CA18det tuning - Knock sensor frequency + more questions

[Ken]

Hello guys.
I have a UK 200sx with ca18det (high-port). Now i am studying in Uk but my car is in Cyprus. I am going back to Cyprus in about a month.
Then i can start my tuning. I haven't tuned an ecu before but i know a bit of theory (i am studying Meng also so i understand)
Specs:
-Ecu 23710-44F02 (according to nissan fast by vin no)
-510cc evo 3 injectors
-afpr + walbro 255
-fmic (i will have one in a month- i am trying to get an ETS Evo X intercooler- custom short piping fmic)
-z32 afm
-gt2871r 52 trim 0.64a/r
-stock internals
-tomei adjustable pulleys(i won't install them until i get bigger cams-or can i gain anything??)
-all regular supporting mods

About my ecu management, i have a greddy emanage ultimate that i got free from a friend+map sensor+lc-1 wideband+mac boost valve
No egt sensor neither aftermarket knock sensor
I will be STREET turning on 98 RON and i am looking forward about 300whp-if possible
In Cyprus there is no E85/E10, neither meth or something. There is 100Ron fuel but is too expensive
I know that i don't have Nistune but i am looking forward to buy one(basically i am looking 4 one over UK) I know it's not needed, both have cons and pros. Plus standalone is not needed, even the ap engineering power fc sucks coz u need extra programs and boxes to tune it+not newbie or user friendly.

a)I've search all the forum and i can't find anything about my question.
I know that the stock knock sensor is non-resonant and has a large bandwidth. So basically it pics up every noise. How do i find that specific frequency where the sensor can hear/pick up the knock/detonation? Is the stock ecu tuned to check a specific frequency? What about the nistune knock sensing feature? I am asking because as i said i don't have money for dyno. The only way to street tune is to have the afrs correct, and when u advance timing to watch where the knock sensor pings. Then pull back 2-3 degrees just to be safe. I also read than you can hit the block with a hummer so u can find the frequency? What should i do?

b)About my hp target...Probably it is about 350hp on the engine. The only thing i can is to see how much air is flowing through the maf and estimate how close i am to the hp i want. I think 1.2bar on the gt2871r will "safe"(but probably it won't flow enough air for my target-at least it's more efficient than the smaller turbos which translates to lower intake temps and less chance of detonation). I am expecting to blow a headgasket mostly, but i don't want want to blow up the pistons:S I can keep the egt's down by injecting more fuel, but even so the pistons have a maximum force they can take. A guy made 380whp with stock internals but with double stock headgasket(decreased the CR) From what i can see in nistune my ecu map has inactive knock map(same as the fuel map) Though the JDM map has leaner afr's and richer knock map(compared to the EDM i have) From this i can understand that probably the EDM map was "designed" that way to run a lower octane fuel. The nice thing also about the EDM map(at least on my car) has no stock afr meter so i runs only open-loop. The EDM(23710-44F02) map i think is already rich enough. Should i rich them more?

300whp are enough for me now. I want to just have a little reliable fun for first time with my car. In the past, i got the car with a 91-92' turbo from a skyline r32 or r31(i don't remember-the one that had t3 turbine housing and t25 compressor housing) After 3 months(only running 0.55bar) i was launching the car, and as i was changing to second gear the turbine wheel was blown and got out of the exhaust(it was broken from the shaft!!-i realized it after i got out the turbo-since it was smoking after it blew up) Then i was running NA for a couple of months(it sucks with 8.5 CP-nooo powwer;p) and then got a used 99' s15 GT28R. After 3-4months it started smoking, then one day WOT at 5th gear the oil seals were completely blown(i did that coz it's not rebuildable) back to NA again, from last September to UK uni, and in June i am going back. What i think i found out the cause is that the GT28R on s15 had the oil restrictor on the oil line, where the GT28R supplied directly from garrett have already have it built-in:s Not much people know about this difference...

So any help before-hand is appreciated as i have to study for my exams right now...

Cheers, Ken

[PL]

Hey Ken,

Sounds like you've already had quite a bit of fun with your CA!

Not sure on the knock frequency of the CA18. There is a formula out there somewhere based on bore size. There would certainly be a band pass filter on the input to the ECU but I've never bothered reverse engineering it.

52t 2871's are usually giving around 220rwkw on 18psi on SR20's without cams (sorry, I do mainly SR20's). Make of that what you will. I guess on a CA at 1.2bar (17psi) you may be struggling to get to your 300hp (223rwkw). But it shouldn't be too far off - cams could get you over the line. My figures are on Dyno Dynamics dynos too - which are less optimistic than Dynojets.

As for detonation detection. Well, you've probably read my spiels on that if you've been through previous posts. I have very little faith in the Nissan system. It may have been worth something when the engines were new and in std trim. But once they're old and rattly with big turbos bolted onto them I think all bets are off. I've had such random results with the factory system that I have no problem turning it off.

Add to that the fact that it's only active for a small portion of the rev range then it appears that even Nissan had little faith in it - I suspect they had a lot of trouble telling real detonation from mechanical noise even during development.

What you say is correct for tuning - if you're doing it on the road then once your AFR's are good, then add timing until it knocks and then back it off a bit. It's a rough way of tuning but it'll usually get you close. Just listen for knock with your ears (and a garden hose or microphone/headphones or whatever your favourite method is). Easy to hear knock on CA18. Nice quiet engines.

Be very careful tuning full boost on the road. I found that things get dangerous very quickly with cars making more than 200rwkw. Not recommended.

PL

[Ken]

Hey Ken,

Sounds like you've already had quite a bit of fun with your CA!

Not sure on the knock frequency of the CA18. There is a formula out there somewhere based on bore size. There would certainly be a band pass filter on the input to the ECU but I've never bothered reverse engineering it.

52t 2871's are usually giving around 220rwkw on 18psi on SR20's without cams (sorry, I do mainly SR20's). Make of that what you will. I guess on a CA at 1.2bar (17psi) you may be struggling to get to your 300hp (223rwkw). But it shouldn't be too far off - cams could get you over the line. My figures are on Dyno Dynamics dynos too - which are less optimistic than Dynojets.

As for detonation detection. Well, you've probably read my spiels on that if you've been through previous posts. I have very little faith in the Nissan system. It may have been worth something when the engines were new and in std trim. But once they're old and rattly with big turbos bolted onto them I think all bets are off. I've had such random results with the factory system that I have no problem turning it off.

Add to that the fact that it's only active for a small portion of the rev range then it appears that even Nissan had little faith in it - I suspect they had a lot of trouble telling real detonation from mechanical noise even during development.

What you say is correct for tuning - if you're doing it on the road then once your AFR's are good, then add timing until it knocks and then back it off a bit. It's a rough way of tuning but it'll usually get you close. Just listen for knock with your ears (and a garden hose or microphone/headphones or whatever your favourite method is). Easy to hear knock on CA18. Nice quiet engines.

Be very careful tuning full boost on the road. I found that things get dangerous very quickly with cars making more than 200rwkw. Not recommended.

PL

Thanks for the response. I had quite some problems with the car;p Basically as u increase the power, something gets out of spec and breaks;p then next is the second weakest in the chain and it goes on;p Even with 19 years and 126000 miles on it it still runs quite good. My father wants me to sell it but i don't think so. You can't get any more fun if you have a low budget and you are still a student(though i work a couple of months in summer)

Since i do Meng, even i am in 1st year i've learned so much. E-manage lets me chose a specific frequency to listen on. Though i've got to find that frequency. Since the engine is not knocking(so far;p) i don't know how that "ping" hears. Though what i can do, is to advance the timing in idle, or lean out until it knocks, and record it so i can find the frequency. Next i can calculate some values and build a low+high pass filter to allow me hear only that range of sounds, and just with isolated headphones hopefully i can hear it myself(instead relying on the Ecu's) I think someone did a similar thing before...i think i'll try it.

CA's run quieter than sr's since they don't have the stupid rockets. If i had sr i would put the sr16ve head on it:p I guess i am fine with 280rwhp and an engine that is running;p Every engine is different because of the fatigue, that's why an iron block is stronger than aluminum one. Aluminium loses strength with every cycle until it breaks. The iron has a fatique limit-a minimum stiffness value for an infinitive value of cycles. Do what are the stock ca18det pistons are made of? Probably aluminum...Also intake air temperature affects how much timing i can advance, but here the problem is more towards the piston strength. Even with no detonation/knock they have their limit..Which means that even if i am not running on the limit, they pistons will lose strength over time(assuming they are aluminum) and then bye bye;p

I can get cams on my hands though i think i can run only 8.8mm lift with the standard springs and retainers so i don't expect much of gains. At least if i find the money i can dyno it, and there i can advance timing until no more torque is produced instead of risking it with detonation. Or i guess i'll save up until i get forged pistons and rods..
Any more advices on street tuning? Should i put a buddy to drive or look the realtime data for knock?;p

[PL]

You can hear knock quite clearly with a microphone under the bonnet connected to a little amp and then fed into some good quality headphones. You don't really need a filter - although it may help if you can get the frequency right.

You usually won't break your pistons unless you get detonation. My mate has been driving his SR hard for many years now and the pistons are still going strong. Now tuned for 265rwkw running 56t 2871 on 18psi. Std pistons.

I think knock is harder to pick on an SR20 because of the ally block. It just seems to ring more. Once you get above about 16psi they just sound really harsh in the headphones. As well as all the rattling and crap from timing chain and rockers!

*Definitely* get a friend to drive. Actually I usually do the driving - and get a friend to tune for me.

PL

[Ken]

Thanks for the tip;p It's not only the pistons, the piston rings might break, even the headgasket...At least i don't have to worry about the valves;pp If ca's pistons were the same as the sr's i wouldn't worry! The nice thing is that with the emanage i can connect some unused wires to drive warning lamps for knock, or high temp, so even after the tune if for any reason i have a problem i would identify it immediately and save the engine! I've seen people with diesel engines blowing a hole through through the pistons coz the truck was overheated;p Imagine running higher boost:pp

As for the headphones method, do you connected them directly to the knock sensor, or do you attach a mic somewhere in the engine??? And how knocking sound hears like??? Over almost the 2 years i have the car i've learned to identify every sound!!

[PL]

I actually mount the microphone in an ally tube and then wrap the tube with something soft (a rag will do) to insulate it from all engine parts. Then stuff it under the inlet manifold pointing towards the engine block.

I've found that this works A LOT better than connecting a mic directly to the engine.

I'm surprised you've had a CA that long and not found out what knock sounds like! It's a random metallic rattling. Imagine you had a jar made of 5mm thick aluminium and you put 2 8mm dia ball bearings in it and sealed the lid. Give it a good shake and you'd have a sound pretty close to detonation.

Alternatively fill your IGN map with 35 degrees in every cell and go for a drive. You'll learn very fast what detonation sounds like, although you may also learn about engine rebuilding! On second thoughts don't do this...

PL

[Ken]

Haha;p Thanks. My engine was running always a bit richer than stock(a bit higher pressure on the afpr) so i don't really know how it sounds. I only maxed out the sto afm so not over 200hp so far...that why i said i didn't have so much fun. Anyway i'll try to find more info and see what i'll make out of it. Any furthermore tips or answers to what i've asked so far are welcome. Until then i'll have to wait till June!

[Zenki]

I'm using the Phormula KS-4 on my CA, it works really fine. It hears knock before you can hear it with your ear, thats my experience. You can also import the signal @ nistune

If you have a small budget, I would go for an CA16 inlet cam

[Ken]

I'm using the Phormula KS-4 on my CA, it works really fine. It hears knock before you can hear it with your ear, thats my experience. You can also import the signal @ nistune

If you have a small budget, I would go for an CA16 inlet cam

I have everything i need except the intercooler so far. I want to go with a decent core but USPS sucks so they have to send it by UPS which makes it to expensive. That's why i've left it last. Even so i don't have exactly such a budget now...About cams i already have Tomei adjustable gears and i can get some used tomei cams at decent price if they still are available(from a guy in Cyprus) Even so i think forged pistons would be the wiser next choice. But then you need new hg, probably u will change bearings, head bolts, etc...it gets too expensive. I don't want to mess with internals since it's not in my budget nor time(i just want to have some fun this summer, and then i'll be forced to sale it if i find someone willing to get it at least the price i paid in parts)

Since i can't go with a KS-4 or similar here's my plan:
1) Make a det-can
2) Configure e-manage to detect knock at 6.9Khz
3) Feed the knock signal to/and inline amplifier to the iphone 4, for recording capabilities and also i have a spectrogram app which help visualize the knock also(not so good in reality but may help)

The friend i'll go with tuning with, is now studying/working as a car mechanic and i remember back in school in the music class he could detect the notes very accurately (the teacher played a specific note in the piano and we had to detect what note is it) So i think i'll give him the det can, and stay with the electronic equipment. In the end we had 1 computer, and 2 humans trying to detect knock via electronic and mechanical input. The ems as accurate as it is, and humans can get better. It will be the first time doing this but how bad can it go with a Meng and a car mechanic??;ppp

Tell me what u think;p

[Zenki]

You can also use an ks3 without tracing

In the first tests I was surprised how early the engine begins to knock before you can it hear with your ear

My first knock detection was a tube from the engine bay into the interior....it also worked, but it wasn't the right way for me to go

[Ken]

You can also use an ks3 without tracing

In the first tests I was surprised how early the engine begins to knock before you can it hear with your ear

My first knock detection was a tube from the engine bay into the interior....it also worked, but it wasn't the right way for me to go

I saw the demonstration of ks-3/4 in their website and it peaks noise also. Knock doesn't happen in a specific frequency. It has a small range. Even so when the car is running it makes so much different noises that's why is so hard to know if you have knock when you may not. At least with both human and electronic detection it raises the chances of detecting it. Even so i don't have any budget to go with more expensive equipment. About pre-ignition which can't be detected by sound, it can be detected easily on dyno. And dyno is the proper thing to do. And i will not advance a lot of degrees. I'll play it safe. At least i not like some stupid mechanics which were running the ca18det in a t25 without attaching the wastegate actuator and blown up the engine. The funny thing is then they say ca18det sucks and sr20det is better. bull*hit;p

[Ken]

Hello. My intercooler is finally on order...i have some questions. I have everything up and running, using the stock ic, running max 5-7psi(depends on ambient temp) on the manifold. From the hotpipe i've got 0.7bar! Or i have a nice leak, or the stock ic is really blocked from the inside. For now i don't care since i will check for leaks after i install the new intercooler + new pipes.

Since i have 44f02 ecu which runs open loop from factory(no stock afr sensor) i got the jdm fuel map translated the afr values from there to the new tablel(z32 afm voltage vs rpms) and auto-afr-tuned. I did that coz when i've changed to the 510cc injectors i ran rich at idle and really lean somewhere between 3-4k rpms.
Emanage reads the knock sensor at 6.73khz and converts the z32 afm voltage to the ca18det one and feeds it on the stock ecu. Normally the output afm at this turbo pressure was getting about 4.93v. This night while studying the logs i saw output voltage 5.04v @ 6822rpm! wtf? Where is the stock fuel cut??? i haven't clamped the afm voltage and injector duty cycle from the stock is ecu is 42.8% This was in 2nd gear 99km/h Afr was 12.1 My timing map needs correction coz i have lot of knocking with the timing map i used this night. I don't know how much it affected having the silencer on.
In addition my compression testing was not very ok. I've used a draper comp tester that u hold onto the spark hole..i got ~9.3(don't remember the unit) but in the fsm acceptable was between 10-12. I poured oil in, throttle body open, warmed engine, but it didn't have any significant change in values from previous measurements i made..Probably piston rings and valve sits..it has over 135k miles(is it normal?)

Timing was set to 15bttc but driving it was very retarded, so i advanced it more through emanage(though it needs more retard at wot) I also think that i am one tooth off in the exhaust cam from the previous timing belt change..that's 2deg on the cam i think..does it play a huge difference?

I can post maps, logs, anything to help me learn a bit of basic tuning! I still don't understand why the stock ecu didn't fuel cut with over 5v...what's is the stock fuel cut afm voltage?

Thanks, Ken

[Matt]

The fuel cut is TP load based (not AFM voltage). If you have increased the TP load beyond what your TP is reaching in the log then it wont do the cut

[Ken]

Thanks for the response. I've read the nistune tuning manual, and i understand what is tp load, but since i'm using a piggyback and not nistune(since i didn't had that chance yet) i have no modifications to the ecu. Even so, the stock ecu will fuel cut when afm voltage or injector duty maxs out, or when tp load only maxes out? I understand that in reality the ca18det afm probably would never reach more than 5v(if it has such restriction in it's circuit) and the stock ecu will read the last cell in the fuel map, but even so where is the fuel cut protection? I'm confused now...what i think now is that may the previous owner chipped the ecu and eliminated the fuel cut:/

update: Another question.
The eur ca18det is running 15bttc at idle and the jdm one 10bttc. In the timing maps in nistune is 21 bttc for the eur one and 23bttc for the jdm one. I've read somewhere in the tuning manual that you use the offset to read real values. This means if the motor is running 15bttc at idle and the map is 21 bbtc this means the whole timing map is +6deg right?

Cheers, Ken

[UNISA JECS]

I'm using the Phormula KS-4 on my CA, it works really fine. It hears knock before you can hear it with your ear, thats my experience. You can also import the signal @ nistune

If you have a small budget, I would go for an CA16 inlet cam

Are you actually logging the phormula knock sensor into Nistune and if so how and what type of signal type is it logging?