944 with the heart of a Mazda millenia

Hi everybody.  I bought a 1985.5 Porsche 944 a few years ago when I lived in Virginia.  I have since moved to Colorado, and she's feeling the altitude.  This feels like a great opportunity to supercharge!  I found these puppies at the pick n pay and I couldn't resist:

If you don't recognize them, these are Lysholm superchargers from The turn-of-the-millenium Mazda Millennia S with a Miller cycle engine.  They are twin-screws roughly similar to the 1200AX's from what I can tell.  The biggest complication I'm running into is how to lubricate them.  On the Mazda, the s/c is plumbed directly to engine oil system.  While tapping the 944 for oil and return isn't too hard, I have no idea what, if any, commercially available nubbins would match up with the supply/return interfaces on the unit. Alternatively, maybe I could convert it to be self-contained. I'm hoping someone in this community has tackled this beast before.

For a second I thought this was going to be the turbocharged Mazda KL-ZE swapped 944 that was on the Hoonigan Bonus channel. Still pretty cool though. I hear those blowers are pretty efficient, certainly much more so than the common Eaton.

Tap for NPT barbed fittings. Seems easy enough? 

I keep looking for one of these locally. Apparently the Millennia wasn't very popular around here.

Nice.

The stock engine might need a little help if you run a lot of boost, I think the 83 engines had forged internals if you find one, a thicker headgasket can help lower compression a bit.  Turbo valves are sodium filled so they handle the abuse a bit better.

Ditch the A/C compressor and the blower might slot in under the intake manifold.

I wouldn't bother with the stock EFI with that sort of setup.  Maybe if you can find a Roque Tuning ECU, but those are rare and support can be spotty from what I understand.

A turbo DME/MAF could be done with the supporting bits, but honestly how old are those parts?

While I'm a huge fan of MegaSquirt and it could certainly work in this situation, I'd vote VEMS instead for a 944 project as it understands the stock pickup and you don't need to add a trigger wheel/add a sensor (though I guess you'll need a pulley to run that blower anyway, so adding a trigger wheel isn't too hard at that point).

I'm looking forward to seeing where this goes!  It should be fun once its up and running!

Fun fact: 'Millenia' is an intentional misspelling of 'millennia' by Mazda. They wanted to create a new word to name the car. 

'Millennium' is a word from two Latin roots for "thousand" and "year". The Latin word for year has two N's 'annus'. If you actually dissect the Latin roots in 'Millenia', which has 1 N, it's literally the Mazda One Thousand Anuses. 

In reply to maschinenbau :

Very interesting, but I don't think that is why they sold so poorly.

Thanks for the interest.  I'll be sure to post pics as it comes along. 

I was hoping someone could identify the oil fittings as something common I'm not familiar with, but Tapping the holes may be the easiest way to go.

I plan to use MS2 as it seems to have a good support base. 

I think I can fit everything in without ditching the a/c, but we'll see.  Don't really need a/c in Colorado Springs (don't have it in the house either)

As far as internals go, I know it's not built for too much.  The boost will be down as I'll be using a 5" crank where the millenia uses a 6".  If I blow it up, I'll do an LS swap.

First off, as you probably know....... The large square opening is the supercharger air outlet, in the mazda it was mounted with this outlet on the top. The brass looking male port, as well as the other smaller male port on the bottom are both gravity drains which return to the oil pan. The larger of the female ports is the oil feed from engine, not sure what the oil pressure would be here, but there are orifices inside the supercharger which control the flow rate. The smallest female port is for a dowel pin for alignment on the engine. There are also two smaller male barbs on top that would go to engine vacuum from what I have heard, but im not 100% sure. There appears to be a cavity after the bearing oil seals, which has a drain and a vacuum port on it, leading me to believe that the designers knew that there would be some leakage past the oil seals.

I am working on converting my millenia supercharger over to self contained lubrication. I am halfway into the dissasembly process.  I will load pictures tmrw of the internals of mine. 

I also own a lysholm 1600ax. It uses a self contained oiling system. However the bearings on the intake side of the charger are sealed bearings. They do not share oil with the gear case and output side bearings. In the case of the millenia charger, they are not self contained, and the oil passages do not lend themselves to sharing oil with the gearcase. The problem with the millenia chargers is that the oil seals fail. If I can find sealed bearings to replace the intake side, I can plug the holes and make the gearcase self contained. 

There is also a 2.66 step up gear ratio inside of the snout, which is not present on other lysholm (or any other twin screw that I know of.) Don't make the mistake of directly comparing this with a 1200ax, to choose pulleys or flow rate calcs etc... This is the reason that the pulley is large (4"). I have heard the stock crank pulley is 5.66".  I have also been unable to verify the boost pressure. Many are saying that it is 1 liter per rev (of the male rotor NOT the pulley!!!) Maybe while I have it apart I can find some way to meaure displacement. My geometrical comparison between the millenia and 1200ax suggests that the displacement is actually 0.835 liters per rev. I get 16psi gauge using that number.

Ill keep you posted on my investigation...

-Tallyho520

Next day turned into next week. 

Removed the throttle body and adaptor from the intake side of the supercharger. TIG welded a long bolt on to the bearing cap, and used the threads and a stack of sockets to pull the cup off, exposing the bearing. 

Check out all of the coked oil inside that compartment. Not good! 

After removing both caps, I was able to remove the nut from each shaft using an impact gun. The bearing support was then able to be "slipped" off by gently prying and slowly working it front side to side (this took a while). The inner race of the roller bearings must seperate from the rotor shafts, and I believe either the race and/or the seal collar is a press fit. I finally got it off. The rotors show the evidence of oil leaking past the bearing seals. 

Unfortunately I have not gotten the other side apart yet. As of now the rotors would not need to be "retimed" if reassembled. I have a strong feeling that it is only these two bearings/seals on the intake side that are the problem. This would mean that the repair/conversion to self contained oiling should be fairly straightforward. 

Pic of the bearing w/ the inner race removed. The seal collar is behind it. This press fits on the rotor shaft, and makes contact with the seal. Seems to have a fixed preload, because the seal rides axially on this seal collar, instead of radially like a normal shaft seal does. More pics to come 

-Tallyho520

Since you have 3 male lobes and 5 female rotor cavities, how are they timed? Every lobe eventually sees every cavity. I get how similar numbers would prefer to be timed after the break in to each other, but that can't happen here.

You are correct ^

The rotors never touch each other. The "timing" refers to getting the rotors aligned with each other in such a way that gear backlash wont allow the rotors to contact. I have more pictures that I can post tmrw, but it looks to me that the gears are keyed to the shafts, and therefore the timing is preset at the factory and you cannot mess up assembly.

I also discovered that the seals actually do have a spring behind them. This discovery is not new, but all the other information on dissasembly of these blowers the pictures are missing, and so the descriptions leave something to be desired. I have pictures I will post tmrw. I also believe that I have found sealed bearings for the intake side that have the proper specs. 

-Tallyho520

I believe Jon Bond still stocks rebuild kits for these units. They're fantastic when they work, but the reliability of the oil seals (as you've found) are lacking and lucky to last beyond 60k miles without major service to the charger. I still have a unit in my basement from when I removed it from it's retrofitting from a VW 9A head that was on an ABA block. BBM developed a more reliable lysolm option about 2 years after I finished that build, I'll dig around some old hard drives to hunt for pictures this weekend. 

I think the correct thread title should be "944 with the "LUNGS" of a Mazda Millenia."

captdownshift:  Do you remember if the oil seal issue is only a problem on the intake side of the charger? I just cleaned my seals and they seem to float in and out very nicely now. Unfortunately,  I didn't really check how well they moved before I soaked and cleaned them. 

So as you know, life gets in the way of good projects.  But this Christmas break I've been able to make a little progress.  Here's an overhead shot as of today:

Yes, my aluminum welding isn't pretty.  I made a custom bracket to hold the s/c, a/c, and alternator (a smaller '90s Nissan quest alternator).  With two ilder pulleys and a 70" double-sided, 6-ribbed serpentine belt from a supercharged ion or cobalt, it will work (the alternator is driven backwards, but alternators don't care).  Since they came with the s/c, I reused the mazda throttle (which has a TPS compatible with MS2) and recirc valve, with much fabrication of brackets.  Not pictured: bigger ebay injectors, volvo FPR.   I used an old oil cooler in front of the radiator to cool the AWIC water.  (I decided on A/W rather than A/A because I didn't want to cut a hole in the hood or plumb all the piping for a front mount.)  Here you can see it plumbed into an old merc aux water pump:

Still to go:

finish plumbing the recirc valve; boost control solenoid (I have one from a GM and several from the mazda)

set up vacuum plumbing, including catch can, IAC

shore up the brackets

probably build a few more brackets

pressure test everything

setup wasted spark from a VW

install WBO2

engine management (MS2) and additional wire harnessing

lube s/c

some other stuff I can't think of right now.  

Anywho, it's fun, even if it takes forever and costs too much.

impressive fit. When I first saw this thread I was like "how the heck is he going to fit a S/C in there" as I looked into the bay of my 924S (same engine) :)

I like it a lot.  I will be following along on the ECU for sure.  I'm acquiring another 944 soon and would really love to ditch the flapper door air meter and upgrade to more modern electronics at the same time.  It has to be better for power, emissions and mileage.

What's your target boost level? If it's truly just for altitude compensation, running 3-ish psi shouldn't require any changes to the ECU because that's just what you'd see at sea level (assuming 5000' or so).

As for the oil fittings, they'll be a Japanese or metric standard.  Start here: Thread types.

irish44j said:

impressive fit. When I first saw this thread I was like "how the heck is he going to fit a S/C in there" as I looked into the bay of my 924S (same engine) :)

It is a bit of a squeeze.  I really wanted to fit the whole supercharger/recirc/throttle body as a unit, as that would have made bracketing much easier, but there's just not enough room without removing the brake booster.  

In reply to Keith Tanner :

Well, it's not just for the atmospheric compensation.  It was fairly gutless at sealevel, too.  I'd like to shoot for under 10psi for now.  If I assume 1.2l per rev of the s/c, I calculate about 8psi.  However, if I use tallyho520's step-down ratio adn 1.0l per male scre rev, that puts my pulley ratio at about 3.5 instead of 1.33, and the boost rises to the upper 20s.  If that's the case, I'm going to be aggressively bypassing.

I pulled apart one of mine using the weld'n'pull method of tallyho520 (thanks).  I used MIG not TIG, which less aggressively heats the base material, and I only got coking on the surface of the plug, not throughout.  It looks like this teensy orifice is the only connection to the rest of the oiling system inside:

So my thoughts are to just grease this rear section and add a zerk to the gravity return tube.  I don't think an appreciable amount of bearing grease will leak out through that orifice.  With all that space up front, I'm not sure I can successfully grease the gears and the front bearings, so I was thinking to fill the cavity with GM supercharger oil.  So the bottom would look like: 

Here's one of the brackets; tapped in IAC and recirc

Concerned about what the actual output is, I roughed up this displacement-o-meter:

I tried cranking it for a rev with the drill several times, and each time I got about 5 cups of air, which is about 1.2 liters.  While my jerry-rigged experiment isn't too accurate,  I think it lends a data point towards 1.2ish liters per rev, rather than 2.2-2.6 liters per rev.

Very nice to see someone doing something with those Milly S blowers.

Believe it or not, those things are shooting 30 psi out at stock boost on the S.  I've always wanted to fab a manifold for my Probe GT track car to use one of those and detune it to something more manageable like 10 PSI. The Mazda KL sounds great N/A and the blower whine on that Milly S blower is unreal, it would be a lethal combination

Glad to see people are still watching this! I have not made any progress on my end, because of other projects. The coked oil on the bearings (in my pics) was not from the welding heat (Mig or Tig doesnt matter here). The coked oil is from normal operation, and this is the source of the seal "failure".  It has been suggested that the coked oil clogs in the seal and inhibits its motion, compromising its sealing function. It has also been suggested that hot tanking the whole supercharger, or using chemicals (maybe marvel mystery oil?) could free up/dissolve these chunks and rejuvenate the seals. I only have the one supercharger, but i should have tried that first :(

Lachlan, I have MS2 experience if you need help with setup or tuning let me know.

And while i applaud your ingenuity creating the displacement test, the rotors don't seal well at low speed, so a single rotation wont give you an accurate sense of true displacement. Be ready to have a new pulley custom made, but it doesn't look like you have a lot of room to go to a bigger diameter. Bleeding excess boost pressure will just mean that you are wasting engine power to drive the blower, and creating more heat in the charge air = less potential power/more risk for detonation. 

Constructive criticism, not trying to rain on your parade.