Could you guys datalog this test:

While idling, jab the throttle very quickly and briefly around 3x, about a second apart.

Then from the datalogs, examine the rise time - the time it takes between - when the log shows the throttle hitting about 15-25%, to the time that the MAP sensor reads about 85-90 kPa.

I have an AEM, and in my case with just one restrictor, it's about 80-100 mS. My problem is that my MAP signal is noisy. If I make the restrictor smaller, the rise time goes to > 150 ms, and it causes a tip-in hesitation.

I'm interested especially with data with the MAP sensor connected to the FPR. I have a 2000 so I don't have an FPR.

 

Well that's interesting.

Do you think a large fuel filter damping the noise on the MAP would also cause a delay in the rise time?

I'm wondering if the fuel filter approach would dampen things without slowing the response as much as a restrictor.

 

Yes and possibly yes. Thus my post.

 

I don't have the specific data you're looking for, but here and here is some data I collected on the MAP noise issue. I think it can probably be solved by sampling the MAP sensor in a different way (either at a different crank angle or by oversampling and filtering) but I haven't had a chance to experiment.

 

Thanks. I don't expect there to be a difference between the snubbers you linked (porous disc), and a simple restrictor (I used a 5 mm long piece of 3/16" brass rod stock with a .035 hole drilled through).

Has anyone experimented with different pickup points on the VICS manifold?

 

Jason, what map sensor are you using on your AEM?

I have a GM 3 bar and i am experiencing some strange issues that im sure i could get tuned out with something like this.

(main problem is free revving holding say 10% throttle (delta tps=0) the RPM will swing as much as 1000rpm)

 

MAP sensor. I was using a GM 2 bar, but now I switched to a Motorola 2.5 bar.
I'm uncertain if I had the same problem with the 2 bar, but I expect I did, but it's just that I am tuning my fuel map now much more finely and thus only noticed the problem now, in prep for running moar boost.

 

I built a circuit which works great:
https://www.miataturbo.net/ecus-tuning-54/my-electronic-map-filter-circuit-48665/

 

Interesting. When I have time I am going to try implementing a discrete time equivalent in the MS2. An exponential smoothing algorithm would be easy to implement and present low CPU overhead.

 

kday if you do it digitally you may want to combine a 2 pole bessel (my filter) with a little bit of hysteresis and median filtering.