DCT fluid flow diagram - Aux DCT Cooler install

[AndyW]

Anyone have a diagram so I can figure out which hose is the goes-in and which is the goes-out? Thx!

[houtan]

I don’t know if this is correct but I found it in my folder of stuff I found researching online. The e9x m3 has a dct cooler working in conjunction with the heat exchanger so it seems we would add one in the same location ( before or after the heat exchanger). A diagram of that setup would be helpful I think.

Also not convinced dct temps are your issue so looking forward to seeing some logs of the boost issue and dct temp.

[mKilgore]

I know this is a bit besides the point of the specific question you are asking, but the BMW DCT is becoming a popular swap among non-BMW cars. I say that because there is more information out there now than in the past couple of years. Check out places like "Seems Legit" Garage. They have made coolers, adapters, hoses, etc. And they also could be a good resource to reach out to for any questions you have. They may not know what the specific target temp is for a 135i ECU to cut power, but overall some of these shops that are doing DCT swaps may have more knowledge than the e9X M3 forums, etc. Just another option to consider.

[AndyW]

Originally Posted by houtan

I don’t know if this is correct but I found it in my folder of stuff I found researching online. The e9x m3 has a dct cooler working in conjunction with the heat exchanger so it seems we would add one in the same location ( before or after the heat exchanger). A diagram of that setup would be helpful I think.

Also not convinced dct temps are your issue so looking forward to seeing some logs of the boost issue and dct temp.

Thx much! I was about 80% convinced this was the flowpath already. I'll verify with an IR temp device before I cut anything.

I actually agree with you on the DCT temp...I don't think it's the issue even though you can see it getting to 255F on the logs. But now that I know that coolant temp will cause the car to go into "pre-limp" as low as 242F, and my water temps got to 248 at one point and there seems to be a strong correlation between the behavior and water temp at or above 241F, I think coolant temp is causing pre-limp.

Why is this coolant temp an issue now after the CSF rad install last year which clearly lowered coolant max temps by 20F? Especially since I just installed a new water pump and thermostat? I have a theory.

1. More heat due to PS2.

2. The EVO III FMIC is massive (and basically locked IAT's below 100F every session on an 80+F ambient day) and blocks about a quarter of the radiator. Can never get something for nothing...

SO why go after DCT temps?

1. BMW clearly did on the M2/3/4.

2. Max temps went like this:

Oil 262F
DCT 255F
Coolant 248F

I could upgrade to 25 row vice 19 row oil coolers (and still may) but I figured I'd address the DCT since it is directly cooled by the coolant. IF I can lower DCT, I will reduce heat load on the coolant system and hopefully stay out of pre-limp territory. At least that's my current theory.

I have another track day on the 2nd and I hope it is HOT. That way, when I log the TQ Lim's it will tell me which system is telling me to slow down.

BTW, do you have any hot day logs at the track I could compare with mine? You have PS2, CSF rad, and DCT, correct?

BTW BTW, I got the 6293 EKP code again. This is after replacing my EKPM3 module and adding cooling. I decided to monitor current with Pro Tool and even at idle, it will cycle between ~12A and ~21A. Seems weird that it would max to 21A on idle. I am going to email Fuel-it about it since the stg 2 pump is new. I hope I just need to flash the module again or something.

[AndyW]

Originally Posted by mKilgore

I know this is a bit besides the point of the specific question you are asking, but the BMW DCT is becoming a popular swap among non-BMW cars. I say that because there is more information out there now than in the past couple of years. Check out places like "Seems Legit" Garage. They have made coolers, adapters, hoses, etc. And they also could be a good resource to reach out to for any questions you have. They may not know what the specific target temp is for a 135i ECU to cut power, but overall some of these shops that are doing DCT swaps may have more knowledge than the e9X M3 forums, etc. Just another option to consider.

Thx much! The DCT "Bible" they have doesn't specify when it goes to pre-limp, but it does say normal is 175+-15. So above 190F is hot.

[houtan]

I don’t have any hot track day logs unfortunately. The last one I did it was around 75 degrees and my oil temps never went above 250 so I didn’t bother logging. I also run around 17ish psi at the track, and more importantly, I am not a very fast driver, so not sure if they would be useful. I am also on the stock oil cooler.

Another thing to consider, which I may do in the future, is get rid of the heat exchanger and just run a thermostat and oil cooler for the transmission and then you could upgrade the radiator to a manual transmission version so it’s solely dedicated to the engine. right now a part of the radiator is dedicated to the heat exchanger.

Interesting to see your oil temps were so high with the mosselman set up.

WRT to the 190 being hot for the dct, I remember reading some information that said the heat exchanger fully opens at 204° which is much higher and what’s considered hot in the DCT Bible. Not sure which one is accurate there.

[AndyW]

Originally Posted by houtan

I don’t have any hot track day logs unfortunately. The last one I did it was around 75 degrees and my oil temps never went above 250 so I didn’t bother logging. I also run around 17ish psi at the track, and more importantly, I am not a very fast driver, so not sure if they would be useful. I am also on the stock oil cooler.

Another thing to consider, which I may do in the future, is get rid of the heat exchanger and just run a thermostat and oil cooler for the transmission and then you could upgrade the radiator to a manual transmission version so it’s solely dedicated to the engine. right now a part of the radiator is dedicated to the heat exchanger.

Interesting to see your oil temps were so high with the mosselman set up.

WRT to the 190 being hot for the dct, I remember reading some information that said the heat exchanger fully opens at 204° which is much higher and what’s considered hot in the DCT Bible. Not sure which one is accurate there.

I think you are correct wrt DCT temp. And 262F is good with the Mosselman setup. I noticed I had similar trans and coolant temps on a hot track day last year. Didn't see any boost cuts. However, Oil temps were 280F and IAT's got to 190F. This was before the CSF Radiator, I believe.

https://datazap.me/u/andyw/very-hot-...zoom=1614-8780

Here's one with the CSF Rad but before the new oil coolers. Cooler day also. First was in the 90's, second was in the 79's, I think.

https://datazap.me/u/andyw/tnia-ridg...zoom=3129-8880

[AndyW]

OK, so the GTR apparently has a similar DCT and the same fluid (FFL-4) and I found a thread here on DCT temps.

"Some observations: the transmission oil not-to-exceed temperature, aka change immediately, has always been 284F. Engine oil not-exceed-temperature is 266F. The change more frequently temperature is 230F for both engine and transmission. For 2009, the interval is 1800 miles, for 2017 3000 miles. If engine oil goes above 275, the car limits RPM to 4000. The manual states that when transmission temp gets to 284, the driver will be alerted, However, the vehicle can continue to be driven until the temperature reaches 295°F. (Scary)"

So DCT fluid is pretty tough, apparently.

[AndyW]

So, I bit the bullet and ordered the stuff to put a 13 row Setrab in front of the FMIC (since it has the most margin now.

[houtan]

Looking forward to hearing how it goes.

[AndyW]

Originally Posted by houtan

Looking forward to hearing how it goes.

Didn't get all the parts in for yesterdays track day, but it's definitely the car cutting power. All three TQ limits at 512!

https://datazap.me/u/andyw/4th-gear-...zoom=4749-4809

[houtan]

Not sure if you already know this but tq limit 512 translates to: mdk_kuehl_mx - cooling. Don’t know which one, trans, coolant, or oil, that is referring to but that’s the translation.

[AndyW]

Originally Posted by houtan

Not sure if you already know this but tq limit 512 translates to: mdk_kuehl_mx - cooling. Don’t know which one, trans, coolant, or oil, that is referring to but that’s the translation.

Ooo...where may I find the magic decoder ring for these?

[AndyW]

Aux trans cooler has solved the issue...now on to replacing my cracking rotors...

Here is a log from my fastest lap of the day, a personal best, actually.

https://datazap.me/u/andyw/hot-laps-...om=11098-11776

It seems apparent that cylinder 4 is my least behaved cylinder from a timing perspective. I actually had a cyl 4 misfire today so this makes sense. Guess the next iteration of my tune is gonna have to dial it back. Or I start using Boostane like JPuehl does...

[houtan]

We need pictures of the cooler set up!

[houtan]

Originally Posted by AndyW

Ooo...where may I find the magic decoder ring for these?

Let me see where the info is in the xdf.

[mKilgore]

Andy, did you run the accusump in the log you posted yesterday?

[AndyW]

Originally Posted by mKilgore

Andy, did you run the accusump in the log you posted yesterday?

Yes. And a VAC baffle plate. I did forget to turn on the Accusump for one session.

[AndyW]

Originally Posted by houtan

We need pictures of the cooler set up!

I know...I was in a hurry to get everything back together so I could get to the track!

[mKilgore]

Originally Posted by AndyW

Yes. And a VAC baffle plate. I did forget to turn on the Accusump for one session.

Ok thanks. I was looking at the log you posted and trying to see if I could locate where the accusump was kicking in. It looks like you have various places where the oil pressure dips to 43-45 psi under deceleration, but I couldn't quite tell if the accusump was kicking in to bring the oil pressure back up, or if the pressure was coming back up because you were on throttle again. It looks like at about time point 1920 there are a couple of blips on the oil pressure that could indicate the accusump was kicking on, but idk. I'm not sure what the oil pressure trace is supposed to look like with the accusump.

Just some observations from an untrained eye trying to learn lol. Regardless, I always enjoy seeing other people's logs from their track days.

[AndyW]

So, based on houtan highly technical and not cartoonish in any way drawing, I made a thing (to quote Jeremy Clarkson)...

So, I realized I needed additional cooling for my track car. Oil was under control with two 19 row Setrabs, and coolant had been mostly under control with a CSF radiator...until I added a Wagner EVO III Race FMIC (https://www.1addicts.com/forums/show....php?t=1810459) and a PS2 turbo. Then, on 85+F days, coolant would get into the 240's ("pre-limp" can start TQ Lim as low as 242F, apparently) and now DCT temp was getting into the 250's(Still don't know the setpoint for TQ limit for DCT temp, but I imagine it's somewhere above 245F, given when power cuts would occur). Since DCT temp was now my highest and I cleverly sussed out that BMW actually adds a decent DCT cooler to it's M cars, I went in that direction. I also knew that since there was a coolant to DCT cooler, anything I did to help DCT temps would also unload the radiator.

First problem is that our little car is cramped. There is little room to mount another cooler if your wheelwells are already full. Since my IAT's were now well under control and I saw some pics on the internet of race cars with DCT's that put their DCT cooler right in front of the radiator, I figured putting a cooler in front of the FMIC would be OK since it had the most "headroom" to spare now. I definitely didn't want to block the coolant radiator more than it already was as that would be counterproductive (although there is enough room to fit the cooler behind the kidneys).

Now to decide what kind. Since there is very little room now between my FMIC and the bumper cover, I had a conundrum. I could stay stockish and get a tiny thin cooler like this:



Or sacrifice the lower "grille" and get something meaty. You would be amazed at how much time I spent figuring out how thick various coolers were and what would fit in the less than 1" between the FMIC and the bumper cover.

In the end, I went with meaty and got a Setrab 13 row. I figured a 19 would block more of the FMIC than was desireable and the 13 seemed to be "right". In hindsight, given how little effect this addition had on IAT's, I could have gone with a 19".

I used some metal strap pieces I had in my parts bin and put M6 rivnuts in the underside of the bumper. Here is the assembly (before I cut out the extended holes to get the spacing where I wanted).



Here is what it looked like fitted up.





I used zip ties to provide some bracing to only having one bolt holding it to the bumper on each side. I didn't want it to "flop".

I reused my horn mounts from the PPK2 kit as zip tie supports...



Getting the Setrab meant I would sacrifice my lower grille, but that also took away the "baffle plates" on the ends of that piece to minimize air going around the FMIC. I solved that by cutting the center out of the grille with a dremel and keeping the side pieces (see final pic below). I did have to modify the bumper cover as the grille clips stick out too far.



Red indicates where i cut off the clips and in the center, cut forward about a quarter inch.

This was actually a "rev 1" pic. I ended up modifying more. Yellow indicates where i cut the clips in half (I was still trying to see if I could "save" the entire plastic grille).

For connecting into the DCT cooler loop, I decided -8AN was most appropriate and it was...-10AN would have been too big for routing the hose.

With some help from the pic above and Seems Legit garage I cut the *lower* trans cooler hose in the *crosscar* section. This is because that pair of hoses already hangs into the cooling fan space and I wanted the hoses to stay in the same direction only slightly offset.

I pushed these into the cut connections on the factory hose:



I built two hoses with the following ends to route to the cooler:





I actually built it with the straight connection for the cars lines and left the other end open so I could figure out routing to the cooler. Unfortunately, I didn't take any good pics of the routing, but basically you snake the hoses to the sides and up though the framing. I used the space behind my brake cooling intakes to have a nice large loop so there is excess for when you unbolt the cooler and pull it down for service or something (the connections are directly behind the bumper in this configuration when it is installed).

Once I had the length mostly figured out, I added some margin and cut. Overall, I used about 9 feet of -8AN twistlock hose. So about 4.5 feet each side.

On the cooler end of the hose I put the swivel joint. I wanted an angled swivel because it gave me more options and adjustability to control the hose routing behind the bumper and would be less stressful on the hose than going straight into the 90 degree fitting. Speaking of that, this is the fitting I used on the top of the Setrab.



I used zip ties to snug the all the hoses together under the car like this:



I was really concerned it would hit the fan, but amazingly, it does not.

Here is the final cooler hookup with connections. Swivel joints make this very easy to install and adjust once installed. I did have several iterations where the swivel joints had some minor weepage but altering and remaking the connections fixed it.



BTW, I did fill the cooler with FFL-4 right before hooking it up. And not much came out of the factory lines when I cut them. When I checked the DCT level I only needed to add about 400ml. I would estimate probably one full quart bottle for this job.

So, with everything installed, it looks like this (yes, that is me barefoot in my garage - don't judge ):



You can see here where I preserved the "wings" of the grille to prevent air bypassing the FMIC. My wife says it looks "almost stock", so I'll take it...

My latest logs speak for themselves on how well this works. Even adjusting for weather, it reduces trans AND coolant temps 10-15F *since any cooling to the DCT fluid means the coolant doesn't have to cool it). And it transfers a LOT of heat. Once session I came off the track, one side of the cooler was 176F and the other was 100F with an IR thermometer. And I estimate less than 5F effect on IAT's.

[AndyW]

Originally Posted by mKilgore

Ok thanks. I was looking at the log you posted and trying to see if I could locate where the accusump was kicking in. It looks like you have various places where the oil pressure dips to 43-45 psi under deceleration, but I couldn't quite tell if the accusump was kicking in to bring the oil pressure back up, or if the pressure was coming back up because you were on throttle again. It looks like at about time point 1920 there are a couple of blips on the oil pressure that could indicate the accusump was kicking on, but idk. I'm not sure what the oil pressure trace is supposed to look like with the accusump.

Just some observations from an untrained eye trying to learn lol. Regardless, I always enjoy seeing other people's logs from their track days.

Well, considering I was getting below 30 psi in some braking zones and/or long left turns, I am going to call it "working".