[derryck]

Indeed...kittyless DPs make a pretty big difference and add a nice sound.

[2Times]

I thought you were saying the cat-back was a restriction.

[AWSAWS]

Please offer us a stage2+ Turbo at stock Turbo prices.

Then this thread will make sense.

[mike@x-ph.com]

Gotta give props that OP is doing his own testing rather than relying on other companies and just copying their work

[Nonecure]

Originally Posted by mike@x-ph.com

Gotta give props that OP is doing his own testing rather than relying on other companies and just copying their work

+1

R&D cost money, but is the only way to develop something better. It looks like he is trying to figure out the N54 the hard way. By documenting what he finds, if he is wrong about anything at all, this group will be quick to point that out. Therefore this forum is a good double check of his findings. I also give him props on ignoring all the negative stuff and stick to learning this engine.

[mike@x-ph.com]

Originally Posted by Nonecure

+1

R&D cost money, but is the only way to develop something better. It looks like he is trying to figure out the N54 the hard way. By documenting what he finds, if he is wrong about anything at all, this group will be quick to point that out. Therefore this forum is a good double check of his findings. I also give him props on ignoring all the negative stuff and stick to learning this engine.

Yes!

[bigdnno98]

Agree. I only know one other vendor that has measured BP and even then, I don't think I've seen it measured on stock turbos and stock DP. This is going in a very good direction. Thanks! And keep up the good work.

[doug@frankenturbo]

Some work-in-progress analysis using Virtual Dyno software. Considering the bone-stock hardware, the car has come up in power quite a lot.

Stock factory map:




JB4 Map1




JB4 Map6 (16psi)




Next up: another form of data measurement...air masses.

Thx

[doug@frankenturbo]

Earlier in the thread I mentioned airflows testing. Given the N54 engine's lack of a mass air meter I'd guess that sounded a little unrealistic to some. But airmass values through the engine are a vital testing metric. So if the car didn't come with an airflow meter, well then we'll just have to put one in.

We have a lot of experience with mass airflow meters. Also known as MAFs, virtually all the VW/Audi cars we support came from the factory with them. And since FrankenTurbo is in the business of moving air, I've made it my business to know these devices really well. And actually they are quite simple to set up "standalone". These VW/Audi sensors operate on a 0-5v gamut, rendering a measurement curve such as this one:




I've chosen this particular MAF assembly because its 60mm inside dimension matches that of the N54 inlets' openings. They too are 60mm, which simplifies this modification. And with a maximum measurement of close to 200 grams/second this should give us the headroom we need while on the stock turbos. For comparison, the VW 1.8T engine this sensor comes from will generate about 170g/s on its stock turbocharger. And that's while feeding a significantly larger displacement than the individual N54 turbochargers do.

We set up the MAF sensor using a simple 0-5v gauge, which will report the values coming back from the sensor.




As you can see, while on the bench the sensor returns a value of 1v, consistent with the scaling graph above. Then, the bone-stock test car became bone stock no longer. We removed the airbox and fit the MAF assembly to the bank #2 turbo's inlet pipe.









The airflows to the MAF sensor are now provided by a pair of BMS open-element filters. This was the most practical way to lay out the intake system for testing. And with the configuration complete, we tested the car on its 16psi boost map. The voltages reported by the gauge were tabulated and converted to mass airflow values.

But then it was time to check on a question that's been on my mind for months: is there an output difference between the two turbos? I've been wondering this ever since we arranged for flow-bench testing of the stock inlets. That testing showed a surprising disparity between the two stock inlets' airflow capacities. Many of you have probably seen the test results -- and the results of subsequent testing of aftermarket ones -- but here they are again:



That's a significant difference. But is there really any way for those results to translate to the engine bay? I was really skeptical. But with the equipment in hand to test, we switched the MAF sensor from the bank #2 inlet to the bank #1 inlet, taking care to lay out the flexible hose and filter so they matched the earlier arrangement.




Here are the findings. Firstly, let's cover off questions of consistency between the runs. We made several passes measuring both banks, then selected the two which had the most similarity for ambient conditions and intake air temperatures. Here are all the data documenting the test conditions:



Virtually identical test conditions. And given that consistency I fully expected to see the airflow values for the two banks to "spoon" together just as cozily. Well, I was wrong. Have a look:




Incredulous, I decided to check the airflow output of that bank #1 turbo when the engine and IATs were much hotter. Even then, it is outworking the rear, bank #2 turbo for much of the rev range.




So the flow-bench testing revealed a design irregularity that actually impacts the turbochargers. And it does so at pretty modest boost levels. Given this data, I consider it a design necessity to have well-balanced intake assemblies. Clearly the troubling imbalance at 16psi would only worsen at higher boost levels.

Thx

[bradsm87]

Great work.


I'd love to see exhaust back pressure measured in the same spot, but with catless downpipes and secondary cats still present.

[sirdaft1]

OP- I originally came into this thread expecting to see the development of a new turbo offering, and gave some flak in earlier posts because it wasn't that. But now seeing the extent of testing that is being conducted, starting with square one... huge props. Love seeing the new data being shared (bank 1 vs bank 2 turbos)

I apologize and look forward to future tests that continue to push this platform forward (and hopefully some hardware to accompany in the near future?!? )

[Jeff@TopGearSolutions]

Originally Posted by doug@frankenturbo

So the flow-bench testing revealed a design irregularity that actually impacts the turbochargers. And it does so at pretty modest boost levels. Given this data, I consider it a design necessity to have well-balanced intake assemblies. Clearly the troubling imbalance at 16psi would only worsen at higher boost levels.

Thx

The rear bank is not only longer so it should need to work harder but it's common on high mileage vehicles to see the rear turbo inlet disconnected or at least partially fallen off, which could skew your results even further. Either way, yes the inlets are imbalanced for sure.

Double check yours.

[Billit335]

Any pics what to look for, sorry if it's been covered already

[doug@frankenturbo]

Time to do an update on the state of modification of the car. As I said in the last post, it received a set of Burger Motorsports' open-element filters. While the airbox was out it seemed a good time to ditch the stock throttle charge pipe with an inexpensive upgrade part from forum sponsors Extreme Power House.






Being all thumbs with a wrench, I quickly ruined one of the stock diverter valves, forcing replacement of both with a pair of our budget-friendly piston-style units.




I was a bit worried that the conductive aluminum throttle charge pipe might affect the MAP/IAT sensor attached to it. And the logs show a minor touch of increased temps. But performance isn't hurt by such a small hit on IATs, so it's a small price to pay for a proper high performance part.

Logs before/after:




And here are the times to speed for before/after:




For those of you asking, hey, where's the added power? I'd say it's too soon to expect a benefit from these first mods. There is still one big bottleneck that needs to be cleared away first: the exhaust. So the throttle pipe offers plenty of solid capability as modding moves forward. And its integrated WMI nozzle bung is a nice plus. The intakes don't change the sound of the engine much, which I also favor. That's because I'm old. And I was even pleasantly surprised by the unobtrusive sound coming from our metal diverter valves. At this boost level they don't announce themselves in the least bit obnoxiously.

Next, that stock intercooler just had to go. One of our suppliers -- a terrific performance parts workshop in Shanghai -- asked us to try one of their N54 intercoolers. I told them I required whatever unit would install the most simply. So they sent us a 130mm cooler that has a taller section at the very front.






Without a doubt this dimension intercooler has a big advantage. Even for me, installation was easy. And it is a lunker compared to the feather-light OEM unit. As a heat sink, I have to think that mass will move the needle a bit. Here are the data:



Nice! IATs are significantly better. And with that improved air temperature management my butt-dyno-sense was absolutely tingling! Big improvement! No doubt. so let's see how this bad boy's time-to-speed looks now!




Ahhh, crap. Butt dynos are worthless.

[sirdaft1]

Originally Posted by doug@frankenturbo

Nice! IATs are significantly better. And with that improved air temperature management my butt-dyno-sense was absolutely tingling! Big improvement! No doubt. so let's see how this bad boy's time-to-speed looks now!



Ahhh, crap. Butt dynos are worthless.

Was this a single gear pull? Curious to see the data from a multi gear pull, as that's when the disparity in performance between an upgraded FMIC vs stock really becomes glaring.

[doug@frankenturbo]

Originally Posted by bradsm87

I'd love to see exhaust back pressure measured in the same spot, but with catless downpipes and secondary cats still present.

Will do.

Originally Posted by sirdaft1

Was this a single gear pull? Curious to see the data from a multi gear pull, as that's when the disparity in performance between an upgraded FMIC vs stock really becomes glaring.

All data we present is done in a single-gear RPM sweep. Unless otherwise noted, the testing is done in 3rd gear on the 6MT transmission of our test car.

[doug@frankenturbo]

We've moved on to a third revision of our Map 6 testing with the JB4 piggyback controller. And with this pass, the hardware limits are presenting themselves. Here are the logs, showing boost unable to meet requests and boost duty scaling very high.



And even the FATs values & Time-to-Speed have suffered:






Interestingly, though, the peak EGTs we're picking up pre-catalytic converters are hardly alarming...




These readings are celsius. And even if you allow a generous 100˚ loss between the rotor exit and the cat entrance, that puts us at only a touch over 800˚c. Those are very reasonable temperatures. So my assumptions about overworked turbos and blazing hot EGTs on this platform could have been completely unfounded. Everything thus far points to acceptable stresses on the components -- at boost pressures that are DOUBLE factory levels.

These are nice cars.

[BMDoublYou]

Crazy info, keep it coming! with so much data to back up your claims, these turbos look like they could really be a solid buy!

[litxus]

Keep it coming, I guess inlets / outlets / fueling and suspension coming in next?

[RWDMIKE]

Nice work here, Doug. EGTs are surprisingly low!

[DennisPacMo]

Welcome to the N54 community Doug. I'm familiar with your products having owned numerous VAG products in the past. I'm excited to see what new and exciting products emerge from having you in the community.

[doug@frankenturbo]

After the false start with an incompatible OBD cable, we're on track now with the MHD Android application. To start, I went with what's cheap...uh, free. I downloaded the MHD flash files linked by BMS in this thread and found the appropriate one for the car's IJE0S version type. After switching the JB4 into Map 0 for passive mode, I loaded the file labeled "Pump" onto the car's DME.




After about a half-hour of flashing, the car was ready to try out. Because the previous logging data showed a shortfall in actual boost, I wanted to try upping the boost command level even higher. Perhaps "doubling down" on the commands might tease a bit more solenoid duty from the control system. Here are the logs of this new configuration:




The standout data points are boost and ignition timing. This latest version of a JB4 Map 6 is clearly asking for more boost than the hardware can develop. And the ignition timing map is very clean with zero corrections. Here are side-by-sides of the Pump Gas map's data compared to those from the stock DME.




So overall the new DME mapping is resulting in better-looking data. And the car feels plenty quick through the 4200-6500rpm sweep. In fact, it's at its fastest level yet:




I was interested, though, in how the new timing map might impact the low end. Reduced timing before 3000rpms is oftentimes good for spooling up the turbos. But it costs torque development. Would that show up in the Time-to-Speed metric?




Yes it does. The car's sub-3000rpm performance has taken a hit. But the new calibration is unmistakably making up for it in the midrange. And the timing pull that was nettling the DME on stock mapping is now eradicated.

Next up: more hardware changes. It's time to explore exhaust mods.