Standard Disclaimer: Your mileage may vary, what worked on my car may not work on yours. I assume no responsibility if you hurt yourself or your car.

Bottom line up front: A superb $1880 intercooler option from ETS

Shameless vendor plug: AWD Motorsports should be able to hook you up with these coolers, make sure to specify the 16 row option

After searching around a bit for an all metal intercooler that didn’t break the bank but could also control IATs better than some of the current options, I found an interesting picture of a 16 row cooler from ETS. Note: No solicitation was involved here, I was just interested in this unique looking design and the price was right. ETS has been in the intercooler business for a while now and recently expanded their coverage to the Porsche and Nissan GTR. The few Porsche 996TT ETS cooler installs I’ve seen have had the usual 12 row configuration with less fin density –visually similar to some of the Bell based intercooler options. I spoke with Michael at ETS a bit about the differences. Basically the 12 row coolers were an early run that used an off the shelf core designed for very high flow while performing decently in IAT control. Realizing they didn’t have a radiator sitting behind the intercooler in the 996TT, they elected to redesign & tighten the external fin pack significantly. Also, they added more charge rows to increase the surface area exposed to the cooler atmosphere for even better heat transfer. To do this, the height of the charge row was reduced to fit 16 rows. With the reduction in row height, the density of the internal fin pack was reduced a touch to provide flow levels similar to that of the 12 row coolers. Note: This is not a knock on the 12 row cooler, this is theory only & I simply haven’t tested the 12 row. Enough theory, on to the hardware!

Some measurements on the ETS cooler:
Weight: 17lbs (each!)
Core Length: 14.5”
Core Width: 8.8”
Core Thickness: 3.5”
Charge Rows: 16
Row Height: 0.28”
External Fin density: 15 fins per inch (each row is independent & staggered)
Internal Fin density: 10 fins per inch (each row is independent & staggered)

Pictures are worth a thousand words:







Since the 997.2 / GT2RS intercoolers are pretty well proven now, I elected to throw these ETS coolers to the fire and test against Porsche’s latest, rather than throwing them a softball by comparing them to stock 996TT coolers (if you want to know how the 997.2s did against stock, see here: https://www.6speedonline.com/forums/...ml#post3001579 )Not going to lie, when I saw the internal density of the ETS coolers, I was worried. Keep reading.


Measurements from the 997.2 cooler:
Weight: ~6lbs each
Core Length: 13.25”
Core Width: 8.625”
Core Width: 3.25”
Charge Rows: 14
Row Height: 0.30”
External Fin Density: 22 fins per inch (staggered but offset from the same set of fins (perforated?))
Internal Fin Density: 19 fins per inch (same staggering method)




Some comparison pics:

 

Installation:

With my ducts already modded for the 997.2 intercoolers, install was a breeze –well, assuming you’re good at removing the bumper . If you’ve got stock coolers, you’ll need to grind down the inner lip on the inlet and outlet ducts like on pretty much any other aftermarket intercooler –this is a good thing as it allows manufacturers to use slightly wider cores. I had to trim my Forge hose slightly at the end tank entry –of course, I’ve had to trim my Forge hoses everywhere else too so this probably isn’t isolated to these coolers. Beyond that, everything else is bueno, including using the original metal straps to hold the ducts together.

The test setup:

2001 996 Turbo, 41k miles, fresh plugs & coils
Low clearance K16/2280 ETT Hybrids @ 1.3 bar (60-130s vary from mid 6s to mid 7s based on where, when & how I do the run)
EPL Tune
2.75” Exhaust with 100 cpi cats
2.5” Inlet pipes
AP Dual Snorkel Airbox
Forge IC hoses
100 octane unleaded (R+M)/2 gas
764 Clutch
997.2 intercoolers, then ETS 16 row intercoolers
~4000ft density altitude –I tried to test on days where temps and pressure were as close as possible. Also used the same location for the pulls.

 

The test:

All done on a closed course.

997.2 Conditions: 31-33C Alt: 30.05
ETS Conditions: 31C Alt: 30.05


Short term gains were tested by doing 60-130 mph pulls with a 3-4 shift in between. I chose the runs shown below because the ambient temps, barometer and starting IATs were nearly identical for these pulls.





At 130mph, the ETS coolers were at 46.5C vs 52.5C for the 997.2s. I tried to execute the 3-4 shift as quickly as possible for both runs, this is where I noticed what I think is a bit quicker boost recovery between shifts with the ETS coolers. This is indicative of higher flow capacity but a bit anecdotal. 60-130 time was 0.3s quicker (most of it probably from the post-shift boost recovery).




Short term gains are nice, but can these coolers perform under some sustained loading? Longer term gains were tested by doing pulls from ~50-170mph to try and capture where the temps peaked. My shift points differed a bit on these as I wound out the ETS setup a touch more. Ambient temps were within 2C on these pulls with the slightly cooler ambient advantage going to the ETS coolers.





Here you can clearly see 17lbs of ETS thermal mass at work. The ETS coolers have a much more shallow IAT increase line as the coolers slowly build heat. Also note the quicker response as I’m back on boost slightly quicker after each shift. 60-160 was a couple of seconds quicker with the ETS coolers. The lightweight 997.2 coolers don’t control the slope as well, and you can see that at each shift, but when you get above ~150mph, that high fin density really goes to work for you and starts to even things up. If the ETS is a great heat sink, how well does it work as a heat exchanger? The temperature recovery time may tell us more…



I’m still working to get more an accurate recovery time comparison as I was usually on the brakes too hard with the ETS coolers for a direct comparison. However, there are a set of runs where I was able to capture the time it took to get from 50C to 45C right after a set of 60-130s that left IATs just above 50C with both coolers in 30C ambient conditions.

ETS / 997.2
93mph 78mph :Average Speed
9.8s 5.2s :Time from 50-45C

This is very limited data but it also points in the direction of the results seen in the 50-170 runs. With the huge amount of thermal mass, the ETS can absorb more heat but it also appears to take a bit longer to recover. The speed should have given the ETS coolers an advantage but the 997.2 still cooled more quickly in this case. I’ll get some more data for this region shortly. (On a side note, I would love to see some recovery times on 5” coolers!)



Thus, so far:

ETS 16 Row Pros vs 997.2s: Better IAT control up to ~160mph, robust, huge thermal mass, less complex duct mods, likely better flow

ETS 16 Row Cons vs 997.2s: $800 more, heavy, probably not as quick a recovery after loading


If you’ve got stock intercoolers and are looking for a robust, relatively inexpensive intercooler option with great temp control for your 996TT, the ETS 16 row coolers fit the bill very nicely. My hunch is that they will bloody the nose of some much more expensive coolers.


If you’ve already got 997.2 coolers, the waters get a bit more muddy. The end tanks will never be an issue if you go with a big build, temps are better thus far and the car is a touch quicker and more responsive with the ETS option, yet recovery time may be longer. You’ll have to decide your goals for the car and if it’s worth the R&R and extra $800.

 

Nice write up Earl. The EPSs are 22 lbs heavier for the pair than the 997.2s. That's a lot. It seems the ETS would probably be a bit better for street cars or drag strips but the 997.2s world probably be better for running hard at the track for 20+ minutes where dissipating heat soak faster is more important....

 

Nice one again.

 

Thanks Earl excellent write up.

 

I wish I could afford to get you a set of the 911tuning 4 1/2" units to test as well, as I have always been curious about the really thick offerings out there.

 

i will buy your 997.2 if you are done with them

 

HI,Thanks for info..

 

Excellent testing and reporting!

 

Earl, I'm assuming some of the boost related differences have to be because the 997.2 IC's were being short shifted, which will certainly cause a loss in RPM's which cause a loss in boost pressure/response.

Looks like a 3rd to 4th, 4th to 5th and 5th to 6th shift. 3rd goes to about 106, 4th to about 134 and 5th to 168 ish. The ETS coolers were being shifted close to redline each time and closer to the powerband.

pdwrhound- I agree that the 22 lbs is not ideal, but you spend a lof of time during a lap at WOT, I would think the IC with the lower IAT at WOT would be better unless you mix in a lot of cool down laps. Benefit ETS.

 

My view point is as follows........

The ETS flow better than the 997 intercoolers as their fin density isn't anywhere near as great

The ETS intercoolers will cool better than the 997 ones to begin with because their end tanks absorb heat (and the 997 end tanks do not)


I believe if you were track racing the ETS intercoolers may produce worse results than the 997 items

My reasoning is that the 997 items greater fin density will keep on cooling long after the ETS end tanks become saturated with heat

To summarise
I believe the 997 intercoolers work by transferring heat away from the intercoolers
Whereas the ETS intercoolers work by absorbing the heat

NOTE - This is nothing more than my poorly informed opinion, I have been wrong on many an occasion before, and could well be wrong again here


Earl, what do you think?

 

I guess the question is what would happen to the performance of the ETS once their high thermal mass is fully heat soaked as they would get after a couple of minutes of wide open lapping. Would their performance decline below that of the 997.2s which will dissipate heat faster during the sections of the track where you are at partial throttle of no throttle at all when under braking? I tend to agree with Ant..

 

did you do any standing 0 to 150 or 160?
I noticed that how they are pre heated will alter the result.... just curious....
not to spoil anything but you know I had a car go thru my hands with those ICs and you need to pass it on because the clips did not fit right and thats after us massaging the end tank.. even then the hoses popped 2 days later on the client... the clip doesn't seat properly.... I would address that if it hasn't been done yet....
btw, Ill do a similar run on his k16g with 5 bar and exhaust just to see how it compares..
good job

 

I did a few other pulls with lower shift points on the ETS and it looks similar to the plot above. The post-shift rpm in all cases were well into the boost threshold with these turbos, but I get what you're saying. I barely thought I felt better response, and the logs showed a slight difference so hopefully I'm not imagining things.

Tough to say on the lapping thing, you would probably start out with more power with the ETS but the lines may well cross a few laps in depending on the track layout.

I tend to agree with this in the ETS is a great heat sink and a good heat exchanger and the 997.2 is a good heat sink but a great heat exchanger.

The flow thing may be true here, but I would caution assuming less dense cores flow better. The 997.2 outflowed a slew of less dense, thicker and shorter coolers (all attributes that should have made them flow better) that, visually, should have easily won in flow (more on that later). It may have been the smoother tube and fin entry design or end tanks helping out, who knows.

Other than strength, I think metal end tanks are a negative. Heat conducts into them quickly (good) but then seems to take a long time to get out. The inlet tank is awful close to the turbo hotside as well. Even with a shield, I think they probably absorb unwanted heat sitting at WOT. With composite end tanks, the heat pretty much stays in the core, where all the fins are. Metal tanks may have been part of the reason for the longer recovery. Of course if you lose an end tank, game over!

Thanks Mark! The standing runs added ~1C to each cooler. The effect of 1st and 2nd gear was almost unmeasurable. Have you seen different results (hotter from a stop?)?

I definately heeded your advice with the end tank clip recesses after you brought up the issue. I tested all 4 connections on the bench with a solid "click click" from each bayonet seating past the clip several times. Installed them on the car and pressure checked them without any problems. Maybe they got some feedback and changed the design?