Skip to content

Build Part 1

Calculation Revisions

Note: This Part 1 of this build has/is being updated with revised calculation. Calculations that take into account what I have found as a result of the actual install of this supercharger. In short, the 3.25" pulley was only good for 2psi of boost. I'm going to continue to revise this page with more accurate numbers as I obtain them, this time through trial and error. You can get still see the original calculations here [[2000 Subaru Impreza 2.5RSC Build Part 1a|Part 1a]]


These pages are ever evolving, as my project is ever evolving. So, check back for updates as they occur.

Another place you can read about this build is on - My M62 Supercharger Build & DIY

Follow along now as I chronicle putting a Eaton/Roots M62 Supercharger from a C230 Mercedes onto my 2000 Subaru Impreza.

I've always been a fan of WRC. But, I never really had a particular WRC car I'd say was my favorite. But, back in 1998 when a buddy of mine took me for a spin in his AUDM WRX while I was living in Australia.. I was hooked. "Today, I become a Subaru Fan." But, alas, they don't have the WRX in N.America. d'Oh!. Thats ok, I don't really need a turbo. I love that car as it sits. I go buy my 00RS off the showroom floor. Ever since that day I drove it off of the lot, I've been thinking about adding Forced Induction to it. haha.. Yeah.. right, I don't need a turbo!! The original Minnam Turbo kit. The old Rimmer SC kit. A few "add your own turbo" ideas. Years and years of research, reading, and absorbing everything I needed to know about adding FI to my car. Now the AVO kit, the FAT kit, WRX Engine Swaps. JDM STi Engies, pieced together WRX turbo bits... the possibilities now are almost endless.

Well, thank mostly to redevil on, Forced Induction on my car is going to become a reality now.... I hope. I'm planning to add an M62 Supercharger from a Mercedes C230 to my car. These next pages chronicle this endeavour.

Build Cost

Build Cost

Turbo vs Supercharger

If your a regular member of NASIOC or, then you've likely read all the Turbo vs SC threads, and heard all the swap sayers, and just turbo it threads. I'm not going to sit here, and explain the differences, or try to convince you of which is better. You can make up your own mind in that regard. I, however, have decided to go with a Supercharger. Here is a brief list of why:

  • A SC can be gradually added piece by piece without turning it on until your good and ready. Unlike a turbo that requires modifications to the exhaust almost from the start. Plus, with just a belt change, or a magnetic clutch, or in my case, a simple piece removal, you can be back to almost stock. for a Daily Driver like my car, this is a good safety fall-back feature.
  • Speaking of exhaust, I already have a very nice header-back SS OBX Exhaust system that I'm very happy with. I don't really want to have to get rid of it, and start on a new exhaust setup.
  • The cost of an SC can be more than a turbo. And used turbos and used SCs run the gambit of prices. But, a used M62 from a merc can be had in the area of$250. Thats really not all that much more than your average TD04 from a WRX.
  • Turbo cross member, and Turbo FSB. Not that it's a big deal changing them out, but.. I have an WhiteLine adjustable FSB that I like. I don't want to have to change it out for a turbo setup, then try to recoup the cost by selling it.
  • I LOVE the feel of instant torque down low. And since I've taken up AutoX lately, torque is my friend. Sure, you can make a heap off torque with a turbo too. That's not in question here. But, as you increase power and add Boost to a Turbo system, you quite often have to pay the price of turbo lag. No longer is the torque down low. Increase boost in a SC, you simply change pulley sizes. And with a roots SC, the boost is still down low where I want it. Ok, that's a simplified explanation, but you get the point.
  • AutoX + Torque = GOOD
  • The turbo guys will tell you that down the road, a Turbo kit has better upgradeability. Better? Why is that? Is it due to the availability of parts? Is it the fact that the Subaru is designed for Turbo stuff? I'm not completely sure. I agree that you can likely make way more power with a turbo, but thats not my goals anyway. I think the term is quite subjective. But, upgradeability of a SC setup is not all that bad either. If sized right, more boost is a pulley change away. Run out of oomph with that M62? The M90 is even more plentiful. Will it fit where the M62 sits? Don't know the answer to that one yet. But, my measurements show it will. So, you tell me, which one is has better upgradeability? Seems like a tie to me.
  • And one of the most fun reasons, it's unique. Sure sure sure.. that's a silly reason. Even a bit stupid. You don't build a FI system just cuz it's cool. But, you can't deny the oooh factor. Like it or not, we're human, and this sorta crap matters to us.

After all that, my decision was NOT an easy one. I might like Superchargers, but I'm a HUGE fan of Turbos. I freakin' love those little suckers. I mean common!! Tap into the exhaust and make power.. What a great getup. Believe me when I tell you I went back and forth between adding a Turbo and a Supercharger. Hey, it took me almost 8 years to make up my mind if you count the first time I looked into the idea. Want more proof? Ok.. Here are the 2 pages of Brain Dumps I created at the same time as I contemplated doing a Turbo vs a Supercharger.

As you can see, I looked at almost every angle, with cost being a driving factor. But, when the dust settled, an [[ Eaton M62]] Supercharger is what I'm going with. I have not picked one up yet, but I'm hoping a good EBay deal comes up soon. Uh... scratch that.. I just won my EBay bid on a used M62!!

Initial Calculations

Before I actually start into any buying of parts, or fabbing of bits, I have to be sure that what I'm planning makes sense. Can the SC I want provide the CFM and boost I need to get my power goals? Will the SC be working too hard to do this? Will I need an intercooler? Well.. lets find out.

The calculations your about to see were not invented by myself. I got all of this from an excellent reference book Supercharged by Corky Bell. does that name sound familiar? Corky Bell is the same guy who wrote Maximum Boost. Arguably the bible of Turbo Systems. Well, he does not disappoint with Supercharged

Pressure Ratio

PR = (14.7psi + boost psi) / 14.7psi (the 14.7psi represents the absolute pressure at sea level)

  • (14.7 + 5) / 14.7 = 1.33
  • 5 PSI of boost has a pressure ratio of 1.34

Another (estimating) way to calculate PR. Divide desired HP by current HP.

  • 220 / 165 = 1.33

That's pretty darn close to what I am hopping for here. And according once again to NASIOC an Most RSs with Turbos and about 5-6 PSI of boost, are making around 220hp. But, the 1.33 represents an ideal scenario. No accounting for heat and VE changes.

Engine Volumetric Efficiency

  • The average modern 2 valves per cylinder engine has a VE of about 80%.
  • The average modern 4 valves per cylinder engine has a VE of about 88%.
  • It turns out that alot of modern OHV engines are getting over 90% VE. Some race NA motors can actually obtain over 100% VE.

Referencing [ this site] I can estimate the VE of an engine if I know it's peak HP and RPM. According to Subaru, the EJ251 has a max 165hp @ 5600 RPM.

  • VE = (HP x 792001.6) / (AP x CR x CID x RPM)
  • AP = Absolute Atmo Pressure
  • CR = Compression Ratio
  • (165 x 792001.6) / (14.7 x 10.5 x 151 x 5600) = 1.00124

Huh? This means that the EJ251 is running at 100% VE? How is that possible? What does this same formula say about HP?

  • HP = (AP x CR x VE x CID x RPM) / 792001.6
  • 14.7 x 10.5 x 1.0 x 151 x 5600 / 792001.6 = 164.7

Duh.. The same values in reverse should have come up with the same numbers. What if I use a more realistic VE of 92%

  • 14.7 x 10.5 x 0.92 x 151 x 5600 / 792001.6 = 152hp

This means that the EJ251 is really only putting down 152hp at the crank? When I look at a Dyno taken of a I-Speed, I note that it has a peek wheel HP of 118 at 4952 rpm. Actually, looking closely at the graph, the peek seems to be at 2 places. 4900 and about 5200. Just so I'm going to stay close to what Subaru claims, and use the 5200 rpm peek.

According to most people on both Nasioc and, the Subaru RS AWD drive line accounts for an average HP loss of about 25%. So, take 152HP and subtract 25% to get 114hp. Pretty darn close to the same numbers. Obviously every dyno provides different numbers, so this is just ballpark figures.

So, what do the experts on nasioc and RS25. think? Looks like 90% is what most of them seem to agree on. Given all of this, I'm going to go with a VE of 90% I think the fact that I have better cams makes this 90% VE very possible.

Blower Volumetric Efficiency

According to the VE graph of a MP62 (which is not the exact same, but very very close), the VE of this blower between 4K and 10K RPM goes between 70% and 90%. As we'll see in the following calculations, I plan to spin this puppy at about 10K. So, I'll use a VE of 90%

Air Flow Rate

Airflow = ((RPM x CID) / 3456) x .90. Recall that the ''estimated'' VE of an EJ251 is 90%.

  • 2500 cc = 151ci
  • 6250 x 151 / 3456 = 273 cfm Theoretical Maximum<
  • 273 * .90 = 245 cfm

According to search results on both NASIOC and, this value of 245cfm is pretty darn close. Most people throw around a value of between 225 and 235cfm. But I do have Delta Cams that should account for the increase. This number can now be used to calculate Airflow at specific pressure ratios. If I want a pressure ratio of 1.35, then

  • 245 x 1.35 = 330cfm
  • 270 * 1.35 = 360cfm - This would be the CFM required for a 100% VE

Now, if I once again look at the Flow Map from Magnusun. I can see that to obtain about 330cfm of air at 5psi, I'll need to spin the SC at about 10,000RPM Perhaps as high as 10,500. Now, I have a few hard numbers to work with.


Now, I can calculate the boost required with the 1.35 pressure ratio.

  • Boost = 0.35 x 14.7psi = 5.145

Once again, this value is darn close to what the RS Impreza community agress is the boost needed to make 220hp. But... once again, no accounting for heat or supercharger VE here.


Lets calculate the heat generated by a Roots Supercharger on an EJ251 with 5psi of boost. Corky Bell made up a magical make believe number of 1075F. This is the max temperature inside the piston cylinder at TDC during the compression stroke that can be created before the onset of Knock. This is NOT a hard and true fact. Just a number that seems to be really close to reality. Obviously things like piston design and quench areas have a big impact. According to Supercharged by Corky Bell, the amount of heat that is added to an intake charge by adding boost can be calculated like this.

  • (PR0.28 - 1 x Tabs) - Tabs
  • PR = Pressure Ratio
  • Tabs = Absolute Ambient Temperature. eg. 90F on an average day is 90 + 460 = 550 degrees absolute.
  • ((1.34)0.28 - 1) x 550 = 47 degrees increase in temperature under ideal conditions.

The average Roots SC has a temperature efficiency of around 55%. An unfortunate side affect of not having an internal compression ratio like a Centrifugal SC. This means it increases the temp by another 55%. This is very much an average though. The DeltT map for the SC of choice should be referenced. The Gen-5 Eaton M62, spinning at 10K RPM & 1.35PR has DeltaT of 60%. I'd call that close enough for my calculations.

  • 46 deg / 0.55 = 84 degrees.

The roots blower will increase the ambient temperature by 84 degrees F at 5 psi of boost. Will it? Lets check my work. If I look at [[ this performance map]] I can see that at 10K rpm & 5psi, the SC adds about 90F. I'd say I'm on the right track here.

I can also calculate the temperature increase from the 10.5CR of my EJ251 engine. I can use the same formula.

  • (CR0.28 - 1) x Tabs
  • (10.50.28 - 1) x 550 = 512 degrees absolute increase.

This means that the increase over 550 degrees absolute ambient (90 on an average day) is 1026 degrees F

Now, given these, I can calculate the temperate of the intake charge after both the SC, and the piston. It's not as simple as just adding them together though. Because the piston is increasing the air that is already heated by the SC

  • (CR0.28 x (Absolute from SC) = In Cylinder temperature.
  • (10.50.28 x (550 + 84)) = 1225 degrees F Absolute.

But, what if I add an Intercooler with an 85% efficiency rating?

  • 84 Degrees F minus 85% of the heat added is 13F.
  • (10.50.28 x (550 + 13)) = 1090 degrees F Absolute.

A Rather huge savings don't you agree. Looks like an IC is likely on the menu.

Calculations with VE Values

I've done a few calculations, and come up with some useful numbers in my project. But, I still have not accounted for the VE of the super charger. Between the above "ideal" scenario, and these numbers, I should get a pretty good idea if I'm on the right track or not. Here is Corky Bells formula.

  • Desired Power = Stock Power x Pressure Ratio x Density Ratio x Volumetric Efficiencies Ratio x Drive Power Efficiency

Volumetric Efficiencies Ratio

Volumetric Efficiencies Ratio = SC VE / Engine VE

  • 90% / 90% = 1.00%

Density Ratio

The Density ratio is the affect of temperature on the air. Despite the poor efficiency of the roots blower, I know I'm going to use an Intercooler. Therefor this number should be quite high (thats good). Remember, I'm using 90 as an average day, and 460 as absolute, and above we guessed that with an IC, the temperature would go up my 13F. Lets round it to 15 for safe measure.

  • Density Ratio = Original Absolute Temperature / Final Absolute Temperature.
  • DR = (460 + 90) / (460 + 90 + 15) = 97%

Refined PR from DR

Refined PR = desired power / (stock power x Density Ratio x VE ratio x Drive Power Efficiency). I'm going to use a drive power ratio of 90%. This is a round average for most modern Superchargers.

  • 220hp / (165 x 0.97 x 1.00 x 0.90) = 1.53
  • Boost = 0.53 x 14.7psi = 7.06psi

Well, after accounting for both the Supercharger VE and the resulting Density Ratio, I can see that to obtain my goal of 220hp with an Eaton Roots M62 blower, I'll need a boost pressure of about 7psi. But, experience and knowledge tell me I'm going to stick with 5psi to start, and go from there.

Pulley Size

Note: 330cfm = 560 meters3/hour

Ok, so I expect that 330cfm with a pressure ratio of 1.34 will give me what I want. Now, how fast do I need to spin an M62 to achieve these values. I've been throwing around 10K above as the value I'd need. What I need to know first is exactly which model of M62 I have. If it's a Gen5,then I can use This Map. If it's a Gen4, I think I can still use the same Map. I've heard that the difference is the bearings, and that the Gen5 can spin higher. The Gen4 maxes at 14K. If it's a Gen3, then I'll need to use a different reference map. The Gen3 makes 330cfm with 5psi at almost exactly 10K. The Gen4/5 makes 330cfm with 5psi at ... uh.. well dang.. the same rpm.. 10K!! Ok.. I admit I didn't see that coming. I guess that makes it easy. I need to spin the M62 at 10K. You know, I should have expected that. M62 afterall refers to 62ci (1L) per revolution. Gen 3 or 4 or 5 should not really have mattered. If the different generations performed differently, then they should have called it something different.

My '00RS has a 5.25" crank pulley and she redlines at 6250. What pulley ratio do I need to make the SC go 10K? The C230 M62 comes with a 3.625" (3-5/8") pulley.

  • 10000 / 6250 = 1.6

This means there needs to be a pulley ratio of 1.6.

  • 5.25 / 1.6 = 3.28125"

I'm going to need a 3.28" (3-9/32") supercharger pulley.

What would happen if I left the 3-5/8" pulley on?

  • 5.25 / 3.625 = 1.448 pulley ratio
  • 6250 x 1.448 = 9050 rpm at the supercharger.

Thats a difference of only 1K RPM! Doesn't seem like much at all! But,it's a difference of about 60cfm. That means instead of the needed 330cfm, I'd be getting only 260cfm. At the low boost pressures I'm looking at, thats a huge jump. What does that mean for Pressure Ratio, Boost PSI?

  • PR = 260 / 230 = 1.13
  • Boost = 0.13 x 14.7 = 1.9psi.

1000 RPM Drop at Redline and I'm down 3 psi from what I'm expecting.

So, I'm looking at a Supercharger pulley in the 3" range. I don't know if I can just go by my exact calculation here, because these all theoretical. I'm going to have to see what happens in the real world to know for sure. This means I"m likely going to have to try about 3 different pulleys. Thats Ok, South Florida Pully HQ has lots to choose from, including a nice quick change setup. But, I'll need to start somewhere? So, I'm going to just role the dice and make a gut feeling guess for the starter pulley. I'm going to start with a 3-1/4" inch pulley. To be frank, I expect this to be a bit big. I'm guessing It'll give me about 3psi of boost. But thats just fine by me. Besides, this is a prototype. I'd rather start lower and work my way up. A new pulley is a heck of a lot cheaper than a new set of pistons.

After all that

So, after all that, I guessed that a pressure ratio of 1.33 would get my power goal of 220hp. This also provided a boost pressure of about 4.5psi. Then, I ran more numbers, and realized that a pressure ratio of 1.48 is needed after accounting for the various VE and DR calculations. This results in a boost pressure of about 7psi. A difference of 2.5psi. Not alot of difference for most SC engines, but a little high for my engine and the generally accepted high boost value on an EJ251.

So now what? Well, it's widely accepted that a stock block and piston EJ251 is happy with 6ish psi or less of boost from a Turbo Charger. Because the 6psi thing is from a Turbo Setup, I can't use it as ''law'' for this setup. 6psi from a turbo is not the same as 6psi from a SC. Mainly due to different heat and density ratios. What matters then is which setup increases the air temperature more? Who ever has 6psi, but lower temperatures not only has a less chance of knock, but also has more air molecules, and thus more power. I'm afraid I don't actually know which setup would make less heat at 6psi. I suspect however, that the turbo setup does indeed have a better heat produced density ratio due to it's internal compression ratio. Again, I'm only guessing here, but I think my SC setup will not have as much power as a turbo at the same PSI.

So, I'm going to take my numbers from above, but keep my boost pressure at a comfortable level. I'm going with 5psi of boost. But all these numbers were not a complete waste of time. It shows me that my goals and 5psi are not way off. But, I'm going to have to settle for more like 200hp. Awe.. gee.. :P Also, it's much easier to add more boost with a pulley change if I think the motor can handle it, but it's pretty hard to pull back boost after the engine has given up the ghost. Next, I'll talk about the parts I'm going to need.


Well, it turns out the 3.25" pulley was indeed not quite small enough. Later on in these pages, you'll see the progress I made, and in part 5, the actual install and startup. What I've found is that the 3.25" pulley is only good for about 2psi of boost on my motor. Actually, I beleive my calculations to be correct, but I forgot to factor in one very important thing. I forgot to account for the pressure drop caused by my intercooler! Oops! It turns out, that the intercooler I chose is far from the most efficient model around. It apparently can cause a pressure drop between 2-3psi @ 12-14psi on a typical turbocharged WRX. So, if I assume a wide margin of error, my goal of 5psi with the 3.25" pulley, minus 1-2psi from the TMIC, comes out to about 3psi of boost. Pretty much spot on what I'm actually seeing. It's a sound theory anyway!!

As we speak, a 2.75" pulley is being fabricated. I base this number on the SC pulley size that RotorFreak (on has on his M62 powered Impreza. I wish I had found that info sooner! D'OH!. So, I need a pulley ratio of 1.9 (+-0.05), which should net me about 7-8psi, minus the 1-2psi from the intercooler, for a total of about 6-7psi of boost. I think I'm on the right track now!