Page 1 of 1
Something strange with my knock module
Posted: Wed May 30, 2018 6:01 pm
by Lyonz
New topic from:
https://rusefi.com/forum/viewtopic.php?f=5&p=29219&sid=4cc9fa7214bbe5c825a965e9196946c2#p29219
russian wrote: ↑Tue May 29, 2018 12:35 am
let's move into some knock-related topic or start a new one?
What does
hipinfo command say in rusefi console?
To answer your question
This is my hipinfo why the last firmware
"2018-05-30 13_52: EngineState: confirmation_hipinfo:7
2018-05-30 13_52: EngineState: spi 1=No/2=Yes/3=Yes
2018-05-30 13_52: EngineState: enabled=Yes state=1 bore=78.50mm freq=7.28kHz PaSDO=6
2018-05-30 13_52: EngineState: band_index=42 gain 1.00/index=-1 output=PC0
2018-05-30 13_52: EngineState: integrator index=-1 knockVThreshold=2.35 knockCount=0 maxKnockSubDeg=20.00
2018-05-30 13_52: EngineState: spi=SPI_DEVICE_2 IntHold@PB11/2 response count=3 incorrect response=0 ok
2018-05-30 13_52: EngineState: CS@PB0 updateCount=0
2018-05-30 13_52: EngineState: hip 0.00v/last=2.65@PC0/max=0.00 adv=0
2018-05-30 13_52: EngineState: mosi=PB15
2018-05-30 13_52: EngineState: miso=PB14
2018-05-30 13_52: EngineState: sck=PB13
2018-05-30 13_52: EngineState: start 35.00 end 135.00
2018-05-30 13_52: EngineState: knock now=No/ever=No
"
Re: Something strange with my knock module
Posted: Wed May 30, 2018 6:05 pm
by Lyonz
puff wrote: ↑Tue May 29, 2018 11:45 am
let's spin off a new topic!
so what do we expect to see in that hipinfo command?
basically, we need some sort of explainer, containing thorough and complete information about the knock phenomenon itself, about setting up knock detection part of rusefi (both, software- and hardware-wise).
1. On a four-cylinder four-stroke enine, knock events occur four times over a camshaft revolution, right?
2. Knock window in rusefi is a sector of a camshaft revolution (relative to initial CPS offset) , where we can expect to see knock in the first cylinder (btw, if knocking occurs just at some certain part of the certain stroke, then what is the need in this setting?)
3. During the knock window, the ADC circuit in rusefi board samples the values from the HIP chip and calculates the average value (can't understand that integration-related graphs and slopes). If that value is higher than those in the predefined table, it triggers a knocking event (which can be observed in tunerstudio).
This is why just knocking with a hammer on the sensor won't necesarily be seen in console, you need to knock at the write moment.
4. So far, Rusefi doesn't have anything in terms of knocking avoidance, does it? No special ignition logic, nothing - just capturing the knock events.
5. predefining the table of baseline 'noise' level requires quite some effort. probably we need some recommendations, where to start from.
6. setting gain - same question
Very good track for future developments.
A question all the same, it also implies to have an installation with camshaft sensor
Re: Something strange with my knock module
Posted: Wed May 30, 2018 9:05 pm
by puff
not necessarily for a 4-stroke 4-cyl engine, unless you have more than one knock sensor (not sure if that setup is even supported in rusefi thogh)
for a simple case, crank sensor of at least some accuracy is imho enough
Re: Something strange with my knock module
Posted: Thu May 31, 2018 4:18 am
by AndreyB
So we have few interesting points here
response count=3 incorrect response=0 ok
this tells us that SPI looks fine and we have communication with the chip
also
hip 0.00v/last=2.65@PC0/max=0.00 adv=0
is interesting line - one the one hand the output voltage is 2.65 on the other hand the firmware did not care for these values so far hip 0.00v and max=0.00. Maybe because RPM were always zero and integration was never started since integration is synched with crankshaft?
I wonder what exactly are you doing? Please describe your test scenario. Is engine running? Is this a board on a bench? What are you doing exactly and what are you expecting exactly and why?
Re: Something strange with my knock module
Posted: Fri Jun 01, 2018 9:49 am
by puff
russian, could you please confirm that
4. So far, Rusefi doesn't have anything in terms of knocking avoidance, does it? No special ignition logic, nothing - just capturing the knock events.
?
Re: Something strange with my knock module
Posted: Fri Jun 01, 2018 11:35 am
by AndreyB
Yes, nothing in terms of knocking avoidance.
Maybe there were some drafts but definitely nothing tested.