Supercharger Thread 90 Series

[baders Junior Member Find all posts View Profile Close baders]

OK, lets do it. Supercharged Prados seem to generate a lot of traffic in the various threads. Here's a place to discuss them exclusively. Please keep it about supercharger or directly related issues. I'll kick it off.

From Glen's build up thread here; http://www.pradopoint.com/showthread...539#post478539

Glen wrote;

Detonation is always a risk and there are some options for reducing them, cooler spark plugs, high octane fuel, lower engine compression via thicker head gasket, meth injection, (de)tuning, probably much more. What I'd love to do if I could find the time, is hack into the vehicles wires to access the knock sensor output and see if I can make it trigger a circuit to illuminate a light up front. That way we could see every time a ping is suspected. Then purchase an exhaust air/fuel monitor and have it's display up front as well. Plus connect the laptop to the xede semi-permanently. As we drive we'll be able to monitor whats happening in real time, and can make changes slowly over time. This couldn't replace a full dyno tune, but by using real world scenarios we can improve the tune.

I think that is a great idea. However, the knock sensors are an input to the ECU, therefore could be many different types of signal. Better to use the output which would be spark retardation ? Theorizing here.

Come to think of it, why doesn't the 5VZFE knock sensor system retard spark enough to eliminate detonation ? Perhaps it is out of the operating range of the devices used as standard ?

[glen_ep Addicted PP Member Find all posts View Profile Close glen_ep]

Being inputs shouldn't be an issue, so long as the signal can continue to the ECU uninterrupted. You might know from the Xede install that we already cut into a couple ECU inputs (MAF and Crack Angle sensor) in order to route them via the Xede and manipulate the signals. This is how the Xede "adjusts" timing and the air/fuel mix - by tricking the ECU with manipulated signals.

Tapping into a spark retardation output is also a nice idea. If we could turn that into a display showing the actual timing (not just Xede's adjustments), then we'd be able to see how close we get to the limits of timing advance/retard. Actually, it'd probably change so dramatically and quickly that we may not be able to attain any use from it - because the ECU constantly adjusts timing on the fly according to many factors, such as RPM, speed, gear, knock history, cruise control, coolant temp and more.

"why doesn't the 5VZFE knock sensor system retard spark enough to eliminate detonation?".

Firstly, detonation/ping is the ignition of fuel before the spark ignites it. Solutions are primarily focused around reducing air intake & engine temps.

I would say there's 3 basic causes for this with our engine:

1) The engine will always have an occasional ping because the ECU cycles through a process of (a) advancing the timing until a ping occurs (according to knock senors), (b) retard timing until ping stops, (c) repeat. This is how more power (timing) can be achieved with higher octane fuels, and how the engine can be protected from ping when other parts of the equation are not ideal (high engine temps - explained next). Knock is not the only controller for timing - others were explained above.

2) There's also a limit to how far the timing can be adjusted - a limit at both ends. So in a worst case scenario the retard limit may be reached, but the ping may continue. It's most likely going to occur with forced induction, with low octane fuel, in middle of summer, running air conditioning, carrying a heavy load, under low RPM but accelerating. Factors which will make the situation worse is a lean fuel mix, poorly functioning cooling system and unsuitable spark plugs, etc.

3) Sensor/electrical failure - the ECU does not know ping is occurring so advances timing to near max levels. *just a theory of mine*

Compare your engine performance between hot days and cold nights. It'll perform noticeably worse on hot days due to the warm air (less air compression meaning less air volume, less fuel, less fire) and a higher engine temp (reduced timing). This affects forced induction engines significantly more than NA.

Here's a very informative document about how the ECU controls the timing:
http://www.autoshop101.com/forms/h40.pdf

[baders Junior Member Find all posts View Profile Close baders]

Thanks Glen that's great info. I suspect point 2 is the case. The pinging I am experiencing is outside the system capability to adjust to. Come to think of it, even in NA form, my Prado has required 98 octane for many years. This could indicate a base tuning problem that I have never really addressed. It's booked in for hub dyno on Tuesday so the truth will be revealed I expect.

I see in the Xede Tuning instructions the following;

Smart Knock and Smart Fuel Control
An exciting new development in interceptor timing and fuel control is currently being tested. The process is described as smart because it uses the nested Map function of the XEDE and feedback control from the knock sensor forignition and an externally added wide band lambda sensor for fuel control to generate a closed loop fuel and timing trim. This smart function allows the tuner to set the aim points for fuel and timing and use the advanced processing power of the XEDE to control the fuel and timing.

I wonder if they have progressed with this function ?

I have ordered an Aten USB to 9 pin converter so that I can hook up my lappy to take a look. It looks like the firmware on my Xede requires Xmap version 3 ?

[glen_ep Addicted PP Member Find all posts View Profile Close glen_ep]

Thanks Glen that's great info. I suspect point 2 is the case. The pinging I am experiencing is outside the system capability to adjust to. Come to think of it, even in NA form, my Prado has required 98 octane for many years. This could indicate a base tuning problem that I have never really addressed. It's booked in for hub dyno on Tuesday so the truth will be revealed I expect.

I see in the Xede Tuning instructions the following;



I wonder if they have progressed with this function ?

I have ordered an Aten USB to 9 pin converter so that I can hook up my lappy to take a look. It looks like the firmware on my Xede requires Xmap version 3 ?

I'm sure you're looking forward to the dyno tomorrow so I'm sorry to be the bearer of this news... Dyno's burn time and a lot of money so I'd suggest two things for you:

1) get the USB converter first and test it, make sure you can connect/download the maps and monitor the engine. Then bring what you've used to do this with you on the dyno day. Reason is, if they have troubles connecting with their gear, the job is over. This happened to me - lucky I had the gear with me.

2) NA 3.4L shouldn't have required 98 for years. There must have been a problem and I wouldn't want to waste dyno time working out which sensor is malfunctioning. Yes they could probably still tune it for 98 and get by, but if you want them to do a quick tune for 91 or E10 as well, better have a go at finding the original problem first.

BTW, what sparks are you using?

Also, before dyno have a look through the engine bay and make sure nothing is near your extractors - when RPM's are held above 4,000 for minutes at a time, they will get glowing red!!

[mjrandom Out of control poster! Find all posts View Profile Close mjrandom]

It doesn't make sense that your standard engine pings. Other suspects are carbon build up on the pistons that stays hot and pre ignites the fuel. Sparkplugs wrong heat range working like glowplugs. How many km on the engine?

[baders Junior Member Find all posts View Profile Close baders]

The engine only pinged using 91. Using 95/98 was fine (when it was NA). It now has 190K on the clock.

They will perform a full diagnostic tomorrow including AFR. I hope this will show any problems. Plugs are good quality NGKs as I recall. Unfortunately the converter won't be here till next week but they assure me they are the Xede experts of Perth

There are a few other guys that have gone the SC route, would love to hear your experiences too (if you are still around).

[Leethal Advanced Member Find all posts View Profile Close Leethal]

As we know the 3.4 l engine has an issue with detonation in the 2000-3000 rpm window when blown, if my memory is correct the knock sensor circuit does not function below 3000 rpm which highlights it. When I get some time I will see if I can locate the information source where I read about the system functionality.
My own experience has been I can create detonation right up to 3000 rpm and once above the 3000 rpm point it stops, or more likely is "controlled" by the engine ECU, I cannot get my engine to detonate above 3000 rpm.
I believe the Exide system tuning uses a MAP sensor to monitor boost, and then enrichens the fuel mixture when on boost through the problem rpm window.
The TRD add on ECU retards timing in the 2-3000 rpm window and drives a seventh injector as well to control detonation.
The retarded timing below 3000 rpm is very noticable with the TRD unit, response is noticabley "doughy" to use a high tech term.

Lee

[glen_ep Addicted PP Member Find all posts View Profile Close glen_ep]

Hey Lee,

I've read conflicting information about the knock sensor functioning - some say only above 3000, some say only below 3000. Would be great to find out this one for certain - it'd really have a big impact on tuning. If knock is not detected below 3000 (which is the critical range) then tuning would involve removing a lot of timing in that range, just to be on the safe side. If the opposite is the case, we would tune for power in that range and rely on the knock sensors & ECU to drop timing if conditions called for it.

Yes timing makes a big difference to performance - when I switch from 91 to 98 tunes (advanced timing) the vehicle suddenly feels so much lighter!!

[Leethal Advanced Member Find all posts View Profile Close Leethal]

Hi Glen,

Yes there is conflicting information out there, however I believe the knock sensor operates above 3000 because all the "fixes" revolve around an issue below 3000 rpm. If the knock sensor was not functioning above 3000 I believe we would be seeing detonation from 3000 rpm as well.
The TRD ECU add on unit I have is very noticable in where it reducs timing, with an obvious increase in power at 3000 plus rpm, almost like a cam coming on song.

I would expect that on a dyno with a good tuner one would see the timing changing below 3000 if it was working, I will contact the people who tuned my Prado to see if they have any information they wish to share.

Lee

[glen_ep Addicted PP Member Find all posts View Profile Close glen_ep]

Whilst I believe there's insufficient documented facts to confirm the knock detection range, I do agree with your logic - the fact that issues frequently occur during low RPM indicates that the ECU isn't adjusting enough in that range, so the knock mustn't be detected there.

Here's the document which claims the detection above 3000 only: http://www.vaglinks.com/OBDII/Gadget...com_U-Tune.pdf

BTW when you say you have a TRD add on to control timing and 7th injector - do you mean the URD kit?

[baders Junior Member Find all posts View Profile Close baders]

Car is in the shop as we speak. Hope to have it back this arvo but may be tomorrow. This is all good info chaps. I will put these questions to the "guru" this afternoon, see if he has any answers. The guys specialises in Xede, Halbech and Motec.

To my mind, the most likely rev range for detonation is below 3000rpm (under load). As per Glens question Lee, what is the "TRD ECU add on" ? I think Glen and I only have the Xede/Chiptoque interceptor ECU, running the 2 bar MAP sensor.

[baders Junior Member Find all posts View Profile Close baders]

Interesting document that Glen. There is "detonation" (bad bad), then there is "spark knock" (much less severe and tunable). Definitely states there is no timing alteration below 3000rpm, but does it refer to the Toyota ECU or URD ?.

[glen_ep Addicted PP Member Find all posts View Profile Close glen_ep]

Interesting document that Glen. There is "detonation" (bad bad), then there is "spark knock" (much less severe and tunable). Definitely states there is no timing alteration below 3000rpm, but does it refer to the Toyota ECU or URD ?.

According to that doc: there is definitely alteration of timing below 3000 RPM, via the Toyota ECU. What doesn't occur below 3000 is the utilisation of live readings from the knock sensors. Below 3000 the ECU uses "an extrapolation of the high RPM ignition timing" - in other words - it uses the historic data from above 3000 to mathematically estimate the adjustments for below 3000.

I'd assume the estimation they're talking about only occurs in regards to knock risk. Below 3000 the ECU would most likely still use live data from other sources to further adjust timing (current gear, if changing gear, cruise control setting, engine temp, MAF and of course RPM).

[baders Junior Member Find all posts View Profile Close baders]

OK, the results are in. Will post dyno run image tomorrow. There was indeed a base tuning error. There were whole columns missing from the Xmap tuning matrices (meaning huge steps in values) and the AFR was leaning way out mid range. Not sure how this came about, as it was supposed to be downloaded with the "3.4 Hilux" map from Chiptorque. Anyways, ping gone and way more drivable.

[baders Junior Member Find all posts View Profile Close baders]

Oh yeah, the fun bit.

147Kw @ 4700rpm.
366 Nm @ 2500rpm.

Remember this is from a hub dyno.