Pages: [1]
Author Topic: Reading N75 article on s4 wiki  (Read 5727 times)
golfputtputt
Sr. Member
****

Karma: +6/-4
Offline Offline

Posts: 367


« on: February 14, 2017, 12:17:24 PM »

https://s4wiki.com/wiki/Wastegate_bypass_regulator_valve

Was reading through this and noticed a few things.

Quote
Notice the Restrictor in Figure 2. It allows only a small amount of air to enter from the green port below. As a, result the pressure above the Restrictor (that goes to the Waste Gate actuator) can be regulated by bleeding off controlled amounts of air through the left side purple port of N75. The bleed-off rate is controlled by how far the electric coil pulls up its central plunger, allowing air to escape under the blue cap at its bottom. That pull-up distance is precisely controlled by a pulse width modulated signal from the ECU. To achieve the intended results, N75 is factory calibrated by a screw above the plunger that sets the proper spring pressure of a spring between the plunger and the blue cap.

mentions that the n75 is managed by a PWM controlled when i believe it's a PID controller correct? Could we correct this to:

Quote
That pull-up distance is precisely controlled by a pulse width modulated [Actually a PID controller] signal from the ECU

just for congruity of information's sake? I know you didn't type this out yourself Nyet but it could help us noobs understand it if the terminology is congruent. Unless i'm off and it actually is PWM.

Also, the links at the bottom of that page are dead:

Quote

Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Online Online

Posts: 12232


WWW
« Reply #1 on: February 14, 2017, 12:24:35 PM »

It is a PWM signal driven by a PWM controller given a desired duty cycle value that comes from a PID.

So, no. Your understanding is incorrect.

PWMs and PIDs have completely different functions and are used for very different things, but are often used together.

Quiz: What is a PWM? What does it do?

What is PID? What does it do?
« Last Edit: February 14, 2017, 12:26:34 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
golfputtputt
Sr. Member
****

Karma: +6/-4
Offline Offline

Posts: 367


« Reply #2 on: February 15, 2017, 10:09:23 AM »

Had to ask someone irl to get a better understanding of this:

A PWM is the amount of time a device recieves a signal or an electrical current to turn it on or off, depending on its function, over a period of time, ex: an LED is sent a signal to be on 50% of the time over 1 min(arbitrary time amount for example's sake) so the LED is turned on for 30 seconds and off for 30 seconds split into rapid increments of the 1 minute course.

A PID is the protocol that the PWM is designed to follow, for example, said LED is looking to put out 50 lumens of light. Through an algorithm, 50 lumens is found to be achieved if the LED is 'pulsed' 75% on over the course of 1 minute. So those 45 seconds of 'on' are split up rapidly between 1 minute. This "setpoint" is the P function (ex: the 75%-50 lumens)(do i have this correct?). The I and D function come in when the desired amount of lumens changes over a longer period of time and the output of the PWM is tryin to follow these desired changes. The I is how much (how many lumens, per example) the output signal overshoots and undershoots after it hits the desired setpoint and the D function is how many times it overshoots and undershoots after it hits the desired setpoint  Do i have I and D correct?

I feel like im almost there?
« Last Edit: February 15, 2017, 10:15:14 AM by golfputtputt » Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Online Online

Posts: 12232


WWW
« Reply #3 on: February 15, 2017, 10:23:15 AM »

Through an algorithm, 50 lumens is found to be achieved if the LED is 'pulsed' 75% on over the course of 1 minute.

If the algorithm is known, a PID is not required, only a fixed calculation of what the DC should be for a given output.

Quote
The I and D function come in when the desired amount of lumens changes over a longer period of time and the output of the PWM is tryin to follow these desired changes. The I is how much (how many lumens, per example) the output signal overshoots and undershoots after it hits the desired setpoint and the D function is how many times it overshoots and undershoots after it hits the desired setpoint  Do i have I and D correct?

No.

Do you know what a derivative is? An integral?
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
golfputtputt
Sr. Member
****

Karma: +6/-4
Offline Offline

Posts: 367


« Reply #4 on: February 15, 2017, 03:26:31 PM »

watched these guys:
https://www.youtube.com/watch?v=SefKQb9y_B4
https://www.youtube.com/watch?v=16Clfh5eBzg&t=944s

and read a bit and came up with this:

Derivative works on the rate of change of error in order to approach the setpoint slower, over and undershoot the setpoint fewer times and to overshoot and undershoot the setpoint to a smaller degree.

so, keeping with the same example, the LED is looking to achieve a setpoint of 75% DC. If the Proportional gain is adjusted properly the setpoint will be reached with an acceptable reaction time and will hunt a minimal of times. The derivative gain can be adjusted to reach setpoint with fewer hunts and less over and undershoot amount. The downside of this tends to be a slowing of the reaction time near approaching the setpoint.

Integral gain works to cover the rest by closing the remaining error gap by adding smaller amounts of DC until the setpoint is reached.

I now understand that as the setpoint changes, the PID functions are used together and separately, depending on how easily the setpoint is to reach. I always thought these had an operative order, P, I, then D and were always computed and reacting in that order, that is not necessarily true. I also thought they were all being computed every time the setpoint changed, that is not necessarily true either.

I'm trying to understand how their root words of derivative (to derive, or to copy/obtain from something else) and integral (integer, a whole number) pertain to the functional operation they perform, as I am expected to fully understand the concept for my employment as of late, unrelated to automotive application.

I'm still getting a grasp on the fundamental mathematics of how these work and interact but I think I have it.
« Last Edit: February 15, 2017, 03:35:20 PM by golfputtputt » Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Online Online

Posts: 12232


WWW
« Reply #5 on: February 15, 2017, 03:41:24 PM »

They are ALWAYS computed whether or not the setpoint changes.
« Last Edit: February 16, 2017, 12:55:12 AM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
AARDQ
Sr. Member
****

Karma: +11/-0
Offline Offline

Posts: 338


« Reply #6 on: February 15, 2017, 11:10:49 PM »

And it's typically thought of the other way, with integral "covering the rest", and derivative very often nulled, or at least used sparingly and with great care -- it can cause control element hyperactivity and ultimately worse control, or a worse experience even if objectively process variable control is "better".  ME7 etc. with its patented I-limiter is a bit different than generic PID controllers and can utilize derivative, but even so a little typically goes a long way.
Logged
golfputtputt
Sr. Member
****

Karma: +6/-4
Offline Offline

Posts: 367


« Reply #7 on: February 16, 2017, 10:51:12 AM »

http://nefariousmotorsports.com/forum/index.php?topic=517.msg4028#msg4028

Is this thread essentially outlining the 'feed forward' open loop tuning process required before being able to accurately calibrate the PID?

Quote
And it's typically thought of the other way, with integral "covering the rest", and derivative very often nulled, or at least used sparingly and with great care -- it can cause control element hyperactivity and ultimately worse control, or a worse experience even if objectively process variable control is "better".  ME7 etc. with its patented I-limiter is a bit different than generic PID controllers and can utilize derivative, but even so a little typically goes a long way.

does this mean Integral gain should be addressed before derivative gain or is the order of operations still P,D then I?
« Last Edit: February 16, 2017, 10:53:11 AM by golfputtputt » Logged
golfputtputt
Sr. Member
****

Karma: +6/-4
Offline Offline

Posts: 367


« Reply #8 on: February 16, 2017, 10:59:39 AM »

also, noticed this under the: Specifying requested boost section

Quote
Note that if you do plan to run closed loop near 2.5 mbar, you are much better off getting a larger pressure sensor and running the 5120 hack.

should it read bar instead of mbar? 2.5 mbar is 0.0362594 psi
Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Online Online

Posts: 12232


WWW
« Reply #9 on: February 16, 2017, 11:42:32 AM »

http://nefariousmotorsports.com/forum/index.php?topic=517.msg4028#msg4028

Is this thread essentially outlining the 'feed forward' open loop tuning process required before being able to accurately calibrate the PID?

More or less. It is tuning the output linearization compensation of the PID controller based on the "open loop" behavior of the hardware during constant wg input. That said, his results still don't actually look quite right to me, but I haven't had the time to figure out where he went wrong.

Quote
does this mean Integral gain should be addressed before derivative gain or is the order of operations still P,D then I?

99% of PID tuning can be done by properly tuning the wg linearization and the I limiter. Typically, the stock D table can be simply be moved to higher RPM such that it is centered around the area where the wg first generally should be opened.

Note that when tuning a PID from scratch completely different methods are used.

Thanks for catching the typo, I will fix it time permitting.

« Last Edit: February 16, 2017, 11:52:37 AM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Pages: [1]
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.129 seconds with 17 queries. (Pretty URLs adds 0s, 0q)