D-jet FI and >2.2L engine
-
lmcchesney
- Posts: 135
- Joined: Thu Nov 20, 2003 8:11 am
D-jet FI and >2.2L engine
I have read the forums/articles regarding using/adjusting the 2.0L D-jet FI system and use in >2.2L displacement. However, I still have not gotten a finished answer. I certainly respect Mr. Green's imput and Mr. Darling's experience with the "Nugent 914 Lite" articles. Here is my thought. Pistons 96mm yeilding a 72.4 surface area versus 94mm piston yeilding 69 surface area with a ratio of 1.05:1.0. Stroke of 80mm yeilding a combustion volume of 579cc vs 71mm stroke yeilding a volume of 490cc an 1.18:1 ratio. With an air:fuel ration of 14.7:1 that has the unmodified injection system injecting 33.3cc in the 2.0L engine vs 39.4cc in the 2.3L engine, or an increase of 18%. It would seem an increase in injector volume such as the mercedes injector or even an increase in fuel supply pressure could overcome the difference. If not, the measurment of A:F via an oxygen sensor in the combined exhaust and adjustment of the potentiometer between cylinder head temp and ECU could compensate. Where is my understanding wrong?
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
Your understanding is almost correct. The that the stock D-jet injectors are already the largest manufactured in Pintel size....anywhere. What fuel pressure did your calculations take into account? Fuel volume used is based upon injector open time x pressure. A simple pressure change of about 3 psi is all it takes to move from 2.0 to 2.2.
The problems you will run into...are not injector problems. They rarely if ever are. Not poking you in the ribs at all....just trying to explain. I call this "carburater" mentality. People think that minor volume changes in FI...are related to minor volume changes in carbed engines...resulting in the need to change "jets". Injectors and jets have nothing in common at all. The injectors will usually be outrun during smallvolumetric changes like you propose, only when the rpm range exceeds the cycle rate.
Here is the problem. If the pintle size goes up....you will get more fuel all the time. Lowering fuel pressure can only be done so much...to correct for too large of an injector, because cycle rate changes at low airflow rates (like at idle and part throttle) can only be so encremental on the low end. You will get a rich spot and lose gobs of HP when the injector is upsized too soon. If you do eventually find the need to go to a larger injector, at that point it would also be better to use a staged, after market variable fuel pressure regulator, so that you can control the injection volume at low rpm.
Generally, as I stated, most people upsize to injectors that can cycle faster. Of course...on D-jet, you only have so much ability to increase injection duration. Mostly this is done on aftermarket systems that have the ability to tailor the cycle rae better.
The reason you are getting incomplete answers...is because the tuning is involved. What exactly do you need to accomplish? The mathematical formulas will not answer your questions. They can get you close in manifold sizing, and give you a ballpark for what the exhaust gases should read when you are close, but most of the active tuning after that, with D-jet...is done by driving, taking notes and making single, calculated tuning adjustments. You also need proper CHT and oil temp guages to tune like this. Its not really hard...just tedious and you must get your mind into the engine to understand why it makes the decisions is does. Also, how the book says D-jet works (what books there are) and how it actually "behaves", are two totally different animals. Iwill try to answer what I can. Ray
The problems you will run into...are not injector problems. They rarely if ever are. Not poking you in the ribs at all....just trying to explain. I call this "carburater" mentality. People think that minor volume changes in FI...are related to minor volume changes in carbed engines...resulting in the need to change "jets". Injectors and jets have nothing in common at all. The injectors will usually be outrun during smallvolumetric changes like you propose, only when the rpm range exceeds the cycle rate.
Here is the problem. If the pintle size goes up....you will get more fuel all the time. Lowering fuel pressure can only be done so much...to correct for too large of an injector, because cycle rate changes at low airflow rates (like at idle and part throttle) can only be so encremental on the low end. You will get a rich spot and lose gobs of HP when the injector is upsized too soon. If you do eventually find the need to go to a larger injector, at that point it would also be better to use a staged, after market variable fuel pressure regulator, so that you can control the injection volume at low rpm.
Generally, as I stated, most people upsize to injectors that can cycle faster. Of course...on D-jet, you only have so much ability to increase injection duration. Mostly this is done on aftermarket systems that have the ability to tailor the cycle rae better.
The reason you are getting incomplete answers...is because the tuning is involved. What exactly do you need to accomplish? The mathematical formulas will not answer your questions. They can get you close in manifold sizing, and give you a ballpark for what the exhaust gases should read when you are close, but most of the active tuning after that, with D-jet...is done by driving, taking notes and making single, calculated tuning adjustments. You also need proper CHT and oil temp guages to tune like this. Its not really hard...just tedious and you must get your mind into the engine to understand why it makes the decisions is does. Also, how the book says D-jet works (what books there are) and how it actually "behaves", are two totally different animals. Iwill try to answer what I can. Ray
-
lmcchesney
- Posts: 135
- Joined: Thu Nov 20, 2003 8:11 am
Thanks
I do appreciate you input. I'm sure it is obvious, that I am not a porsche mechanic or otherwise, just a did well in physics and have an interest and a 914 project. Thus, if I can be sure I do understand your toughts, the 2.0L D-ject system can be adjusted to accomplish fuel management of the 2.3L, but will take fine tuning adjustments. The capability of is there though. I believe you had previously written that you would be involved in a >2.0L project. Has your experience been as predicted? are ther flat spots with the adjustments made? Do you believe with the D-jet FI you are able to realize the full HP of the engine?
Thanks ahead of time.
Thanks ahead of time.
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
I havn't been there yet with the person who is contemplating this. I have tuned D-jet on numerous engines marginally larger than 2.0 though.
Does the engine realize ultimate HP? Hard to say really. Yes and no.
Look at a stock 2.0 for a "for instance". If you are staying within the stock manifolds and plenum, there is no other injection that will really produce any better HP. We are just speaking of the injection system here. There may be some whose encremental adjustments offer slightly better throttle reponse, but I doubt it. Those I have spoken to who havee used other injection systems slaved to the stock manifolds, plenum and engine, who claim to see miraculous results.....have never had a properly tuned D-jet as a base line, to KNOW that a newer injection system actually was an improvement.
The advantages of newer aftermarket injection systems will actually be better adjustability through transitional driving conditions...meaning, that the sheer adjustability of SOME aftermarket systems may allow smoother running at varying weather conditions, odd temperature ranges...etc. But from what I have seen, most of the aftermarket systems are usually missing a bit from the metering package. They generally are MAP or TPS based....or both. Thats not as accurate as a mass airflow plate, or the MPS of D-jet. But...the aftermarket systems do have better ability to set baseline fuel curves and adjust ranges of temperature sensors. You can also make adjustments or ballasting on D-jet temp sensors.
Others will quote all the problems of D-jet...but lump them together, usually citing dependability as a performance issue. Thats not the case. The dependability problems with D-jet were the harness and connectors. Everything else was quite solid. Building a new harness is not that hard. Certainly less hard than some of the restoration efforts many of us put into basket case cars we buy
.
The factory level of adjustment on most D-jet was kind of crude in my opinion. It ran smooth, but was not really optimized for better HP or throttle response.
In short...the limitations that keep your baseline engine configuration from reaching what it could....are not within the electronic part of the system. They are within the manifolding, plenum, heads and cam. With those limitations, any other injection system would have the same limitations...and some of them would have more.
That being said, the manifolds of D-jet were not actually that bad. They are just a little dated...compared to what some of the modern injection systems in new cars come with. But they are certainly as good...and sometimes better than what comes with some of the aftermarket systems. D-jet can be tuned to give an easy 10%+ better HP and throttle response on a stck engine. With some porting, better compression, a better cam, better balance, and much better ignition...it can achieve upwards to 20-30% better than stock. But...then the limitations of the manifold and the ranges of enrichment will stop you from going further. This is not to say that thingscannot be added to correct these issues. But you may start approaching the point of dimenishing returns for that effort. Ray
Does the engine realize ultimate HP? Hard to say really. Yes and no.
Look at a stock 2.0 for a "for instance". If you are staying within the stock manifolds and plenum, there is no other injection that will really produce any better HP. We are just speaking of the injection system here. There may be some whose encremental adjustments offer slightly better throttle reponse, but I doubt it. Those I have spoken to who havee used other injection systems slaved to the stock manifolds, plenum and engine, who claim to see miraculous results.....have never had a properly tuned D-jet as a base line, to KNOW that a newer injection system actually was an improvement.
The advantages of newer aftermarket injection systems will actually be better adjustability through transitional driving conditions...meaning, that the sheer adjustability of SOME aftermarket systems may allow smoother running at varying weather conditions, odd temperature ranges...etc. But from what I have seen, most of the aftermarket systems are usually missing a bit from the metering package. They generally are MAP or TPS based....or both. Thats not as accurate as a mass airflow plate, or the MPS of D-jet. But...the aftermarket systems do have better ability to set baseline fuel curves and adjust ranges of temperature sensors. You can also make adjustments or ballasting on D-jet temp sensors.
Others will quote all the problems of D-jet...but lump them together, usually citing dependability as a performance issue. Thats not the case. The dependability problems with D-jet were the harness and connectors. Everything else was quite solid. Building a new harness is not that hard. Certainly less hard than some of the restoration efforts many of us put into basket case cars we buy
.The factory level of adjustment on most D-jet was kind of crude in my opinion. It ran smooth, but was not really optimized for better HP or throttle response.
In short...the limitations that keep your baseline engine configuration from reaching what it could....are not within the electronic part of the system. They are within the manifolding, plenum, heads and cam. With those limitations, any other injection system would have the same limitations...and some of them would have more.
That being said, the manifolds of D-jet were not actually that bad. They are just a little dated...compared to what some of the modern injection systems in new cars come with. But they are certainly as good...and sometimes better than what comes with some of the aftermarket systems. D-jet can be tuned to give an easy 10%+ better HP and throttle response on a stck engine. With some porting, better compression, a better cam, better balance, and much better ignition...it can achieve upwards to 20-30% better than stock. But...then the limitations of the manifold and the ranges of enrichment will stop you from going further. This is not to say that thingscannot be added to correct these issues. But you may start approaching the point of dimenishing returns for that effort. Ray
-
Bleyseng
- Posts: 994
- Joined: Fri Apr 21, 2000 12:01 am
Va914, for a stocker run the 96mm p&c's and a Webcam FI unit if you are willing to fool with the MPS. Some people have idle trouble some don't.
As for the 2.2L with Djet, the correct A/F mix is 13.7 to 1 on these engines not 14.7 to 1 at partload. I think you can make it work if you use a mild cam like the stock oem unit BUT you could have too much heat that only a external oil cooler could handle. A 2056cc motor requires a few adjustments to run well, mostly adjusting the MPS or running a 73 2.0l unit. IF you have access to a dyno and a Wavetek inductance meter you can adjust the MPS. I have done this to optimized the hp on my 2056cc to achieve 95hp at the rear wheels on my 914. Thats about 110-115hp at the flywheel.
As always you have to be concerned with the extra heat a larger motor will produce along with the correct A/F mix. The MPS especially the 73 unit has the adjustment range for the displacement increase.
In the long run, using Megasquirt might be a better bet due to its tuning abitlity witha computer vs the mechanical djet.
The runners and plenum have been sized to run a 2.2l motor from what others have figured out.
Geoff
As for the 2.2L with Djet, the correct A/F mix is 13.7 to 1 on these engines not 14.7 to 1 at partload. I think you can make it work if you use a mild cam like the stock oem unit BUT you could have too much heat that only a external oil cooler could handle. A 2056cc motor requires a few adjustments to run well, mostly adjusting the MPS or running a 73 2.0l unit. IF you have access to a dyno and a Wavetek inductance meter you can adjust the MPS. I have done this to optimized the hp on my 2056cc to achieve 95hp at the rear wheels on my 914. Thats about 110-115hp at the flywheel.
As always you have to be concerned with the extra heat a larger motor will produce along with the correct A/F mix. The MPS especially the 73 unit has the adjustment range for the displacement increase.
In the long run, using Megasquirt might be a better bet due to its tuning abitlity witha computer vs the mechanical djet.
The runners and plenum have been sized to run a 2.2l motor from what others have figured out.
Geoff
-
Dave_Darling
- Posts: 2534
- Joined: Thu Apr 27, 2000 12:01 am
Now that's just a silly statement. MAP == MPS. Or rather, it is a more reliable and cheaper equivalent of the MPS. They both measure the same thing, but I think the solid-state device does it better. For instance, have you ever tried running a D-jet MPS in a vertical orientation? Go over a bump and the engine hiccups significantly. The iron core moves, which changes the mixture. I bet the same thing is happening, on a much smaller scale, with the sensor mounted normally.raygreenwood wrote:...But from what I have seen, most of the aftermarket systems are usually missing a bit from the metering package. They generally are MAP or TPS based....or both. Thats not as accurate as a mass airflow plate, or the MPS of D-jet.
There is no such thing as a "mass airflow plate". The moving-plate (CIS or "barn door" air flow meter) measures air volume, not air mass. The FI has to fiddle and take other measurements and make some guesses and infer what the air mass is. The hot-wire or hot-film anemometer air flow meters are much more accurate at determining the mass of incoming air.
--DD
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
Dave, its not really a silly statement at all. The MPS and Map are not the same. The MAP has no provision for sensing ambient air pressure differential. It also does not have the range of assisted enrichment signal of the copper plate. The MAP is also not adjustable in general.
Though the metering method of the map is more accurate, it is not as sensitive as the MPS. Dependability of the map in 1:1 terms, is about the same. They last about 80K miles by my experience.
The hiccup problem is simply a matter of adjustment. 2-3 ncm more tension on the mainspring, and closer adjustment of the outer stop of the copper plate, will get rid of that problem. Also, you must correct the main fuel mixture adjustment for that extra tension. Also, correcting the radial alignment of the metering pin is a must when seting these up. They can drag on the coil armature for the same reasons when in the horizontal position. Checking the tightness of the ceramic bearings is also a plus. I replace mine with Torlon. It allows more tightness of the leaf spring, with even less actual friction on the rodholder. Besides....and I'm not try ing to be an *sshole...but if the settings on your MPS cause it to have enertia problems in the vertical position...don't install it that way! Thats like complaining that yor windshield washer tank leaks when you install it upside down!
The mass airflow plate comment...was a misnomer on my part. Apologies. I was refering to the "barn door" in the L-jet. Yes...it is a mass air flow sensor in several respects. It is the mass of the air...not vacume, that moves the door...of a set cross section, against a progressive rate spring of known calibration. It is NOT a mass air sensor, in the sense that the "volume" of air comin in, is not fully corrected for temperature by the ambient air temp sensor. It does not have enough range in L-jet. The mass of the incoming air colume is "assumed" based around a factory mean temperature setting used as the baseline. So... Is it a pure mass air flow sensor? No. I am in agreement there. The hot wire or heated film systems are truely the best....as long as they have an ambient air temp sensor that can cross check them. By the way...in the same inverse respect, hotwire systems are not true mass airflow sensors either, due to the fact that they make their fecision based upon, temperature and velocity only. The MAP cross checks depression...but that is a load sensor. It has no method of crosschecking external atmospheric ressure with temperature and velocity in a channel of known volume...which is a true mass air flow meter. Only factory systems have this ability to my knowledge....I'm not aware of any after market systems thats do this.
By the way, calculating air mass is a function of calculating its density. Less dense air has less mass. The reason the barn door has mass air flow measuring qualities, is that it was designed with the idea that a certain amount of air mass within the specific volume (that volume being constant within the unchanging cross section of the intake) ..would be required to move the wighted/sprung flap X amount of degrees. Of course air volume and density...and therefore, air mass....vary with temperature. Hence the reason that its only measuring true mass at one target temperature (the same as the one it was calibrated to at the factory. Unless, the stock ambientair temp sensor is doing more through the range than it appears to.
Bleyseng: heat should not be a real problem...with D-jet anymore than any other engine. If its running too hot, its an efficiency issue. Check overall advance. Keep the compression up. It actually lowers the heat generated...and keeps the advance need down. Now...if its not over heating, but it is keeping the CHT in the wrong range, ballast the CHT. Its also not very hard with a clean sub-miniature relay, to allow ballasted and unballasted readings from the same sensor, according to ambient. Sounds like you have done very well in your tuning of D-jet. The range of enrichment...or the encrements of it, are the same in this system, no matter what the displacement (short of out running the manifold capability). If enrichment is not enough through part throttle, because of displacement changes, even when the vacume signature has basically been maintained, its a fuel volume adjustment that needs to be made. Usually a bump upward in fuel pressure.
One of things I have ranted about before, is the propensity of people to only tune D-jet along factory guidlines....which never encompassed ANY performance tuning. Baselining with wavetek meter is necessary when starting from scratch on an unkown system....and when nothing about the MPS original settings is known. Outside of that...it has to be done on the road. The dyno can get you close, but it is not the same as the road. It does not take into account (fully) gear ratios, driving habits, shift points , enertia shifts of the body, cornering with power on and off,hills, wind, variable temperatures and weather etc. etc. Ray
Though the metering method of the map is more accurate, it is not as sensitive as the MPS. Dependability of the map in 1:1 terms, is about the same. They last about 80K miles by my experience.
The hiccup problem is simply a matter of adjustment. 2-3 ncm more tension on the mainspring, and closer adjustment of the outer stop of the copper plate, will get rid of that problem. Also, you must correct the main fuel mixture adjustment for that extra tension. Also, correcting the radial alignment of the metering pin is a must when seting these up. They can drag on the coil armature for the same reasons when in the horizontal position. Checking the tightness of the ceramic bearings is also a plus. I replace mine with Torlon. It allows more tightness of the leaf spring, with even less actual friction on the rodholder. Besides....and I'm not try ing to be an *sshole...but if the settings on your MPS cause it to have enertia problems in the vertical position...don't install it that way! Thats like complaining that yor windshield washer tank leaks when you install it upside down!
The mass airflow plate comment...was a misnomer on my part. Apologies. I was refering to the "barn door" in the L-jet. Yes...it is a mass air flow sensor in several respects. It is the mass of the air...not vacume, that moves the door...of a set cross section, against a progressive rate spring of known calibration. It is NOT a mass air sensor, in the sense that the "volume" of air comin in, is not fully corrected for temperature by the ambient air temp sensor. It does not have enough range in L-jet. The mass of the incoming air colume is "assumed" based around a factory mean temperature setting used as the baseline. So... Is it a pure mass air flow sensor? No. I am in agreement there. The hot wire or heated film systems are truely the best....as long as they have an ambient air temp sensor that can cross check them. By the way...in the same inverse respect, hotwire systems are not true mass airflow sensors either, due to the fact that they make their fecision based upon, temperature and velocity only. The MAP cross checks depression...but that is a load sensor. It has no method of crosschecking external atmospheric ressure with temperature and velocity in a channel of known volume...which is a true mass air flow meter. Only factory systems have this ability to my knowledge....I'm not aware of any after market systems thats do this.
By the way, calculating air mass is a function of calculating its density. Less dense air has less mass. The reason the barn door has mass air flow measuring qualities, is that it was designed with the idea that a certain amount of air mass within the specific volume (that volume being constant within the unchanging cross section of the intake) ..would be required to move the wighted/sprung flap X amount of degrees. Of course air volume and density...and therefore, air mass....vary with temperature. Hence the reason that its only measuring true mass at one target temperature (the same as the one it was calibrated to at the factory. Unless, the stock ambientair temp sensor is doing more through the range than it appears to.
Bleyseng: heat should not be a real problem...with D-jet anymore than any other engine. If its running too hot, its an efficiency issue. Check overall advance. Keep the compression up. It actually lowers the heat generated...and keeps the advance need down. Now...if its not over heating, but it is keeping the CHT in the wrong range, ballast the CHT. Its also not very hard with a clean sub-miniature relay, to allow ballasted and unballasted readings from the same sensor, according to ambient. Sounds like you have done very well in your tuning of D-jet. The range of enrichment...or the encrements of it, are the same in this system, no matter what the displacement (short of out running the manifold capability). If enrichment is not enough through part throttle, because of displacement changes, even when the vacume signature has basically been maintained, its a fuel volume adjustment that needs to be made. Usually a bump upward in fuel pressure.
One of things I have ranted about before, is the propensity of people to only tune D-jet along factory guidlines....which never encompassed ANY performance tuning. Baselining with wavetek meter is necessary when starting from scratch on an unkown system....and when nothing about the MPS original settings is known. Outside of that...it has to be done on the road. The dyno can get you close, but it is not the same as the road. It does not take into account (fully) gear ratios, driving habits, shift points , enertia shifts of the body, cornering with power on and off,hills, wind, variable temperatures and weather etc. etc. Ray
-
Dave_Darling
- Posts: 2534
- Joined: Thu Apr 27, 2000 12:01 am
No, it isn't. It's a matter of the design of the sensor. The iron core "bounces" when the sensor is moved rapidly up and down. Changing the springs and such will not change that fact. The iron core moving will change the mixture. That is another fact--that is, after all, exactly what it is supposed to do. Try it on your own car after you get it perfectly adjusted. Better yet, try dropping the MPS vertically from about 2" onto the ground or something else solid. A hiccup will result.The hiccup problem is simply a matter of adjustment.
I do not run my MPS vertically, I was using that to show what I think may very well be happening even when the MPS is mounted correctly. That is, the measurement instrument has inertia and the measurement itself will be adversely affected by movement. Such as the vibration that goes through the whole car's chassis when the engine is running and the car is moving.
You don't need the adjustment built into the sensor. In fact, I think it's better not to. The sensor should simply sense, that's it. You can (and do!) write into the code provisions for extra enrichment at WOT. You adjust in software, if you have a decent aftermarket setup.The MAP has no provision for sensing ambient air pressure differential. It also does not have the range of assisted enrichment signal of the copper plate. The MAP is also not adjustable in general.
--DD[/code]
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
Dave I concurr that aftermarket systems can work around the limitations of MAP. Wasn't saying D-jet was better by any means...simply that MAP and the MPS are not one and the same. The map performs a single function. As you noted, it simply "senses". The MPS also has to perform "corrections", for atmospheric differences at part throttle...mainly becasue there are no other sensors in the system (Like an 02 sensor) that can help in those situations. MPS is also a MAP sensor..but does not measure absolute pressure. Its differential pressure.
Where programming cannot work around in some of the aftermarket systems, is the gap in sensitivity MAP has...most especially in twin butterfly systems. MAP and TPS combined cannot account for some fluctuations at part throttle...because they have nothing to measure them with. You can and will have conflicting high vacume readings from a mostly closed butterfly at high rpm from MAP. The TPS...most of them...do not have the encremental steps to tell the system any different...so you get lean spots. Yes...a system with 02...and good programming can get around that. This is just an opinion...but from everythig I have observed....the programming and trial and error...is no more or less intense than what it takes to make D-jet perform better...but thats just my opinion. It is simpler though...just to sit down in front of the PC...no argument there.
Yes...the enertia problems with the MPS can be gotten around. I did that quite some time ago. If you take a look at Brad Anders site, he makes note of the limit that the armature (rod) may penetrate into the coil. It can only go so far, before its outside the parameters of operation of the system...in either directon. That position is easy to find. Mine has an adjustable stop for that position ( I installed it). The "hiccup" you speak of...is not just a minor change in fuel mixture. It is the rod...actually moving past the limit point. That cannot happen with a stop screw. Your suspicion that it happens in the horizontal...is actually almost as much of a problem as vertical . I dropped in a post some time back (spring I think?) about a 3 point bearing arond the iron rod ( I made one). Its almost friction free...and prevent axial mis-alignment. Its actually what the creamic whisker bearings on the inside were designed to help prevent. Unfortunately...they have little tension on them...and they are located too far forward toward one end to be of any real help.
The flat locating leaf spring attached to end of the armature...are also made to help keep alignment and damp movement on the rod. Unfortunately...they also are a detriment to part throttle enrichment sustainment. They have too much spring...in short. They have about 5-7 ncm of tension between them. By putting on a stop screw for the rod at the back end (mainspring end) and increasing the main spring pressure about half that much....and removing the rear leaf (the one closest to the main spring)...The MPS is now more sensitive to part throttle vacume...but has just a hair more tension to allow damping. Yes...I can make it hiccup with a "drop" to a hard surface...but you will not experience static shocks like that while driving...unless you drop off a cliff. The extra mainspring tension (adjusted with shims)...cause initial fuel settings to be richer...across the board...because it drives the rod farther out out of the coil. A simple fuel pressure bump reset all of that. Because they are compressed under more spring load now...the aneroids respond quicker when throttle is opened. They move at a faster rate.
Like I said...I don't care much for factory tuning or literature. Its not that they don't think any of this possible...they don't even mention it. Its 23+ years of fiddling and experimenting....to get to this stage.
I find D-jet fun to run...not because it superior in any way...but because its a challenge to get it to do more than it was layed out to be. I like the originality of it. Within the limits of its manifolding...its capable of quite a bit. Its also not for everyone...because the tuning can take a bit of effort.
Yes..I know you used the example of standing the MPS on end as an illustration....I was just having fun. It took me 20 minutes to come up with that snappy windshield washer come-back ....hee hee
Oh....not sure if I'm in sunnyvale on the 1st or Austin.....Its still up in the air. Ray
Where programming cannot work around in some of the aftermarket systems, is the gap in sensitivity MAP has...most especially in twin butterfly systems. MAP and TPS combined cannot account for some fluctuations at part throttle...because they have nothing to measure them with. You can and will have conflicting high vacume readings from a mostly closed butterfly at high rpm from MAP. The TPS...most of them...do not have the encremental steps to tell the system any different...so you get lean spots. Yes...a system with 02...and good programming can get around that. This is just an opinion...but from everythig I have observed....the programming and trial and error...is no more or less intense than what it takes to make D-jet perform better...but thats just my opinion. It is simpler though...just to sit down in front of the PC...no argument there.
Yes...the enertia problems with the MPS can be gotten around. I did that quite some time ago. If you take a look at Brad Anders site, he makes note of the limit that the armature (rod) may penetrate into the coil. It can only go so far, before its outside the parameters of operation of the system...in either directon. That position is easy to find. Mine has an adjustable stop for that position ( I installed it). The "hiccup" you speak of...is not just a minor change in fuel mixture. It is the rod...actually moving past the limit point. That cannot happen with a stop screw. Your suspicion that it happens in the horizontal...is actually almost as much of a problem as vertical . I dropped in a post some time back (spring I think?) about a 3 point bearing arond the iron rod ( I made one). Its almost friction free...and prevent axial mis-alignment. Its actually what the creamic whisker bearings on the inside were designed to help prevent. Unfortunately...they have little tension on them...and they are located too far forward toward one end to be of any real help.
The flat locating leaf spring attached to end of the armature...are also made to help keep alignment and damp movement on the rod. Unfortunately...they also are a detriment to part throttle enrichment sustainment. They have too much spring...in short. They have about 5-7 ncm of tension between them. By putting on a stop screw for the rod at the back end (mainspring end) and increasing the main spring pressure about half that much....and removing the rear leaf (the one closest to the main spring)...The MPS is now more sensitive to part throttle vacume...but has just a hair more tension to allow damping. Yes...I can make it hiccup with a "drop" to a hard surface...but you will not experience static shocks like that while driving...unless you drop off a cliff. The extra mainspring tension (adjusted with shims)...cause initial fuel settings to be richer...across the board...because it drives the rod farther out out of the coil. A simple fuel pressure bump reset all of that. Because they are compressed under more spring load now...the aneroids respond quicker when throttle is opened. They move at a faster rate.
Like I said...I don't care much for factory tuning or literature. Its not that they don't think any of this possible...they don't even mention it. Its 23+ years of fiddling and experimenting....to get to this stage.
I find D-jet fun to run...not because it superior in any way...but because its a challenge to get it to do more than it was layed out to be. I like the originality of it. Within the limits of its manifolding...its capable of quite a bit. Its also not for everyone...because the tuning can take a bit of effort.
Yes..I know you used the example of standing the MPS on end as an illustration....I was just having fun. It took me 20 minutes to come up with that snappy windshield washer come-back
....hee hee Oh....not sure if I'm in sunnyvale on the 1st or Austin.....Its still up in the air. Ray
-
Dave_Darling
- Posts: 2534
- Joined: Thu Apr 27, 2000 12:01 am
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
-
lmcchesney
- Posts: 135
- Joined: Thu Nov 20, 2003 8:11 am
-
Dave_Darling
- Posts: 2534
- Joined: Thu Apr 27, 2000 12:01 am
The problem with simply viewing the engine and related systems as an "air pump" is that you wind up forgetting the evil details while getting a good grasp on the big picture.
The main "evil details" in this case being the stock manifolding and the stock fuel metering system. D-jet deals very poorly with "hot" cams. The intake does not develop enough vacuum at idle (in particular) and so the FI "thinks" the engine is running under significant load. So you get a lot of fuel, leading to a very rich mixture.
Messing with the various adjustments can let you "tune around" that to some extent, but there is only so much that you can do. Most of this kind of tuning you do will compromise some other part of the RPM/load range, as well.
In addition, the aggressive cams and large valves will change the volumetric efficiency curve of the motor. This will change the actual amount of air going into the motor at the various vacuum levels, which means that the assumptions built into the FI are no longer necessarily correct. Again, you can tune around this to some extent, but only so much.
How big can you go? Well, Ray and Geoff have both been successful with the Web #73 grind. It's still pretty mild, but is more aggressive than any of the stock grinds. I'm not sure if anyone has gotten D-jet to work with that cam and larger-than-stock valves. I could see it going either way, frankly...
--DD
The main "evil details" in this case being the stock manifolding and the stock fuel metering system. D-jet deals very poorly with "hot" cams. The intake does not develop enough vacuum at idle (in particular) and so the FI "thinks" the engine is running under significant load. So you get a lot of fuel, leading to a very rich mixture.
Messing with the various adjustments can let you "tune around" that to some extent, but there is only so much that you can do. Most of this kind of tuning you do will compromise some other part of the RPM/load range, as well.
In addition, the aggressive cams and large valves will change the volumetric efficiency curve of the motor. This will change the actual amount of air going into the motor at the various vacuum levels, which means that the assumptions built into the FI are no longer necessarily correct. Again, you can tune around this to some extent, but only so much.
How big can you go? Well, Ray and Geoff have both been successful with the Web #73 grind. It's still pretty mild, but is more aggressive than any of the stock grinds. I'm not sure if anyone has gotten D-jet to work with that cam and larger-than-stock valves. I could see it going either way, frankly...
--DD
-
raygreenwood
- Posts: 11907
- Joined: Wed Jan 22, 2003 12:01 am
I think Dave is hitting it on the head for a handful of reasons. This is just from a lot of tuning. Bear with me now. I will try not to get too philosophical.
VW/Bosch, set up the manifold volume versus engine size geometry to make the MPS sensitive...within a certain range of fluctuation...so it would match the range of electronic fuel mixture adjustability that the ECU was designed with. Bear in mind that the development time spent AFTER this system went into production, was pretty much arrested...as within 7 years, (probably only 4 while L-jet was starting to be developed)...D-jet was superceded by L-jet...for a lot of reasons. So there are a whole handful oflittle things that seem illogical to us...as there is a god deal of adjustability built into the system...and it can...somewhat..use larger valves and a few other parts. Its a good chance that they could have done more at the factory, but just never got around to it before it was superceded. In short, the volumetric geometry should be srespected to a large degree..or else you may not be able to adjust for it.
Some for instances...that seem to work well:
The 1.7 has these volumes:
cyl. vol.: 419.7cc
runner vol: 330cc
plenum: 1 liter
stock valve diameter (minus stem): 29.6mm
runner diameter: 33.375mm
So the runner volume is about 80% of the cylinder.
The plenum is is just at 60% of the displacement. Its a little more than two complete cylinder fills worth, and about 3 runner fills worth.
In your mind...like stop action photagraphy...you can see that the movement of air from the TB through the plenum, through the runner, through an open valve...will be...moving...under tension...like a cord. That "taughtness" or tension of that cord...is modulated by restriction at top or bottom. Wether cord gets stretched too tight in between...is a measure of wether volume in between the valve and the plenum is correct. The MPS meters off that volume. Add to this that depending upon the RPM, 2 or more cylinders may be in variuos stages of the same process.
Overlap of intake to exhasut valve...orwhen they are both slightly open at the same time....can help to stop a dead slamming of the air column...or "cord" I desribed above.
Enough of that. The stock valve cross section was a little small for the runner diameter. It was a slight restriction. But that slight restriction kept runner pressure up ..and buffered shocks in the MPS. It kept harmonics down. It also kept adjustability of the MPS...a little twitchy. Timing affected it a lot, because better or worse ignition timing, affected the heat and velocity of the exhaust charge, which affected the level of draw through on valve overlap. You can see this, and the affect it had on the MPS, by watching the fuel pressure guage..and advancing the ignition timing. It commences to swing wildly...as the vacume flucuation begins a rapid harmonic over-extension of the mps armature...giving rapid over-enrichment...and subsequent over-leaning by rapidly fluctating injector cycling.
This is why the web # 73 runs quite well with stock valve sizing. It adds just a hair more lift and a bit more duration, effectively making the stock small valve act like a larger valve. Of course it drops velocity a hair...so web went back to the early 411 camshaft valve timing...which placed the intake valve opening point a little closer to the exhaust closing point. It had more overlap. Velocity went up. Nice design.
Duration ismore important to the D-jet than lift. Its like putting your thunb over the end of the hose. Your thumb being the valve. Lift it too little...you get a restriction, velocity is high...but volume low. Lift it too high...volume is great...but, velocity is way low. So put it in the middle...and make it the right diameter...and flow is modulated.
Longer duration with valve diameters whose cross section area is in calibration to the runner size, maintains velocity, limits the type of turbulence that can shear fuel from air...and helps to insure that fuel/air is entering for as much of the induction stroke as possible. The 2.0 has a stroke about 7% longer than the 1.7. I think the web #73 is about 10% longer duration than stock. That makes sense. The web #73 runs better on the 1.7 because the extra lift helps make up for the small valve diameter. It runs even better I think, on the 2.0. Its extra duration more closely matches the extra stroke. If you are keeping induction volumes within ratio to the engine volumes...I would think you could move your duration on your cam, with analmost 1:1 improvment with your stroke. Of course, at some point, because you only have 360 degrees to play with, that will start to affect lift and overlap.
I think 48 mm is too big for the intake valve. Then again, that will depend upon displacment...and that may drive you to make custom manifolds and plenums that are in ratio to the displacment. If it generates a runner whose cross section is similar to that valve...it may work out.
I know this didn't answer your question...but there is very little info available to say what "math", VW used in setting up the volumetric ratios that operate within the parameters of the MPS. Sorry for the length. I hope this at least helps the thought process. Incidentally...the volumes of the 1.8 and the 2.0....are pretty much in line, though the 1.8 is L-jet. The 2.0 uses larger diameter runners, larger volume plenum...and a larger TB. The valve on the 2.0 is 42mm-7.9mm for stem=34.1mm. The runner...I think is 2-3mm larger?..That puts it just at 35-36 mm. TB's seem to follow closely with a WOT sizing of about 46mm on a 2.0 ( not exactly sure of that meazurement but its close)...minus the shaft of 8.35mm= 37.75mm. That ratio holds up pretty close...and thats not counting cross section of the idle circuit bypass. So TB is about intake valve diameter. A 1:1 relationship through the manifold at WOT. Ray
VW/Bosch, set up the manifold volume versus engine size geometry to make the MPS sensitive...within a certain range of fluctuation...so it would match the range of electronic fuel mixture adjustability that the ECU was designed with. Bear in mind that the development time spent AFTER this system went into production, was pretty much arrested...as within 7 years, (probably only 4 while L-jet was starting to be developed)...D-jet was superceded by L-jet...for a lot of reasons. So there are a whole handful oflittle things that seem illogical to us...as there is a god deal of adjustability built into the system...and it can...somewhat..use larger valves and a few other parts. Its a good chance that they could have done more at the factory, but just never got around to it before it was superceded. In short, the volumetric geometry should be srespected to a large degree..or else you may not be able to adjust for it.
Some for instances...that seem to work well:
The 1.7 has these volumes:
cyl. vol.: 419.7cc
runner vol: 330cc
plenum: 1 liter
stock valve diameter (minus stem): 29.6mm
runner diameter: 33.375mm
So the runner volume is about 80% of the cylinder.
The plenum is is just at 60% of the displacement. Its a little more than two complete cylinder fills worth, and about 3 runner fills worth.
In your mind...like stop action photagraphy...you can see that the movement of air from the TB through the plenum, through the runner, through an open valve...will be...moving...under tension...like a cord. That "taughtness" or tension of that cord...is modulated by restriction at top or bottom. Wether cord gets stretched too tight in between...is a measure of wether volume in between the valve and the plenum is correct. The MPS meters off that volume. Add to this that depending upon the RPM, 2 or more cylinders may be in variuos stages of the same process.
Overlap of intake to exhasut valve...orwhen they are both slightly open at the same time....can help to stop a dead slamming of the air column...or "cord" I desribed above.
Enough of that. The stock valve cross section was a little small for the runner diameter. It was a slight restriction. But that slight restriction kept runner pressure up ..and buffered shocks in the MPS. It kept harmonics down. It also kept adjustability of the MPS...a little twitchy. Timing affected it a lot, because better or worse ignition timing, affected the heat and velocity of the exhaust charge, which affected the level of draw through on valve overlap. You can see this, and the affect it had on the MPS, by watching the fuel pressure guage..and advancing the ignition timing. It commences to swing wildly...as the vacume flucuation begins a rapid harmonic over-extension of the mps armature...giving rapid over-enrichment...and subsequent over-leaning by rapidly fluctating injector cycling.
This is why the web # 73 runs quite well with stock valve sizing. It adds just a hair more lift and a bit more duration, effectively making the stock small valve act like a larger valve. Of course it drops velocity a hair...so web went back to the early 411 camshaft valve timing...which placed the intake valve opening point a little closer to the exhaust closing point. It had more overlap. Velocity went up. Nice design.
Duration ismore important to the D-jet than lift. Its like putting your thunb over the end of the hose. Your thumb being the valve. Lift it too little...you get a restriction, velocity is high...but volume low. Lift it too high...volume is great...but, velocity is way low. So put it in the middle...and make it the right diameter...and flow is modulated.
Longer duration with valve diameters whose cross section area is in calibration to the runner size, maintains velocity, limits the type of turbulence that can shear fuel from air...and helps to insure that fuel/air is entering for as much of the induction stroke as possible. The 2.0 has a stroke about 7% longer than the 1.7. I think the web #73 is about 10% longer duration than stock. That makes sense. The web #73 runs better on the 1.7 because the extra lift helps make up for the small valve diameter. It runs even better I think, on the 2.0. Its extra duration more closely matches the extra stroke. If you are keeping induction volumes within ratio to the engine volumes...I would think you could move your duration on your cam, with analmost 1:1 improvment with your stroke. Of course, at some point, because you only have 360 degrees to play with, that will start to affect lift and overlap.
I think 48 mm is too big for the intake valve. Then again, that will depend upon displacment...and that may drive you to make custom manifolds and plenums that are in ratio to the displacment. If it generates a runner whose cross section is similar to that valve...it may work out.
I know this didn't answer your question...but there is very little info available to say what "math", VW used in setting up the volumetric ratios that operate within the parameters of the MPS. Sorry for the length. I hope this at least helps the thought process. Incidentally...the volumes of the 1.8 and the 2.0....are pretty much in line, though the 1.8 is L-jet. The 2.0 uses larger diameter runners, larger volume plenum...and a larger TB. The valve on the 2.0 is 42mm-7.9mm for stem=34.1mm. The runner...I think is 2-3mm larger?..That puts it just at 35-36 mm. TB's seem to follow closely with a WOT sizing of about 46mm on a 2.0 ( not exactly sure of that meazurement but its close)...minus the shaft of 8.35mm= 37.75mm. That ratio holds up pretty close...and thats not counting cross section of the idle circuit bypass. So TB is about intake valve diameter. A 1:1 relationship through the manifold at WOT. Ray