MyE28.com

Tiit wrote: Mar 23, 2026 12:44 AM It's hard to tell what's on the picture.
I wouldn't worry too much "gasket" matching. For example, m52b28 intake runners fit the head very poorly, yet it runs quite efficient.

That is a close up shot looking back into the intake port from the combustion chamber side, valve removed of course.

I have seen a similar mismatch on the M52TUB28. The ports are squared off in the bottom corners and the runners are a rounded oval/rectangular shape. What I don't like is the variation between cylinders. Some are blended pretty well as cast, but others have a step.

At a minimum I will port a B25 manifold to improve this transition and keep it handy for the next time I have one of these B25 intakes off of either car. The B23 manifold is actually pretty well off here as all runners approach the intake ports at the same angle, and I have already cut a long, gradual radius into the bottom of each runner to transition into the larger B25 ports. This should make it easier for airflow to follow the port floor and make the turn through the short side of the intake port. Perhaps I will ignore the step, it would be more of a problem if it was a step in the other direction with a protrusion sticking out into the runner.

B23 is the top pic, B25 is the bottom:





The B23 does have a more consistent angle on the ports compared to the B25. I do like that aspect, each runner has a straight section before the flange. The B25 has a curved runner feeding 1 and 6, and varying lengths of runner on 2-5. 1 and 6 actually have a pretty good angle on the port, where the rest require a turn at the flange. BMW cast in a bit of a scallop on these runners but it's a bit of an afterthought. A smooth slope downward would be more suitable for such a high velocity area.




The scallop is hard to get a good picture of but it is significant. There is no good reason not to do this more gradually and blend it back into the runner.

Third gear is fun.

The shenanigans continue. This is a TiAL F38 adapted to the td manifold wastegate flange:







Not too big a deal. If I were to make another one I could do it better but this will work, and it's got a 1.3 bar spring set loaded up. If I get 1.3 bar over atmospheric it will be hitting 210kpa absolute pressure. If I get a pressure ratio of 1.3 I should see about 190 kpa absolute.  Either way is fine for boost on spoolup, and I can start tuning the onset of boost creep to start kicking in above 5000 or so RPM. 

I may end up not needing the internal gate at all, in which case I'll probably spring for a GT3071R with a slightly larger turbine. I don't know if I want to go as large as I did with the GT35R but I'm very nearly there already with the current turbo. 

These open dumps sure are loud. I had forgotten quite how loud. Everything fits together nicely:



Since I'm silly and I had a segment of aluminum tubing of just the right diameter, I figured why not see if it survives as a dump pipe. This jagged remnant is all that was left when I got back from the test drive:



For boost, it is hitting 20.9 psi. Wastegate opens in third at 3700 RPM. By 6400, it's down to about 17. I am clearly fighting some sort of inefficiency in this system that is either causing boost to overshoot on spoolup or taper off toward redline. The internal wastegates do the same thing. I would like to nail this down before I attempt to push it any harder. The td manifold may be starting to limit what I can do here, though I don't know if I can say for sure at this point. In any case, this thing is juicy. Not short on power by any means.



These are 42lb injectors at 85% duty cycle. By my calculations, these injectors at that duty cycle are delivering about 393 rwhp of fuel with optimal ignition timing and AFR. I'm a little richer than optimal and the timing is conservative, so it's probably in the 350-360 rear wheel range.

Up next, I will be constructing a dump pipe to route the noise out of the engine compartment and hopefully take edge off. I also need to get the det cans back out and see what I can do with timing. It's only getting 3.5 degrees at 220kpa, nothing at all compared to what I used to run back in the days when it was on the ragged edge. I suspect I can split the difference.

 

How comfortable are you with higher duty cycles? Whats your limit, 85-90%?

I have 42lb injectors as well and am targeting a similar duty cycle at 180kpa on m30b35. But I found out mine are actually like 39lb as bmw stock fuel pressure levels are lower than my bosch injectors. 

Sounds like a nice setup. How do you feel the engine is for response at your elevation(like 4500ft right)?  In general that seems to be the trade off with a bigger turbine capable of higher rpm flow.

I'm curious to see the difference in my ongoing setup with a reduction in piping volume and pressure drop through the intercooler. 

if you put that datalog segment into virtual dyno, you could totally estimate the power level at the wheels. Just need to setup the car profile and know weight of the car and tire size. There is some estimate with the frontal area and cD also.

 

I am all too comfortable with higher duty cycles. The highest I've ever seen was 130% on a datalog. I don't know how that works but that's what it said. I think 90% is perfectly fine. It's nice to have some margin.

Elevation is 5700' here. Response is not what it could be, mostly on account of the cam. It uses a later intake closing after BDC than the B25 cam and gives up some dynamic compression because of it. With the relatively hot cam and a big turbine it comes on later, wastegate opens between 3500-4000 depending on the gear. I have no complaints though, it's not hard to get it moving. If I do end up going with a different turbo I probably will go larger on the hot side, it's not at the point right now where I wouldn't consider going further. Even the GT35R wasn't too bad, it made boost from low revs it just wouldn't fully spool until at least 4000.

I downloaded virtual dyno just now, I'll play around with it and see what it says. I've got a pretty good idea where it's at with my previous builds and how they dyno'd at whatever boost level. They're always pretty consistent. I always laugh when I see posts on various forums about how this guy or that guy is claiming 450rwhp out of his B25 at 14 lbs of boost. It just doesn't work like that.

Also, this water/air intercooler arrived yesterday. I have always been a fan of overkill when it comes to intercooling, adding this in series with the air to air in front of the car will surely get the job done:



The core size is 4x9x3.5" thick, 2.5" inlet and outlet. I think I can fit this easily in the area where the stock airbox goes in a 528. If this works as well as I think it will, I can probably use this as the sole intercooler on the 88 and quite possibly get the A/C operational. For this car I would like to keep the front mounted air to air for redundancy and to keep IAT down even at high boost. My original build used two air to air coolers in series and by the time the boost reached the manifold it was never more than ten degree above ambient. That sort of density recovery is exactly what I'm looking for, and low IAT is highly beneficial in avoiding detonation on those deep triple digit top end pulls.

Definitely curious to see the results of your work! Not many W2A setups, but there are some huge benefits depending on the application, especially with the low pressure drop and relatively consistent IAT's.

From my research the larger core size of the IC is really beneficial for long duration as the water begins to heat. But the heat exchanger and pump may have as big of an affect on the temps all depending how long you are boosting. If long 5+ minutes of boost this becomes very important.

Very curious to see how your hybrid cooling setup could look!

Here's a note I saved regarding VirtualDyno smoothing:

Misuse of the smoothing feature can mask incorrect or inflated runs.The smoothing function in VD can be used to determine how accurate your power output is based on the consistency of power from Smoothing 1 to Max. A good rule of thumb is, if your power drops less then 12-15WHP from smoothing 1 to Max, then the run is accurate and will most likely make very similar numbers on a real Dynojet dyno, usually with 0-3%. If your power drops off 20-25WHP or more from smoothing 1 to Max, you should be aware that the VD chart may not be accurate and your power could be off 5%, if not more, from a real Dynojet dyno.

I ordered the rest of the parts for this W/A today. Heat exchanger will be a KX85 radiator and the pump will be one of these: https://www.ebay.com/itm/284937822421.

It seems to be rated for continuous usage and flows about 4.6 gallons per minute. It is cheap and reliability is going to be interesting. The Johnson CM10 will be plan B if this things conks out. Again, with this being a secondary intercooler, it will still perform as a heat sink if the pump dies and the front mount will still be doing most of the work. I can afford to take some chances and see what happens.

I extended the wastegate dump tube today, now it routes all the way down to where the sway bar and crossmember live. I can probably put a little kick out to the side and divert that noise away from the cabin somewhat. It is still pretty damn loud. I'm too old to be putting up with that kind of shit.

It fits very well. I don't have the heat exchanger yet but that's not going to be any trouble. Will stick it right behind the kidney grilles.

There were a few extra bungs in the end tanks so I drilled and tapped one of them for a GM IAT sensor. I made a bracket to mount the core from the bottom which secures to the engine mount stud which is almost directly below. I may make another bracket to secure the core from the top against transverse movement but the boots also help secure this thing.

I need to make a new pipe to relocate the blow off valves and IAC feed but thats all done and ready to weld.

This is a third gear pull with the new setup. IAT is starting to move only slightly by 6000 RPM, and mind you this is with just the water to air core installed. There is no water in it, no pump, and no heat exchanger yet. The air to air in front is doing most of the work, and this thing is just operating as a heat sink. It will be absolutely excessive when the system is completed. Pressure drop is negligible and the car is ripping.

I also bumped the timing up a bit. I felt like it wanted more after installing the external wastegate. This should be significantly more efficient, lowering exhaust manifold pressure and increasing turbine efficiency. I tried running the det cans to see what I can hear but once that wastegate opens it drowns out absolutely everything. I'm not too concerned, this is still less timing than I was running with the GT2871R back in 2008 and it's seeing ambient IAT and it's running a larger turbine with an external gate. Should still be very much on the safe side.

Very encouraging to see that result! Really cool to have that dual cooler setup... the w2a will really help on those longer and repeated triple digit runs!

Great progress. do you like this setup more than the super eta 88 turbo combination? Is it more laggy but better max power overall?

The B25 is pretty sweet. It feels stronger at high revs than the same head, cam and intake does on the super e block. The 2.7 makes better torque off boost though and has more hit in the midrange. I wouldn't go out of my way to build a super eta, but if I had a perfectly good super e bottom end I would use it. At least for me this is a rare treat these days. Between scored cylinder walls, upper cylinder wear, cracked decks and whatever else people have managed to do to these over the years there's just not a whole lot of good ones out there. The jam here is that oversize super eta pistons are very hard to come by while .50 over M20B25 pistons are readily available. That's why this car has a B25 in it, short of having custom pistons made the 2.7 route was just not an option.

Lag isn't much of a factor, these ball bearing units get into the boost without delay. Boost threshold depends on the turbine size but even the big turbos are very responsive above that point. There's no perceptible difference between the 2.5 and 2.7 in that regard, it is likely that the bump in compression helps the 2.5 overcome the slight deficit in displacement.

Hey Dan...u mention the b23 IM in post above...is that the same as the euro e323/m20 manifold we recently discused (and i recently picked up)?

jblunceford wrote: May 29, 2026 2:51 PM Hey Dan...u mention the b23 IM in post above...is that the same as the euro e323/m20 manifold we recently discused (and i recently picked up)?

Good deal...thanks for explaining.

I got the euro manifold in last week and have started cleaning it up. It's in really good shape other than the usual salvage yard oxidation on outer surface. I'll be putting it on as soon as I pick up a late year M20 throttle body.

Yesterday I replaced torn dog bone bushings and installed staggered set of 16" wheels and fresh tires. Now running a 205/55 and 225/50 set up on 7 1/2 and 8 1/2 wheels. This really tightened up feel of the rear end, which i knew felt.a little sloppy...didn't realize how much lateral slop those rotten dog bones created.

I took this thing out for some exercise yesterday to really put the intercooling to the test. I've also been incrementally increasing the ignition timing. I'm still conservative compared to what I was running back in 2008/2009 when I had no qualms about replacing head gaskets at a moments notice.

The best evidence was a 2-3-4 pull. The speedo doesn't indicate correctly anywhere above highway speeds but 6200 rpm in 4th with the 3.25 rear end would have been about 140 mph. IAT was completely flat through second and third and began to increase through fourth gear. The total IAT Δ was +5.1 degrees from 40-140. IAT was climbing steadily by the top of fourth but not quickly. Could have grabbed fifth and kept pulling, its got a lot more in it. Afterward I was pleased to discover nominal cooling system pressure with no coolant loss, so the tune is still on the safe side.