An BMW M42.5? Transplanting a M42 head onto a M43 in preparation for a DASC
The following description is a project completed this spring, installing an E36 16-valve M42 head onto my E36/7 8-valve M43 engine. There were no major problems with the transplant - The hybrid has notably better high end power compared to the plain M43. The primary motivation for this was that I want to fit a Downing / Atlanta supercharger kit in my 1998 Z3 1.8 L. The M43 was a.f.a.i.k never imported into the US and the DASC kit was developed for the M44, however there was a very nice magazine article describing a DASC install in an E30 M42. The only real problem the author encountered with the M42 install was a mixture on the lean side at high revs.
The strategy was to mix and match the most durable parts from the M42 and M43, in order to prepare the engine for forced induction. The priority was not to create an engine that would take more RPM's, but make it last with the DASC mounted and do what was possible to prepare for a higher boost at a later point in time. The most important thing for me was to re-use the M43 block with the nicer oil squirters and knock sensors (the M42 block simply did not have the casting "knobs" that the knock sensors were bolted on to.)
During the process, the engine was taken completely apart and everything was thoroughly cleaned, all gaskets, O-rings and seals were replaced. The chain and tensioner was also replaced but chain guides were re-used. I could not stand the temptation of painting the block BMW "M" blue and the timing case and cover white to make it easy to spot leaks.
ECU, DISA delete, lambda control
The original motronic DME 1.7.2 system was kept in place and no modification has been done, other than a DISA valve delete. The reason is that I am don’t want to encounter problems with an older ECU not being able to work with the newer DSC, EWS, ABS, BC etc. If the DISA valve is removed, the ECU will report an error (code 18, changeover valve / DISA butterfly). To overcome this, a large resistor was placed across the pins. The DISA valve draws 160 mAmps which corresponds to 2 Watts. A resistor around 60 Ohms will fool the ECU to think the valve is there. It will become hot, so don't wrap it up in too much insulation tape. The DASC solution, where the DISA valve is placed in a plastic bag and stripped to the dipstick tube, is a quick fix; however I do not like the aestics of it. Edit: I redid the resistor and mounted it with two pop-studs to a small strip of sheet aluminum to act as a heatsink (do use a little paste).
After the swap, the larger injectors that fit the E30/M42 manifold made the engine run excessively rich - so rich that the ECU initially couldn't adjust the lambda into range, resulting in an error code (201 Oxygen sensor control). This is where you thank Bosch that the ECU is adaptive - After clearing the adaptation data, and some driving, the ECU was able to adjust the mixture and achieve the correct lambda. I have heard that it will also do so over time, without the need to reset the adaptive data. The air shrouded injectors that are used in the M43, M44 and DASC manifolds are too big for the M42 manifold. I am unsure if the M42 in the before described article had an adaptive motronic – that would be interesting to know.
I saw that it was possible to get a set of adapters that would allow the M42 injectors in the DASC manifold - The good part would be that they have a higher flow rating, talking against it, is that the air shrouding will give a better mixture. Buying 4 new injectors is not part of the budget, so I might go for the adapters. On another note, if you are using air shrouded injectors, from what i understand it is crucial to install a check valve, so the boost wont simply blow backwards through the air shrouding, reducing the boost about 0.5 psi.
Coolant hose delete
I decided not to keep the hose that heats the M43 throttle housing and M43 manifold and simply routed it from the plastic flange (DO spend the few cents on replacing the O-ring that seals between the flange and block) and into the head.
The only real modification necessary, was the addition of a threaded hole in the bottom of the aluminum timing chain housing. I suppose that it was an upgrade of the old E30 oil pan, where disaster could happen if the gasket between the block and pan gave in and let the pump suck in air, leading to a loss in oil pressure. Using the oil pan as a guide, a hole was drilled and threaded. Edit: This is not the best solution: The small thickness of the aluminum timing chain housing will not offer a strong thread. It is mostly cosmetic. Perhaps a nutsert will fit.
The M42 donor engine had about 100.000 miles on it and it was evident that some work had been done on it - The main problem was that the aluminum head and timing housing had a lot of damaged threads. It is a good idea to get a set of 6 mm helicoils and replace when the engine is out. It is disappointing that all of this could have been avoided by tightening by torque. Replacing all the bolts with stainless hex head ones gave the engine a nice finishing touch. EDIT: Get a set of 6 mm time-serts, they are much nicer than the helicoils.
Crankshaft, Pistons and rods
The M42 pistons were cleaned and polished and fitted with new rings. the main bearings were replaced and checked using plastigage before it was assembled. The crankshaft was not ground, but simply fitted with new main bearings.
The Con rod bearings were similarly replaced and checked using plastigage. The piston pins were reused and the retaining clips replaced.
In this manner, the M42 Pistons, connecting rods and crankshaft was transplanted as a whole set into the M43 block. The only thing I feared, was whether the counterweights would interfere with the more protruding oil squirters of the M43, however, there is plenty of clearance.
Preparing the M42 head:
The head was taken completely apart, leaving just the valve guides and seats in place. The ports were mildly ground, in order to remove burrs. The brass valve guides were trimmed, removing the 1-2 mm's protruding above the intake and exhaust. The valvetrain was cleaned in an ultrasonic bath and polished using a buffing wheel and polish. The valves and seats were ground by hand using grinding paste. There is a difference in the valve stem diameter between the early and late M42 engines, so if you can, get the later one which has 6 mm stems.
The E30 manifold is much smaller and shorter than the M43 and the later M42 which incorporate water heating and DISA. The only real problem is that it is slightly too wide to clear the plastic electrical junction box that sits on top of the firewall. From time to time, this result in squeaking sounds that seem like you have mice living in your cabin heating system.
The junction box that sits in the middle of the M43 manifold had to be taken apart and taped up in order to the wires to fit underneath (this has to be done in order to fit the Downing / Atlanta manifold later on, so I might as well do it now). This was easily done and reusing the M43 crank and cam sensors meant that there was no worry about the coding of the plugs / impedances etc. However, I think the M42 sensors would have worked just fine.
The Throttle housings TPS plug was identical, however the coding of the idle control valve was different. The M42 ICV only had two wires as opposed to the M43's 3 wires. An external Bosch ICV from a Peugeot 206 proved to fit and was used to connect the manifold to the air intake.
The PCV valve was cleaned with brake cleaner (was totally gunked up) and mounted on the Downing / Atlanta PCV plate, which is an adaptor that allows the manifold mounted PCV to be connected to a hose running to the manifold (after the throttle plate).
After putting everything together, the idle was all over the place, at first I feared that it was the ICV that was not compatible with the ECU, however, using a can of diesel start to locate the leaks and applying 40 feet of insulating tape took care of the leaks and idle is now perfect.
Exhaust & manifold
The exhaust manifold was replaced with a SS header - A bung was fitted for the oxygen probe and a clamp was used to assemble the two pipes.
A 10W40 oil was used for the breaking in alongside with a new oil filter - The starting up procedure to generate oil pressure was nerve staking and it took a while before the hydraulic lifters kept quiet, since this oil probably is on the thicker side. This project was made on a budget and the prioritizations made meant that i would not spend 120 USD on a can of 5w30 Castrol. The first 600 miles, RPM's were kept below 4000. After the first 600 miles the oil was changed again and then again after the next 600.
The engine seem generally noisy – defiantly more ticking sounds than the M43 – However, I always seem to find and notice more and more sounds as time goes by – I start to question if I should have taken the lifters completely apart or replaced them or if the valve guides should have been exchanged… It can drive one insane .
The oil filter housing of the M43 was fitted onto the M42 timing housing and the M43 oil pressure relief valve was moved over as well. The check valve that sits in the engine blocks main riser channed (that feeds the head) was fitted. I guess this valve keeps the head from draining once the engine has stopped in order to have more oil present in the head at the next start. Neither the M42 or M43 has the oil restrictor of the M40, If you use an M40 bottom end, be sure to remove it, also be certain, that the check valve is installed in the right orientation (allowing upflow, not downflow).
From reading other peoples experiences, it seems like the hydraulic lifters require somewhat thin oil for optimal performance and I intend to use the Castroil that BMW recommends for the M42: the 5W30. In the meanwhile I have head people praise the 10w60 EDGE formula RS, however I am a bit doubt which one is the most suitable. After trying both, I prefer the 10w60, which really suppresses timing case noise (and the M42 ticks quite a bit compared to the M43) – I was worried whether the 10w60 would be thin enough to work well with the hydraulic lifters, but it seems fine at least in the summertime (10 - 20 degrees C).
Fitting of Air / Fuel meter
The exhaust header was fitted with an extra hole and Bung for a wideband lambda sensor, in order to eastablish a “baseline” before fitting the D/A kit. The meter used was an Innovate LM-2, which was connected to the lambda probe and via the Innovate inductive clamp to the rev signal. Initially, both of these turned up problematic: The lambda sensor would start to warm up (Counting Wxx -> Wxx) but would never exit the warm up mode. It turned out to be caused by a snapped retainer in the oxygen probe lead – once stripped together with zip-ties, it heated up fast and worked flawlessly.
Like wise, the readings from the clamp were very erratic and would only show the correct number of revs below 2500 revs. I ended up throwing out the inductive clamp – It simply is no match for tapping the rev signal from the diagnosis plug. Don’t be lazy – just fit the wire.
The wiring loom for the LM-2 was routed through the firewall and then a hole in the roof of the glovebox – in this manner, the LM_2 and wires can stay in the closed glovebox when not in use (since it had to be connected all the time the wideband is fitted in the header).
(pressure sensor, T – barb)
The DASC kit was assembled with only a minor mishap along the way – the vacuum actuator stud broke and I had to repair it with epoxy and a small brass pipe. You should wait until the end of the install to mount the Y bracket that supports the rear of the compressor – you need to be able to get your little hands in there to route the fuel lines and all the wires.
Luckily, I sourced an M44 throttle body very cheaply, making it possible to throw away the external idle regulator valve and hoses – The wires that go onto the TPS are very short, so be gentle and do not pull them too hard.
There is no standard way of connecting the M44 throttle housing to the M43 Air Flow Meter – cutting up and old M42 intake rubber boot seemed like a goos solution – using miles of insulation tape.
Initially, idle was not very good – this was due to the engine taking in unmetered air throught the extra nipple on the throttle housing for the evaporative canister – Once plugged, idle was rock steady and en engine sounded really nice, without any whine or evidence of the upgrade.
The AFR meter reported the engine running somewhat rich (10-12 AFR) which was expected using the stock M44 injectors.
I did not take long for the ECU to adapt to the larger injectors and work towards an AFR of 14,7- When looking at the logs, the median is clearly 14,7 AFR and there is no systematic slant towards a higher AFR with revs and boost. I interpret this as the fuel delivery system being adequate to support the boost – I have to monitor this more closely and see if I can make it a bit more rich by adjusting the fuel pressure regulator – I assume that this will only work once it exits closed-loop operation