes, then either the MAF scaling is off, or it's just noise in the AFR signal. So I figure he means that with the right AVCS setting you should aim for a richer or leaner AFR (I can't guess which).

 

And speaking of noise in the AFR signal, there's a lot of it, in every graph I've seen (other than those in dyno charts with smoothing applied). In that series of graphs up above, with all the jagged lines, there's more variance in the noise than in the underlying signals... That guy's test results would be a lot more interesting if he'd done several runs at each setting and averaged the data, to get a better signal:noise ratio. Of course it would take more work, but it would produce more meaningful results.

[LittleBlueGT]

I had always assumed that Clark's point is that with different AVCS settings, the optimum AFR changes. Not the observed AFR. If the observed AFR changes, then either the MAF scaling is off, or it's just noise in the AFR signal. So I figure he means that with the right AVCS setting you should aim for a richer or leaner AFR (I can't guess which).

 

And speaking of noise in the AFR signal, there's a lot of it, in every graph I've seen (other than those in dyno charts with smoothing applied). In that series of graphs up above, with all the jagged lines, there's more variance in the noise than in the underlying signals... That guy's test results would be a lot more interesting if he'd done several runs at each setting and averaged the data, to get a better signal:noise ratio. Of course it would take more work, but it would produce more meaningful results.

 

I have tried from 11:1 to 13:1 in the spool-up range (all w/ alky) and noticed no appreciable difference.

 

I understand your point though.

 

Sometimes I think people choose not to divulge little secrets just to keep a mystery going, like they are the best tuner or something. Many times I think those "secrets" are nothing.

 

I helped a guy tune his FPG, then he went to clrak for one of his special tunes, no real difference, except after the clark tune the motor went.

[NSFW]

Always try for maximum VE, timing is a crutch, as in if you can get better power with a different AVCS setting do that, even if it means you have to take some timing out.

 

Where does the "timing is a crutch" idea come from? For any given conditions (load, afr, compression, rpm) isn't there an 'ideal' timing that will produce the best torque? (Peak cylinder pressure at 20 (?) degrees ATDC, IIRC.) Though we have to run less at high loads due to detonation.

 

If a new AVCS setting means you can get better power with less timing, that just means you had too much timing for the new AVCS setting.

[Infamous1]

 

I helped a guy tune his FPG, then he went to clrak for one of his special tunes, no real difference, except after the clark tune the motor went.

Wow, I hope that was Ron....

[LittleBlueGT]

Where does the "timing is a crutch" idea come from? For any given conditions (load, afr, compression, rpm) isn't there an 'ideal' timing that will produce the best torque? (Peak cylinder pressure at 20 (?) degrees ATDC, IIRC.) Though we have to run less at high loads due to detonation.

 

If a new AVCS setting means you can get better power with less timing, that just means you had too much timing for the new AVCS setting.

 

You are thinking exactly the same way I am. Maybe it is just getting lost in the type.

 

There is an "ideal" timing for every...................

 

What I meant was tune AVCS for best VE, then run the best timing you can (which, unless you are on E85, is likely not past MBT). One should theorize, "if I run this AVCS setting I have to run less timing, so I won't do it".

 

Am I clear now?

 

 

infamous, it was ron.

[Legend]

What I meant was tune AVCS for best VE...

 

For best "VE"? This term always throws me off. What do you mean? What are you measuring/logging?

[NSFW]

OK, I'm with you now. I just misunderstood the "timing is a crutch" thing.

 

Legend, VE basically means "how much air does the motor pump, per revolution, for a given manifold pressure." It can be expressed as a percentage, where 100% VE means the motor pumps 2.5L per two revolution, or you can just look for changes in load (or MAF at a given RPM) for equivalent pulls.

 

See the 2nd screenshot on this page:

http://www.codeplex.com/ssm/Wiki/View.aspx?title=Pictures

That shows it as a table of load values, with MAP on the left and RPM on the bottom.

[LittleBlueGT]

For best "VE"? This term always throws me off. What do you mean? What are you measuring/logging?

 

Not so simple to log. There is a thread on it in RR, I am sure you have seen it.

 

There is the possibility that better "VE" actually wastes some air out the exhaust valves w/ too much advance.

 

I really just tune for best power.

[Guest]

AVCS map for JDM STI twinscroll looks quite different than any USDM AVCS maps... (more here)

 

http://freon.shackspace.com/car/ecu%20and%20tuning/camadvance/2006jdm-sti-twinscroll.png

[BlackHole]

There is the possibility that better "VE" actually wastes some air out the exhaust valves w/ too much advance.

 

If this happens it should be pretty apparent from the AFR going lean.

[mickeyd2005]

If this happens it should be pretty apparent from the AFR going lean.

 

 

Not really. LBGT meant air-fuel mixture going out. The amount of excess fuel relative to oxygen content is still the same.