Ignition rev 6 and Kaboom?
#61
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Yes, thats how I felt when I got the results. Its an automatic (tiptronic) so it looks like a really bad down shift to second while in manual mode. It being so recent makes me want to run (not walk) away but with all the discussion here, I thought i'd get a few more opinions before dismissing it all together.
#63
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
When someone pointed out a single range 6 over rev and suggested a glitch, I thought how could that be, pretty sure it's not a glitchy software problem ... but perhaps it's a sampling and aliasing problem.
Should be simple enough to test the idea.
From Fourier analysis, we know we need to sample the highest frequency to be able to faithfully re-create all lower frequencies. Let's pick 9000 rpm just to be safe (if we get a 9000 rpm money shift in a 997, we're not gonna be happy).
So 9000 rpm / 60 sec / min = 150 rotations per second .
By Nyquist we need to sample at least 2x that frequency or 300 rotations per second. So to get a sample rate in milliseconds, 1000 /300 = 3.33 millisecond sample rate.
Ok, the sensors on the engine that read the analog rotation and convert it to a digital signal to be transmitted to the DME and logged need to sample about every 3 milliseconds to faithfully re-create an analog 9000 rpm event in the digital world.
But what if the system only samples at let's say 4 milliseconds, we could faithfully only recreate an analog 7500 rpm event. So if a higher rpm is reported we cannot be certain it is real versus simply an aliased signal; perhaps a "glitch" in this regard (for example, an 8800 rpm event may be very real and recorded, we just don't know with certainty this is the case - unless the engine blows - then we know with certainty
). And that's the problem with aliased signals, they cannot be distinguish from real signals.
To be fair this is rather simplified example and I do not know the sampling rate into the DME. I would hope it is sufficient to not have aliasing effects at the top of range 6.
Apologies for the "left turn", returning to the subject at hand ...
Should be simple enough to test the idea.
From Fourier analysis, we know we need to sample the highest frequency to be able to faithfully re-create all lower frequencies. Let's pick 9000 rpm just to be safe (if we get a 9000 rpm money shift in a 997, we're not gonna be happy).
So 9000 rpm / 60 sec / min = 150 rotations per second .
By Nyquist we need to sample at least 2x that frequency or 300 rotations per second. So to get a sample rate in milliseconds, 1000 /300 = 3.33 millisecond sample rate.
Ok, the sensors on the engine that read the analog rotation and convert it to a digital signal to be transmitted to the DME and logged need to sample about every 3 milliseconds to faithfully re-create an analog 9000 rpm event in the digital world.
But what if the system only samples at let's say 4 milliseconds, we could faithfully only recreate an analog 7500 rpm event. So if a higher rpm is reported we cannot be certain it is real versus simply an aliased signal; perhaps a "glitch" in this regard (for example, an 8800 rpm event may be very real and recorded, we just don't know with certainty this is the case - unless the engine blows - then we know with certainty
![Frown](https://rennlist.com/forums/images/smilies/frown.gif)
To be fair this is rather simplified example and I do not know the sampling rate into the DME. I would hope it is sufficient to not have aliasing effects at the top of range 6.
Apologies for the "left turn", returning to the subject at hand ...
#64
Rennlist Member
#65
Instructor
Join Date: Mar 2015
Location: Lake Oswego, Oregon 97035
Posts: 104
Likes: 0
Received 0 Likes
on
0 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Originally Posted by Johnny DB
A tiptronic will let you money shift when in manual mode???
#66
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
I have heard of Range 1s and 2s on Tiptronics, theoretically induced by the sheer inertia of the pistons when accelerating fast and hitting redline, but Range 5s and 6s is a new one for me.