SAFC vs. E-Manage

[Sheltie2306]

The switch between close loop and open loop should depend on two conditions, TPS and rpm.

With OBD2 scanner connector, rev the engine up at stationary until you see "fuel system status" goes from "close loop" to "open loop", that's the cut over rpm. For TPS, its a little tricky. What I do is to get someone to operate the OBD2 scanner. Drive the car at slow speed but high gear, so that the car don't accelerate too much in the process. Press the throttle slowly and watch the "Absolute Throttle Position" and "fuel system status" reading. At the point where "fuel system status" goes into "open loop", take the "absolute throttle position" reading, and that's it. Hope that help. I don't think your TPS cutover point is that high of 95%, it's normally lower, may be about 60%-70%.

[Vtim]

that's a very wise move my friend

[Sheltie2306]

I managed to find a very blured pdf e-manage manual from the Internet. Looks like the fuel map is rpm/TPS, but ignition map is rpm/pressure(MAP), am I right?

 

For the fuel map, it can't be that you can only tune at the last row. The close/open loop switch over should be at ONE particular TPS and ONE particular rpm only, and not vary or inter-related to each others. That means it's a "either or" condition. Suggest you use the method I post to determine the switch over points of your car. So, you eventually find the close region should be a square on the upper left hand corner of the map, and the rest are open loop. The area should be quite big. Anyway, I am just guessing based on my experience, never tune e-manage before

Edited June 17, 2005 by Sheltie2306

[Slowpoke156]

The stock ecu use MAF signal to decide the injector duty cycle. Piggyback will alter the MAF signal and thus indirectly alter the injector duty cycle, isn't it that way?

 

Yes it is. I made my statement to further clarify his point, which can be misread that MAF and injector duty cycle are one and the same.

 

E-manage, like all other piggyback "can work" in both close loop and open loop. However, the tuner should neutralise the fuel trim as close to 0% as possible in close loop region, say if you have a larger than stock injector, so that the long term fuel trim will not fluctuate in the open loop region.

 

All piggybacks can work in open and closed loops. BUT... on cars with lambda and newer ECUs, some piggybacks WILL NOT be able to tune the closed loop. By "tune", I mean both ignition and fuelling. Not much point in having one without the other.

 

That's probably the tuner try to advance the timing too much such that eventual net timing is still far beyond the safty margin even the original ecu has retarded to its allowable range.

 

And do you know the actual timing the ECU is giving out in the first place? You don't, because no one does. That is why it gets worse if you fool MAF signals as well because there is now an added parameter where critical conditions can occur.

 

I never tune unichip before, but from conversation with unichip tuner, it should be the same as e-manage.

 

It is not the same, because I have the full equipment amd access to Dastek's resources and I know what it can and cannot do.

 

Yes, agree. But I had seen piggyback that can seperate map region based on TPS and rpm, so you can clearly know where's the open loop and where's the close loop.

 

I don't understand what you're saying. How do you tell if the system is in open or closed loop simply by looking at a map?

 

When we floor, isn't it WOT

 

Again, that statement was made in response to the original post that WOT is REQUIRED for open loop operation, which it is NOT. Please read the original post again.

Edited June 17, 2005 by Slowpoke156

[Jamesgetz]

v interesting pts on the SAFC. there's almost a SAFC religion out there whereby followers of SAFC advocate that it works and the dyno test results which they have is gospel truth.

 

wat about ECU re-mapping by flashing the maps stored in the stock ECU? are the gains likely to be credible (assuming the tuner is knowledgeable)?

[Zrun]

If your primary concern is dyno results, I think safc, unichip, e-manage will perform equally well. It is because dyno testing only shows WOT performance. However, for extensibility, such as bigger injectors, NA->turbo, turbo mods, SAFC will not be able to handle these safely.

 

ECU remapping is the best way, short of standalone. With all tuning, you need experience and time on the dyno to get the proper gains.

Edited June 17, 2005 by Zrun

[Jamesgetz]

cool, thx for the advice. it's just that SAFC followers are so convincing abt their 'religion' that i sometimes get swayed over...

 

a few poss options which i'm tempted to try:

1) SAFC

2) Autodesigns tuning box (can tune up to 128 pts if i remember)

3) ECU re-mappping

 

1 good thing abt SAFC is that it's easy to sell off. depreciates by maybe $100 to 200 max by the time you sell it. AD tuning box is a relatively new product and only AD can tune it(think monopoly...). i feel ECU re-mapping with proper tuning will give the best gains but the only prob is that it's not transferrable (since it's tuned to the specific car)and it's also the most exp option -> about $600+++

[Sheltie2306]

Hey, thanks very much, this is really a good sharing session, few things to clarify:

 

>All piggybacks can work in open and closed loops. BUT... on cars with lambda and newer ECUs, some piggybacks WILL NOT be able to tune the closed loop. By "tune", I mean both ignition and fuelling. Not much point in having one without the other.

 

The piggyback itself has no concept of "open loop" and "close loop", it's the tuner that need to be aware of which region he's working on. This is also related to another point later. Basically, in close loop region, the original ecu will try to gain control of the A/F, so this issue is more pertaining to fuel map, not so much on timing map. I don't think there's "close/open loop" for timing map. The ecu is in control thoughout.

I ever read from somewhere that piggyback like perfect power can actually alter the O2 sensor signal, and thus can tune the A/F in the close loop region too. Any idea about this?

 

>And do you know the actual timing the ECU is giving out in the first place? You don't, because no one does. That is why it gets worse if you fool MAF signals as well because there is now an added parameter where critical conditions can occur.

 

I would say that's the short coming of piggyback in terms of timing control. We can use the obd2 to check on the actual timing coming out from the ecu to have a gauge of where we are instead of adding timing blindly.

 

>It is not the same, because I have the full equipment amd access to Dastek's resources and I know what it can and cannot do.

 

Any chance to share the Dastek document? I just want to have a better knowledge of unichip?

 

>I don't understand what you're saying. How do you tell if the system is in open or closed loop simply by looking at a map?

 

Normally, we should not be able to tell the region from the piggyback map. However, the close/open loop is normally based on either/or of TPS and rpm. It's the tuner's responsibility to figure out the cut over tps and rpm for the car, and from there to determine where's the close and open loop region on the map. And yes, I came across this piggyback system that you can actually key in value to state from which % TPS then the piggyback will switch over to a different map region. So pratically, you have two map regions, a close loop and a open loop that you can see clearly from the map.

 

 

>Again, that statement was made in response to the original post that WOT is REQUIRED for open loop operation, which it is NOT. Please read the original post again.

 

Sorry, I misunderstood you initially. Anyway, I agree with your original post, and that's why I suggested zrun to check out the open/close loop TPS and RPM cut over point in another post.

Edited June 18, 2005 by Sheltie2306

[Sheltie2306]

By the way, if unichip cannot tune fueling in the close loop region, how do you handle a larger than stock injector? Does it has some kind of overall injector scaling function?

 

Or does unichip just let the original ecu run into large negative fuel trim and try to alter the fuel map in the open loop region to compensate for that? If that's the case, it looks like there's limitation on the size of injector that this method can handle.

[Slowpoke156]

The piggyback itself has no concept of "open loop" and "close loop", it's the tuner that need to be aware of which region he's working on. This is also related to another point later. Basically, in close loop region, the original ecu will try to gain control of the A/F, so this issue is more pertaining to fuel map, not so much on timing map. I don't think there's "close/open loop" for timing map. The ecu is in control thoughout.

I ever read from somewhere that piggyback like perfect power can actually alter the O2 sensor signal, and thus can tune the A/F in the close loop region too. Any idea about this?

 

You are spot on about the piggybacks not having an idea of what they ECU is doing. All they do is add or subtract from the given value. That is why it is dangerous if you do not know what you're doing. Even famous tuners make mistakes. Not sure about perfect power. I gather it should be the same as what E-Manage does, maybe different in implementation, but the idea behind it could be the same.

 

I would say that's the short coming of piggyback in terms of timing control. We can use the obd2 to check on the actual timing coming out from the ecu to have a gauge of where we are instead of adding timing blindly.

 

A word of caution. OBD2 tools are not very precise. In timing controls, you are talking about the engine firing twice every full rotation of the crankshaft (for 4 stroke 4 cylinder engines). Depending on how timing is derived from the crankshaft (how many missing teeth), ignition timing is a precise art. At 7000rpm, you are talking about 14,000 firings per minute, or if you like, 233.33 firings per second. There is no consumer OBD2 tool in the market that has that resolution. Even the engine diagnostics that most car workshops use are at best 5Hz. What happens in between in anyone's guess. You still think you have a useful tool?

 

Any chance to share the Dastek document? I just want to have a better knowledge of unichip?

 

No can do. They will revoke my priviledges if I do that.

 

Normally, we should not be able to tell the region from the piggyback map. However, the close/open loop is normally based on either/or of TPS and rpm. It's the tuner's responsibility to figure out the cut over tps and rpm for the car, and from there to determine where's the close and open loop region on the map. And yes, I came across this piggyback system that you can actually key in value to state from which % TPS then the piggyback will switch over to a different map region. So pratically, you have two map regions, a close loop and a open loop that you can see clearly from the map.

 

That shounds like a good piggyback. I recommend you use it rather than everything else than. But first, you will still need to know the cutover points. So gotcha!

 

Sorry, I misunderstood you initially. Anyway, I agree with your original post, and that's why I suggested zrun to check out the open/close loop TPS and RPM cut over point in another post.

 

No problems dude. Have a good weekend... or whatever's left of it.

[Slowpoke156]

Larger than stock injectors? The ECU will reduce duty cycle automatically. But physically, the car may have problems running at low rpms because all injectors have optimum RPM ranges and if the duty cycle is way too low, you could run the risk of running too rich or lean and the ECU will be mucking up the entire engine trying to get the injectors to shoot correctly. Unichip will still be unable to help.

 

My recommendation is to use the chip to control additional injectors instead. It's a smarter method and reduces the risk of screwing up closed loop control.

 

And the only reason why you would every need to oversize your injectors will be because of forced induction. Otherwise, it's simply a waste of money. Normal cars run at max. 50% duty cycle, which is the generally accepted optimal duty for injector life and service. No matter what you do or feed the car with, you will never run out of duty cycle, short of forced induction.

[Sheltie2306]

Well, I guess that's reason why the JB unichip shop strongly recommend additional injector instead of large injector

 

Yes, with larger injector, the ecu will resort to negative fuel trim to neutralize it. My approach is to use piggyback to reduce fuel in the close loop region, at the same time, watch the long term fuel trim, until it's near 0%. In that way, we will have more room to play in the open loop region as the long term is about 0%. The piggyback don't need to add fuel like crazy to achieve the desired AF. With this approach, we can cater for larger injector size if required too, or else the CEL light will turn on the moment fuel trim goes below -25% (or somewhere, can't remember very well). What do you think?

 

Regarding the timing that you mentioned in the earlier post. Say, if we are holding at 7000rpm, it should sample and read off the timing at certain interval. I think this value should be feeded by the ecu, either from the timing table, or whatever. It's not "measured" by the obd2. And normally, the timing don't fluctuate too much if the rpm is stable at certain value, I feel that we can trust the value pretty well.

Edited June 20, 2005 by Sheltie2306

[Thumbs]

>>There is no consumer OBD2 tool in the market that has that resolution. Even the engine diagnostics that most car workshops use are at best 5Hz. What happens in between in anyone's guess. You still think you have a useful tool?<<

 

So how can ignition timing be determined?

[Slowpoke156]

As mentioned before, unless you're into turbo/supercharging, there is no need to change the injectors. They are already under-utilised as it stands, why bother?

 

Ignition timing is never fed by the ECU. It is sensed, via the hall sensors on your crankshaft. The time difference between 2 pulses will in a way "tell" your ECU the position of the pistons and hence firing timing. This is the same signal which is tapped and use by the piggybacks. ECUs DO NOT generate timings. Piggybacks use the sensed value and then add or subtract from it. For example, if it wants your ECU to advance timing by say, 2 degrees, it takes the value between 2 pulses and calculates the crank angle, then add accordingly. This is why it is also important to match piggybacks with the original ECU's speed as much as possible. The current Unichip is a 32bit ECU, hence the same speed and resolution as most BOSCH ECUs out there in the market. I am saying OBD2 tools are not accurate simply because within the sample time interval, the ignition timing could have varied by +-5 degrees. If you depend on it for tuning, the values are going to be far out.

[Slowpoke156]

If you are not using the OBD2 tools for tuning, then you can use it to "determine" timing. Otherwise, do the tuning on a dyno.

[Thumbs]

Why so complex... can't a timing light be used?

[Sheltie2306]

Yes, I am tuning TC. My IDC used to hit nearly 100% until I upgraded injectors.

[Slowpoke156]

Then it's a different ballgame altogether. Say so earlier mah!