R34 tuning - How much does an air filter affect things??

[PL]

OK tuning nerds and engine geeks,

I have a tuning story to share and I invite your comments/suggestions. It's a bit long.

R34 RB25DET. Friend of a friend bought this car and so I got the call when it came time for tuning. He did quite well - got it cheap at auction and it's a manual car in that nice metallic blue that you see R34 GTR's in. Turns out it had a GTR pump fitted, big exhaust and dump pipe, FMIC, some sort of HKS turbo and a set of NISMO 480's. No AFM cos they must have been running a D-Jetro or something previously.

Anyway, after an initial check and Z32 AFM fitment we took it to the dyno and after 2 hours hard work I had it somewhere decent. Must be a small turbo cos boost wouldn't hold any more than 12psi at high rpm. Producing an underwhelming 225rwkw.

So anyway, I took it out on the road for the customary road test (with trusty "knock ears" still fitted) and it went fine if boost ramped up slowly (ie: like it would be on the dyno) but any sudden boosting would result in knocking so hard it made me wince! Nasty.

I took out some more timing and sent him away with instructions to come back and see me, so we could hook up my LM-1 and the Knock Ears for some serious road testing and logging. Given some time to reflect on the situation I was thinking it had to be AFM related. He had only fitted a used Z32 AFM, so this was immediately suspect. During road testing when boosted up suddenly it wasn't seeing enough load - resulting in lean mixtures and high timing figures. So I removed the AFM and cleaned the sensing element thoroughly.

Before re-fitting the cleaned AFM I tried my brand new "reference" Z32 AFM on the car without air filter. It was rich everywhere and knock was largely gone. So then we whacked the cleaned AFM back on and got similar results. So then we pulled over and I got the owner to re-fit the air filter. I was sitting watching the figures with the engine running and idle was rich at around 11.5:1. To my amazement when the air filter went back on the AFR's slowly climbed to 14.5:1 (over 5 to 10 seconds). So I got him to remove the filter again and it went back down to 11.5:1! I noted the figures -

Filter off = AFM 1.3V, inj 1.0ms, AFR 11.5:1
Filter on = AFM 1.23V, inj 0.95ms, AFR 14.5:1

That's not much change in AFM voltage hey! But it made a big difference to idle mixtures.

We then drove the car under load a bit more with filter on and we were back to lean mixtures and detonation.

Air filter was one of those HKS foam things (what I call "gravel traps" due to their poor filtering ability). It was in good condition but it looks like it's causing a restriction even at idle. Who knows - it may even be the reason for the low boost and poor power figures...

I've instructed owner to go get a K&N and I've booked the dyno for more fun and games tomorrow night.

I'll let you know the results.

You guys are great listeners.

PL

[TM_S13]

Now that's a curious thing to read on...

Quite "weird" at the same time, though after reading the results and values kind of makes some sence...

I'm not too fond of K&N air filters due to the fact that some of them come from factory a bit too much oiled wich in turn gets into the AFM filaments and stuff etc you know what I'm going to say.

The turbo not holding boost or spooling late might also be due to a weak, or poorly fitted, actuator too... I've seen it happen before.

Let us know how it goes after the new dyno session with the new air filter
I got curious now, so that I might even remove the air filter on my car just to see if I spot any diferences too hehe.

Thanks

[HermaN]

This is rather interesting! I personally don't like the HKS filters due to poor filtration as you said, even the new 3-layer ones in all honesty. Hence I use an A'Pexi induction kit which is a dry cone filter and 100% maintenance free. As TM_S13 said above, don't like oiled filters when used with a MAF application either, because if over-oiled it can foul the AFM and send it way off until cleaned up again.

Please do keep us updated regarding this. Really interested to see how it goes with the new filter.

[bachig24u]

PL, Different materials or simply changing the intake pipe shape, size, or curve. Same kind of result, the airflow over different areas on the sensor gives different results. It all comes down to air speed over the sensor.
By simply fitting a restrictor plate with a small 1"" hole cut in it you can richen mixtures dramatically. Lots of geeky fun by using a plate with an offset hole cut in it to simlulate pipe bends.

I usually don't take the AFR's under 1500rpm in any pipes larger than stock as gospel unless I'm using gas analyzer.

[Legionnaire]

I'm not sure if the problem is in AFM, as injection duty seems consistent with AFM output - slight change in one causes slight change of another.

It is AFR that seems wrong for the PW change. Purple R34 nismos flow 480cc/min at 3 bar of fuel pressure, so
1 ms PW delivers 8 cubic millimeters of fuel,
at 0.95 ms pulse 7.6 cu. mm. are injected.

Assuming fuel density of 0.72kg/litre, results are
5.76 miligram of fuel @ 1ms pulsewidth
5.472 milligram @ 0.95ms PW.

Now calculating airmass for delivered fuel and actual AFR, we have
5.76 * 11.5 = 66.24 mg of air
5.472 * 14.5 = 79.344 mg of air

Going back to the beginning, his means

Filter off = AFM 1.3V, inj 1.0ms, AFR 11.5:1 = 66.24 mg of air
Filter on = AFM 1.23V, inj 0.95ms, AFR 14.5:1 = 79.34 mg of air

So either filter has some hidden induction device in it, or something is wrong with AFM,as it shows lower voltage at higher airflow, WTF

This is really strange. I don't know what is with this engine, but I generally don't trust AFR readings at and near idle, as due to misfire, badly prepared mixtures and slow gas velocity AFR can jump all over the place. AFM readings may not be very useful either, as I believe that at idle AFM input is either disregarded or is not relied heavily upon by ECU. What this kind of behaviour could mean is significant VE drop AND turbulence caused by air filter, which resulted in both jumping AFM output voltages and lower load cells being accessed. This theory also explains why the car idles poorly, is lazy in creating boost and struggles to make adequate boost pressure. On the other hand, low VE engine should not be as knock prone as higher VE ones.
PL, did you happen to look at the turbo? What kind of HKS is it?

My 0.02c.
Regards, Petros.

[bachig24u]

Are the engine temps the same in both examples?

[PL]

OK. Well I'd hoped to have a definitive answer to this. But I'm far from it...

Unfortunately the owner found a loose hose clamp - so it'd been losing air under boost all along. First run on the dyno run it showed a 2 to 3 psi improvement across the range.

So we embarked on some air filter testing. He'd bought a cheap "K&N clone" and fitted it. With this one in place I did a run and AFR's were quite rich. We removed it at idle and idle AFR's went from around 12:1 up to 14.5:1. The tuner at the dyno shop held the new filter in one hand and pointed to the bin...

Then I did a run with no filter and AFR's leaned out a bit. Then we borrowed a small K&N from the dyno shop and AFR's were pretty much identical to the run with no filter fitted. I also fitted my new Z32 AFM with no filter and AFR's came up the same.

So then we tried his original HKS foam filter and AFR's went quite lean. This really threw me as a pattern was developing until that point. I aborted that run as it went lean enough to cause detonation with the IGN timing figures being used.

We then aborted the dyno session and the owner is going to source a new genuine K&N filter. We would have grabbed one from the dyno place but they had none the right size.

So that's it. Inconclusive really....

I'm thinking that some of the AFR differences between runs with different filters may have been influenced by underbonnet temps, heatsoak etc. Coolant temp was quite consistent between runs.

Very annoying. I really was hoping to get some real answers here. On the upside, it's now making around 245rwkw - up from 225. Due to the extra boost. Once tuned we should be able to improve on that a bit too.

PL

Paul R34 filter tests edit.jpg

(86.8 KiB) Downloaded 3660 times

[bachig24u]

So the moral of the story, when in doubt up the boost to hide the flat spot ???

[PL]

No, the moral of the story is to spend 30 seconds checking all the clamps before you tune!

PL