EFI vs Carb/ engine build choices

[jayb]

Thanks for the replies guys (even FElony)
As far as CR goes I think I'll shoot for 10.5 static. It's a 4x 4.180 bore. 1U crank offset to 4"
Gonna flow the heads (BBM) before deciding on a cam. Nothing off the shelf. Solid roller. Prolly won't stray to far off the norm.
Oiling through the pushrods to Harland Sharp (new) rockers.
351C RPM air gap through Jays adapter.
Should make a '66 scoot right along.

Marc, cam and compression should be selected together.  Don't order the pistons until you know which cam you will be using - Jay

[Falcon67]

Short answer is "EFI don't care".  You build the motor, then you'll have to tune the fuel system and timing to make it live.  That's it.  Either can cost you time on the dyno or a lot of passes on the strip to get dialed in.  EFI just costs more up front.  IMHO it can actually be more work than a carb - same feed back required using a wide band sensor - except you have to get your hands dirty taking the carb apart to change jets as needed.  For a street car, once you do get EFI dialed in with the fuel curves and timing - if it's a closed loop system it'll tend to take care of it's self.  If its for bracket racing, I'd stick with a good carb.  If using EFI for brackets, I'd choose open loop.  Don't need to be changing fuel calibrations as the day goes on.

[turbohunter]

Yeah Jay I'll get all the info first.
Chris it's definitely more work and money for efi but I've committed to it in order to learn. I know next to nothing about injectors and modern gear.
I'm actually going to follow Jay's idea with his efi thread and fire the engine on carb while having the gear in place to go efi then switch it over. That way I'll have a baseline in my head about the engine and where it should be.

[My427stang]

The only thing I'll reiterate is that overlap does matter to EFI, the lower the RPM and more often the engine will operate there, the more you want to limit overlap.

It doesn't have to be dramatic, you can run a decent cam, but as overlap increases, at lower RPM the O2 sensor will not "see the truth"  and report to the ECM that the engine is rich when it's really not, that goes away somewhat with RPM.

My opinion, with a good system you can tune for almost anything, but to uncomplicate things, limiting overlap to mid 70's will help you on the street

What does that mean in numbers?  Along with proper cam choice in general, with a 300 degree hyd roller, I'd spread to 112-114 and pick whatever ICL you want to match compression and use (earlier usually better on the street IMHO)   For a 290 degree or less SFT, I'd go as tight as 110.   The less lobe you have (and more lash), the tighter you could go because overlap is reduced

[turbohunter]

I know it might be counter intuitive for the street, but I am toying with the idea of solid roller.

[My427stang]

I am not a solid roller fan, but admittedly haven't used one since good pin oiled lifters came out (I know, get with the program, it's only been 15 years LOL)

However, typical solid rollers and solid flat tappet advertised duration is at .020 rise, hydraulics are at .006.  Not as huge of a difference as you'd think comparing the two because of lash make up .006 ratings are within guessing distance of a lashed .020 duration, but the shape of a solid lobe usually minimizes overlap a bit so it shouldn't hurt.

What's the planned gear ratio / tranny / street use plan when complete?

[turbohunter]

TKO 600 with the .68 5th gear to 4.10s (Truetrac) in the rear end.
I went with the taller 5th gear because I plan to do some freeway driving up to PCH and a few great roads.
Figured I wanted a 4 speed around town with a nice over drive for gitten where I gotta go.

[My427stang]

That's a great combo, I love the tall 5th TKO in mine with 4.10s and 275/60s

Is it a stroker FE?

[Falcon67]

Good point on the overlap.  I do know that a lot of the farther out builds tend to use F.A.S.T units.  Most of the street friendly/DIY friendly units might not be able to handle very radical cams and such.  Even with a carb, that really screw up low speed metering.

[jayb]

I would like to qualify Ross's statement about overlap.  If you have a self-tuning EFI system that relies on the O2 sensor to determine the fuel required then yes, a big cam will tend to fool the sensor and make the EFI system not work properly.  However, Marc is going with a REAL fuel injection system, not one of the tunes itself versions.  In that case any cam can be set up properly.  After screwing around with these systems for several years I can say with confidence that any cam can be used with a fully tunable EFI system.  The trick is to tune while driving the car, and use the O2 sensor as a guide only, not in closed-loop mode.

My Galaxie is a good example of how this can be done.  Back in 2008 with the EFI system first installed, the engine was making 935 HP and had 286@0.050" cam lobes on the intake, if I recall correctly.  After tuning on the dyno at wide open throttle and at idle, I got the engine in the car and hooked up the computer.  My son the computer geek rode along, tapping the keys.  We went to a straight stretch of road and I held a constant speed (constant RPM and vacuum, or MAP) so that the indicator on the VE table (fuel table) was stable in one spot.  Again, this was open loop mode, so the O2 sensor was not doing anything to change the fuel delivery.  While reading the O2 sensor I told my son to lean out the map until we were at about 14.7.  The engine was sputtering badly long before we got there, it was so lean.  Clearly, the O2 sensor was being fooled by the cam events.  Next we went the other way, richening that spot on the map until the engine started sputtering with too much fuel.  Now we had a range.  I figured it was probably an A/F of 10:1 on the rich side, and 16:1 on the lean side, so we went most of the way back to lean, so that the engine was still running nicely.  The O2 sensor was reading about 11.2:1!  No way was that the real A/F ratio, I'll bet we were close to perfect stoich at that point.  We tuned by listening to what the engine wanted.  And it was easy.

We repeated this procedure for several more points on the VE map, different engine speeds and vacuum readings, and when finished I extrapolated between these points to fill in the rest of the map.  I made further adjustments later, but the car was perfectly drivable under all conditions at that point.  No feedback control from the O2 sensor was ever used because as stated earlier, a big cam will fool the O2 sensor, to say nothing of what would happen to the car if the O2 sensor went out of calibration or failed completely. 

It was really fun to tune the car like this, by the way.  You get an instant reaction from your inputs.  You can save all your tunes as files, so that if you screw something up you can go back to the previous version.  Close to unlimited flexibility, makes tuning the engine a blast.

Marc, on another note I am glad to hear that you are thinking about a solid roller cam.  Go for it!

[My427stang]

Jay, you are 100% correct and that is what I tried to convey early in this, that you can do anything if willing to manage sensor input in open or closed loop when appropriate if the system allows.  However it is not just a cheap system that can trick an O2, it is ALL systems if you use that sensor in a closed loop option, the question is, does the system have the ability to allow you to tune around it and will the vehicle behave the way you want it to when you do?

Anyone who tunes seriously uses the open loop regularly when needed, but it is limiting the function of the system by design, and to be honest, using a purely open loop speed density system will run great, but is limited in it's ability to adapt if you have an engine that doesn't need to run open loop.

Let me qualify what you said too, because I actually run a more "real" system than a FAST SD by running a MAF system with both WBO2 and NBO2 and left and right bank inputs. The modified OEM system I use is more adaptable than any speed density system because it additionally measures intake airflow, something a speed density system cannot do. The system is also modified to only adjust a/f tables using WBO2 in those times that I allow to and uses NB for normal operations.  That NBO2 is limiting in itself in its narrow range of responsiveness, and I may actually dump them now with a recent upgrade to the software, but having the input for both styles and the combination of logs that can be combined with load, speed, airflow, ambient temp and pressure, etc, really works well. 

In the end though, open loop is not as responsive by design as closed loop and depending on the use of the car, may not be as happy as yours have proven to be for your use.  It is essentially creating a map or table for a given range as you discuss, if you want a multi-dimensional map and want the system to adapt it, that O2 sensor is a very good thing (until you have a characteristic of the engine or environment that you cannot do it, but on a pure streeter, IMHO the goal should be to get back to closed loop if/when you can)

I do agree with you completely though, when you have a system, and that system has the ability to be manipulated for your use, and you learn how to work it, it becomes very usable. 

I am also not a cammer guy, but I am interested in what your equiv overlap at the valve is with a 286@.050 SOHC cam and how they are installed compared to a wedge, my hunch is a quite a bit less duration at the valve with the low rocker ratio so therefore less overlap. (Although I bet the overlap on a hemi does have more authority)  Not that it matters with the way you use your system, but if i remember correctly, you mentioned that the cam has to make up rocker losses.  Regardless, I cannot see a a 286@.050 being the right cam choice for an EFI wedge with a 2.63 final drive, so hard for me to recommend open loop operation for a 600-ish hp wedge.

For the crowd, you are seeing the most common EFI argument on the net unfold between experienced guys though   technique vs technique.  I try to avoid open loop and get into closed loop as often as I can because of the rpm range and power levels I am tuning.  Jay is building for MUCH more power and RPM, needs those tools and is experienced to tune to it.  Who is right?  The answer is yes LOL  but bottom line, neither could do what we are doing with a handheld controller

I also need to echo Jay's tuning fun comment, file save, reload, try something new again, sometimes even in a parking lot on the way home, undo and go back to last Thursday's load, all in the course of a ride across town without getting dirty.  It's cool as cool can be

[jayb]

On the SOHC, rocker ratio won't enter into the overlap value because overlap is in degrees.  For that cam I mentioned above, overlap was 87 degrees.  The SOHC lobes are large compared to a normal FE wedge lobe in order to make the valve follow the same profile as a normal engine, even with the low rocker ratio.  So what the valve sees is essentially the same as what a wedge engine valve would see with an equivalent cam.  And you are right about the sensitivity to overlap with the SOHC Ross, pretty easy for that intake charge to go out the exhaust during overlap with the canted valves and hemi chamber.

I think one major reason I don't like to run in closed loop is I have seen sooooo many of the wideband O2 sensors go bad, either not working completely or being significantly off, and not able to be calibrated.  I'll bet I've tossed out 20 or 30 of those Bosch wideband O2 sensors over the years, and a few NTKs also.  Funny how some of them seem to last a long time, and some don't.  This is mostly in dyno situations so one could argue that they are being abused, but it still seems excessive to me. 

The only reason I think I would run in closed loop was if I was worried about variations in the fuel.  On my cars I normally use a single table, with the top of the fuel table (e.g.wide open throttle) tuned for race fuel and the rest of the table tuned for pump premium.  If there were significant variations in the pump gas I was using, the closed loop system would have an advantage.  But since I only drive my cars in the good weather months, I don't worry about that too much.

Also, I would strongly disagree with your statement that open loop is not as responsive by design.  That's just not true, IMO; in either case the engine is operating somewhere on the fuel map and will respond the same in either case.  Can you provide an example of why a closed loop system would be more responsive, to back up your argument?

[My427stang]

On the SOHC, rocker ratio won't enter into the overlap value because overlap is in degrees.  For that cam I mentioned above, overlap was 87 degrees.  The SOHC lobes are large compared to a normal FE wedge lobe in order to make the valve follow the same profile as a normal engine, even with the low rocker ratio.  So what the valve sees is essentially the same as what a wedge engine valve would see with an equivalent cam.  And you are right about the sensitivity to overlap with the SOHC Ross, pretty easy for that intake charge to go out the exhaust during overlap with the canted valves and hemi chamber.

I think one major reason I don't like to run in closed loop is I have seen sooooo many of the wideband O2 sensors go bad, either not working completely or being significantly off, and not able to be calibrated.  I'll bet I've tossed out 20 or 30 of those Bosch wideband O2 sensors over the years, and a few NTKs also.  Funny how some of them seem to last a long time, and some don't.  This is mostly in dyno situations so one could argue that they are being abused, but it still seems excessive to me. 

The only reason I think I would run in closed loop was if I was worried about variations in the fuel.  On my cars I normally use a single table, with the top of the fuel table (e.g.wide open throttle) tuned for race fuel and the rest of the table tuned for pump premium.  If there were significant variations in the pump gas I was using, the closed loop system would have an advantage.  But since I only drive my cars in the good weather months, I don't worry about that too much.

Also, I would strongly disagree with your statement that open loop is not as responsive by design.  That's just not true, IMO; in either case the engine is operating somewhere on the fuel map and will respond the same in either case.  Can you provide an example of why a closed loop system would be more responsive, to back up your argument?

Sure, I do not mean responsive to your foot, I mean responsive to conditions, sorry for the confusion.  Certainly no intention to say that it doesn't respond to foot pressure.  I also do not mean that open loop cannot work well with tuning, because it does and I run open loop myself when required, but if you could map for all situations, there wouldn't be need closed loop and WBO2 sensors on modern vehicles needed because a/f mixtures would be where you want them to be and remain in a narrow range.  Remember, just like you, I am talking for a type of build and use that I build, so I am not saying you are wrong in any way, I am just saying that you can make good street power in closed loop, and that the WBO2 dilution isn't just a cheap EFI problem it is a problem with too much overlap even on adjustable systems that let you tune around it.

As far as no change in duration with rocker ratio, if I have 286 degrees at .050 lobe lift with a 1.5 rocker, at that lift point I have .075 of lift at the valve, with the same .050 lobe lift and a 1.76 ratio, you'd be at .088 at the valve.  If you measured rotation at the valve from .075 to .075, you'd get more duration from the wedge.  So essentially the wedge rocker creates duration at the valve, which increases overlap on both lobes. Inversely, at the valve, the time that a valve is open for a given lobe duration is less on a SOHC. Heck I thought I read that in your book, either that or I read it in the older magazine article with the Coon parts

Again, please do not let me sound like I am saying you are incorrect in using your equipment the way you do. It is a standard and proven tuning method, especially for the power you are making.  I just wanted to point out some rules of thumb and point out the fact that at some healthy levels you can run closed loop without incident, and that my original point wasn't far off, sometimes you need open loop, however my technique is to return to closed loop too

I do like the way you tune though, especially the idea of changing parameters in portions of the curve for fuel.  Just don't go WOT with cheap gas right?

[turbohunter]

Very much enjoying your discussion guys. Thank you.
All you guys.

[jayb]

Ross, I'm still not clear how you are coming to the conclusion that closed loop is more responsive to conditions, unless you are assuming that the fuel map and the temperature corrections are not right.  Seems to me both would be equally responsive to changes in conditions.  Can you provide a concrete example of what you are saying?  I'm not trying to be argumentative, I'd just like to know your rationale.

Also, you are correct on the fact that more duration @0.050" is required for an SOHC with the lower rocker ratio, than would be required for a wedge.  Probably about 12 degrees more.  However, overlap is generally given in degrees from intake seat opening to exhaust seat closing.  The duration at 0.050" given in your example doesn't enter into it.  And the point at which the seat opens is not dictated by rocker ratio.  So, for the cams I use the advertised duration is given at 0.020" valve lift, which is the lash value, and is calculated the same as it would be for a wedge.