Jump to content

MJ's 3.2 Engine Build


Recommended Posts

Thought it was about time I finally put up my engine build details, carried out over a year ago by the talented Richard Chamberlain and team up in Cambridge. Meant to do this a lot sooner but never got around to it but thought it could be helpful for others pondering similar rebuilds or options. Richard and his team (CTR Developments) did a great job on my engine, very professional and up-front and  honest throughout and his costs were very reasonable.

So here's the info I accumulated over the build (though may be gaps due to time since):-

Dec 14th 2016

 Friday will be taken up removing the engine and gearbox and taking the rolling car back to the farm. I plan to strip the top end of the engine over the weekend but before I start I'm going to do a cylinder leakage test. This helps me pinpoint any potential issues with the valves or cylinders. I don't expect to find anything wrong but once the engine is in the stand the test takes about 30 minutes and its a quick way of seeing if there is anything untoward. I'm also going to check the cam timing on both sides as I strip it down. I don't know what I will find here but I'm trying to find an explanation for the poor power so I rather hope the cam timing will be miles out! You gave me the cam installation information so i know what the timing should be so I will quickly see if this is an issue. Despite what a lot of people tell you cam timing does not need to be correct to within a thou. In fact, we have done a lot of work on timing and you have to swing the cam quite a long way before you see a dramatic change in power and torque. You can "influence" the power and torque (mainly just changing the revs at which you see the maxima) but its pretty restrictive. However, if the timing is miles out then the power will be killed so it could be the source of the power loss. When I build the engine I will make sure the cam timing is averaged over the three cylinders and this average is the correct figure. Obviously, there is nothing we can do if one cylinder is way out because that's all part of the accuracy of the cam grind but at least we will know and we can then decide what to do.

 Dec 16th, 2016

 Just a quick photo to show that work has commenced! Exhaust and bumper are off and the engine should be out by midday today. Out of interest, the wiring for your LED number plate lights had melted against the silencer so when we put it back we will use some proper heat shrink sleeving and we will clip it up so it can't touch.

20161216_144055_resized.thumb.jpg.0f8f7bcd0e10dfc0fc593ec32dbd5f5b.jpg

 

 

I also checked the oil pressure sender unit. It appears to be faulty so if you are OK I'll buy a new one and then we can see what the pressure really is! Again, if you are OK I think we should change the shells in the rods when we do the rod bolts. The reason for this is I see the car has done 180k miles and I think its daft not to do so. However, its more money (circa £250) so please confirm if you are happy for me to do so.

59e32a89c7d04_camtimingcheck_prebuild.thumb.jpg.b313e51158fcdeb7f7f84f96fec4e180.jpg 

 

 

 

 

 

Just a quick pic showing me checking the cam timing when I stripped the engine

 

Dec 20th 2016

 Piston machining went well last night and we now have enough valve to piston clearance to increase the compression ratio to around 10.7:1. I'll machine the barrels later today and then the "bottom end" modifications are complete. 

I have attached a photo taken this morning of me measuring the new valve to piston clearance on your special flat 4!!

 

 

 59e32a7f8d2dd_valvetopistonclearance.thumb.jpg.7ae31e537839f1732615210d4709e8ab.jpg

 

 

 So, after a bit of a struggle we have got the 12 old studs out. It was an excellent decision to change them as two are cracked! I will keep all the old stuff so you can see everything. I have also removed the rods so I can change the rod bolts and the big end shells were also ready to replace. The good news is the crank looks perfect. Tonight's task is to finish machining the pistons. It's a pretty difficult job because the pistons have a weird top surface which means we have to do a bit of 3D machining. What could possibly go wrong?????
I'll update you in the morning.

Jan 6th
Progress has been OK ish over Xmas. I have built the bottom end with new big end shells, ARP rod bolts and 12 new deliver studs to replace the 12 old ones. All your pistons are machined. I have stripped the heads and started work on the port to throat interface but have not been in the workshop for two days because I have an awful cold.
So jobs to do are finish the heads, machines the barrels to give the right comp ratio and fit and wire up motec. I'm expecting that by the end of the month we will be ready for mapping.
I'll let you know when the heads and barrels are done.

Jan 6th

Yes the headers have moved - thanks to some prodigious heat and some long bars. I wasn't able to check alignment as the heads were not on the engine but at least I know they will move. (this was due to finding my TT race headers were very badly aligned with the ports! - Richard re-aligned them for me with heat)

 Jan 16th, 2017

 The engine is now built up to the barrels and pistons. I have machined the barrel face to increase the compression ratio and machined the pistons to give clearance between the valves and pistons. The compression ratio is now 10.7:1. The rod bolts have been replaced with ARP ones, the big end shells have been replaced and I have replaced the 12 old head studs with the latest spec dilivar ones from Porsche.

 The heads are almost complete. I have spent a fair amount of time getting the transition between the port and throat right and I'm pretty happy with the result. I'm afraid I have had to change a few valve guides because I damaged them when doing the porting but at least this allowed me to get the valve seat and guide exactly concentric!

 At the moment I am working on the loom. I have had to make a new injector loom as the original was wired up such that it was impossible to run the engine in fully sequential mode. The Motec ECU allows this so its silly not to make use of the feature - even if it does mean a new loom! I have also have to add a cam phase sensor and a new crank trigger as the existing trigger was not good enough and the engine does not have a phase sensor in standard form! Don't worry, they are not too expensive! Lastly, your exhaust did not have a lambda port so I have made one and welded it on. We will run the engine under closed loop lambda control which will be more efficient.

I hope to have the heads complete and the engine finished within a week and the loom and ecu fitted a few days later. So, we are still on for a mapping session by the end of the month. Lets hope it all goes well!

Jan 17th, 2017

We will run the engine in on the rolling road. It wont take long - about 60 minutes at various revs and power settings. Its just to get the rings to re seal in the bores and to bed the shells in. I'm going to run the engine on a semi synthetic oil whilst on the rollers and I suggest you change it, including the filter, after a 1000 miles or so. After this I would run the engine on a fully synthetic oil and we use Castrol Edge 10W/60. I remember you are taking the car to the RS track day at Oulton. I think this is mid March. If you think you won't have done 1000 miles by then I'm happy to run the engine on Castrol Edge from the get go. It will just take a little longer for the rings to re seat but it won't do any harm. 

  I've got a list of all the extra stuff but from memory the main parts are the bearing shells, the head studs, the crank trigger and the cam phase sensor. I buy the latter two from the US as their product is really good and I would struggle to make it for the same price.

There will also be an extra labour and parts cost as we have had to make new injector and lambda looms which has really surprised me. Lambda wise your loom has been cut ( I assume because your new exhaust does not have a lambda port) but it did not support the new 5 wire sensors and I had forgotten the very early Motronic units did not fire the injectors sequentially. Basically, the Porsche injector loom has 4 wires and the new one will have 12! We make all our looms from the correct aircraft spec lightweight wire and cover them in Raychem heatshrink so they look the part! We also use the correct  motorsport derived DTM connectors. These are robust and waterproof so ideal for this application.

 

Jan 22nd, 2017

 I have finished modifying the heads - after a bit of a struggle! In the end, I had to replace the guides because I could not get into the ports to modify them without damaging the guides. Its a shame because the ones in your heads were pretty good. I will give you these at cost in an attempt to help out! However, I had great difficulty in matching up an inlet valve; in fact I could not get it to seat as I wanted. I thought it needed further investigation and to my surprise I found it slightly bent. Not really enough to hurt performance but too much for it to seat correctly. I've swapped it for a new one and all is good. Very slightly bent valves don't hurt performance but are a disaster for an engine. Every time the valve closes it flexes and over time the head will drop off with catastrophic results. I'm not sure why you had a bent inlet - its the exhausts that usually bend do to an over rev - but at least its no longer in your engine.  The good news that the heads look good and flow much better at small valve lifts (compatible with your cams) so I think the work will be beneficial on the rolling road.

 My plan is to build the engine Wednesday and if necessary, Thursday, and refit it into the chassis on Friday. Once its running, maybe Friday or Monday depending on all the usual "aggravation factors" I'll arrange with our mapping man to work his magic. With luck it should be ready for you at the end of the first week of February. I'm pretty sure it will make a little more power than when we received it!!!!!.

Jan 25th 2017

I thought I would send you some photos before we close the engine up. I'm afraid they are not very good but I don't pretend to be a photographer! They attempt to show the machining on the piston, the barrels to give higher compression, the combustion chamber and sealing face, the inlet port and lastly the exhaust. Having looked at the photos they don't look too good to me but I can assure you they are very good in reality!

 

As you know, we found a bent inlet valve in number 1 cylinder. This can only have happened due to it hitting the piston and interestingly, when we did the "as delivered" clearance check there was not much clearance between the inlets and the pistons at or around top dead centre. I strongly suspect that an over rev, probably changing down a little too early, has caused the valve to float and hit the piston. The engine is now running considerably higher compression than before and under normal circumstances this would bring the valve even closer to the piston. Hence we have machined a small clearance pocket into the piston to give extra clearance. I'm going to dummy build the engine today to check both the new compression ratio and the clearances and I'll make a decision on the final clearances once I have this information.

 

One word of warning. You have had a bent valve, most likely caused by an over rev. Under ideal circumstances we should change all 12 valves as if one has hit a piston its probable others have too. None of the other valves measure bent so I believe that under the circumstances of trying to save money we continue and use the remaining 11 valves. However, it has to be your decision. Please have a think about it and confirm, or otherwise, that you are happy. I will be finally "closing" the engine tomorrow so I really need to know today. I'm not sure of the price of 3.2 valves (I have not paid for the replacement inlet yet!) but I suspect it will be over £1000!

20170125_081350_resized.thumb.jpg.b2be5d9cc913469282e830e764fc9e7a.jpg20170125_081337_resized.thumb.jpg.4048c4f949622141e8bfde6fdbe2a7be.jpg

 

20170125_081409_resized.thumb.jpg.7543b02b0bf4815028dcc8701f8d3df3.jpg

 20170125_081535_resized.thumb.jpg.bf9edafe78e1fa1875069d0cf283a029.jpg

January 27th, 2017

 Good news. The engine is back in the car and it runs. The oil pressure gauge works (5 bar when cold and then drops at idle when hot) and Motec is controlling the engine and the idle. So, now all we have to do is get it mapped! Hopefully, we will be able to get a date some time next week but I'll let you know.

 

For your record, the compression ratio is 10.7:1, the minimum inlet valve to piston clearance is 1.4 mm (its a little tighter than I'd want for a road engine but still 50% more than when you delivered to us) and the minimum exhaust valve clearance is 1.5 mm, again a little tighter than optimal for a road engine but perfectly OK. In comparison, for a full race air cooled 911 engine with the valves shimmed up to virtually coil bind at max lift we would run clearances below 0.75 mm as measured cold. This is a bit on the edge but racers want every last ounce of power! 

 Motec is set up to run in fully sequential mode. To enable this to work we have re wired the injectors so that each one is uniquely "addressable", we have fitted a cam phase sensor so the ECU knows if number 1 cylinder is on a compression stroke or an exhaust stroke and we have fitted a new front pulley and sensor so the ECU knows the angular position of the crank. We have removed the old Bosch flap valve mass air flow sensor and fitted a different air filter. The engine is now mapped using the throttle position, we had to fit a proper throttle pot, and engine rpm. We have fitted a lambda port in the exhaust and will run the engine in closed loop lambda format so that the ECU will alter the fueling to achieve the target lambda. The lambda sensor is one of the new 5 wire wide band versions.

 We have used a Motec M84 ECU. This comes with limited logging but does, I think but will confirm after mapping, allow us to store the maximum rpm seen by the engine. Additionally, we have installed a ignition amplifier and a 1 bar MAP sensor. The ECU sits on top of the standard Bosch ECU and both are under the passenger seat. We have made our own loom to connect the M84 to the engine. This loom uses the proper Raychem heat shrink and lightweight aircraft spec wire. We only use motorsport specification DTM connectors.

 

From a mechanical prospective we have changed the Porsche rod bolts and fitted ARP ones. We have shortened the barrels to increase the compression ratio and re machined the head fire face. We have pocketed the pistons to give the inlet valve more clearance and have ported the inlet and exhaust ports giving the port to combustion chamber area special attention. We finished the heads off with a CTR "road spec" 3 angle valve job. The road spec means that the seat is slightly wider than one for a race engine as it has to last a bit longer! One inlet valve was bent so we replaced it. We have used the original valve springs and caps as these had been changed in the previous rebuild. We had to replace all 12 valve guides as with them in place it was not possible to achieve the port shape we required. We replaced the final 12 head studs with the latest 993 spec Porsche Dilivar head studs; the other 12 having been replaced in the previous rebuild. We had to use new cam chain idler sprockets as the shorter barrel coupled with a old chain meant the chain adjuster was not working in its ideal position. The special larger idler put the tensioner geometry back to the correct place. Of course, the larger idler meant we had to machine the pivot arms for clearance -  there never seems to be a straightforward job! Lastly, we aligned the silencer with the exhaust primaries. Just as a note for you I measured the exhaust primaries and they are pretty small which I hope won't be an issue.

 So, all we have to do now is map it. Lets keep our fingers crossed for a good result. In all honesty, I can't think of anything else we can do to increase power without spending a lot of money (on new barrels/pistons and cams and induction and.........) so I'm certain this will be as good a 3.2 as there is no matter what power people say they have!

 

 January 31st 2017

 Well, I have some news. I took the car to the same Rolling road this morning and Dave has mapped it. All is well and it makes 248.5 bhp or some 50 bhp more than when we last tested it. This is at the same rolling road so the increase is real. It's also important to say that this is a real figure without any alterations to the temperatures or pressures so I know that if you take it elsewhere you should see a slightly higher figure.

I was a little disappointed with the numbers but I have just driven it around the block and believe me - you will be amazed how it goes now. I have the power graph superimposed on the original and it makes more power everywhere and keeps going for longer. On the road it's really smooth but with a lot more noticeable power. I'll email you the graph when I get to a pc.

 

 February 01, 2017

Torque wise the "new" engine makes 25 ft lbs more at 2300 rpm and continues to impress all the way up the rev range. At 5100 rpm the new one is making 26 ft lbs more and at 6000 rpm it makes 15 ft lbs more. Maximum torque on the new engine is 230 ft lbs whereas before it was 192 ft lbs. A power comparison is max power now 249 bhp at 6100 rpm and before it was 196 bhp at 5600 rpm.

I have attached two graphs, one showing the power and torque now and the other with the new figures superimposed onto the old.

dyno_prebuild.JPG.14ca7d4be571f208bd1c318479071eca.JPG 

dyno_postbuild.JPG.51b6d1759f2cbbba3ab62f04f16c0a5f.JPG

 

All I can say is the car is so much stronger throughout the rev-range and free revving than it was originally (despite a top end rebuild only two years prior!). Most noticeably on track where you can really push it. I was very happy with the work that Richard and his team did and wouldn't hesitate to go back for yet further improvements with more time and money in the future. Another dyno might show even higher figures but his is used for bona-fide race cars and teams so expect the results are more believable than some optimistic readings elsewhere in the country.

MJ

Link to comment
Share on other sites

Wowsers Mark, I didn't realise you had issues after the first rebuild.  And Motec fitted too!  Do you think it was an over Rev that caused it?

Interesting has mine lost circa 40bhp and 25torq due to at least one bent valve

Link to comment
Share on other sites

I do think it was an over-Rev and a subsequent broken rocker/bent valve that caused my woes at a certain dyno session, that some of you may remember, as the car had 231bhp after the first rebuild, so something caused the big power drop and I'd certainly never done it myself on road or track. 

Still, all good now :)

 

ps ..realised I worded the sentence about the graphs wrongly above. Should say first pic is dyno carried out just before the rebuild and second one is the overlay of pre and post builds. 

Edited by MarkJ
Link to comment
Share on other sites

Richard C's recommendation,  plus it meant if I plan any further future mods its ready for them. I know there are other alternatives to Motec nowadays but its a tried and tested system and there are knowledgeable guys about to tweak and tune them, which you cant readily say for the cheaper alternatives.

Link to comment
Share on other sites

Thanks Alex, I could have fleshed it out further with my side of the conversations but it would be mostly stupid questions :D

Richard was great at communicating the process throughout and gave it the personal touch, always appreciated when your pride and joy is out of sight. 

Link to comment
Share on other sites

I knew the cam info was posted somewhere, which was why I said "remind me" :) 

DC 20/21. 

According to DRCams site, that is fractionally hotter than a Supercup.  I made 250-260hp depending on dyno and exhaust before I did the intake.  If you did the 964 intake you could get to 280hp.  Just pushing you off the slippery slope.  No need to thank me ;) 

Link to comment
Share on other sites

25 minutes ago, Richard Bernau said:

I knew the cam info was posted somewhere, which was why I said "remind me" :) 

DC 20/21. 

According to DRCams site, that is fractionally hotter than a Supercup.  I made 250-260hp depending on dyno and exhaust before I did the intake.  If you did the 964 intake you could get to 280hp.  Just pushing you off the slippery slope.  No need to thank me ;) 

I mentioned the possibility of using the 964 intake at the start and he was willing to fit it if I sourced the parts but he wasn't convinced it was worth the outlay/effort. Not sure if that was based on experience or a hunch though. Could a plastic intake really make that much difference in bhp?

Reading back through some of my 'dear diary' I wonder why RC thought the exhaust primaries were a bit on the small side? Considering they are TurboThomas Race headers? I know they were badly aligned with the ports which pee'd me off considering I had them fitted at TT's workshop!

Link to comment
Share on other sites

He replaced my Maf with a new throttle body and cone filter so not sure if/how that can be configured with a 964 intake but something to consider for the future. Guess it's no longer a reasonably priced update either as you can no longer get 964 stuff cheaply. 

Link to comment
Share on other sites

If you have already replaced the AFM, then you are part way there.  You need the head to intake adapters and an MSD rpm switch for the resonance flap, then you reuse everything else from the 3.2 setup (vac system, fuel system, idle stabiliser, TB boot etc).  A 3.2 TB (maybe get it bored your as your next step?) would be an easier fit as the 993 TB stumped me as parts of it are reversed from 3.2 and you have to stretch everything to fit as it bigger.

Link to comment
Share on other sites

As RB suggests, I intend to bore an old 3.2 TB to make life way easier when hooking up a 964 intake. Vacuum lines and all throttle position sensors/plugs are all correct. I had made some adaptors to a let a 964 plastic inlet bolt directly to a 3.2 two bolt head. Can send you details if you wish.

Chris.

 

Link to comment
Share on other sites

I had mine bored out and a new butterfly fitted by a chap on e36coupe.com for circa £120 back in 2008/9.

Surely fitting Motec is a stepping stone to ITBs? (adding to aforementioned slippery slope)

Link to comment
Share on other sites

Shrish, I have after your previous recommendation, tried to find the chap to whom you refer and I failed. I've looked at others BMW forums, again no joy.

I intend to do the job this winter, so I should start looking again!

Chris.

 

 

Link to comment
Share on other sites

pretty sure Richard at CTR could bore out a tb, they are old school engineers and have all the machinery. But as he runs a race team I doubt he takes on small jobs unless part of a greater picture.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...