535is Homebrew Turbo (ex 633 Callaway) - 2024 Update

[mitch5]

After some further testing:

I seem to have injector pulse as I do get fuel smell and wet on the spark plugs after cranking.

I confirmed 12v power on green lead to coil. I confirmed fuel pump operation with the test mode.
I just tried the coil in test mode and was not able to get a spark from the main coil lead in any test mode setting. Coil set to 'one' and all possible Coils (A-F).

I imagine I need to look at something on the board but I am really not sure where to start. The hardware manual is jibberish to me...

EDIT: Peter at PF tuning thinks the board or harness has issues after I mentioned the test mode not working. Going to send him the board to check it out.

Ms2 is great once you get all the bugs sorted but you made the right decision by sending it out, i made the mistake of building mine and after spending 50+ hours diagnosing a bad vr signal, i had to finally send it out to someone who knew the board well. Turned out to be a faulty transistor and a short. only cost $80 to have shipped and fixed.

After seeing what you have gone through with trying to get your arfs on target and fuel issues sorted, it makes me happy to see you switching to ms2. Now you can tune all those parameters with only a few clicks. Only thing left is to get the electronic boost controller setup.

[tschultz]

Ms2 is great once you get all the bugs sorted but you made the right decision by sending it out, i made the mistake of building mine and after spending 50+ hours diagnosing a bad vr signal, i had to finally send it out to someone who knew the board well. Turned out to be a faulty transistor and a short. only cost $80 to have shipped and fixed.

After seeing what you have gone through with trying to get your arfs on target and fuel issues sorted, it makes me happy to see you switching to ms2. Now you can tune all those parameters with only a few clicks. Only thing left is to get the electronic boost controller setup.

Thanks for the thoughts! It's annoying having the car sit there and waiting for UPS to deliver the package, but oh well.

The idea was to have a track car and learn along the way, which has been the best part. I feel like I really understand M1.0 now, and I understand the differences with it and M1.3 a bit better. I am also understanding a bit of the theory of engine modifications from the perspective of things like AFM flapper vs MAF, spark timing control and alpha-n vs speed density fuel methods. I will be curious to understand charge IAT's with my current layout and see how a different airbox or air ducting can reduce these.

I haven't converted to the electronic boost control but I will have to explore this as I get more familiar with MS2.

[tschultz]

Waiting for Peter F to get me my MS2 unit back, i decided to review some photos from D4C. If you didn't already see, there's some video footage in the 2016 D4C event video here: https://www.youtube.com/watch?v=X_01-wKaOe0


But I also got some photos of the car on the long return route home. We took US 550 from Durango along the Million Dollar Highway (some of the footage in the above video is from that route) to Ouray. It isn't the best driving road as there isn't much passing space but it is very scenic and there are quite a few cliffs without guard rails.

The two sharks at Wolf Creek Pass. The other is a true euro 633CSi, very early production in 1976. He actually showed me something with the original headlights-- they had the VIN stamped into the glass. He had only one remaining original headlight. Avery rare color too.



On our extended drive, we made a stop in Purgatory... let me tell you it is not so bad with a group of BMW enthusiasts!





Looking over the historic mining town of Silverton



My headlight got busted on the way up Wolf Creek Pass. Poor euro lens



We did some dirt road driving. The dirt road and billowing dust was the most awesome I have ever driven. Didn't get good pictures of it, but got a neat one of my friend's M5.





Intercooler is visible. But i sure did get a lot of rock chips on my spoiler this trip!



Overlooking the Land's End observatory of Grand Mesa.

[marc79euro645]

Nice to see some other sharks on this board. Yours looks nice,very similar to mine(also black euro).
Nice scenery too.

[tschultz]

Nice to see some other sharks on this board. Yours looks nice,very similar to mine(also black euro).
Nice scenery too.

Thanks-- I don't think I have seen pictures of yours! I imagine it is very similar to my '80, but do you have M90, M30B34 or B35 in your shark?

[marc79euro645]

Mine is an m104 out of an 84 /745, basically an m30b34 with piston oil squirters.I'm using the stock kkk27 turbo with the m102 intake and bypass setup. I'm running ms2 for injection and spark. I'll have to post some pics of mine someday, it looks good at a distance, up close the flaws are obvious. I've debated spending the $2-$3k it would take to get the bodywork done, for right now I'm ok with it being a rat rod.

[tschultz]

Waiting on getting my ms2 unit back from Peter F, who found an incomplete connection to the spark output pin. He's going to repair it and install a mapdaddy sensor for on the fly barometric correction when I drive into the mountains.

In the mean time, I decided to mess with my air box. I really didn't like the cone that I had on the car, but I didn't have many alternatives as it didn't fit in the car with the Callaway kit. Fortunately, with the megasquirt I am removing the AFM and using an IAT sensor instead. When I started looking, I realized that i can finally put a stock airbox back on the car. I got this one and have had it sitting around with the intention of installing.

The main reason being some infared images I saw on this board back when I joined the forum. I have saved them for reference:

Its thermal sensor is 320x240. The camera is professional level designed mainly for industrial work. My buddy uses it to do diagnosing of electrical connections and equipment in commercial and industrial buildings.
Its range is -40°C To +2,000°C, thermal sensitivity 0.08°

The second picture illustrates how good the stock airbox is. I'm not sure if the car is idling or in heat soak, but either way, it works very well.

I adjusted the turbo inlet height and position of the elbow/coupling so that it would align a little better. I mocked up the air box using a bracket I had laying around, the positioning is slightly forward of stock, but the hood still closes and now the air filter is more secured than ever before.



I also like that it makes the engine bay look more stock. Now when I pop the hood, it will be less obvious that the car is turbocharged.

Here you can see the diverter valve and piping, along with the placement of the IAT sensor. I didn't even have to adjust the hole where the 7th injector used to be. The bore was very close to an NPS tap drill, so all we had to do was to tap it 3/8" NPS. I think this works better than a bung as it places the sensor closer to the center of the intake pipe.



The engine bay is looking so much better than it had been. This air box helped, along with removing the extra wiring for the Callaway Microfueler.
Oh yeah, I also removed AC components in the engine bay as I realized I wouldn't be converting it ever. I did this when I put the 60-2 toothed wheel a few weeks back. I had to remove the radiator, but otherwise it was pretty simple. I haven't removed the parts under the dash, but I still removed probably 45 lbs of weight.

[marc79euro645]

I would like to go 60-2 setup. Right now I have early dizzy for vr pulse, which is kinda iffy when battery low or motor cold, so obviously crank vr would be twice as fast and i hope more dependable. I've looked at all the threads about it, but still unsure as to ms2 setting.Any tips appreciated.

[tschultz]

You have to take out the radiator and remove the fan and pulley. I'm sure it is easier on an e28 since the hood opens the other way. Otherwise, add the VR sensor and bracket in place of the old Motronic 1.0 sensor (who knows what that does anyway?). Since the 60-2 wheel is located via a pin, there is only one way to install it.

Wire it up to the pins for speed or reference sensors if you have the stock 1.0 harness and use the 60-2 settings from the MS FAQ here:


That seems to be all there is to it. From what I understand, megasquirt uses the crank signal to determine injector firing.

[elcrazon]

I don't know if you need enough clearance to have to pull the radiator, though I did mine when the engine was out of the car, so I can't say for sure. Make sure to install the sensor bracket before getting the 60-2 wheel torqued down, there's a mounting bolt you can't reach with the wheel installed. As for the MS2 settings, I think this might be in the FAQ a number of pages in, but the recommended setup these days is to use rising edge trigger, which gives a tooth #1 angle of about 84*, which you can verify with a timing light and fixed timing enabled.

[mitch5]

Before you get to far into the megasquirt tuning, i would suggest moving your iat the furthest point away from the engine heat as possible. Mine is mounted right after the intercooler behind my valance. You will almost certainly get heat soak from mounting above the valve cover. In short when you turn your car off, heat from the engine bay will radiate and heat up the sensor. The sensor will think the air is 130f when its actually 80f and will lean the engine out, this creates a big head ache when starting the car for the first few minutes. You can tune to compensate this but it way easier to move the sensor.

[marc79euro645]

Thanks guys, good info and tips. Now I just need to get the balancer and vr. I've changed the balancer before with the radiator in on my 6 series. I used the old balancer to make a crank holder (welded a 2x2 square tube to it), and a big cheater bar. 330 ft lbs of torque on the crank hub is serious torque.
thanks again
marc

[tschultz]

Before you get to far into the megasquirt tuning, i would suggest moving your iat the furthest point away from the engine heat as possible. Mine is mounted right after the intercooler behind my valance. You will almost certainly get heat soak from mounting above the valve cover. In short when you turn your car off, heat from the engine bay will radiate and heat up the sensor. The sensor will think the air is 130f when its actually 80f and will lean the engine out, this creates a big head ache when starting the car for the first few minutes. You can tune to compensate this but it way easier to move the sensor.

Thanks for the comment, I will monitor this and see if it is something I need to change.


I got the Ms2 back yesterday and had some time to mess with settings. It fired up with Elcrazon's tune but I had to adjust Warmup inrichment and afterstart enrichment. It also seemed that my idle settings needed some adjustment.

My brother and I adjusted the the spark timing enough so that it wouldn't totally die and got to a point where it was able to idle and not hunt. A few things I need to resolve though:

1. Noisy signals. The CLT, TPS and AFR #'s all jump around a lot when the car gets running. I realize this was happening before megasquirt as the AFR's never seemed to settle out. Any comments here? I checked grounds at the battery/chassis, head/firewall and head/wire harness (the only 3 major ones I know of) and they all look good... Any thoughts here?


2. No tach signal on instrument cluster. I read this on E30's about jumping the engine speed output to the cluster:

the engine speed output and gives the instrument cluster the signal for the tachometer. This must be jumped to pin #32, the fuel consumption output, whichgives the cluster the signal for the econogauge

I also read about an inline resistor. Anybody been through this?

[elcrazon]

Oh yeah, sorry, my warmup and afterstart enrichment aren't very good at the moment because I'm still playing around with the map table and it runs well enough on my car at the moment.

How noisy are your signals? The TPS can be a bit jumpy if you're hooked up to the non shielded wires in the harness that run by the injector wiring. You can smooth some of it out with the lag settings. Where do you have the o2 controller grounded to? I also pulled the grounding for the idle control circuit on the board off the low current ground plan and moved it to the high current ground plane when I thought I was having major voltage issues from the idle valve. That turned out to be caused by a bad main relay, but I think it did smooth out a little of the sensor voltage fluctuation.

For the tach signal I pulled the signal wire out of the harness near the motronic plug and connected it to a c103 connector I cut from the old 35-pin harness.

[tschultz]

Oh yeah, sorry, my warmup and afterstart enrichment aren't very good at the moment because I'm still playing around with the map table and it runs well enough on my car at the moment.

How noisy are your signals? The TPS can be a bit jumpy if you're hooked up to the non shielded wires in the harness that run by the injector wiring. You can smooth some of it out with the lag settings. Where do you have the o2 controller grounded to? I also pulled the grounding for the idle control circuit on the board off the low current ground plan and moved it to the high current ground plane when I thought I was having major voltage issues from the idle valve. That turned out to be caused by a bad main relay, but I think it did smooth out a little of the sensor voltage fluctuation.

For the tach signal I pulled the signal wire out of the harness near the motronic plug and connected it to a c103 connector I cut from the old 35-pin harness.

Thanks for the help. I added some lag factors for now until I can get it sorted. It seems the TPS and AFR have the most noise so I will have to try to shield the TPS wiring or rewire power and ground to my AFR sensor.

I adjusted the warmup and after start enrichment and was able to get the car to start better at my elevation. I tried it this morning in 55F and it started right up. I did some reading and changed the Idle value to "Use last vale" instead of use table. This made a big difference in removing the idle hunting that I had.

Now my idle returns to about 1100rpm and takes about 6-8 seconds to activate CL Idle mode and drop to my target of 850-900rpm. I found this engine with motronic always seemed to idle higher than my 535i, so I'm ok with it.


Regarding the tach, does this sound right?
From my research of following the quote in my last post, I need to tap into Pin 11 of my motronic harness (white wire, fuel output) and connect it to the white/black wire for the tach input:



I'll try to share a video by this weekend. I am working on the base map without EGO control. With windows up I could hear some detonation/pinging with a hybrid of your tune and the FAQ ignition timing. I suspect it is related to 85octane we have here. I have removed 5-8 degrees and will continue refining. The plan is to have it ready in about 3 weeks for another track day.

[elcrazon]

You should be able to adjust the CL idle engagement settings to get it to enter CL idle faster, that sounds quite slow at 6-8 seconds. I moved away from using "Last value" for the idle valve due to an issue where it would get confused and just leave the valve all the way open, not super cool to be coasting down the street with the engine "idling" at 2500+ rpm.

I think the VE map on the tune I uploaded is a bit low across the board, I think most cells should probably go up at least 2-3 points and the high vacuum side needs to be bumped up more. And yeah, the timing is a bit aggressive since I'm running a b34.

For the tach, no, you need a black wire. I think it's basically the same thing as the ignition output. The fuel output stuff isn't going to work with megasquirt unless you get a board to actually produce the right signal for the cluster.

You need something like this:

or this:


Pin 1 on c103 to Pin 21 on a 35 pin harness or pin 6 on a 55 pin harness. Though the basic output circuit might work fine, you could need a conditioning circuit on the Megasquirt side. I have the goathumper adapter board that has something on it, but I don't remember exactly what it does and the board has terrible documentation.

[tschultz]

Thank you el crazon! That should work for me hopefully. I have another car project I am working on so I didn't have a chance to try it out last night, but I will!

I am still using 'use last value'. Going to test output mode and verifying valve duty was helpful. The old percentages were off and my new numbers are a little higher, like closed duty 36 and open duty cycle 80 (where idle speed stops going up). Part of my idle hunting was related to fuel cut settings. The activate below map value was too high, so it would go into fuel cut mode rather than closed loop idle.

After some more tuning I have found that due to my altitude, my VE table has to be much different than yours. I have EGO control off and have found the car was rich throughout the range, even in boost (AFR's of 10.8 down to 10.0 displayed). My spark map has had to be retarded also as I had been getting pinging, probably related to the advance and 87 octane fuel. I want to tune on this fuel so that in a higher stressed environment like the track I'm not on the edge of detonation. Obviously higher intake temps will be common so I'll start with the safe tune and then buy the 91 octane for the track day. I have read max power isn't always made at max advance anyway.

I have been working to refine both maps and have been able to drive the car to work and start it in the morning in cold 45F temperatures and drive home from work in 85F temps. I did have heat soaking like as suggested a few posts back, but I want to keep my eye on it before changing the sensor location. My theory is that a location in the charge piping may be better, but on actual hot days the basic idle settings will still see the high temps and the idle will still be lean. Apparently the GM sensor has somewhat poor response time at low airflow, so if I was stuck in traffic in 100F weather, I think the engine would start to idle lean. I read about MAT correction to help compensate the ECU into adding fuel when intake temps are high. I think this has merit because when actual intake temps are high the megasquirt won't have a tendency to remove fuel. I just hope I am using the feature correctly...


Will post VE and ignition map maybe in a week or so. I need to keep adjusting it before enabling the EGO control.

[tschultz]

I have been driving the car, but still haven't figured out the tach. The C103 signal is connected and should be bringing the output to the tach. I don't know where my tach output pin is as there appear to be a few different outputs possible. If I change my 'tacho output' from IGN (JS10), the car won't start.

I have been getting pinging with 87 octane as I approach 120-140kPa. I assume the TCD chip was super retarded as I never had this issue before. I keep removing a few degrees at a time and my map seems very conservative compared to others I have seen. I am very close to getting the full premium (91 oct here) and tuning on that fuel instead.

This is with the TCD turbo, 41lb injectors (req fuel=8.7), and a small intercooler.




I haven't really been able to tune above that 140kPa range yet but am fairly happy with the other zones (at 70-80 F). When I drive home from work, the uncorrected EGO values are very close to my targets since this is what I have tuned. However, I get to the next morning and on my way into work, the whole range of the ve map seems to be richer than the day before. It is definitely related to temperature. I want to get a handle on this so the EGO correction doesn't have to do so much work. I have found a few threads elsewhere with similar issues but no real solutions. I have been playing with the MAT corrections and this is what I have so far.
I'm not sure it is what I want as it seems I can adjust the table slightly and AFR's come close to target in vacuum but then I think the correction compounds and removes too much fuel under boost. I may just set this table to 100%, I'm not sure yet.




I had an issue the last few days where I was losing coolant. I was hoping this didn't mean headgasket Last night I noticed one of the water pump hoses wet with coolant. Not any dripping, but evidence of coolant turning to the green goo on the end of the hose. I also found some wet spots in the engine bay, again no drips. I'm hoping readjusting that clamp fixes my leak. I really want to get the car to the next track day Oct 16!

[elcrazon]

Can you post your AFR table? My approach has been to tune the VE map for 14.7 across the board and then let the O2 sensor make adjustments as necessary based on my AFR target. Though, I'm not running boost yet and I think the boosted area of the VE table should be hard coded to whatever AFR you want to make there as the O2 sensor is pretty useless there. Are you using Multiply Map? I have 42lb injectors with a similar ReqFuel and have to have much higher VE in the idle region to get a decent idle AFR.

I think you're going to need a lot of retard in there boosted region if you're trying to run on anything less than 91 octane. When I talked to Todd about my setup he didn't recommend anything less than 91, though our gas here in Cali isn't very good to begin with.

As for your MAT correction, I was under the impression that since it modeled the ideal gas law it shouldn't be changed much unless you're really sure that's the issue. Maybe you're just having issues with your temp sensor placement as previously mentioned in the thread.

[tschultz]

Sure. I am set to multiply also:



Recall I am at higher elevation than you, so my 80kPa is close to atmospheric. I specifically wanted this region to be about 13:1. I also have lower grades of fuel available. 91 is the highest you can find in most gas stations around here, but it sounds like I may just have to always use premium.

My understanding is the MS controller references the VE map as the base fuel value (along with req_fuel). Then EGO control is allowed depending on the circumstance and preset settings in case of fluctuations between ve map and the target AFR. For example, my ego control isn't allowed over 80kpa so once I get in that zone, I am relying on the ve map. The idea is to get the base map target zones close before any corrections.

For the AFR Map, you want to tune the different zones for the application. Notice my 15.5 zones are cruise in low load, so lean is acceptable to save fuel. Approaching positive pressure I target 13:1 and above I aim for 12:1. But I have my ego control disable above atmospheric. I'm not sure what other users do.

I'll keep investigating mat corrections. The iat sensor placement seems to be related mostly to hot restarts. I have found it to run a bit lean in these conditions until closed loop idle is able to interfere and bring everything near the target. Temperatures are cooling down and I haven't had hot restart issues, but it is still on my mind.

I can send you my closed loop idle settings if you want to try them. I also have idle advance set at 15* so it never has to change timing in the idle region. I suspect my altitude has a little to do with my low ve values. Recall high elevation means less air. There's about 15% less air pressure here, and less air requires less fuel for stoich.

[tschultz]

The good news is that I was able to get the tach working by connecting the coil to the black wire to the cluster. I did have to disconnect the blue/black (that leads back to the MS unit) as it would cause the car to die when I connected the wires.


The bad news is that yesterday I noticed my BOV and turbo all of the sudden louder than ever before. I'm not quite sure why but I hear them both more clearly even with the windows up.

I just drove around the block and afterwards noticed the dipstick popped up and oil spray all over that side of the engine. Upon further inspection, the breather screen I have is smoking much more than I ever remember before. One last check and pulling the oil cap I no longer have vacuum at idle in the head,itt is also smoking out at idle.

I think I'll look at the breather tomorrow. The car is running really well otherwise. Thoughts and ideas appreciated.

[tschultz]

So another update.

The leaking I was getting was related to the dipstick. For some reason I have a lot more blow by all of the sudden and the breather I had on the valve cover wasn't venting well enough so the dipstick kept popping up to relieve the pressure.

I removed the breather and just ran a length of hose out to atmosphere. The dipstick isn't popping up any more, but it just seems to be smoking more than it should be out of the valve cover breather. I don't really want to plumb this breather back into the intake as I don't want the oily mess in the intercooler and charge piping...
I'm not getting any smoke out the exhaust, nor am I getting milkshake color in the oil or coolant.

I took a video to show you guys and see if you have any further thoughts: https://www.youtube.com/watch?v=XAiu5jg3e68



My blow off valve and turbo noises coincidentally seemed to get a little louder with this development. I didn't used to be able to hear the turbo spooling up at all, only a slight high pitched whine at 6+psi if the windows were open.

Either way, you can distinctly hear the BOV and turbo spooling now with the windows up (right atmospheric). Here's a short clip: https://youtu.be/QOxR25wHvE8


I have driven it the past week to work and will plan to drive it next week to work to verify there are no issues. It is running well but I am sort of nervous about the blow-by/smoking at idle...

[Shadow]

Oil coming out the breather does that mean your rings or shot?

[marc79euro645]

Plumb it in after the intercooler. Your car sounds good to me!

[George]

There is little gained by running the breather directly back into the intake. Its done on the OEsfor emissions. That ship has long sailed for this type of car.

You'll gain a little bit of vacuum if it is run pre-turbo but as you said - its an oily mess. You might want to consider an air / oil separator. When done properly they can be heated to reduce moisture build-up and they will drain excess oil back to the block. With that said, I run my mine straight to atmosphere with a 3/4" braided hose directly over the valve cover. The hose occasionally has some moisture build-up but is never oily.

This is an air / oil separator I designed for Subarus a few years back. The top was Plexiglas as it was a prototype.