What does the guy who runs the ship get himself for Christmas?  Da Bom mota.

How about a 3.9 liter with laptop programmable engine management, the best components there are, plus time to install it!





I built the motor in spare time over the summer of 2006 but never had time to install it.  I pretty much tripped over all the parts, like the 3.9 P&Cs that were "donated", the CMW heads that I traded some other parts for a, and the blown up 993 case, which I had welded together and may still well regret using.

Christmas Eve 2006.  I rolled the last of the customer cars out of the shop.

Christmas day I sat down with a hacked up 993 harness, a wiring diagram, and an empty AMP 55 pin connector.  By the end of the day I had it all wired up.  That's a prototype harness, now that it works I'll be making one with proper jacketing.

Next I final prepped the motor and installed it.  Tuesday and Wednesday I was sick.  Thursday the relatives showed up.  Friday..well...plumbed the fuel system and installed the wiring harness.  Saturday, odds and ends and on Sunday, New Year's Eve, I finally got up the cuts to fire it up.  IT STARTED!

Specs:





Here's the key. CMW Billet D-Port heads,  Retail $5,000 a set.  Ouch.



Those are chormolly "Holley" retainers.  No titanium, that's a wear item.





56mm valves in a 104mm bore.  Not much room for more valve!  Note the groove cut for the copper stepped rings.



Carrillo rods,better make sure you notch the oil pump on a 3.6!  The rods must have been designed for a 3.2 shorter stroke.





OK, so there is the motor ready to have the case halves put together.


That's the CMW flat done piston.  Target compression ratio is 11.8:1.  The initial ratio came in at 11.0:1.  We had it up to 12.5:1 at some point and had to shim it down.





You can see where we had o remachine the piston along the perimiter to knock off about 1.5mm.



The Whistler. It checks compression ratio based on the resonant frequency of the combustion chamber, which is proportional to volume.  Program in the temperature, the engine size and the number of cylinders and it does the rest.  Very important to check calibration.  Note that 0.003" is a 0.1 change in compression on this motor.  That's 75 microns. A human hair ranges from 40 to 120 microns.  Get it? 





Here it is checking another set of heads on the motor.  They were SLIGHTLY low at 11.4:1.  What is interesting is that you actually get a 0.25 bump in compression when you torque the heads from about 5 ft-lbs to 35 ft.lbs due to the shortening of the cylinder of approx 0.010" (ten-thousandths or a quarter of a millimeter).


So the compression was so low (actually 11.0:1) I manually cc'd the heads.  Well, DUH the chamber volume was like 90cc and it was supposed to be about 84. When all was said and done I  had to have th pistons AND  cylinders machined. The pistons  came too close to the heads (you need some slop so they don;t hit when the tilt at TDC and the rods stretch) and I had to shorten the cylinders to get the compression back up. Well the machine shop had to take off like  0.025" to get them all even, which brought the CR up to 12.5:1 then I had to shim them back down to 11.8:1.  What a mess.  Damn good thing I had the Whistler!

You would think when you spend four grand for a set of pistons and cylinders from the same manufacturer everythign would work.  HA!  WHen you are dealing with htese tolerances then it's not just a matter of bolting it together.



There are the .25mm (0.10") shims.  $15 each but the guys spent $66 overnighting them to me form California.  Thanks

Sealant on the cylinder.  When I got the case bored I didn't want to go to the trouble of cutting the internal O-ring.  This works fine. 



This is actually the sister motor.  Note the copper sealing rings.  They are stepped so they have an inverted "T" shaped cross section.  $100 each.  Ouch.



Together.



Now that is a direct route to the combustion chamber.  45mm ports that flow 325 CFM each, with the 56mm valves.  Some big honkin' cams and life is gonna be just fine.





As I said, this motor is a mutt.  I got the P&Cs for a song, but three of they cylinders (the far ones) got redone by Millenium, and as a result ended up a different length, which sent me into the semi infinte loop of compression checking, clearancing, and basically beating my head against the wall.



Now just slide in the cams and a few nuts and bolts....yeah....all that happened in August.






Nasty weld, well, in hindsight....oh well.  I definately didn;t want to inflict that on a customer!



Now that's a pretty sight. A pair of 3.9s.  Shame we can't hook 'em together and get 800+ HP out of 7.8 liters!



Fast forward to Christmas Break 2007.





I admit it.  Teh wiring and the plumbin is a bloody mess.  All in good time.  I have to pull the motor anyway so what the hell.



There is the DTA control box.  Nice.  None of the knobs worked until we took the connector off and put the wires int eh right places.  That was them. Hell if I can wire a 55-pin connector you'd think they could get a 15 pin connector right!





These are our headers, in 1-7/8" OD.  The supertrapps will give way to coastfabs or spintechs. These were theived from Lyd's car.










Also:  Talks to the AIM Pista dash/logger.  Now I need to tie in the oil pressure and fuel pressure.  The AIM MXL Pista can get the temps and 02 readings right from teh can bus from the DTA P8 Pro.
Now if you thought the engien compartment wiring was ugly....this is really scary!  THink of all teh weight I saved by deletign the clock. Sadly it was  because I needed a quick way to run the wires to the dash/logger!



Little tiny wing for keeping the butt planted.  Yeah, it works.

Then we took it to the dyno.  Made what we thought, though there is definately room for improvment.  We got a totla of 3 runs before we had to go.  As you can see the fuel mixture is still off as we shoot for 13.2.

The first graphs are to the pavement, the second and third are adjusted to 17.5% losses, which is what we see on this dyno.  Comaprisons are from variosu motors we have done with the 993 VRam as a baseline.