carbonwood

I finally completed my homebrew supercharger kit these past few weeks. Installed an M62 from a 99 Mercedes Benz SLK. The engine runs well and the blower produces such a pleasant whine. However, I seem to be low on power. My car has never seen a real Dyno, but I datalogged some pulls and VirtualDyno shows only a gain of +25 hp resulting in the engine producing just 125hp. This is odd, because I'm making 8psi of boost on an M62 supercharger with water/meth injection. It certainly feels faster than stock. Engine pulls pretty well in the low end, but I can feel the torque drop off after 4200 RPM. There are Miatas with similar setups that make +180hp. And there are Miatas with M45 superchargers on 6psi that make 150-170hp. With 8psi at the manifold, I'm sure I should see more than a gain of just 25hp.

Engine management is a SpeedyEFI Speeduino ECU. Installed RX-8 Yellow injectors pulled from a junkyard car. I verified ignition timing with a timing light and it's spot-on. I performed a compression test on the engine, and it's at 175psi across all cylinders. At WOT, AFR's are at 12.0 ~ 12.2. I've attached my spark table and VirtualDyno graph showing a supercharged pull + naturally aspirated pull that was recorded before the s/c was installed. The n/a pull shows peak power of about 100hp, which would have me believe VirtualDyno is somewhat accurate. I'm not sure if I'm missing something. Is it possible that VirtualDyno is way off? Where's all my power at?

yossi126

iTrader: (6)

That intake looks restrictive, and the blower's port too. Maybe start from there?

SpartanSV

Post the tune and logs including the N/A one.

der_vierte

Stock exhaust? Cat?
And no IC...Intake doesn't look good aswell.

Changed timing belt? Maybe a cam is a tooth off?

carbonwood

Exhaust is Racing Beat headers, no cat, and ISR catback. No intercooler because I'm on water/meth injection. Injecting pre-blower.

Not sure why everyone is saying the intake looks restrictive. It's all 2.5 inch piping going into and out of the supercharger.




The inlet reaches past the firewall into the wiper cowl for Randall-style induction. The air filter sits inside the wiper cowl so it is a little small, but that cannot possibly account for a 50hp restriction... right? I have a couple larger air filters laying around so I can try to do a pull with one of those and see if it makes any difference.

I did change my timing belt about a year ago. Everything lined up perfectly when I did it. When I get home I will double check cam timing. Maybe it jumped a tooth or something. I'll also upload the tune and datalogs.

RunninOnEmpty

iTrader: (5)

Have you checked for a clogged cat?

Assuming a hotside M62 is similar enough to an MP62, and that's a stockish 1.8, 8psi is roughly a 1.615 pulley ratio (sc pulley / crank pulley). If your OEM crank pulley is that standard 130mm, your SC pulley is roughly 80mm. If it's larger than that (like 5mm increments), then there might be something else going on with the boost readings.

Thread drift - any reason you put the check valve between the catch can and VC? I figured you'd want it between the IM and catch can. Then the can doesn't see boost, and the oil vapor isn't soaking the flapper on the check valve. I'm probably wrong though.

carbonwood

Engine has no cat, has Racing Beat headers, and ISR exhaust system all the way back. This is a 1.6L. I'm pretty sure the MP62 and SLK M62 use the same rotor pack and are identical in function. They both move 62 cubic inches of air per revolution. However the SLK M62 pulley is about 95mm. The 1.6L crank pulley is stock, standard 130mm. My pulley ratio is about 1.35.
I hadn't considered putting the PCV valve between the catch can and the manifold. Not sure it would make a difference as I don't think the catch can seeing boost is an issue. Haven't had any problems with the location of the PCV valve. I may reevaluate my thinking, though. Thanks for bringing it up.

Stealth97

iTrader: (5)

That’s a really cool concept with the blower clutch. Curious if you have a bypass? With the clutch disengaged and no bypass that’s going to be massively restricted compared to NA- the vacuum of the engine will turn the blower.

Couple things I can think of… is your boost reading accurate? With a pulley that size on the blower I’d be surprised if it is really pushing 8psi. Secondly, intake and outlet restrictions on these blowers make a massive impact. Your outlet manifold does not look like it would flow well.

EDIT- with that blower pulley I really, REALLY doubt it’s making 8 psi. And that right angle in front of your blower on the intake site has got to be choking it.

these blowers don’t care how many RPM you turn, fixed amount of air per revolution. Torque is going to drop off as RPM increases… especially with that pulley ratio.

curly

iTrader: (25)

I’m guessing that’s kpa, not psi like labeled, which would make 140=5.8psi. You might check to make sure your other VD numbers are correct.

carbonwood

Yep it's kPa. Datalog of that pull shows a peak of 145kPa which is 6.3 psi. Let's consider the pull is at 145kPa/6psi instead of 8psi. That still doesn't explain why the engine is only producing 125hp. Even at 6psi, power should be higher than a gain of just 25hp.

I also double checked my cam timing. Crank pulley lined up with TDC mark. Removed cyl #1 spark plug and confirmed #1 piston to be at the top of it's travel. Intake and exhaust Cam gears are lined up with E/I marks respectively. There are 19 belt teeth between top marks of cam gears. However, the belt seems to have a very slight amount of extra slack in it, so I will reset tension on it.

carbonwood

Yes I have a vacuum operated recirculating bypass valve from a Mini Cooper S.

In this thread, OP has an NB 1.8 with exactly the same supercharger and pulley ratio as me.
https://www.mx5nutz.com/threads/my-m...-build.348425/


His Dyno pull shows his engine is producing 170hp without any intercooler or water/meth injection. He's at 7psi with his less restrictive NB cylinder head. My engine is a 1.6, so taking into account the restriction of the 1.6 manifold and cylinder head, 7/8psi sounds about right to me.

oreo

Cats on the nice build.
I really like your custom bits, and everything is nicely laid out.

Regarding your dilemma, it's hard to do anything but guess at the problem.

1. First issue could be related to your measured boost, compared to your buddies. If he is accurately measuring 7psi at a particular RPM, I would expect that you should
be measuring more at that same RPM. Your water injection will reduce temps a bit, reducing measured boost, but your engine is smaller so you should definitely be seeing more boost than him.
Things to check; belt slip, malfunctioning or poorly set bypass valve, restrictive supercharger intake, leaking hoses on supercharger outlet, poor supercharger rotor sealing.

2. Too much water injection? I presume you have already played around with this. Tchaps has an interesting build titled "Mmmmm Lysholm...." where he goes into detail
about how much injection his engine liked, which may prove to be beneficial reading. Actually his whole thread is an interesting (but long) read.

3. FWIW, I found with my SC setups, I got the most power running 12.5:1. They were higher boost, however intercooled using 94 octane.

Good luck! Searching for missing power is never fun, but finding it is.

About 3 yrs ago, I was helping a mechanic who had a 1.6 custom SC miata, which was not making the power expected either.
We tried a ton of things but never were able to get the car to where it should have been. It was frustrating.

I notice you have a temperature sensor between the SC and the intake manifold. What kind of temperatures are you seeing?

carbonwood

Thanks for your insight. I think it's possible the bypass valve is leaking a little bit and I may be experiencing belt slip around 5000 RPM. I'm going to order a new bypass valve since they're pretty cheap. The one I have is ratty and old. I'm also currently using an off-the-shelf Continental belt. To remedy the potential belt-slip issue I might consider a Gates RPM belt. Do you have any experience with one of those? They're supposed to have tight tolerances and be made of superior materials suitable for superchargers and high torque applications... Or so their marketing department would have you believe.

I have the basic Snow Performance water/meth kit installed in my car. When I got the kit, it came with a large nozzle, and VERY large nozzle. I installed the smaller of the two, which was Snow Performance's Size 4 Nozzle, appropriate for 400hp cars... The IAT's were great! (100F in boost and steadily decreasing when the WMI pump is running.) But the engine did not seem to like the large volume of water/meth. I ran with it for about a week while as I waited for the Size 2 nozzle to come in the mail. Engine likes the new nozzle and my current IAT's in boost hit a maximum of 120F when it's hot out. If the weather is cool outside, IAT's are around 100F. If the water/meth in the tank is also at a cool temp, I noticed it had a substantial effect on IAT's. Anyone ever throw ice cubes in their meth tank? lol

carbonwood

So I swapped to a larger air filter, removed the exhaust tip silencer, changed spark plugs (NGK BKR7E gapped to .28), confirmed cam timing is correct. Attached is a fresh VirtualDyno pull with files of datalog and tune (Speeduino tune). Still seem to be missing a bunch of power. In the datalog, you can see manifold pressure drop off after 4000 RPM. Is that indicative of belt slip or is that just caused by the stock cams?

*Please note, my tune has "incorrect" req. fuel and VE numbers for RX-8 injectors. The values have been scaled to accommodate higher resolution of the VE table to help with idle.

oreo

You will see belt dust if your belt is slipping.
I suppose it could be the belt itself, but for most custom installs, it's a problem with belt alignment.
Sorry, I have no experience with Gates RPM belts.

The bypass can be tested pretty easily. You just want to make sure it closes and seals properly when in boost
(just test that the arm on the bypass is fully closed with the engine off, and that the stop on the bypass is set correctly).

Tchaps

I ran some very very *** packet maths, and I think you could expect to see more boost on average from this setup.

You are running a pressure ratio of approx 1.5PR currently.

With your pulley ratio of 1.35, SC capacity of 1.016litres (sorry I am metric, mostly ) a volumetric efficiency guess of about 90% for both sc and engine at peak eff and your 1.6l engine, PR came out at 1.7.

But that doesn't account for heat added to the air charge, so you would expect PR to actually be higher than 1.7 (I estimated up to 2.1PR with an outlet temp of 100degC)

Also your boost curve to me suggests an intake restriction. I don't have a data point for a 1.6, but I know from reverting back to my stock 1.8 BP4W head on my car that boost climbs as rpm increases because my sc has a higher VE than than the engine. You are operating your M62 quite nicely in its efficiency island, so I would expect to see a similar trend. An undersize air filter will restrict more at higher flow rates, so as rpms increase it will reduce output more and more. FWIW roughly for every 1kPa of intake pressure drop I see a reduction in output pressure of 2kPa. Reducing my intake restriction by about 3kPa netted me a 10hp power increase or about 3.3%

So I would look at your air filter and bypass. Try running just a bellmouth on the intake pipe (just for a quick test) to see if you see an increase in pressure ratio and power.

Also I have seen it stated from the old NACA papers that WMI does not significantly increase charge density, like an intercooler. So if at the moment your system is achieving a PR of 1.5, but with the charge density of 80-100degC air, you would not expect to see the 50% increase in hp a 1.5PR at atmospheric temperatures roughly would achieve. Logically I think there is sense in WMI not increasing density. It works by evaporating a liquid (dense) into a gas (not dense), the evaporated water and meth will take up more space and exert more pressure in the intake tract as a gas than a liquid, counteracting the less space the air takes up because of its reduced temp.

carbonwood

What kind of math did you do to get those numbers? I just want to make sure I understand what you're saying. Do you mean my pressure ratio is actually lower than it should be when you account for air density/temperature?

If you're correct about my PR being off, changing the intake to eliminate or reduce the restriction should improve the efficiency of the supercharger right?

Also, I think it's worth mentioning I'm currently residing in Tucson AZ. Altitude here is 2300 ft. My MAP sensor measures atmospheric pressure is a maximum of 91kPa. Atmospheric pressure is somewhat low and air is thin here. And I know my MAP sensor is accurate because it reads 99kPA at sea level.

I will revise my intake and start plans to mount an intercooler so I don't have to rely so heavily on WMI. In the meantime, I went ahead and ordered a 76mm supercharger pulley. It will put me at a 1.71 ratio. Not entirely sure how much more additional boost it will produce but we'll see in 2-3 weeks.

Tchaps

It was just a really basic calculation:

Capacity of sc per rev: 1.016litres
Volumetric eff: 90%
Pulley ratio: 1.35

Capacity of engine per rev: 1.6litres/2=0.8litres
volumetric efficiency at peak tq = 90% (bit of guess)

(1.016*90%*1.35)/(0.8*90%)= 1.7PR

But that assumes the air is still at ambient temp. I made a isentropic eff calculator and I can't remember the maths behind it, but I plugged in your PR, flow rate and a guess at eff and it came out at 100degC

I then just used p1/t1=p2/t2 to get to 2.16PR, p1=1.7, t1=20+273, t2=100+273

It is likely a gross oversimplification when I benchmarked my engine with the above I got a PR of 2.5, and really the highest I see is 2.1PR currently, so it seems to estimate a bit high. It will depend on all sorts of things like cam timing and relative VEs of the engine and sc and how they coincide, and its probably just not that simple, pressure loss in pipework etc.

Your 90kPa ambient pressure will absolutely drop your boost levels - so if you are making 145kPa at 90kPa pressure that is a PR of 145/90=1.6PR - starts moving you closer to my estimation.

Yes any intake restriction basically lowers your ambient pressure further, so say you have 10kPa intake restriction at WOT - that means your PR to get 145kPa is actually 145/(90-10)=1.8PR

Things start getting into the realms of realistic deviation from the basic calcs above.

Out of interest do you see a difference in boost pressure with and without the meth?

Tchaps

I don't know how much or if you have read any of my build thread on here but I used some engine simulation software and so far on my engine the sim vs dyno results have been very close.

I set the model up for your 1.6 engine with 9.4CR, supercharger size, pulley ratio, lack of intercooling and 0.9atm pressure. It does make a few assumptions like the flow coeff of the 1.6 head is the same as a 1.8 BP4W and also it uses the 1.8 BP4W cams, but I corrected the model to use 1.6 engine valve sizes.

Red ringed are the average absolute pressures in bar for 5k and 7k rpm - about 0.4bar of boost. I think this is basically exactly the boost you see? This is using your 90kPa atm pressure and then up to a 3kPa intake restriction at 7krpm, which would replicate a pretty good intake setup. My sim predicts the boost on my setup very closely also, so apologies for the last post thinking you should be making far more boost, I won't be bothering with the *** packet maths again. Maybe you could get a pressure reading from your intake setup to confirm what pressure drops you see? Would be very useful info. Do you see AITs of about 80degC without the WMI spraying?

The torque curve in my sim is peaking a bit later - I think the actual 1.6 cams are not as big as the 1.8 BP4W,aswell as the ports being smaller which may well explain this. 158Nm is 116.5lbft which I think is pretty close to your peak torque from your virtual dyno testing. You just seem to loose out in the top end power compared to my sim, which reckons 115kW / 154whp, again probably a portion of this better top end in the sim will be the bigger cam and port size in my sim.

One thing I noticed on your virtual dyno plot is that at peak torque there is a bigger difference between your NA and supercharged torque lines than at redline. If you were making the same +23lbft over NA you do at peak torque at redline you would be at around 144whp. A clue that maybe your intake is restrictive, as this would affect power more and more as rpms increase.

So I don't think you are a million miles off where you should be, but maybe some careful optimisations can yield improved results. Hopefully that helps a bit more than my initial estimations.

Tchaps

Sorry for spamming, but I got hooked and went through the sim again updating a load of efficiencies, port sizes, cam profiles etc for your setup with the 1.6l engine.

When I incorporated these changes into the model it reproduced your virtual dyno information very closely. This is at 1.35 pulley ratio, 90kPa atmosphere with an intake pressure drop of 4kPa by 7krpm (a decent intake setup).

Boost levels (absolute) stayed around 0.35 to 0.4bar. Peak torque was moved earlier and slightly increased to 121.7lbft @4krpm, torque at 7krpm was identical to your vd testing at 103lbft. With the smaller ports and cams now in the sim it caused it to replicate your power curve very closely, reducing top end power. Sim says 137whp @ 7krpm. If you notice in your VD graph hp and tq do not calculate correctly. It says 124whp / 103lbft, but 103lbft @ 7krpm is (103*7000)/5252= 137whp, not 124whp. I think this is a quirk of how VD smooths the lines, I think it does them separately rather than smoothing torque then calculating power from that or vice versa.




Long story short, I don't think your power figures are unexpected any more. Move to the seaside or put a smaller pulley on it