20G hot vs cold, will not make more boost hot

[synchronic]

Hey guys, I wanted to post this here to see if anyone else has seen this same phenomenon. Our resident dyno queen is a 4G63 Eclipse GST. The turbo is a Mitsubishi 20G running on the stock manifold. It is an oil-only CHRA, no water.

Boost was perfect and predictable with every setting on the 40mm WG001. Port C, D, C+D, C+A, D+A were all consistent results when the car was first run. After heat soaking the car with F3 mode on the Dynapack, going in and out of boost, the turbo was glowing. Subsequent runs then started to make lower boost. I had to isolate if the problem was the wastegate, or some other problem in the system.

When cold Port C+A was making 20 psi of boost. When heat soaked, the exact same combination, without any changes whatsoever, was only making 16 psi. The boost also spooled up slower and didn't have the same curve as the cold setting. This difference between cold and hot turbo conditions was replicated by letting the turbo cool with a fan and then running the car again. Identical results each time. Engine temp was also consistent between cold and hot turbo runs. But this still doesn't isolate the wastegate from being the problem.

So, the next logical step was to run tests minus the wastegate in the equation. Luckily, the setup is somewhat of a slow spooling application. There is a maximum boost that the turbo will make at a given RPM with the wastegate completely closed.

With pneumatic pressure at the top of the wastegate to always keep it closed, I ran tests cold vs hot to isolate the turbo. With the turbo being cold and the wastegate held shut, the turbo made 20 psi at 5100 RPM. With the turbo heatsoaked to the point of glowing, the turbo only made 16 psi at 5100 RPM with the wastegate held pneumatically shut. It was also observed that the boost pressure was erratic when the turbo was heat soaked. Boost was moving between 15-18 psi in F3 Dynapack mode with the RPM held steady at 5100. The boost also seemed to come up very slowly compared to when it was cold. And, with the WG held shut and the turbo was cool, the turbo would make more boost as you increased the RPM. 5100=20 psi, 5200=22 psi, etc. When the turbo was hot, it would not make more boost as RPM increased, 5100=16 psi, 5200=16 psi. And this is with the wastegate held shut. The more RPM you drilled into the turbo, the more unstable it became as well. It would actually start to surge with more RPM.I believe that this test completely eliminated the wastegate from the equation as being a problem.

Anyone else see this before?

 

[The_EE]

I have not seen that as a problem but I dont have a dyno to test on.

I run a GReddy Mitsu 20g, same setup as yours, only oil cooled.

Would you happen to know the max boost for these turbos? I have never found the info and do plan on starting a thread if I cant find it after my search on here. I am tired of looking. I know 22psi is cool, I know its a 44lb turbo and at 22psi I am seeing only 40lbs of airflow, so I know I could likely do a little more boost.

Getting 20 psi by 5100 rpms seems quite a bit late to me. I am seeing full spool with a tubular manifold by 4300 rpms at the latest. What are you tuning with?

 

[thilaksharma]

isnt this just for the synapse wg

 

[Boo5t_Addict]

Take the turbo apart and see what you can find. Maybe being hot to the point of glowing is causing the center section to have much more friction than normal. If I remember from high school physics correctly, metal not only expands as it heats up, but the coefficient of friction increases as well. i would look toward the center section. Or drop in another 20g and see if you have the same problem, you could just have a faulty turbo.

 

[synchronic]

Figured this one out for now. Fuel pump was going out slowly, coupled with some injectors that were sticking. Fuel pump changed, and in need of a tune, so an AEM EMS is going in the next week or so.