[FaRKle!]

I've been playing with intercoolers over the past half year, and at my last track day at Laguna Seca I ran the

View post on imgur.com

. This past weekend I ran ATM's intercooler along with Wagner's EVO 2 Comp at Thunderhill (east). Both days were pretty hot (~95F) and I was able to get some good sessions in that were very comparable. All of these intercoolers will work with both the N20/26, N55, and N47 engines, but you may need to be aware of slightly different inlet/outlet locations compared to the OE intercooler when using aftermarket turbo to intercooler and/or intercooler to intake pipes (in particular ATM and CSF).



I used the Torque pro app in conjunction with an OBDLink LX bluetooth OBDII reader to log data at 4Hz. The car only reports temperature in 1C (1.8F) increments. From the data read, the "charge air cooler temp" is the temperature of the air after the intercooler (sometimes seen as IAT elsewhere). This was compared to the ambient air temp to see how well the intercoolers were cooling the pressurized air.

The car is a '17 328dx wagon with JR Tuning's stage 2 flash. The turbo will boost up to 29-30PSI. The turbo to intercooler and intercooler to intake manifold pipes are stock, since both are of the flexible rubber hose type on the 328d, and it makes swapping intercoolers much quicker (I was able to do the swap in under 30min at the end of the first day).


Here's a plot showing the difference in performance between the intercoolers. The ATM is consistently lower than the Wagner by 7-11F. I was surprised at this result given the EVO 2 Competition's description. One thing I noticed that ATM does on their core is that they have rounded header bars inside the end tanks to make the flow into the core smoother.


Here's the raw temperature data showing the charge air temps and ambient air temps. These two sessions were really well matched ambient temperature wise.

Over the next week I'll be logging commute data with the Wagner (got the ATM's data last week) so I can do a comparison on performance in "every day" driving.
Daily Driving Performance Link

[n00bkiller944]

Very cool! Are they priced fairly similarly as well?

[FaRKle!]

Originally Posted by n00bkiller944

Very cool! Are they priced fairly similarly as well?

I paid $755 for the Wagner and $629 for the ATM. I've seen the ATM another 10% off (~$560) during better sales.

[Jpup]

In your opinion, how does the CSF compare to the Evo 2 and ATM? Hard to compare via the temps since it was tested at a different time but would be curious to hear your opinion given you have driven with all 3.

[mike@x-ph.com]

Nice comparison OP, thank you for taking the time and sharing it.

You forgot to compare pressure air drop and the weight between the 2.
Both should easily be in favor of the Wagner since it is a tube and fin core.

[FaRKle!]

Originally Posted by Jpup

In your opinion, how does the CSF compare to the Evo 2 and ATM? Hard to compare via the temps since it was tested at a different time but would be curious to hear your opinion given you have driven with all 3.

It's heavier (25.8lbs, vs 24lbs for the ATM, 18.6lbs for the EVO 2 Comp, and 7.8lbs OE), has a higher fin density (accounting for the increase in weight over the ATM since it shares the exact same dimensions, ATM came out first), and is more restrictive than both the ATM and Wagner. This made the lag more pronounced on it, which was most noticeable in daily driving.

Temperatures-wise I can't really compare it with the other two since the conditions were so different. I suppose you could try and do some hand waving by seeing how far over ambient the CACT got with it vs the ATM and Wagner, but that was at a different track, with non-tuned car at the time.

Someone from my other thread asked how much of a difference the IC can make. Looking at data from the same day (so same IC), in the morning when it was cooler (78F ambient, 118.4F CACT) my max speed on the front straight was 6.75mph higher than the max speed I reached later on in the day (95F ambient, 129.2F CACT). Across all laps between that morning session and this afternoon one, the morning speeds were 6.25mph higher on average.

[FaRKle!]

Originally Posted by johnung

I've been researching intercooler upgrades for the N55 so I am very curious about your test results. My understanding is that the ATM is a bar & plate design while the Wagner Competition2 is a tube & fin design. I have the impression that the Wagner should have better heat recovery during repeated pulls and, maybe, less pressure drop. Curious what your findings are?

The Wagner does flow better. This isn't observed on the track (since RPMs are kept high enough that you're never really starting from 0 boost), but is most observable in daily driving (such as accelerating from a stop). Looking at the track data, the boost between the ATM and Wagner was the same, and one didn't seem to really rise/fall significantly faster than the other.

In daily driving though I clearly noticed a difference in how low my mpg meter would drop going from a stop before the system was pressurized. The Wagner drops to about mid to high 8mpg, the ATM low 8mpg, and the CSF would do mid to high 7mpg. This correlated with turbo lag, or how long it felt before the power came on.

From the graphs above, the Wagner tube and fin design doesn't show any benefit with regards to heat recovery, and even if it did, it's baseline level is still as high/higher than the highest ATM reading.

[johnung]

Originally Posted by FaRKle!

I paid $755 for the Wagner and $629 for the ATM. I've seen the ATM another 10% off (~$560) during better sales.

Thanks for all the great IC information. Your cost on the Wagner has me puzzled. I've seen prices around $800 for the Wagner EVO2 "Performance" which is a bar & plate design, and around $900 for the Wagner EVO2 "Competition" which is a tube & fin design. They look very similar. Both have the same dimensions of 520mm x 215mm x 145mm. (Stock IC is 510mm x 130mm x 110mm)
Is it definitely the Wagner "Competition" model that you tested? Did you get some amazing price for it new, or did you find it used? Thanks.

[johnung]

Originally Posted by FaRKle!

The Wagner does flow better. This isn't observed on the track (since RPMs are kept high enough that you're never really starting from 0 boost), but is most observable in daily driving (such as accelerating from a stop). Looking at the track data, the boost between the ATM and Wagner was the same, and one didn't seem to really rise/fall significantly faster than the other.

In daily driving though I clearly noticed a difference in how low my mpg meter would drop going from a stop before the system was pressurized. The Wagner drops to about mid to high 8mpg, the ATM low 8mpg, and the CSF would do mid to high 7mpg. This correlated with turbo lag, or how long it felt before the power came on.

From the graphs above, the Wagner tube and fin design doesn't show any benefit with regards to heat recovery, and even if it did, it's baseline level is still as high/higher than the highest ATM reading.

This is really great stuff! Thanks for posting. My car may never see a track. All of my mods are to make it a great daily driver/ street performer. So I'm most curious about your daily driving data and daily driving impressions of each IC.
What I think I hear you saying is that the Wagner Competition2 (tube & fin) is noticeably more responsive in daily driving because it flows better so the boost increases quicker from a standing start. I believe that the smaller stock IC may be the only other tube & fin design that you have tested. How does the stock IC compare in terms of flow and responsiveness? (I am thinking that maybe BMW uses a stock tube & fin IC to lessen perceived turbo lag in daily driving.) The Wagner Competition1 and Competition2 models are the only tube & fin upgrade IC's that I have come across on the aftermarket. Have you seen any others? I'm curious to read your comments on the IC's and their daily driving characteristics. Thanks again!

[rbreding]

Interesting info farkle. Thanks for the writeup. Don't you just wish the f30 didn't have electronic steering.

[IK6SPEED]

Originally Posted by FaRKle!

I've been playing with intercoolers over the past half year, and at my last track day at Laguna Seca I ran the

View post on imgur.com

. This past weekend I ran ATM's intercooler along with Wagner's EVO 2 Comp at Thunderhill (east). Both days were pretty hot (~95F) and I was able to get some good sessions in that were very comparable. All of these intercoolers will work with both the N20/26, N55, and N47 engines, but you may need to be aware of slightly different inlet/outlet locations compared to the OE intercooler when using aftermarket turbo to intercooler and/or intercooler to intake pipes (in particular ATM and CSF).



I used the Torque pro app in conjunction with an OBDLink LX bluetooth OBDII reader to log data at 4Hz. The car only reports temperature in 1C (1.8F) increments. From the data read, the "charge air cooler temp" is the temperature of the air after the intercooler (sometimes seen as IAT elsewhere). This was compared to the ambient air temp to see how well the intercoolers were cooling the pressurized air.

The car is a '17 328dx wagon with JR Tuning's stage 2 flash. The turbo will boost up to 29-30PSI. The turbo to intercooler and intercooler to intake manifold pipes are stock, since both are of the flexible rubber hose type on the 328d, and it makes swapping intercoolers much quicker (I was able to do the swap in under 30min at the end of the first day).


Here's a plot showing the difference in performance between the intercoolers. The ATM is consistently lower than the Wagner by 7-11F. I was surprised at this result given the EVO 2 Competition's description. One thing I noticed that ATM does on their core is that they have rounded header bars inside the end tanks to make the flow into the core smoother.


Here's the raw temperature data showing the charge air temps and ambient air temps. These two sessions were really well matched ambient temperature wise.

Over the next week I'll be logging commute data with the Wagner (got the ATM's data last week) so I can do a comparison on performance in "every day" driving.

Yes, ATM’s are the best secret out there that gets very little exposure on forums.

[FaRKle!]

Originally Posted by johnung

Thanks for all the great IC information. Your cost on the Wagner has me puzzled. I've seen prices around $800 for the Wagner EVO2 "Performance" which is a bar & plate design, and around $900 for the Wagner EVO2 "Competition" which is a tube & fin design. They look very similar. Both have the same dimensions of 520mm x 215mm x 145mm. (Stock IC is 510mm x 130mm x 110mm)
Is it definitely the Wagner "Competition" model that you tested? Did you get some amazing price for it new, or did you find it used? Thanks.

I definitely have the competition model. The weight is a giveaway that it's not the performance bar & plate version. I paid $755 by purchasing the intercooler from X-PH's eBay store. I got their forum member discount, and combined it with a $100 off eBay coupon. If there's no eBay coupon to stack with, you can get a better forum member discount by buying directly from X-PH.

Originally Posted by johnung

What I think I hear you saying is that the Wagner Competition2 (tube & fin) is noticeably more responsive in daily driving because it flows better so the boost increases quicker from a standing start. I believe that the smaller stock IC may be the only other tube & fin design that you have tested. How does the stock IC compare in terms of flow and responsiveness? (I am thinking that maybe BMW uses a stock tube & fin IC to lessen perceived turbo lag in daily driving.) The Wagner Competition1 and Competition2 models are the only tube & fin upgrade IC's that I have come across on the aftermarket. Have you seen any others? I'm curious to read your comments on the IC's and their daily driving characteristics. Thanks again!

I wouldn't say that the Wagner is very noticeably less laggy than the ATM (if you're concentrating on it then you can notice it a bit). I also can't really compare with the OE intercooler, because I never ran it with the Stage 2 tune, although I definitely remember lag with it. The only IC where the lag really bothered me was with the CSF (which I did run with the Stage 2 for a short bit). You'd get nothing from a start, then a huge surge and jerk when it suddenly pressurized. That wasn't pleasant for daily driving.

Originally Posted by rbreding

Interesting info farkle. Thanks for the writeup. Don't you just wish the f30 didn't have electronic steering.

The steering isn't as great as others, however I found that adding a front strut bar definitely added more feeling into it. That mod was TOTALLY worth it (and cheap too!). Adding Superpro camber adjustable bushings didn't do anything really for the steering feel though.

[johnung]

Originally Posted by FaRKle!

I definitely have the competition model. The weight is a giveaway that it's not the performance bar & plate version. I paid $755 by purchasing the intercooler from X-PH's eBay store. I got their forum member discount, and combined it with a $100 off eBay coupon. If there's no eBay coupon to stack with, you can get a better forum member discount by buying directly from X-PH.



I wouldn't say that the Wagner is very noticeably less laggy than the ATM (if you're concentrating on it then you can notice it a bit). I also can't really compare with the OE intercooler, because I never ran it with the Stage 2 tune, although I definitely remember lag with it. The only IC where the lag really bothered me was with the CSF (which I did run with the Stage 2 for a short bit). You'd get nothing from a start, then a huge surge and jerk when it suddenly pressurized. That wasn't pleasant for daily driving.



The steering isn't as great as others, however I found that adding a front strut bar definitely added more feeling into it. That mod was TOTALLY worth it (and cheap too!). Adding Superpro camber adjustable bushings didn't do anything really for the steering feel though.

I know what a front sway bar is but I'm not familiar with the "front strut bar" that you mention. What is it? and where did you get it?

[FaRKle!]

Originally Posted by johnung

I know what a front sway bar is but I'm not familiar with the "front strut bar" that you mention. What is it? and where did you get it?

You can see my strut bar (by KC Design) in my log post (about halfway down).

[FaRKle!]

Part II: Daily Driving Performance

To evaluate "daily driving" performance I logged my temps/ect. during my weekly commute. I divided the commute data into morning and evening data sets. The intercoolers were run in back to back weeks, but on average, the ambient temps the week with the Wagner were a bit cooler (for the plots the two closest ambient temp runs were chosen).

My commute is about 18 miles with about half city streets and half freeway. The freeway has a hill in it that rises about 486ft in 2.2mi one way and 531ft in 2.5mi on the other. On the initial part of the freeway in both morning and afternoon commutes traffic is moving slowly, including going up the hill. I only hit actual freeway speeds about the last 3-4mi on the freeway.

Morning Commute Temp Plots




In the morning temps you can see that the ATM has a consistent temperature advantage over the Wagner. That said, in city streets the advantage isn't much. Of note, large spikes in the CACT temps during city street driving is due to no airflow while sitting at a stop light. I wouldn't read too much into the magnitude of it, or the difference in it between the ATM and Wagner since it falls so quickly once the vehicle gets moving again.

Evening Commute Temp Plots




Ambient temperatures in the evening were higher, especially where I live. City street commute speeds leaving my office are usually pretty slow, and the temps show both intercoolers performing roughly the same. My guess is the speeds aren't high enough to push enough airflow through the ATM to show an improvement over the Wagner that's seen everywhere else, such as the second city street section.

One interesting thing to note, is that where the CACT over ambient delta is rising is when I'm in traffic going up the hill, but the peak is when I'm going down the hill coasting (well, coasting and braking as there's still lots of traffic). Temps don't start falling till I reach constant higher/normal freeway speeds, instead of the 15-40mph fluctuations before that. The Wagner doesn't show any advantage in dissipating heat over the ATM in lower speeds or higher ones. Once again, any sharp, short spikes are the temps while sitting at a stop light, which dissipate quickly once moving again.

A Note on Lag
Measurements elsewhere have shown that the Wagner EVO 2 Comp should flow a bit better than the ATM. Paying attention to lag while accelerating from a stop, and also watching my fuel economy meter seem to confirm that. Accelerating form a stop my fuel economy meter drops to about mid to high 8mpgs with the Wagner. With the ATM it's in the low 8s. The lag is slightly more with the ATM, but it's not large and annoying, and if you're not paying attention to it, probably won't notice a difference from the Wagner. How much of this is due to actual internal restriction, and how much is due to pure volume (the ATM is larger) is a guess to me. I did notice that the ATM uses rounded header bars in the end tanks on the core to help flow (the Wagner has nothing special on the core).

It looks to me like the ATM is the superior intercooler.

[SeanWRT]

OP, thanks for the great info.

One thing you misunderstood about the pressure loss - you can never measure it by post intercooler boost reading.

They always read the same boost, because that's what DME is targeting. Where it makes the difference is pre intercooler boost that you don't get to measure. Turbo would work harder for more boost to make up for what's lost going thru a bar and plate core, so the post intercooler boost target would be met.

In fact, the mpg reading you came up with is a very interesting and effective way of showing lag which ultimately means pressure loss. High mpg means more fuel gets burned, means more air goes thru IC and gets into cylinder.

Another point is that you compared them with a high boost car. Intercooler performs very differently to different boost level. There is actually such a thing as a too big IC for the boost you're looking to make. I'm not familiar with diesel engine and how diesel motor boost translates to flow rate lb/m and how hot they come out of turbo. Someone really needs to conduct similar tests with a N20 on 22-20psi or N55 on 18-17psi for a proper comparison. These boost level are what pump gas can support at top end at maximum. Above that, ethanol and meth are logical choice over enlarging your intercooler.

Thanks again for the great info.

Sean

[FaRKle!]

Originally Posted by SeanWRT

OP, thanks for the great info.

One thing you misunderstood about the pressure loss - you can never measure it by post intercooler boost reading.

They always read the same boost, because that's what DME is targeting. Where it makes the difference is pre intercooler boost that you don't get to measure. Turbo would work harder and make more boost to offset what's lost in going thru a bar and plate core, so the post intercooler boost would be the same and meets the boost target.

In fact the mpg reading you came up with is a very interesting and an effective way of showing lag which ultimately means pressure loss. High mpg means more fuel gets burned, means more air gets into cylinder.

Thanks for the feedback. I would imagine that if you have higher pressure loss (and thus lag) you would notice it in the post-IC boost rate of rise. When I originally looked at my track data, I was looking to see how quickly I reached max boost when getting on the throttle again. In that data set I didn't see any difference. It was probably there, but not resolvable at 4Hz (and I definitely couldn't feel it).

Originally Posted by SeanWRT

Another point is that you compare them with a high boost car. Intercooler perform very differently to different boost level. There is actually such a thing as a too big IC for the boost you're looking to make.

Absolutely! Unfortunately the N47 we get is a single-scroll turbo, and thus lag is more noticeable. When I ran the CSF IC, which is the same size as the ATM, but more restrictive due to non rounded internal header bars and a more restrictive fin pack, I noticed LOTS of lag in daily driving. That IC was too large and restrictive for my single turbo.

[Enabled]

Variable geometry turbos as in our N47s are superior even to twin scroll turbos. We have infinite scroll regulation.
The VNT/VGT turbos just get very expensive to build for them to withstand gasoline EGTs. Porsche started using them on their gas cars, and greatly reduced their lag.

Lag that is introduced by changing the volume of the tract to the intake can be improved by tuning. Factory flash turbo control was designed for the factory setup. I doubt anyone is tuning that.

[SeanWRT]

There is another Wagner vs ATM comparison with a different result. Interesting.

https://f30.bimmerpost.com/forums/sh....php?t=1141360

[FaRKle!]

Originally Posted by SeanWRT

There is another Wagner vs ATM comparison with a different result. Interesting.

https://f30.bimmerpost.com/forums/sh....php?t=1141360

Yup I'm well aware of his testing of the Wagner and CSF ICs.

And I have to say, after my own personal experience with all of those (excluding the EVO 2 Perf), I'm a skeptic of his reviews and level of independence from the manufacturers. That said, there are also testing methodology differences.

For the test you linked in particular, the data runs are pretty short. A couple of single gear pulls, followed by a three gear pull. How much cool down time was between each of the pulls? With my on-track data you know I was out there the full 20min session and going WOT often. My dataset is also a lot more comprehensive. My summary for each IC is that from a whole day of track testing, looking at all the laps, finding the trends, and presenting differences with clear signal. For the every-day commute driving I looked at a whole week's worth of commuting both in the morning and afternoon/evening. Additionally, for the direct comparison plots, the ambient temps in my tests are much closer than the other test's (and you can see the ambient temp's values throughout the entire plot). So that's two full weeks of data for my tests vs one day for the other's. I don't feel the other test's test cases are representative of actual every-day driving or track driving.

Another reason I'm skeptical is that the test was done at Wagner Tuning, in their presence, and using their equipment. That opens the door for manufacturer influence. My test was completely independent. I've spent $3671.84 (intercoolers, track days, metrology) of my own money doing intercooler tests this year. There's also the amount of time I've spent looking at the data, bringing my lift with me to the track to be able to swap ICs, and other soft costs.

Lastly, I have no financial stake in the results. The tester in the other test is a vendor that happens to sell both Wagner and CSF intercoolers.

[CSF Cooling]

There are other companies out there who have made objective videos proving the performance of the CSF radiator.

Would also venture to say that more pro-am M2/235 guys are using CSF than any other system on the market.

Watch the video to the end, and you can see with the intercooler + stage 1 remap gained 33.1 whp.. Don't see other companies with data like this to support their products.

https://www.youtube.com/watch?v=oKNE1oGqmbw

[FaRKle!]

Originally Posted by CSF Cooling

There are other companies out there who have made objective videos proving the performance of the CSF radiator.

You should link them. Aside from my comparison vs my stock IC, all I've seen are highly suspect tests where the results provided financial incentives for the tester or had manufacturer influcence. Everybody knows an aftermarket IC will do better than stock, but a question answered here is how do a couple aftermarket units compare against each other.

Originally Posted by CSF Cooling

Watch the video to the end, and you can see with the intercooler + stage 1 remap gained 33.1 whp.. Don't see other companies with data like this to support their products.

https://www.youtube.com/watch?v=oKNE1oGqmbw

How much of that 33.1hp is due to the remap and how much is due to the IC? The 2.3hp increase at stock with just the IC is pretty much within the noise IMO. More importantly, how close does the airflow being sent through the IC on the dyno match that in real world driving (on the street and track)? That's a huge contributing factor. The dyno results really don't say anything about your IC compared to how much it promotes the remap.