Help - Search - Members - Calendar
Full Version: Interesting SDS EFI results
914World.com > The 914 Forums > 914World Garage
Mark Henry
I sold the 2.0 junkyard dog (a 2.0 I made from scrap parts) it was stock except for the C-25 Scat cam and installed my bone stock 1.8
Here’s what’s happening.

I haven’t even adjusted my values and I still have an OK reading on my narrow-band (Haltec AF 30) meter. I have a wide-band on the way from SDS, so I’ll have a direct comparison in about 1-1/2 weeks.

My duty cycles are down about 10% across the board, I get no more than 40% (WOT) duty cycle. At highway speed I’m no more than 20% DC. I’m still using the 2.0 injectors…I think I’ll swap them out for 1.7’s and see what happens.

The main interesting point is the 2.0 always lurched a bit at 1500 to 2500rpm, it was a bit of a PITA in stop and go traffic, never at normal rpms. My manifold pressure was always all over the place. I have since installed this engine in a bug with 44mm Webers and an A-1 header and it now has none of the lurching.

The stock 1.8 has no lurching at all and the manifold pressure is now rock solid! Power is down from the 2.0, but that was expected. It is just as smooth as the old L-jet except for start up, as I have no cold enrichment.

Cam choice is very important!

When I get the 2.6 done (looks like late summer now) it will have a special MassIVe cam so we’ll see how it compares in the drivability department.

Jake and I will be comparing notes to figure out the best combo’s for his and my customers.
DNHunt
Mark

I've had the same problem with MAP readings on my 2.0. I'm beginning to think there is a lot of turbulence in there. I'm thinking with it being fairly flat and the tB on top that the air is just tumbling in.

Dave
Mark Henry
It could be part of the problem...when I first installed the SDS I took my MAP off of the side port and it was all over the place. I moved it back to stock and it cleaned up the signal quite a bit. But it was still not perfect.

With the 1.8 my MAP is now perfect.

I wish I had my IR TB's to try and see what's happening...maybe I'll try them on the 1.8 when I get them.
DNHunt
I've tried 4 different sitesand the stock place is best. Now I'm trying different sized restrictions in the vacuum line to the MAP sensor. The minimum seems to be an opening of about 1mm in diameter. Much smaller than that and the change in MAP is too slow and the engine leans out on acceleration. Larger than that and it bounces all over. So the trick is finding a fairly small resistriction and then compensating with acceleration enrichment.

I'm kind of going overboard here. The car drives great with no tinkering with the MAP signal but, the A/F ratio goes all over so I suspect it is leaving a little pwer on the table.

Dave
phantom914
QUOTE (DNHunt @ May 17 2005, 05:34 AM)
I've tried 4 different sitesand the stock place is best. Now I'm trying different sized restrictions in the vacuum line to the MAP sensor. The minimum seems to be an opening of about 1mm in diameter. Much smaller than that and the change in MAP is too slow and the engine leans out on acceleration. Larger than that and it bounces all over. So the trick is finding a fairly small resistriction and then compensating with acceleration enrichment.

I'm kind of going overboard here. The car drives great with no tinkering with the MAP signal but, the A/F ratio goes all over so I suspect it is leaving a little pwer on the table.

Dave

I'm not sure where you mount you're MAP sensor, but if it is not close to the manifold (meaning a long hose), another factor in response time is whether you place the restriction close to the sensor or to the manifold.


Andrew
DNHunt
The sensor is on the firewall so, the vacuum line runs about 18 inches. The restriction is at the manifold. I'm getting it closer. The process is put in a restriction then go datalog and look at the MAP sensor then try a different one.

Dave
Jake Raby
I'm doing 4 installs as we speak... one of them is on my 912E that I pulled the prototype Kit Carlson set up from.

I tested one engine yesterday with I/R and found great results against carbs... I now have my 16 channel logger up and going and can easily see differences in the entire engine.

I am creating an entire library of this testing as I go for the site. Stay tuned for some very cool results.

SDS electronics rock!
MecGen
Hi Guys
Map sensor HUH...

I have absolutly no experience with the MS. Is there not a tuning solution ? Can the MS be fined tuned ? I did some OBD1 Tuning but mostely OBD2. I have some serious help in this department.

Speed density tuning was always harder, due to the cam duration / valve / intake / setup. I always thought flat engines in general where not inclined to give a solid vacume with a cam.

I don't mind looking at your scripts, but they might be top secret ?

Something to look into, Jap cars use a "muffler" in the map hose, I am thinking Suzuki V6, I think I saw one yesterday. It acts as a buffer for vacum pulses, of course the PCM is tuned with this. It should help stabilize the vacume signal at min and may cure a hesitation at best.
Time to start studying..

Later
Joe

crash914
guys, I am running a small fuel filter in line. My MAP hose is tied into all 4 ports, then into a single hose, fuel filter, then in to the MS.....

I don't know if it made any difference, but it works...
Mueller
QUOTE
I'm not sure where you mount you're MAP sensor, but if it is not close to the manifold (meaning a long hose), another factor in response time is whether you place the restriction close to the sensor or to the manifold


Nothing wrong with a long hose as long as it is strong enough not to collapse or change shape with the changing signals. (and the diameter is small, you wouldn't want to run a 1" hose smile.gif )

The physical placement of the MAP sensor has no real bearing on the readings.....a hose up to several meters long can be used.
914efi
This is the main problem with the SDS system. It can't tune for cams with high idle MP. What is your MP at idle? If it is high (-5" to -10") and you tune that MP spot to idle well, it won't run well when it hits that MP value at a different engine condition. I notice exactly what you describe, under very low load you get a bucking(too rich). I have tuned it out pretty well by making compromises, but it will never be perfect without a 2D matrix like MS has. I have mine on a 911 2.2E. It will work better with stock cams that have a low idle MP and that never hit that value again during driving conditions, except maybe when you are off throttle.

You can get the idle TP set so that it will idle without looking at MP, but as soon as you move off that TP value, it will jump to whatever the MP is telling it to do and you will have another jumpy type of situation.

I think this is why you see a lot of SDS systems on cars with turbos, they have a more predictable MP vs fuel curve.
Mueller
QUOTE
have tuned it out pretty well by making compromises, but it will never be perfect without a 2D matrix like MS has.


should be 3D for the MS smile.gif

DNHunt
I'm intriqued by the muffler in the vacuum line. My MAP signal flucuates most the nearer I get to atmospheric pressure (WOT). I suppose this shouldn't be that much of a surprise. This also where pulsewidths are the largest.

MS and I suspect other speed density EFI's use an equation to figure pulse width that is the product a constant (injector flow rate, engine size) X a value from a lookup table (MAP vs RPM) X MAP + any enrichments. It is the second use of MAP in the equation that is the reason I'm hung up on getting a stable MAP signal. It is the only variable that is used twice.

At some point I'll have to compromise.

Dave
phantom914
QUOTE (Mueller @ May 17 2005, 08:56 AM)
QUOTE
I'm not sure where you mount you're MAP sensor, but if it is not close to the manifold (meaning a long hose), another factor in response time is whether you place the restriction close to the sensor or to the manifold


Nothing wrong with a long hose as long as it is strong enough not to collapse or change shape with the changing signals. (and the diameter is small, you wouldn't want to run a 1" hose smile.gif )

The physical placement of the MAP sensor has no real bearing on the readings.....a hose up to several meters long can be used.

My point was that with a long hose, the placement of the restriction could be another factor. For instance, with a restriction placed near the manifold, you can end up with a slight canister effect which will slow the response due to the volume of air between the restriction and MAP sensor. Maybe. I didn't mean that the steady state response of the sensor would be affected, just the transient response.



Andrew
lapuwali
QUOTE (914efi @ May 17 2005, 09:49 AM)
This is the main problem with the SDS system. It can't tune for cams with high idle MP. What is your MP at idle? If it is high (-5" to -10") and you tune that MP spot to idle well, it won't run well when it hits that MP value at a different engine condition. I notice exactly what you describe, under very low load you get a bucking(too rich). I have tuned it out pretty well by making compromises, but it will never be perfect without a 2D matrix like MS has. I have mine on a 911 2.2E. It will work better with stock cams that have a low idle MP and that never hit that value again during driving conditions, except maybe when you are off throttle.

You can get the idle TP set so that it will idle without looking at MP, but as soon as you move off that TP value, it will jump to whatever the MP is telling it to do and you will have another jumpy type of situation.

I think this is why you see a lot of SDS systems on cars with turbos, they have a more predictable MP vs fuel curve.

This really isn't a problem with SDS specifically, but with any speed-density system. Megasquirt, Haltech, even D-Jet all fall into the same boat. Most aftermarket systems are speed-density, in fact. If you're running a wild cam, you really have to use alpha-N (TPS instead of MAP as the load sensor), which SDS will do just fine (as will most other aftermarket systems).

With some systems (Link will do this, as will Electromotive), you can use both manifold pressure and TPS as load sensors, using TPS at low engine speeds where MAP tends to fluctuate unpredictably; and using MAP at higher engine speeds, where MAP tends to be fairly stable even with really hot cams. As far as I know, SDS doesn't do this, nor does MS (though there's some work in progress to change that).

In practice, alpha-N alone works OK on the street, if the MAP sensor is used for barometric correction. Bosch MFI is basically an alpha-N system with baro correction.
crash914
here is my data log....you can see the map sensor value..There is some noise, but not too much.
crash914
here is another picture of my map signal...
914efi
With respect to using the TPS only mode of SDS, I have done this but find it very annoying to drive in this mode. As soon as you begin to climb a slight hill where you don't move your foot on the gas, the engine bogs because it can't sense the load change unles you are moving your foot. This may be a personal matter, but I do like the MAP load sensing, even with its problems.

Crash, Could you explain the megasquirt data that is shown?
crash914
Yea, purple (magenta) line is the map signal. This is with 4 throttle bodies tied into one hose with an added fuel filter to try to smooth out the signal....I was showing that I only get around 3Kpa variation with steady state cruise. This is a 2666 with a Jake special cam...so the vacuum signal is pretty wild..

If anyone wants to help me with the rpm, DC1 spike, I won't mind.. wink.gif
MecGen
That depends ...what is DC1?
Sorry I am unfamiliar with the terminology (sp)
The RPM spike is usually in the secondary ignition
its way to fast to have actually revved that quick...
Later
Joe

beerchug.gif
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2024 Invision Power Services, Inc.