9krpmrx8

iTrader: (46)

You should have went Isuzu V-6 bro.

Brettus

iTrader: (2)

/\ nah man .... Saab turbo all the way !

Carbon8

iTrader: (4)

So it was impossible

Hope to see it finished soon, nice to see people still venture outside the box!

Harlan

iTrader: (3)

Removed the cursed engine today and used the time to check manifold/omp clearance. It's a tight fit but should be ok. I think this engine is becoming a lamp, maybe a clock. I tried too much with too little experience and it was a total failure, but I did learn a lot from it, some of which I'm willing to share.

Ceramic coatings work fine on rotors, and can be applied at home as long as you follow the directions.

Don't screw with bearing coatings at home, leave it to the professionals.

You can uncap yellows at home the tip dimensions change slightly and heat shrink is required to keep the rubber grommets from falling off.

Exhaust wrap covered with foil HVAC tape stands up well to heat and makes a great thermal barrier. I originally used fiberglass, but am changing to lava rock because I did see some degradation.

Here are some pictures of the old and new wrap. The new stuff isn't covered with foil tape yet, I wanted to check clearance first.

slash128

Yay another PTP! If you have any questions I'm happy to share my experience. I can tell you you'll def want some shielding between the up pipe in that last pic and the OMP lines/pump. My forward line blistered and melted pretty quick. Ultimately I replaced all the lines with braided SS. I do like how you flipped the OMP connector so it's pointing up, didn't know you could do that. I shielded the whole thing and haven't had any problems, but may give that a try just as an added measure.

ScallopedRotors

Here you go.

Rx-8 OMP Shield

slash128

Thanks but no thanks.

Harlan

iTrader: (3)

Went through your build thread slash, looks great!

Right now I'm working 70 hours a week on nights at work, so I'm getting done what little I can. I have my knock detection/failsafe box hooked up for baseline data. So far it can detect knock, stop the ignition, and detect rpm. The problem is that it's creating some signal noise which causes misfires at high RPM. I think the ecu is miscounting teeth on the trigger wheel at high RPM. Looking for a solution either in code or in wiring. Once this is resolved and my new engine gets back I'm ready to boost/tune.

slash128

Thanks Harlan! 70 hours a week is harsh when you have a project like this waiting at home

I am interested in your knock box. I am battling with timing getting pulled myself. I haven't figured out if it is actual knock or the sound of my open wastegate dump triggering it... Doesn't matter how much fuel I add the timing is always pulled the same.

Brettus

iTrader: (2)

Low rpm or high rpm ?
If it's low , I had the same thing and gave up with the idea that it might be knock ,so I disabled the knock sensor .

Harlan

iTrader: (3)

If it's above 7k then it's not the knock sensor. Above 7k the combustion chamber is too big and the burn time too short to prevent knock, so the ecu ignores it. Also high rpm knock is less dangerous because the short duration prevents hot spot formation, so it doesn't start to snowball cycle to cycle. This is also why coolant and oil temp begins to spike at higher rpms. I may have a trick to boost power and get rid of knock at high rpm, but I'm not willing to share until I know I'm not crazy.

The knock box is my biggest push right now. With it the rest of the build falls into place. I'm still planning on releasing the source/schematics for it. I may sell a kit for it as well, and offer support with the kit. It's too early to think that far ahead, the thing barely works and still has some issues.

Oh and I now have a S2 transmission on it's way to me. So when the engine gets here and the transmission gets here, and I get some time off work, things should get exciting.

slash128

New parts are fun! Would love to see your knock box schematic when you are ready.


How low are you talking? My timing starts getting pulled around 5K-5.5K RPM all the way to 7K where the max knock retard RPM is set and them it jumps right back up.

logalinipoo

I've also been waiting for some details or something to buy.

Just thinking a little. After 7K the ECU ignores the sensor because the chamber is too big and duration is short. Does that mean that the knock can't move across the chamber fast enough to be a problem? or does that just mean the detection system is too slow to correct it in time?

Brettus

iTrader: (2)

no .... mine was 3000 through to 5000 ...... nothing above 5000

Harlan

iTrader: (3)

Neither. Knock breaks down the boundary layer and allows more heat to transfer to (and stay on the surface of) the rotor and housing. At low RPM this can cause either preignition or just faster burning of the mixture. Either way it's like adding timing in an uncontrolled manner. This makes the knock worse, and the cycle snowballs until the knock is bad enough to start breaking things on its own or the preignition has raised housing temperatures and pressure and that breaks things.

At high rpm there is less time for this heat transfer to take place even with a little knock so it's mostly harmless. With a stock tune this is true, with a turbo, well lets just say it's questionable.


This is also why we lose so much power at high rpm in the form of heat to the coolant and oil. Just cruise at 6k and you'll see your coolant temp start to rise rapidly. Minor knock is happening and that is causing more heat to transfer to the housings.

I think I have a solution that will decrease knock at high rpm, but I need my knock box working first because it's... a little out there.With the right split you should be able to push more timing for both NA and turbo, and get better power too. We'll see.

I'll post my current source to the knock box thread, as well as the schematics for it. It's pretty darn simple right now, but it may have to get more complex if I can't resolve the phantom signal.

TeamRX8

iTrader: (25)

Nothing is possible when your basic premise is flawed ...

dannobre

iTrader: (3)

If you tune a rotary anywhere knock like a piston engine you are doomed to fail...listening for knock in a turbo charged rotary is like trying to use birth control weeks after sex....not going to have a good outcome

Harlan

iTrader: (3)

Hate to break it to you, but the stock ECU already hears knock and adjusts the tuning accordingly. It works just fine and engines don't get damaged until you push beyond what the ECU can control. If you disconnect the stock knock sensor (and dummy it out or leave it hanging) then monitor for knock you will find some inaudible knock, and some low level audible. Just driving around normal during the winter with premium gas and a stock tune I sometimes hear knock. Every time you start up the engine knocks, it sounds like a little belt squeak, but it's really knock. And it's natural because the ecu is just blindly putting in timing and fuel until the engine catches.

What we need to prevent is the bad knock. The stuff that causes engine damage quickly. That stuff snowballs fast, but if you catch it early and cut the engine life is fine.

dannobre

iTrader: (3)

A lot of what the stock knock sensor hears isnt knock....and if you log knock retard you will see that. The S2's changed the knock setup from the S1 strategy for a reason. You need to tune for safe timing and then slowly move up slightly as you can. You cant tune with too much timing and then hope the knock sensor will save your engine. It wont...and you will have another rebuild that you didnt need to do....possibly with a lot of broken parts to replace.
The Renesis is different than prior Rotaries....but dont throw out the basics. ..they still apply

slash128

I don't know what's right or wrong but I have logs with timing pulled, up to 5*, and other logs with nothing pulled, same loads, same fuel, etc. I've beat my head against the wall. I'm not the brightest but it seems the sensor is doing something but I get nothing in the logs as far as sensor knock retard, just the timing itself. Would love some insight.

Harlan

iTrader: (3)

The stock knock sensor is looking for a set amplitude per RPM. It does not care about frequency and can get fooled by any number of noises, but that doesn't mean it's worthless. If it is pulling timing on some logs and not on others I would start searching for anything that could be causing knock. Sure it could be spurious, but why take that chance?

As for knock retard and logs, it's been previously discussed, the AP does not log actual knock retard, it logs one variable and tells you almost nothing about what is going on.

https://www.rx8club.com/series-i-eng...-200224/page3/

If you put a 100k ohm resistor and a 10k pf capacitor in series and connect to the knock sensor signal (the yellow and red wire) you can then record what the knock sensor is hearing (or make yourself a simple pair of det cans) and then see knock clearly on a spectrograph. Or you can even filter light knock until the sound is obvious. If you want, send me a recording and I can show you what's knock and what isn't.

dannobre, I agree that you should always slowly advance timing to maintain safety, what I'm saying is that it's easy to completely override the stock knock sensor by advancing timing and never realize that the safety net is gone. I'm not making something that will replace stock knock control, I just want to make a safety net which can stop the engine before things go out of control and require a rebuild.

slash128

Thanks Harlan. Fazda pointed me to that thread as well. I need to spend some time reading.

Harlan

iTrader: (3)

S2 Transmission arrived today. Still waiting on engine. Knock box is still in a semi working state, gonna need to ditch the LCD display in order to get it completely functional, but no huge modifications to code. Think I should be able to have a knock based water injection system too, but no promises.

Harlan

iTrader: (3)

Back on the road. The S2 transmission feels a little clunkier than the S1. I haven't had a single grind, but the motion does not feel as smooth. Probably just need to get used to it.


Still getting knock control to a workable state, I may have a cheap off the shelf option waiting on parts. One way or another I will have a working knock based fail-safe before going into boost.

Engine breaking is progressing slowly. Just driving to work and back. I set up the throttle duty cycle to limit rpm to 5500-6000, but I don't know if it's working as intended because I haven't had rpm up that high! I think I'll set up the throttle duty cycles when I'm done to be both an RPM limit and allow for flat foot shifting/ launch control. Need to find out under what circumstances the ECU uses each table before I can do that.

My OMP wiring modifications are working fine, no CEL/limp mode, so the ECU is reading the sensor. I think the OMP sensor is some kind of rotary encoder and not a potentiometer as I first though.

Oil filter kit is working like a champ with no leaks with stock oil pressure. I did have to give up on the higher oil pressure mod because it blew out filter o-rings when I had the last engine in the car. Here are some pictures of that, I'll take more the next time I have my bumper off wish I had documented it better in the first place.

Also have more maf/IAT info I'll be sharing once I have time to confirm.

Also, a shout out to Rotary Resurrection for building me a great engine!

slash128

Good to see you back at it!