I've read through several threads now, but I didn't find one with my exact problems.

I just did a single turbo swap, it's a streetported engine now running 1000cc primaries/1680 secondaries, VMIC, HKS twin power blah blah blah. I'm not sure how many of these things make a difference but I figured I'd mention them just in case. There is alot of other things too but I figured these were the most important to mention.

When it was still on stock turbos, I had it tuned pretty rich (most of the time it would be around 10 or 11 but would sometimes drop to the 9s), but it ran fine. The knock was pretty high but since I'd never seen anyone else's to compare and I had it reading by the bar instead of by numbers, I didn't know it was too high. Come to find out, all the **** in the back of the knock sensor had melted out so it was REALLY sensitive. The car has been at my tuner's shop (SR Tuning here in Colorado Springs) and he is VERY thorough and won't tune the car with high knock. All the while on stock turbos and with the same sensor on the new turbo, it would read high ALL the time, but almost max out the bar during acceleration.

I just put in a brand new knock sensor from Mazda today, and now it reads around 50 or 60 while idling but it got up to 134 during high rev today. My tuner has talked to Ari Yallon to get some rotary knowledge and Ari strongly suggested against tuning on a bad knock sensor. I've read about it possibly just being engine noise or any number of reasons, but I'm really worried my engine could be toast. Although it has run great since the rebuild in March.

I have one friend that has super low knock even during boost, and another friend that has knock more like mine, when he boosts but not at idle or low throttle.

Any other ideas? Am I really looking at a new engine already or am I just getting worried about nothing?

 

I have been trouble shooting high knock readings as well. I am leaning towards injector issues. Running the same 1680-1000 setup. However I ran this set up before and had very low knock. I have checked the igniter, coil packs, and most other items the manual suggest's. I have even pulled timing but that made no differance. Checked crank sensors, tps, plugs, wires.. VERY tired of this issue. But I agree with Ari that the knock readings need to be correct before any tuning can be done or a new motor will be in needed. Manual says high knock cause by: timing, coil packs, igniter, thermosensor, crank sensor, tps to name a few.

 

Good grief, haha. Well I guess I'm not short on possible causes then haha. My tuner has my manual right now, he's reading up on this tonight. I failed to mention before that we had the guys at Rotary Performance make a map for me with corrections according to all my mods...injector size, map sensor (3 bar btw), exhaust size, turbo, IC, all that. So we're starting off with a pretty specific map, and Frank (my tuner) is going to call Ari tomorrow to ask him about this stuff. Thanks for your input! I'll let Frank know tomorrow if he hasn't already read about it.

Oh, my idle is also hunting again, like it did when I first put the engine in. He said if he dumps in more fuel it smooths out the idle but then it smokes pretty bad.

 

when you are at idle were is the knock?

You are correct, so many things thats can cause hunting as well. From what I have learned is best to start at one possible cause, eliminate it and move on. Sooner or later it will be found. Best of luck and post up when you find the issue.

 

even on a "good" knock sensor it's hard to tune with them, although your numbers are pretty off... I would trust EGT gauges and plug reading more than a stock knock sensor though.

 

At idle it looks to be around 50, but because its hunting its hard to tell.

Frank also suggested an EGT but I don't have one right now. Had I known how useful they are for tuning I would have gotten one. I do have a wideband on datalogit on the car though.

Thanks guys, I'll post up when we figure it out...IF we figure it out HAHA.

If anyone else has any input too, please post.

 

I agree 100% but my point was if the knock sensor was reading correctly and he is seeing 100+ readings something is very wrong. Now that I see him saying he see's 50+ at idle I am guessing a false reading. Exhaust is hitting or you are picking up other noise.

 

Yeah we're thinking false reading too. Here's a fun fact...my tuner swapped my computer with my buddy's and swapped my comlete map over too. Somehow it fixed the idle but the knock stayed about the same.

 

what about key on, is knock zero?

 

I'm not sure what knock is with just the key on, I haven't looked yet.

We discovered tonight that knock is only around 1 or 2 at idle, and even though it still jumps up to around 120 when we rev, if you keep revving higher it drops back down to about 70 or so. But we've moved on from that. I drilled a hole in the manifold tonight and put in a bung for an EGT so we can tune with that and the wideband.

After I got to the tuner's shop, we swapped in another friend's computer and his idled better than mine too. We swapped mine back in and it was hunting again. After I put the EGT bung in, we put the first guy's computer back in and he's going to tune on that. And we put MY computer in HIS car and it's having problems too, but not the same problems I was having. On his, after he comes down from high rev it slowly dies. And his car is idling around 400 now instead of whatever it was on his computer. So I'm not sure what the hell is going on now, haha. I'm thinking about buying another PFC and trying it to see what happens. But for now we're going to tune on my buddy's computer and see if we can get his running right on mine.

And it's also running like it has a stuck seal now. It sounds like a helicopter, and it was running like that yesterday on my computer and on both my friend's computers, so I know it's the engine or something on just my car.

So my question now is what is a good EGT to tune with at boost levels?

 

well I may have a pfc for sale in the next few weeks as I will be swaping to AEM.

I just keep thinking about those 1000cc inj you are running. They are the green and black short fat ones, right? I had a hard time geting mine to seal and not leak. Also the inj wiring. If you have a bad connection (short) that is allowing the inj to fire partial (like 50% of the time) you will obviously have knock and run on one rotor. I know I had this problem when I first fired it up.

 

I might be buying one from a buddy on the East Coast actually. How much are you thinking of asking for yours?

I'm not sure if there is any green on my injectors but there is some black and they are FAT for sure. I had a problem getting them to seal right at first too, but that was because the guy that ripped me off on the fuel system (yeah I'm posting your name now sen2two!) sent me a primary rail for an RX-8. Yeah, he sure did. Not to mention there were pieces missing, one broken, etc. I'll never buy anything from him again, and I suggest no one else does either. He tried to blame the missing parts on USPS...yeah OK.

ANYWAY, once I got the proper RX-7 primary rail, I got them to seal fine. Did you have a problem getting yours to seal because they were bumping against each other? Keith (at KGParts) told me I might have that problem but I didn't, they almost touch though. I also wired in an FJO injector driver, and I made sure to figure out the proper way to wire it in before I did the job too. But I think I'm going to check inj duty tomorrow anyway. I didn't have the "stuck seal" problem until yesterday though, so I'm not sure about that theory. Thanks for your input!

 

not sure on a price as I was going to wait till after the tune on the AEM to make sure all goes fine. As far as your wiring I was thinking more on the lines of the injector clips. I had a short in mine that was causing only one rotor to fire (no fuel to the other) it was happening only part of the time. Just a thought.

 

My car had a knock issue that was fuel related and causing the extra fuel to detonate under initial acceleration.

Never ever ever did it go even close to 100 though, 100+ is freaking high!!!

My symptoms were 1 to 2 at idle and then at initial acceleration the knock would go to 45 - 50 and then once I began to boost the knock would go down to 15-20.

I feel like it was fuel related and that the excess fuel was detonating and causing the false reading... never know though

 

Over rich misfire can also register as knock under boost.

High knock (noise) under no boost is normally caused by engine/drivetrain parts rubbing or hitting other parts. Shake the crap out of your engine and exhaust system listening for such noise.

 

if you swap two PFC's and both have different idle learning procedure calibrations (as they inevitably will), then it is no surprise that their behavior at idle is different, even if you have the same maps.

 

Update-

The car is idling and running GREAT now. As it turns out, it DID have a stuck seal and it popped out today and since then it's been running great. The knock is still high at certain RPMs but we're tuning with wideband and EGT now, and according to those it's not likely to have detonation. So Frank tells me anyway haha. I got in and he told me to run it up but watch the boost...well my dumbass didn't watch boost, I was watching the RPMs to see when the turbo kicks in. So I ended up taking it all the way up to redline and a huge plume of black smoke came out the exhaust just as I was letting off. The engine and exhaust had a ton of carbon buildup so I'm chalking it up to that. Then at idle, it started surging again. The vacuum line from the filler neck to UIM had popped off, and when replaced, it idled fine again, around 800 RPMs. We ran it again and Frank took it up to 16psi and let off. We have the boost controller set to 10psi, so something is not working, the wastegate isn't opening. I think it's because I have a vacuum line going from the top of the wastegate to the UIM near the throttle body, around where the other line was blown off. So obviously if it's the same psi there, shouldn't it be equalizing on each side of the diaphragm in the wastegate therefore not letting it open? I could be wrong, and probably am, but I sure hope not. At this point, that's the only thing standing in our way of finishing the tune.

And when I checked knock today, it was on the setting where it shows the bar, and at idle it was SUPER low, but Frank told me it still gets high throughout the RPMs.

 

10 psi spring pulling 16. I dont think any location should have that much effect. I have heard the debate for ever on were it should be pulled and the comp. cover or even the elbow coming of, (even coupler) is a good choice , but again I do not think it should have that great of an effect. I would make sure the gate is installed correct, vac to what and controller to what? What gate is it? My last Tial woukld pull 2-3 psi over spring rate.. Also found that my psi readings were 1-2 over what 2 differant dyno readings were. So could be PIM settings.... Take it easy until you get it all working correctly. I know its hard but you will be happy in the long run.

 

I didn't really follow what you just wrote there, haha. I don't know the specs on the spring that's in the WG. Here is my vacuum setup...

-top of wastegate to UIM by the throttle body
-side of wastegate to COM on boost controller
-NC of boost controller to IC piping (on turbo side)
-bottom of filler neck to nipple on turbo (as directed by internal comsuktion engine)
-side of filler neck to UIM by throttle body
-nipple on LIM behind secondary fuel rail to FPR
-end of UIM to boost gauge
-end of UIM to boost controller solenoid
-UIM (by throttle body) to charcoal can solenoid
-UIM (by throttle body) to BOV
-oil injectors to small filter (from stock rats nest)

FYI, I'm running an AVC-R boost controller, SARD FPR, HKS wastegate, Tial BOV, Garrett GT3767S DBB turbo, Rotary Extreme monster VMIC

 

Your boost controller is hooked up wrong. See the AVC-R manual pdf page 47, available on the Apex'i website.

Remember that pressure on the top port pushes the wastegate shut (increases boost). Pressure on the side port of the wastegate pushes it open (decreases boost). There are two forces working against each other, and it is up to the solenoid valve and boost controller to determine which one will "win" and set your boost to some level. The solenoid valve regulates how much air (from the pressure source connected to the COM port) goes to the wastegate itself, outputted through either the NC or NO ports on the soldenoid. The "normally closed" NC port would go to the top of an external wastegate. It is "normally closed" because you want pressurized air, regulated through the solenoid, to push down on the top of the wastegate so that it opens later and increases your boost. So the higher the solenoid duty cycle, the more the solenoid valve opens (because it is fully closed at 0% duty cycle), and the more force closing the spring--thus higher boost. In that configuration the side port is not hooked to the boost controller directly, it just needs straight pressure from whatever source you decide to go with so that there is force to push open the spring.

The "normally open" NO port would typically go to an internal wastegate actuator. In that case, the more air the actuator sees, the lower the boost, but no lower than spring pressure. Thus a "normally open" solenoid valve allows 100% of the air to the wastegate actuator at 0% duty cycle. As duty cycle increases, the wastegate actuator sees less air and boost increases.

By using one universal Denso solenoid (it is used on Greddy boost controllers as well) with either a normally open (NO) or normally closed (NC) configuration, 0% duty cycle from the boost controller itself will correlate with the lowest pressure possible on either an internal or external wastegate, and the AVC-R software doesn't need any internal software changes to account for the different ways that internal and external wastegates work.

 

Can you please simplify this for my simple mind? LOL

I'm kinda confused now. So the NC goes to the top of the WG? In the directions in that PDF it says the top of the WG goes to COM on the boost controller valve. And it looks like the NC is supposed to go to the IC piping AND the side of the WG with a T fitting.

 

Actually I was incorrect, I got the first two mixed up, they're actually set up like this...

-top of wastegate to COM on boost controller
-side of wastegate to UIM by the throttle body

However I was also mistaken because the wastegate IS opening, it's still not cutting off at 10psi like it's supposed to though. I removed the hose from the COM port on the boost controller valve (basically removing it from the top of the wastegate), letting the spring control it, and it did cut off around 10psi. So we're just going to tune to 10psi for now and figure out what the problem is in the mean time.

 

yeah that was a long and admittedly technical explanation of how that all works (NC vs NO port), but COM is what goes to the top of the wastegate, not the NC or NO ports, my mistake, just wasn't thinking when I wrote that. Still works that same basic way though.

Try getting NO and the pressure source to the side of the wastegate from the same place (use a Tor Y splitter). A lot of other boost controllers call for that in the instructions. Or swap the pressure sources: side port of wastegate from before the intercooler, NO port from the UIM. I'll skip the technical explanation why.

 

i'm interested in the technical reason why

 

Well I figured out the problem today. Turns out I THOUGHT I had the nipple on the NC port when all along it was on the NO port, so I swapped them and now it's working fine. Unfortunately, this was AFTER we finished tuning it for 10psi and getting it off the dyno. The car is pretty streetable, which surprised me. I thought boost lag was going to be horrible but it's not as sluggish as I thought it was going to be. I was going to tune it for 10psi low boost and 20psi high, but I don't have water injection hooked up yet and I'm not running methanol so I'll just stop at 15psi for high right now. Now I have to go back and make sure his tune will be good for 5psi more. He actually has it running pretty rich still as a safety net, so I imagine it should be fine. Thanks guys!