do the cones of the down pipes ,were they connect the manifolds, have any purpose or was that for easy construction ? The cones go further beyond the weld , they are constructed as a 36 mm injector in a 41 mm tube.
I have seen a thread of someone grinding that away but can't find it anymore.
I'm wondering , as we have a flat HP curve above 4700 rpm with FI , if those are a possible bottleneck ?
We build a new exhaust with bigger tubes but not yet the 4 down pipes.
so i think of building 4 downpipes in 50 mm or at least grind away those cones overshoot inside to keep the 41 mm diameter as in the headers.
My guess is that the cones also allow the exhaust to move around a bit without bending or breaking when the engine moves. This is also in conjunction with the springs and the flanges having a bit of give.
If you are going down the exhaust road, tubular headers outperform log exhaust manifolds every day. They introduce there own sets of problems however.
The "bottleneck" for the power is in the 90 degree intake right before the cylinder in the heads - no way around this without modifying the intake and the heads. If you want increased power, you have to go FI.
'93 850Ci - Mineralweiß Metallic
2001 740iL - Titansilber
ALPINA B7 -Alpinweiß III
...the price of cool ain't cheap!
perhaps i was not clear .... i have FI ( supercharger) ) , 0.5 bar at 6000 rpm. But HP curve stays flat after 4700 rpm at 410 HP... up to 6000 rpm boost is still raising and A/F stays good… so something holds it flat ? I tought ignition is retarded or the exhaust is at his max flow ?
The construction of cones at the manifolds is indeed to allow a bit of rotation angle but that has nothing to do with the cone going inside the 41 mm downpipe. That overshoot does nothing except reducing the gas flow section from diameter 41 mm to 36 mm.
just drilled out that cone in a down pipe with a holesaw : you get 41 mm instead of 36 mm ! that's a improvement of 30 % flow section…. can be useful for FI applications…. i think .
see pic : i just "cleaned" a bit the oem overshoot cone to make it visible , next to the drilled one.
SAM_2625.jpg
PS the outlet of the cast manifolds is also 41 mm
Last edited by Belgiumbarry; 01-03-2019 at 12:31 PM.
what cams are you running? That would be what I'd look at to solve the flat spot
'91 Dinan 860 Stage III (new 6L engine)
'91 Dinan 850 TT stage III (brand new engine) 21st Century Tech meets 18th Century Dinan...
'91 850i 6sp (mint) (sold)
'90 Dinan 750iL TT stage III (Guido - The Beast)
'94 850 CSi The Detroit Auto Show car (restored to factory perfect) (sold)
'96 850Ci, The George Carlin car
''73 3.0 csi, '08 535i, '03 X5 4.6is
...and a few other non BMW cars
the oem cams i suppose.... at least what was in it when i bought the car.
Pitty i'm 10 years ? to late with my FI project.... i know there were/are others, but no replys anymore …. did they change cams ? down pipes ? …. or happy with the 400...410 HP it gives ( and not forget 600 NM torque) ? it sure already drives quite faster than oem but not 450 HP as "we" expected. I have the malchance to have a son with a tuning shop ( modern cars) but there is a dyno i can use whenever i want.... so i know what i have under the hood…
whatever it is , there is sure more to gain when we can postpone that dip / flat to higher rpm's…..
you running an intercooler? What is your charged air temp at those points?
'91 Dinan 860 Stage III (new 6L engine)
'91 Dinan 850 TT stage III (brand new engine) 21st Century Tech meets 18th Century Dinan...
'91 850i 6sp (mint) (sold)
'90 Dinan 750iL TT stage III (Guido - The Beast)
'94 850 CSi The Detroit Auto Show car (restored to factory perfect) (sold)
'96 850Ci, The George Carlin car
''73 3.0 csi, '08 535i, '03 X5 4.6is
...and a few other non BMW cars
yes , with A/A intercooler. We even made a special design , one inlet , dual symmetrical divided cores/outlets , so to have a equal load on the 2 banks .
I didn't check air temp inlet so far... that's why i asked in previous thread what the temp sensors in the inlet manifolds can do.... but it seems according replys nothing ignition related ,( right ?? ) only for A/F mix calculation , ….and in my case, A/F stays good.
I can measure it next dyno run , as i know now the Ohm's they give according temp from the sensors.
i just wonder if anybody got really 450 HP out of a SC 850i with the oem "downpipes"... not to mention the complete exhaust line. The turbo guys won't have that problem as they must have a new custom downpipe(s) and sure don't use anymore the oem ones.
PS i use the term "downpipe" as for the tube after the turbo , in the SC case i ment the tubes from manifolds to the flanges down from where the horizontal exhaust line starts.
i removed today my downpipes … ( the 4 i drilled are spares) … to show me a even greater mysterie : 3 had the overshoot cone to 36 mm , 1 was a butt weld , keeping the 41 mm ! They all look the same age , don't think one was ever swapped. So in factory they could make them full flow , and more... why different ones on a single engine ??
To be correct : the butt weld has a overlay weld material inside….. so it is 39 mm i think. My drilled ones are better : 41 mm smooth all the way .
Removing those bolts was a bit difficult…. keeping the nut underside fix and turning above till all felt on the ground .... Reassemble is still a mystery for me how to put spring/nut on the bolts …. need somebody with baby hands and a long arm.
Removing that material should help with flow dynamics, keep us posted.
Regards,
Brian
Cave Creek, AZ
This how I and my kid did it, and you have to have a partner, I did not pull exhaust system so I could not get the second pipe out.
Make sure you have you heat shield on top already pulled.
Front exhaust pipe (1-3) cannot be removed, and you pull out the other one (4-6). Reassembly requires the front pipe to be in the perfect not explained here position and then you get the other one in and twist and maneuver until they both are ready (PIA). Key problem is the bolts and springs. Exhaust pipe for 4-6 gets bolted up first. There are such problems with space and getting them lined up I just shortened the spring. This gives you an extra bit of thread to help you. Taped the spring to the nut and stiffened up the extension with more tape. Person on top has to hold the flange with one finger while you maneuver the socket and nut from below. Transfer that single finger to the shortened spring. Snake the stiffened socket set up and get it lined up. Person on top does the tightening. You also turn the steering rack to the left to get additional maneuver space. Then you do 1-3 the same way
Enjoy, If you do headers, Pull the motor or you drop your front suspension
Last edited by rjjablo; 01-06-2019 at 02:13 PM.
i can believe you rjjablo... seems impossible. O well, i must do it... either way.
i was thinking if ...i place the bolts at the underside... spot welded to the loose flanges ? so i could get them aligned with the manifolds holes and just had to push them upwards. Than springs and nuts above. Same result but far more easy …. if those were not in the way of spark plugs…. some are in the center of plug wires… must check it, if wires/caps can clear those springs i think it's a possibility… ok, for renewing plugs one needs to dismount the nuts/springs, in that case not difficult …. i hope not the bolts , so one has to lower the loose flange with bolts ….
maybe…. worth testing ! requires only some holes drilled in the heat shields… peanuts.
i'm amazed that it works with the springs above and not in the way for pulling plugs ! Now to find the best solution , either bolt or nut welded on the loose flanges…
here mockup on the spare engine... seems so easy ...when one can reach it
SAM_2626.jpg
I assume the flanges are cast iron. Nuts I was told are some kind of beryllium copper. Mine were magnetic so I am not sure about that. Might want to test weld before you do it.
new idea ...no welding.... as the holes in the loose flanges are already 10 mm i can drill it to 10.2 ( or 10.5) mm and thread it M12 and use helicoils to have the wanted M8 in the loose flanges.... that would also keep the threads in perfect center-center distance…. as i see the bolts have not many clearance in the manifold holes...
Only problem could be to wiggle with the loose flange so the bolts start in the M8's threads ... cannot estimate it.
my idea doesn't work. I can get 1 bolt started but not the second.... no "feeling" in the heavy loose flange to get it aligned.
So back to oem but with the springs above. Went quite well.. as i was alone , i had to make some helping strips above to push the bolts down while working below.
SAM_2628.jpg
drilled holes in the heat shields to let the bolts/spring protrude.
SAM_2629.jpg
now i hope to have some improvement on exhaust flow....
Our Stage 3+ kits on bone stock E31 M70 (500+ hp) do not exhibit this. Thus I wouldn't suspect the exhaust being the cause of the flat power you described.
Perhaps could be the famed 4200 rpm EML bug. Are you sure you have true WOT above 4200 rpm? Data logging the DME's (not the EML) and/or physical observation of the DK butterfly angle at WOT should reveal.
Otherwise, I would have a look at the spark curves. Keep in mind that dyno HP curves look like a frown. Spark curves should look like a smile. Are you managing fuel + spark natively in the DME's?
Rgds...
Many thanks John for reply ! last dyno was done with the Wokke V2 EML chip.... but i find it to "aggresive" on a FI 850 ... so i will swap for a V1 to test .
Great idea tophysical look at the DK butterfly angle at WOT on dyno … i will sure do that . I have indeed read alot about the famous 4200 rpm dip...
DME's have chips from Chargertech ( the kit supplier , a Raptor V charger ) , and i installed dual A/F gauge… A/F stays "good" all the way to 6000 rpm on both banks.
That is our mystery : from 4700 to 6000 rpm the boost raises another 0.1 bar , the A/F stays "good" but we get a almost flat curve at 410 HP.
We are looking at the mapping ( TunerPro) and bought a eprom reader/erasor but must study more... so far we don't understand it enough to daire to make changes...
Do you ,or know, DME mapping ( or chips) for a 6...7 PSI supercharger FI i can buy that worked well without the bug ?
regards
Barry
Yes. Give it a try. We have a big collection of EML chips. Some of them have the bug. Others don't. Unfortunately unable to report on the Wokke versions you mention.
We have data loggers to look at this but on a budget, an old mobile phone + some duct tape works wonders here. Strap it in pointing towards the DK motor and go for a few WOT blasts. We did this in the early days and would also tape a toothpick on the DK motor so the video would capture better the butterfly angle. There is no easier/quicker way to confirm the RPM bug than this.
Good news.
With belt driven supercharger, boost should be linear and directly proportional to engine RPM. Thus, higher RPM should always produce higher boost. The question, though, is where is this boost? Pre- or post-butterfly? If your DK motors are only opening 75 degrees, then higher boost before the DK's won't make the difference that it should - - which could explain your flat HP issue.
I personally wouldn't spend any time with fuel/spark maps until confirming that you have 90 degree angle on the butterfly at engine WOT. Not saying not to tune the maps later. Just that it's a waste of time until the bug is completely ruled out.
The bug is not in the DME chips. Just the EML. Basically the WOT signal from the accelerator pedal is sent to the EML which is in turn supposed to split that signal and forward it onwards to the DME's. In the "bug'd" EML's, the WOT signal doesn't get to the DME's.
Hope this helps.
Rgds...
thanks for replys John ! i will sure do the visual DK motor test on the dyno as soon as the car is ready … i must say my boost gauge is connected to the inlet manifold , so post butterfly. I have another connection before the intercooler , we used it once to get a idea of pressure drop over the cooler .. was approx 0.1 bar .
I'm reassembling the engine after i got a water leak in the metal tubes/ water pump O rings in the valley… and while is was in there , i removed all aux heater stuff ( Webasto ) , took some time to find the replacement hoses … but with help from here on the forum we found the right # 's ! all is in , done a water pressure test …. so now still the inlet manifolds and we can go testing again .
Happy to know the "bug" is in the EML …. IF it's that bug that gives us the flat curve.... ?
also timed the cams…. don't know what that can do ? 1-6 was 9° behind and 7-12 6 ° due chain stretch....( crank ° )
Last edited by Belgiumbarry; 01-15-2019 at 06:00 PM.
Glad you're getting things back together. Timing the cams should make an overall difference in performance but will not likely solve the flat power issue at higher RPM's. Quick review of the facts...
- Boost is 0.5 bar at 6000rpm
- Boost is 0.4 bar at 4700-ish rpm's (implied from what you wrote above)
- Producing 410hp which remains flat from 4700-6000
- AFR (air fuel ratio) is "good"
In a perfect world (perfect tune, fuel, timing, AFR, IAT, etc) 0.4 bar would make ~420hp whilst 0.5 bar would make ~450. Without seeing the dyno sheets - - is it possible for you to post those preferably showing also boost and AFR's? - - I suspect it's down to one of the following:
- 4200rpm EML Bug - Discussed in previous post.
- Spark Related Issues - Could be timing related. Could also be spark blow out but if AFR's are good to redline perhaps not. This is why seeing the AFR data would be helpful.
- Supercharger Compressor - It could be that your SC head unit is simply running out of it's efficiency zone. Do you have a compressor map that you could share?
Rgds...
John, your quick review of facts is right .
A/F must still be tuned but i tought it was "good" and could not explain the flat curve : from 2000 rpm ( 12 ) it gets richer to 3500rpm ( 11) , than it leans to 5000 rpm ( 13 ) to again richer at 6000 ( 12 ) . Better would be it stays around 12 and not that lean peak of 13 ?
Exact boost we must check again , now i see we have filmed the gauge on the dyno under a to great angle…. so difficult to give exact numbers.
I asked for a compressor map in the past but did not get a answer….. it's a Raptor V 30 .
http://raptorsc.com.au/supercharger-models/
It could be we are at the limit of that compressor.... and having a intercooler/pipes with 0.1 bar pressure drop. The pulleys give at 6000 rpm the max 18000 charger rpm according fabricator. I will film the gauge at a right angle next time to be sure what we exactly have.
Here are the curves from last dyno : ( HP and torque values to be taken as "x 0.75 "as we only drive 3 sets rolls and not the 4 sets )
BMW 850i - crank pully - WOKKE EML - HP.pdfBMW 850i - crank pully - WOKKE EML - NM.pdf
thanks again for your interest and help !
Barry
Bookmarks