Pull boost or pull timing.

[Dell'Orto]

After the BHG repairs i took it to the shop yesterday and had it all re checked and tweaked against the ECU, Dizzy to match up to 19.5 Degrees.

Did they explain why it had failed?

[method]

Might try repositioning the knock sensor.

?

That's got to be the dumbest idea imaginable.


What CR are you running.?

Generally a high CR (N/A) requires a small ammount of base timing, a lower CR requires more although the overall timing is about the same.

On a car with a distributor, there's base timing (static), maybe between 5 and 10 deg depending on motor.
Then there's centrifugal advance that might start at 1000rpm and be "all in " by 3000.(in the order of 25 deg max depending on motor)
Then there's vacuum advance which is dependant on load. Maybe 10 deg max.

So at full load, full rpms you'd have static + centrifugal, probably never any more than 30 deg.
Low rpm, low load you'd have static + vacuum, 15 to 20 deg.
Low RPM, high load, just static advance.


Anybody who says that they've got X degrees advance at XX rpm is wrong unless they've measured it with a timing light. Computers tell lies.

Steve

Well as i said its a bluetop engine, they have no knock sensor from factory. I had to weld a nut on to fit the knock sensor so the GZE ecu would be happy.

CR is 9 to 1 off memory for the standard TVIS bluetops with MAP sensor.

And my car has a CAS, no dizzy. Its running a 4AGZE ignition system run by the ECU.

Yeah 4AGZE 365cc injectors and STOCK 4AGZE ECU, i relise injectors would be close to maxing out, am running slightly more fuel presusre to compensate. Getting AF readouts done this weekend.

[1598cc]

get your a/f's done first.

check the knock sensor isnt giving flunked readouts.

if its a map sensor 4agze ecu, from my experiences they always seem to go flat after 5k rpm, not sure why. afm's setups didnt

[method]

get your a/f's done first.

check the knock sensor isnt giving flunked readouts.

if its a map sensor 4agze ecu, from my experiences they always seem to go flat after 5k rpm, not sure why. afm's setups didnt

How do i check knock sensor?

Yeah, got af readings at suprafest and they were alright.

[~SlideWays~]

On my old 4agze (MAP) AE86 I always had the 5500rpm problem really bad, I fixed it by stuffing around with the ign timing. Retarded it from memory.

Also its hugely effected by the water temperature. ie when its cold the 5500 problem will be worse. So if you're always running cold or even just below temp it'll back off at 5500. Possibly check the water temp sensor is working properly as well.

That was just running 10psi still supercharged and vr4 fmic, so I doubt its a boost pressure problem.

Few things to try at least.

[method]

Thanks a lot for that.

Im not even sure if my water temp sensor works. Have pretty good water temps, they arnt low. Use electric fan and thermostat to control it.

[noob question] Is the water temp sensor on the back of the block by those plugs?

[~SlideWays~]

Thanks a lot for that.

Im not even sure if my water temp sensor works. Have pretty good water temps, they arnt low. Use electric fan and thermostat to control it.

[noob question] Is the water temp sensor on the back of the block by those plugs?

Pretty sure it'll be one of the sensors in the alloy housing bolted to the back of the head. Which one... Im not sure.

Those MAP gze's are very sensitive to operating temp so I'd get a good water temp gauge and see what its doing.

On my old one I could literally tell if it was going to rev passed 5500rpm just by looking at the temp gauge... I stuffed around with this a lot because I initially had problems with getting a thermostat setup to open and close at the right temp. Either too cold or too hot until trial and error got it perfect.

[MAGN1T]

Using a stock ECU and map sensor, it won't see boost so the AFR will be a bit screwed up. Changing the fuel pressure screws it up even more. Stock injectors should be good for 250 or so horses.

Having a CAS instead of a distributor, the principal is the same really, the timing is taken care of electronically.

Anyway, you'll be leaning out at the top end and probably too rich before it leans out.

Steve

[Crampy]

Using a stock ECU and map sensor, it won't see boost so the AFR will be a bit screwed up. Changing the fuel pressure screws it up even more. Stock injectors should be good for 250 or so horses.

Having a CAS instead of a distributor, the principal is the same really, the timing is taken care of electronically.

Anyway, you'll be leaning out at the top end and probably too rich before it leans out.

Steve

He's running a 4AGZE ECU, so it will see boost. Hence the reason why he said he can go up to 13psi and be just below the fuel cut. A Stock Bluetop ECU and MAP sensor will fuel cut at something like 3-4psi.

It just looks like the good ol' MAP sensored 4AGZE ECU flat spot at 5500rpm. Heard of it many a time before.

[method]

Lets all LOL at this ugly graph.

AF is sweet at peak torque around 12. so long as its not leaner than 12.5 im happy.

Power dropps off, i assume ECU pulls timing then engine runs rich due to not fully combusting.

[l1ttle_d3vil]

[noob question] Is the water temp sensor on the back of the block by those plugs?

water temp sensor for the ecu is the green one on the back of the head.
the brown sensor is for the cold start injector timer.
there will probably be other sensors around which will be for fan and dash gauge too.

it would more than likely be running like crap all over if the temp sensor wasn't operating properly though, not just one flat spot. you can check if the sensor is working properly by what resistance it has at certain temps...

[Caveman]

that dyno curve looks familiar, this is my car while running the 4agze ecu



Apparently a new knock sensor can fix this problem, however I recon go all out and get the proper after market ECU and have an 12 second evo killer

also true to the colder plugs. I was running slightly colder plugs, dont remember heat range however they were also platniums. I should have given them to you but they were in the motor I just sold over $100 for good plugs

[mjrstar]

I can't figure out why hp and torque are inversely preportional to boost, but i guess the ecu must be pulling some serious timing out of the equation.
I think my car is just running out of top end breathing ie cams valves etc..
it's a stock 1gge with singe turbo. i have put a 3'' exhaust on it since this was done so power may be up, i don't really know if the 2 1/2'' crush bent exhaust was holding it back or not.

[method]

Ahh crap, must be the ECU thing then

I think the only reason its not dropping off as much as yours is because im running 15* advance on base timing. When i was running the 8 you could feel it drop off real bad.

[MAGN1T]

It's running way too rich.

Leaning it out, after dropping the boost and retarding the timing would make it go a bit better.

Or maybe I should rephrase that, Torque takes a dive at the same time the AFR goes too rich.
Maybe the ECU is picking up knock due to too much timing, then flooding it. Or maybe it's just running too rich and loosing power. It's always best to start with low boost and get it running right before increasing boost.

Steve

[method]

It's running way too rich.

Leaning it out, after dropping the boost and retarding the timing would make it go a bit better.

Or maybe I should rephrase that, Torque takes a dive at the same time the AFR goes too rich.
Maybe the ECU is picking up knock due to too much timing, then flooding it. Or maybe it's just running too rich and loosing power. It's always best to start with low boost and get it running right before increasing boost.

Steve

Ive now seen 4 different dyno plots with the same problem. Its clearly a ECU problem.

AF is fine, its 12.2 at max torque, it only goes crazy rich when the ecu takes the power away and by that time everything the ECU doing is $&#$%, not just AF ratio.

Will try playing around. may try ECU in diagnostics mode.

[ChaosAD]

Its the map gze ecu. aparently the afm one doesnt have the problem.

Many theories as to what causes it and how to fix it. Some poeple say a new knock sensor, others a bigger i/c.

But an aftermarket ecu will definately fix the problem

[Caveman]

It's running way too rich.

Leaning it out, after dropping the boost and retarding the timing would make it go a bit better.

Or maybe I should rephrase that, Torque takes a dive at the same time the AFR goes too rich.
Maybe the ECU is picking up knock due to too much timing, then flooding it. Or maybe it's just running too rich and loosing power. It's always best to start with low boost and get it running right before increasing boost.

Steve

You say in your sig after market computer is for people who dont know how to tune stock ones lol no tuning with this ECU, all you can do is make sure all sensors are 100% and even then it seems to be 50/50 running this style of setup. Or save yourself fuss, tune it properly with a programmable controller and get more power with more reliability.

The ECU is flooding the cylinder and pulling the timing, its trying to save the motor from what one can only presume is knock picked up from the knock sensor. On my plot you can clearly see the boost going through the roof when the timing is pulled as all the energy is going to the exhaust gas and the internal gate cant devert enough energy. Methods running an external so no boost problems for him.

Get one of the boys at ECE to make up a megasquirt

[MAGN1T]

The ECU is flooding the cylinder and pulling the timing, its trying to save the motor from what one can only presume is knock picked up from the knock sensor. )

So you agree that water injection has a very good chance of fixing it?

Steve

[Caveman]

The ECU is flooding the cylinder and pulling the timing, its trying to save the motor from what one can only presume is knock picked up from the knock sensor. )

So you agree that water injection has a very good chance of fixing it?

Steve

Not sure actually, I have done quite a bit of experimenting with a similar setup of mine and was directly involved in building methods setup and the thing that fixed my setup was relocating the knock sensor off the block. It was an extremely risky thing to do however in my lack of money I would try anything. As far as im aware the knock sensor needs noise or the car will go into limp so you cant have the knock sensor just hanging in space. Also it needs extremely good grounding or the car will go into limp. From this I hypothesized that the ECU is over reactive to engine noise and mistakes it as knock killing all the power. It could be a resonance phenomenon over 4ages as it appears to happen on bluetops, redtops and smallports at that specific RPM. I will not recommend this to method (but he'll try it anyway) as it completely removes any knock protection from the motor and is an engine failure waiting to happen. My motor running this setup was running 10.3:1 compression at 9psi. I used it for motorsport and it survived 100+ laps of pukekohe plus various hill climbs and other club sport without catastrophic failure, although after a year of abuse, increasing amounts of blue smoke became and issue for following cars... (probably more due to the constant high RPM than increased power output).

As you suggested, water injection will improve the efficiency of the motor by allowing a higher compression ratio (i.e. more boost) and decrease the thermal load on engine components however im not certain this will be a complete solution. Unfortunatly I feel it lyes in the ECU its self and unless someone has managed to interface and rewrite the standard ECU (not sure if this is possible for 4agze) then a replacement or at minimum a piggy back ECU is the only true solution.

However once youve swapped a supercharger for a turbo, changed both manifolds and even a new exhaust the fuel and ignition maps should be modified anyway, therefore after market seems like the only option.

One more thing ill suggest is ive heard of cases where a brand new knock sensor has cured the problem. Im hessitant to suggest this as a new knock sensor can be over $200, the source of this information was another car forum and an after market ECU would be fun to play with.