retarding ignition timing?

[StuartM]

matt: the fuel will be delivered at the wrong time by the amount you set it out by. but thats no biggie if its only 2deg out.

my point is that the ECU does NOT KNOW that the timing is out and therefore you can expect some error in the electronic spark advance. because the ESA is additive to the base timing, you modify the total advance (given). But exact timing of the spark in an ESA system is not soley additive to base timing; TPS, manifold pressure, measured airflow, coolant temperature among other things can effect the delivered ignition timing.

In the case of my sister's curren having the dizzy a few degrees out was enough to make it idle rough and lose power, this is because the spark was being output at what the computer thought to be exactly (for example) 12deg but it reality was less, ie the ESA was all to sh1t.

[matt dunn]

and when i say 'some error' dont knee-jerk and think your car wont start, you probably wont ever even notice.

so you loose performance by the amount you can never feel,

But you gain the performance by having it retarded
by the amount that everyone on here that has a 20V and has done it and has noticed by not loosing the extra advance that the knock sensors take away.


Why dont you give us a clue? Tell us what else effects timing that you put out by retarding the timing?

[StuartM]

If the air temperature gets too high in a good turbo ECU, the timing would be retarded to prevent pre-ignition. If you set the base timing more retarded than expected by the ECU then you will lose power than you otherwise would, does that make sense?

Why dont you give us a clue? Tell us what else effects timing that you put out by retarding the timing?

Done. I dont have time for flame wars.

[StuartM]

Better to have the timing retarded by 2 degree's manually than the computer detecting knock and you loose 8 degree's.

anyway, i gotta agree with that. never had knock in a car except for the old 4k with base timing of 20 deg, with vacuum it was like 60+ deg

dont want to be off topic but what did you do to de-compress your 20v engine for turboing?

[RedMist]

matt: the fuel will be delivered at the wrong time by the amount you set it out by. but thats no biggie if its only 2deg out.

I've never seen a sequential injection standard ECU. Althouth the 20 valve has 4 injector FETS and 4 seperate drivers its still a batch fire injection system. It doesnt matter if its 2 or 200 degrees out of sync.

my point is that the ECU does NOT KNOW that the timing is out and therefore you can expect some error in the electronic spark advance. because the ESA is additive to the base timing, you modify the total advance (given). But exact timing of the spark in an ESA system is not soley additive to base timing; TPS, manifold pressure, measured airflow, coolant temperature among other things can effect the delivered ignition timing.

Actually the ECU doesnt have a clue, or care as to what the base timing is set to. No error will occur unless closed loop knock control operates out of bounds. There is an electronic advance which is calculated based on the factors you have given then there is the mechanical advance set at the CAS. You simply add the two to get total advance.

In the case of my sister's curren having the dizzy a few degrees out was enough to make it idle rough and lose power, this is because the spark was being output at what the computer thought to be exactly (for example) 12deg but it reality was less, ie the ESA was all to sh1t.

So I gather you are saying that the computer, now getting its signal a few degrees different from stock has another reference? It somehow realizes its not set to 10 degrees by referencing it against that other signal, then purposely turns it all to shit?

Sorry but there was other mitigating factors in the fault with your sisters curren.

As to your correction in regards to the sensor type, I think you'll find that CAS can either be vane hall effect or variable reluctance. The 20 valve uses Vane.

[Mr. Mainstream]

muh i dunno if its worth mentioning but after we put the BPD engine in my 4WD323 we set the timing to factory and a few days lata i advanced the timing slightly by bringing the dizzy forward and OMG the difference, pulled alot harder and came onto boost bout a grand earlier, pings slightly between 6-7G on 95 but the difference was crazy

[StuartM]

Sorry but there was other mitigating factors in the fault with your sisters curren.

I've never seen a sequential injection standard ECU.

tell me what exactly was the problem with my sister's curren? and why do you have a hard time beliving me? oh, and why would these mitigating factors disappear when i set the timing correctly?

you havn't seen many modern ECUs, so what? I dont like your belligerant attitude. If you know what was wrong, please tell me, otherwise your just acting like a know-it-all.

no hard feelings mate, but im on this forum to learn, not argue.

[sergei]

BTW I can check if 20v is sequential, grouped, or simultanious (I'll come back to it later), but I'm more biased that 20v is sequential system, the clues are:
High impeadance injectors (relatively slow, inacurate on high rpm low load)
4 separate injector drives (why not make one common, it is cheaper anyway).
I have pdfs here stating that there are toyota ECU which run sequential injection. It states there that

...A sequential system will have one injector per injector driver.

My check involves in hooking up 2 channel oscilloscope to each injector while car running and checking injector timing. But I need motivation to pull oscilloscope out (as it is a bench unit)...
[/quote]

[StuartM]

So I gather you are saying that the computer, now getting its signal a few degrees different from stock has another reference?

yes, all measurments are relative and AFAIK the ESA is relative to the timing signal received by the distributor. When the base timing is adjusted the output of air temperature-based-retard and coolant-temp-based retard/advance will not be perfect, for example, in a turbocharged engine you may require less air temp-based retard because you retarded the base timing.

for a perfect spark advance curve to exist the value added by ESA would need to re-calibrated and i dont think modifying the spark table ROM would be very practical. for what its worth, on a 4A-GE ECU you will probably not notice much error with the timing changed, but bear in mind it wont be perfect.

[StuartM]

BTW I can check if 20v is sequential, grouped, or simultanious (I'll come back to it later), but I'm more biased that 20v is sequential system, the clues are:
High impeadance injectors (relatively slow, inacurate on high rpm low load)
4 separate injector drives (why not make one common, it is cheaper anyway).
I have pdfs here stating that there are toyota ECU which run sequential injection. It states there that

...A sequential system will have one injector per injector driver.

My check involves in hooking up 2 channel oscilloscope to each injector while car running and checking injector timing. But I need motivation to pull oscilloscope out (as it is a bench unit)...

hey this sounds like a fun idea! surely it will be documented somewhere though?

edit:
anecdotally; the early 3S-GE ECUs used two banks of two injectors in parallel driven by pulling the signal to ground at the ECU end and the 1994 3S-GE in the curren uses 4 seperate injector circuits.

[85AW20v]

Hey Sergei - got a chainsaw to make the doorway big enought o get the car to the bench

[RedMist]

tell me what exactly was the problem with my sister's curren? and why do you have a hard time beliving me? oh, and why would these mitigating factors disappear when i set the timing correctly?

I stated there were other mitigating factors. I don’t know what the issue was, but it certainly wasn’t the ECU detecting that the timing at the CAS was incorrect and for some reason purposely introducing a misfire. Perhaps it was just as simple as the timing needed setting?

you havn't seen many modern ECUs, so what? I dont like your belligerant attitude. If you know what was wrong, please tell me, otherwise your just acting like a know-it-all.

I have currently a 20 valve (4 drivers), 16 valve redtop (2 drivers), a Link LEM 4 (2 drivers), a genboard (4 drivers, up to 8,), a Mitsi V6 1600 ECU (about to open it up). And have had in the past many more copies from the above and a Link Plus to examine. I'm also a computer systems engineer which explains my fascination of the computing side of my race car.
I'm sorry but just because I don’t agree in your synopsis doesn’t mean I'm belligerent. And no, I don’t know it all. I however do know about setting base timing on a stock or aftermarket ECU, the trigger side of EFI, and basic computer logic.

no hard feelings mate, but im on this forum to learn, not argue.

Unfortunately it appears the complete opposite. Stuart in order for you to learn you need to take things onboard and either logically prove or disprove them. If you manage to disprove them, then a logical, non name calling, response is required so that others may learn from either your success or failure.

There are many here that have reset base timing to non stock without the symptoms you describe, myself included. Hence I find it very hard to believe your statements.

[RedMist]

So I gather you are saying that the computer, now getting its signal a few degrees different from stock has another reference?

yes, all measurments are relative and AFAIK the ESA is relative to the timing signal received by the distributor. When the base timing is adjusted the output of air temperature-based-retard and coolant-temp-based retard/advance will not be perfect, for example, in a turbocharged engine you may require less air temp-based retard because you retarded the base timing.

I agree with you in part. You will be retarding or advaincing throughout the map, and this may have a possible effect on optimal performance somewhere in said map. However if the octane rating of the fuel is at question then all high load values of the map should be retarded. If not by shifting base timing then how do you suggest it done?

for a perfect spark advance curve to exist the value added by ESA would need to re-calibrated and i dont think modifying the spark table ROM would be very practical. for what its worth, on a 4A-GE ECU you will probably not notice much error with the timing changed, but bear in mind it wont be perfect.

Couldn't agree more. As far as I know nobody has yet cracked the DENSO chipset.

[sergei]

As far as I know nobody has yet cracked the DENSO chipset.

But you are wrong, as I own 2 cars with 2 ecus wich have been "cracked" (one by BLITZ another by Mine's)... It is again case of my lack of motivation to pull out the ROM chip and read what is on it and possibly reverse engineer it (well I'm not really capable to do it right now).
All right, I will pull out osciloscope on saturday (if it is going to be sunny day), and keep you updated...

[RedMist]

As far as I know nobody has yet cracked the DENSO chipset.

But you are wrong, as I own 2 cars with 2 ecus wich have been "cracked" (one by BLITZ another by Mine's)... It is again case of my lack of motivation to pull out the ROM chip and read what is on it and possibly reverse engineer it (well I'm not really capable to do it right now).
All right, I will pull out osciloscope on saturday (if it is going to be sunny day), and keep you updated...

Serious? I thought that both the Blitz and Mines were replacement Chipsets using their own codeset but the original Denso pinouts? IE they hadnt attempted to crack the Denso code at all. But as I havent cracked either the Denso, Blitz or the Mines chipsets I don't know. I just hadnt heard of it being done before.

Would be fantastic to also know for sure if the 20 valve drivers are phased. I had always believed them to be batched through seperate drivers much as the Link Plus is. And before you say the Link Plus is sequential.... it is... until about 5k RPM, where its batched.

[sergei]

well, they are in fact replacement chipsets, but obviosuly adapted to work with DENSO board, I always believed that DENSO chipset is just modified intel (8051) ....
As for the link plus, after 5k it is pointless to have sequential, as injector duty cylce will be way longer than valve opening time.

[RedMist]

well, they are in fact replacement chipsets, but obviosuly adapted to work with DENSO board, I always believed that DENSO chipset is just modified intel (8051) ....
As for the link plus, after 5k it is pointless to have sequential, as injector duty cylce will be way longer than valve opening time.

I hadn't thouht of the Denso chipset as a rebadged Intel. Quite feasable, however I dont think its an 8051 as the Denso board has two processors a 42 and a 64 pin (in the 20 valve 175000-4623). And if it were intel... wouldnt it have been cracked by every power chip company in the world by now? But rebadged prominant processor manufacturer... almost certainly. Which still begs the question... why isnt there a powerchip for these ECU's?

Interesting in regards to sequential. When you think of the valve timing, and duty cycle, sequential must offer bugger all benefits. at 85% duty cycle it must be firing against a closed valve a majority of the time. Even at 50% with a perfectly timed injector fire, you're still firing against a closed valve for quite a period of time. Would be interesting to swap the injector drivers around and re-dyno..... Sounds like something to do on the Bunderson next dyno visit... although it only has two drivers.

[sergei]

sequential is only beneficial on low load low rpm, or very big injectors....

[Grrrrrrr!]

the advantages of sequential injection are emissions and fuel economy at low duty cycle... ie around town driving and open road driving. at full throttle the benefits are neglible.

[ollieboy]

You know what I find outrageous? People who have only been posting on TS since the start of the year and don't know people on here and make arguments with people who are incredibly smart and have been posting for a long time. I don't mean to say the more you post the more you know but the less time you are on TS the less you know the people and the people who know stuff. Take Matt Dunn for instance, a person who has been a member of TS for years and has contributed huge amounts of useful information to people in that time and the crowning moment "He actually knows what he is talking about." And yet people choose to argue with him. There are a few others on here that need a mention as well but I am tired and can't remember.

Sorry its off topic but I just felt it needed to be said a long time ago.

Ollie