Project Armo "330i" M-tech 1

[Skarpa]

Great update. Chassis is really coming along and I bet you can't wait to get it into paint....and nice touch with the valve cover.


MJ

Thanks! I'm indeed anxious to get the body painted. I like sheetmetal work and I Iike assembly stuff but the various painting and prepping jobs are not my favorite. Although I quite like applying seam sealer. It's somehow relaxing and satisfying.

EDIT: Actually, why hide it - I really, really like applying seam sealer

If you know what I mean

All in all I'm happy with how the engine turned out. I'm satisfied with the modifications I made both functionally and aesthetically. Time will tell how it actually works. I'm excited and slightly nervous to see that since this is the first engine I've built.

[jeenyus]

All in all I'm happy with how the engine turned out. I'm satisfied with the modifications I made both functionally and aesthetically. Time will tell how it actually works. I'm excited and slightly nervous to see that since this is the first engine I've built.

That nervous feeling is a good feeling though! It means you're outside your comfort zone and probably learning something. This build has been extremely motivational and I appreciate you sharing it with the forum. I've been curious as well, some of your welds are FLUX welds, aren't they? they can't all be MIG with some of the splatter left behind on them.
'

[Skarpa]

That nervous feeling is a good feeling though! It means you're outside your comfort zone and probably learning something. This build has been extremely motivational and I appreciate you sharing it with the forum. I've been curious as well, some of your welds are FLUX welds, aren't they? they can't all be MIG with some of the splatter left behind on them.
'

This whole project has been all about stepping out of comfort zone and learning things since I had zero experience in bodywork, welding, machining, rebuilding diffs and engines etc. before starting. I like learning stuff and I enjoyed building the engine but it's a bit different than other work because pretty much everywhere else it's easy to assess your work as you go by and redo something if it doesn't pass the inspection. Of course I did the same when building the engine but still you don't get the final peace of mind until the car runs and drives. It's also the single most expensive part of the build even though my engine is fairly budget setup.

Bodywork is all mig welds with solid wire. Exhaust was tacked together with flux core stainless wire and then tig welded. Otherwise all the pipework is tig welded. My current mig isn't that great and I' ve had problems with it, especially when needing to weld overhead. Could be part of the reason. The mig I used earlier, when I was still sharing a shop with workmates in a different town was lightyears better (and more expensive) than my current one.

[jeenyus]

Your welds still look fantastic and it was incredibly motivating to see the progression through the build. Thanks again for sharing. Sent from my Pixel 2 using Tapatalk

[Skarpa]

PS. Here's the deadline for the project:

[Skarpa]

I continued work on the motor. I made a bracket for the O2 sensor plugs. We'll seeif they fit there or foul with the firewall.




I didn't know that the O2 sensors have different plugs depending on the model year and of course all my O2 sensors had the wrong plugs but my brother went through his shop and found a couple of sensors with the right connector. I addition I have plenty of sensor with the wrong connector but the connectors should be interchangeable with the correct tool for separating the pins. The O2 sensors are a bit cramped because I wanted to locate the sensors, flexible joint and the exhaust system connections all in the "downpipe" portion of the headers to avoid having anything under the car that might catch rocks, curb edges etc. I made heat shields to protect the O2 sensor cables where they are closest to the other header.


This was the limit how much I could do with the engine in the engine stand. Installing the flywheel, clutch, gearbox and other stuff I needed to get the engine off the stand. I made a rig for the engine using a pallet.



I also needed to come up with a lifting beam that allows lifting the engine without taking off the ITBs.


My plan was to attach everything that I needed for test running the engine. In case I find something I need to work on it's a lot easier to do now than when the engine is in the car.




Gas tank:








All the classic luxury cars have hardwood veneer dashboards, dont they? I did the bare minimum of chassis electrics to run the engine properly. That included the ignition switch, start button, oil pressure light, charge trigger wire with the warning light and the fuel pump. I removed the spark plugs and disconnected the fuel pump to make it easier for the starter to turn the engine and ran it until I had oil pressure. Then I poured E85 in the tank and ran the fuel pump until I had pressure on the injectors and most of the air was rid of the injector manifold. The exciting moment dawned.

https://youtu.be/W_X9G-saGhE

The engine started at once and ran fairly smoothly but there's something wrong with the idle. It start's ok but then the revs start rising rapidly after a short moment. My laptop stopped working recently and I didn't find the correct driver for my inspection cable to use it on another laptop so I didn't have the access to fault codes or the live data. But I need to check them as soon as I'm able and also check the throttle flaps, possible leaks on the intake side and the operation of the ICV.

[gnmzl]

Once again, I am impressed. I've fantasized about building a rig like this, never did it though. Well executed! Can't wait to see this in the e30 and ripping

[Skarpa]

Once again, I am impressed. I've fantasized about building a rig like this, never did it though. Well executed! Can't wait to see this in the e30 and ripping

Thanks! Running the engine in a test rig was a fun little project that I've been thinking about in the past and got a chance to try out now. And immediately the work paid off as I possibly need to remove the plenum from the engine to sort out the idle issue. Will be a lot easier than in a car.

[rzerob]

Do a smoke test to pin point your vacuum leak. With your capabilities you should have no issues making a smoke tester.

[Skarpa]

Do a smoke test to pin point your vacuum leak. With your capabilities you should have no issues making a smoke tester.

I might do that. Currently I tend to lean towards a vacuum leak in my speculations. Perhaps through some unused hose connection that I've forgot to block or somethong like that or throttle flaps not fully closing. The engine picks up revs so eagerly that it points toward a major leak. With the surplus of fuel that E85 requires for a cold start it wouldn't be a surprise that the engine quickly revs up if it's allowed enough air.

[Skarpa]

Small update: No vacuum leaks and the throttles close completely. When I blocked the ICV, the engine stopped running unless i opened the throttle a little. So the too high idle is controlled by the ICV. I'll remove the ICV and test that it functions as it should. Otherwise we're most likely looking into software problem or something funny with the TPS or the air mass meter. I got INPA running on another laptop for a while but the next time powering up it barely started, had lost the driver for the cable and wouldn't accept it again. Time to dig deeper for that one ancient laptop I have with Windows XP in it.

[Skarpa]

All right, I've done quite a lot of work to improve the running of the engine. I started with the basics and removed the ICV from the engine, cleaned it and ested the operation. The ICV didn't feel stuck but it sure was quite dirty:

Cleaning it made it feel like it was moving slightly more freely. I didn't test the operation before cleaning but after cleaning the ICV moved nicely to both directions when electrified.

Next I did the mechanical set up of the ITBs. The S50 ITBs have a signle operating axle on top of the throttle bodies. This is operated via the throttle cable. The throttle bodies are split into pairs. Each pair is connected to the operating axle with a length-adjustable link:


The system is cleverly designed to have a progressive opening of the throttle flaps. With small openings the movement of the flaps in relation to the pedal movement is smaller and increases progressively when approaching the wide open -position. Factory workshop manual assumes that the adjustable screws for flaps closed -position are set at the factory and never tampered with after that. Mechanical synchronization of the throttle flaps is done by adjusting the length of the connecting links to have all the flaps close at the same time. This is tested by placing a strip of paper on each stopper screw and letting the throttle flaps close. When correctly set, it's not possible to easily remove any of the paper strips. Then the position of each throttle flap is measured at the wide open -position and the limit screw is adjusted by comparing the average value with the value given in workshop manual and adjusting accordingly.


Here's a link to the workshop manual:
https://workshop-manuals.com/bmw/3_series_e36/m3_s50_conver/2_repair_instructions/13__fuel_system_(s50)/54__throttle_and_operation/5_ra__throttle_body_basic_setting___exhaust_synchr onization_(s50_b30)/

Unfortunately my throttle bodies had lost the factory setting of these limit screws and they were all the way loose. So I began my set up with the information that on small throttle openings, all the throttle flaps should be open the same amount. I wedged the throttle lever into minimally open position and measured the opening gap of each throttle flap with a feeler gauge. Then I set them all to same value. After that I adjusted the fully closed -limit screws with the paper strip trick and checked that all the flaps close at the same time. Finally I set the fully open -position according to the workshop manual.

In addition to the mechanical synchronization there's a vacuum synchronization setting to ensure that all the cylinders get equal amount of air when idling. Each throttle body has a bypass channel around the throttle flap and an adjusting needle that limits the air going through this bypass channel. I carefully cleaned all the bypass channels and the adjusting needles and replaced the O-rings. I think these were the last seals in the engine that I hadn't replaced. Having a look at the needles before cleaning you can easily see which ones have been closed and which ones open:



While I was checking adjustments I double checked the cam timing. Having read M5X engine projects it's not uncommon to have have changes in the cam timing after the engine has been running. When originally adjusting the camp timing I was aiming for symmetrical opening of 1,0 mm of both valves at the TDC. I think I left the cams at 1,05 mm intake and 0,95 mm exhaust. However, now that the engine has been running the timing was 0,85 mm for intake and 0,9 mm for exhaust so there's a small change. Also, having read a bit more on cam timing, the symmetric lift for both cams doesn't seem to be the optimal setting. M5X engines seem to like slightly higher lift for intake cam at the TDC so I advanced the intake cam slightly and was satisfied with the values of 1,25 mm intake and 0,9 mm exhaust.

Meanwhile I got the INPA running on my laptop and had a look at the engine sensor values. There were to sensors that stood up. The TPS was giving a reading of 4,4 V at idle and 0,9 V at full throttle. Pazi88 confirmed that the values are opposite to what they should be. It's not surprising when you have closer look at the throttle bodies. Even though M52 and S50 have the same TPS sensor, the M52 throttle opens counter--clockwise when looking from the sensor while the S50 throttle opens clockwise. No wonder the idle was lost and the acceleration enrichment didn't work when the ecu thought the throttle was wide open when idling. Luckily the TPS is a basic potentiometer with a symmetric operation so switching the pins in the engine wire loom for the sensor operating voltage switched the operation of the sensor to desired.

The second sensor the I noticed was the intake temperature sensor. The ecu claimed the intake temperature to be 38 degrees celsius while the reality was around 15 degrees. While back I thought I checked the sensor to be same in both engines but that's just not true. I think I was fooled by the US M3 on the compatibility list..

S50 plenum is made of aluminum and has a threaded emboss for the sensor while the M52 plastic intake manifold has a smooth hole for the sensor with an o-ring and a retaining clip. I addition the sensors seem to have completely different scales. I removed the sensor connection from the M52 manifold, cut it up and turned it into an adapter piece in lathe:
0



Then I drilled the sensor hole in the plenum into bigger size, tapped it into 3/8" pipe thread and glued the adapter piece into it:


I'm sure I could have found a sensor with the correct thread and suitable scale but a stock part is always stock part and makes the maintenance easier when you don't need remember where each sensor comes from.

After the changes, the extremely high idle was dropped to more reasonable (but still a little high) level and the acceleration enrichment started to work:

https://youtu.be/tpoK6i21lII

[Skarpa]

I made a long post above and it's unapproved. Anyone know what I need to do to get it approved anf how to avoid this in future?

[Skarpa]

Duplicate post

[Skarpa]

Idle was still pretty rough and the O2 sensors never heated up. Well, no wonder. They are just centimeters away from open header ends. So I mounted the front piece of the exhaust system. The V-bands were not at hand so I had to improvise with aluminum tape and industrial connector bands.


After adding the exhaust system the O2 sensors heated up and the idle stabilized at around 1300 rpm. I got to make the idle vacuum synchronization. The ITBs have vacuum measuring ports for that. I built a vacuum manometer that measures the relative vacuum of each intake port compared to the rest:






The system comprises of transparent hoses fixed to a piece of board. The hoses are all connected at the bottom end and filled to about 1/3 height with dyed ethanol spirit. The height of fluid column in each hose tells the relatve vacuum level of the respective cylinder. In the beginning all the adjusting needles were fully closed. When the engine was running I started slowly openind the needles on the cylinders with higher vacuum until I had all the fluid columns at the same level. This further smoothened out the idle but the rpm stayed at 1300. That's a mystery I'll continue investigating once the engine is in the car. I've ironed out all the obvious faults and made the initial adjustments and the engine is running pretty well. There are a bunch of potential reasons for high idle: lack of load on alternator, no airbox in place, no power steering pump, leaking improvised exhaust connection etc. But my test rig has served its purpose and I see no point in trying to make everything as in a real car. Now it's time to focus on the body which I off course have been working on in the mean time as well. More on that the next time.

PS. Have a look at my previous project post as well. There was an issue with forum software post approval and it just became visible.