AEM AEM EMS SERIES 1 knock

[myturbo]

Alright I'm needing some help/info on setting up knock calibration on series 1.

Car: 1g Auto awd,2.0 stock block, supertech dualspings,S2 cams,frh intake manifold, hx35, fmic, full 3" exhaust etc. Aem 3.5bar map and iat. Dw 1200cc inj, walbro 450, base for 43.5 ETC sensors and base timing all set. No leaks, good compression etc

I have been told different things by different people so..

From what I gathered, Safe tune, do a few pulls logging knock volts (?) The peaks is my Cal for that specific rpm x1.25.

How do I know if that is my normal noise voltage or Knock ?

Should I be looking at knock raw ?

Right now my noise sits around .5-.6v with a few small spots Spike to 1-1.2volts. this is on 10 psi or less 5° timing and afrs in low 10s ( I have added & removed timing & boost 8-16 psi without change)

If someone could help me and or correct me it would be very appreciated!

 

[DSMPT]

this is on 10 psi or less 5° timing and afrs in low 10s

If like you wrote, then probably the spike is not knock.
Maybe there is some difference between series 1 and series 2. but normally 1.0-1.2v is just catching some noise from vibration or some gravel hit the oil pan or so. I use series 2 with 2G/EVO1-3 knock sensor. AEM said lower than 1.5v is not considered as knock. I don't pay attention if it is lower than 1.5v. I also use Plex knock monitor and the Plex knock monitor is not detecting any knock if the voltage is lower than 1.5v, actually even around 2.0v neither.

 

[silvreclips]

I can't comment on setting up the software, but it may be helpful for you to make a pair of detcans. While your setting up your aem you can also use the detcans to verify if you encountered knock or not.

 

[Mello]

The way I did my knock calibration was by logging knock volts, doing a few easy, low load, full rpm range pulls on a conservation safe tune with a conservative ignition timing table to baseline the normal engine noise. After, I opened the pull in AEMdata & used the histogram with knock volt vs rpm to see the engine's nature noise volts output at the different rpms & set the knock threshold values about 15%-25% higher; a little beyond the normal noise level. When you knock the spikes are really high & will be over the threshold.

 

[myturbo]

The way I did my knock calibration was by logging knock volts, doing a few easy, low load, full rpm range pulls on a conservation safe tune with a conservative ignition timing table to baseline the normal engine noise. After, I opened the pull in AEMdata & used the histogram with knock volt vs rpm to see the engine's nature noise volts output at the different rpms & set the knock threshold values about 15%-25% higher; a little beyond the normal noise level. When you knock the spikes are really high & will be over the threshold.

Should I be wot ? Or lower throttle ?

 

[Mello]

No WOT needed for data logging your engines normal niose. I did 3 or so gradual pulls on a flat road from around 1500 to redline with only enough throttle to get to redline. Than I ran the log through AEMdata (or what ever the logging software is called) & used the histgram anaysis tool to get all the volt values to line up to the same rpms over the rpm range. You're trying to get a normal engine noise threaholdso you can input, at the different rpms, what is normal & to ignore from "this is to much volts" & is knock. Your graph for the volts should get higher (more) as rpms increase. 4g63's are not a very noisy motor in general and I think my threashold volts was around 2 volts at redline (maybe 2.25v) sloping up from like .2 volts at idle. My curve is not straigth & around 2500 rpms noise jumps up & then sort of flatness out to redline.

 

[myturbo]

I don't see what the histogram does? Just shows my high voltage

 

[myturbo]

If I did this correctly these are my values. I did add some timing (7-9° total while WOT) in and get afrs in high 10-11.1s.

I just went through 4 2nd gear pulls (auto) took the avarage at each rpm point in table highest volt* X1.25,from each log and set that on my Knock Cal tables.

Sounds good ?

 

[Mello]

Sorry, use a scatter plot not histogram. You have to zoom in on the plot to scale the axes. Your knock sensor graph looks believable. Attached is a scatter plot I made as an example but I wasn't logging for setting up knock. I was logging for some other reason but it had the knock volts channel. This is what my knock sensor look like also.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[myturbo]

I should have been using the highest peaks in logs correct ?

 

[Mello]

Yes you want your threshold above the normal noise level. I'll double check after work but I think AEM suggests 25% more (knock volts * 1.25).

 

[myturbo]

Alright I believe that's figured out then, now I'm working on idle.

When in the idle tab say I have target at 1200 across the board. When I adjust the idle% vs target table. I'm never adjusting at 1200rpm itas always above my target. Say target 1200 I may need to adjust in the 2500 on %vsTarget.

Idle in gear actually needed adusting around 70% below 1500 to get it to stay at 850-900 , which is very vibrating. Probably due to cams and solid mounts.

I would think this is wrong somehow and I would be adjusting at 1200. Reading aem instructions they are very vague

Idle learned value is always 0 , no matter how much I adjust it up or down

 

[Mello]

A leaned idle always 0 would indicate the EMS isn't entering the idle mode. Is yoiur car running well enough to get to operating temp? At temp will is run without you having to help it? What rpm is it trying to run at when at temp? What is the value for the option "Idle FB Off Below RPM"? What is your throttle percent & what value is showing for "Idle off If TP over"? Did you do the throttle setup wizard to calibrate the TPS?

 

[myturbo]

This should make it easier

 

[myturbo]

Advanced settings i sorta copied from an evo in hopes it was a solid fix

 

[Mello]

I need you to make some changes. Go to the advanced idle option-Advanced idle. "Check" <Idle FB Dir Invert>. Change <Idle FB Minimum> to as close to -20% you can. It should end up with a true value of -19.92 i think. Do the same to <Idle FB max> except positive (+20%). Now change the <Idle FB off Below> to 1800rpms. Redo your <Idle% vs RPM> graph. Make it 15% @ 500rpms & 50% @ 3000rpms, then hold the shift fey while using the arrow keys and highlight every value inbetween, press ctrl + M, press ctrl + L, so the software will calculate everything between 500 & 3000rpms in a straight line. Save & load these changes & see if you get idle learned to respond when you're at temp. Most of these changes are the base values for the AEM EMS 2G startup cal. This is for a 2G correct? Your avatar has a 1G but your "CAR" profile is 2G auto???? Stop coping stuff unless you know what it does & why.

 

[myturbo]

It is 1g auto , 2g was like 8yrs ago , I tried to edit my profile, but it won't save details. I would think what you said should be pretty close to what I would do for a 1g aswell . So doing this is the learned value has a feedback, I should adjust the idle% vs target then ?

 

[Mello]

Those changes should work fine on a 1G. Learned idle is the output showing the difference between your settings & what the EMS needs to correct to The stepper motor table to make engine idle at the target rpm. We are trying to make the EMS control idle & always 0 indicates the EMS is not doing anything, most likely because it doesn't know it's trying to idle. We want learned idle to be changing when we are idling. Well deal with adjusting idle once the EMS knows it's idling.

 

[myturbo]

Ok uploading now to test, by the way it looks the factory idle throlle stop isnt even used & is just rpm monitored for idle?

 

[myturbo]

Entering that , if i adjust idle setting to get a stable idle 500-3000. i always get -20 or +20.

when i get idle all set and good, it will randomly have high idle. if i rev it rpms will hover around 1700 unluess i adjust the %, Which in turn messes it all up again .

Now i know you didnt tell me to adjust these settings, but i figured i should see what happens. would it be more helfull if i uploaded my tune ?

 

[myturbo]

anyone think this is real knock ? never had it spike like this. But due to afr/ign/boost im in doubt

 

[Mello]

Is this a log from logging to the PC or internal from the EMS? It's hard to tell but something made EMS activate knock retard.

 

[myturbo]

Internal log, I don't mess with the quick pc,logging. Doesn't seem accurate enough

Just weird the knock voltage Spike at these variables. Maybe up top I didn't do a good job on knock cal ?

 

[Mello]

Possible. Two volts at high rpm's is little on the quite side but every motor is different. You are correct on internal logging for knock cal. PC logging will not have a fast enough sample rate to capture engine noise or knock events well.

 

[myturbo]

Well I suppose I'll log it again and see how it looks. If it still looks high, I'll raise my knock cal around that rpm and up. Then see what another pull looks like.

For the idle what learned value am I shooting for ? . It makes out and get unstable