ATP GTX2871r - Boost spike issue/ ATP customer service.

I purchased an ATP GTX2871r IWG with an ATP catless DP back in 2015. For me and on paper, that turbo has a great transient response/spool for track/AutoX days and enough top end flow and power to keep a smile on my face on the street. The turbo is nice and happy as long as we keep the boost low, once we start pushing >21 PSI, thing starts to spike during up shifting @ WOT. We've seen spikes anywhere from 3 to 6PSI. See email convo for more technical details and how this spike affects boost control and tuning. Anyhow, the turbo having some issues isn't my biggest concern.. shit happens. It's the way ATP handled this issue, nothing but dancing around it and trying to put the blame on the install/hardware/tuner (installer error LOL). Right now, they stopped responding to our emails. I used to be a huge fan of the ATP kits as they used Garret turbos, and I really like these guys....best bang for the buck imo.. well, they used to be. Anyhow, attached PDF has all the emails back and forth, logs and explanation from Mr. Finkle. You can click on the individual logs and files within this PDF to view.

Edit 4/19/2018: I still get questions about this turbo often. The issue was solved by porting the hotside pretty good a couple of years back. I had it professionally ported by force performance down in TX, but I'm sure anyone with tools/knowledge can do it no issue. This turbo is problem free, controllable and makes 400+whp with ease. Attached picture of hotside.

 

Thanks for sharing. Sorry to hear about your frustrating experience. You and Rob tried.

 

The 3071's have to be ported on the hot side to stop creep, maybe look into this.

 

We will definitely give it a try. I'm sure we'll figure it out one way or another.. but the point of this is kit is that it's designed to work out of the box and not having the customer to experiment with solutions. ATP has failed me personally and I wanted to share my experience so the community is aware of this particular turbo and ATP's customer care dept.

 

I understand the frustration and hope things work out for you Sari.

I have a few things that come to mind.

-Aren't there quite a few of these turbos out in the market and running? An issue on 3 of them, but not all of the others strikes me as odd and likely not a design fault. It could still be a hardware fault that is specific to a certain production batch or something possibly, but a design flaw like the 3071 wg porting, is unlikely if there are plenty of these out and running without issue. (I may have a very incorrect assumption of how many people are running this kit)

-ATP is far from the only turbo manufacturer to run into hardware issues. BNR had wg actuator issues, CS has creep issues if catless, for example. That's just in the last year or so, and I am sure there are others. I know pitchforks always occur in these situations, but perhaps that is not the best way to resolve things.

If this is in fact a design flaw similar to the 3071 issue and BNR wg actuator, then obviously it needs to be addressed and I appreciate the efforts to accomplish that. Perhaps it is worth trying to get other end users and tuners involved to try to figure out if this is an issue universal to all gtx2871 or if it is an anomaly happening to a few. More, repeatable data may help convince ATP as well as help pinpoint the issue.

Good luck getting it figured out man. If I can help in anyway, you and Rob know how to get in touch.

 

Thanks buddy, I appreciate the support. To answer your questions:
I know of 4 people that are running this turbo. Myself, Jeromy, someone on IG that I lost contact with, and another guy who just installed it. I honestly don't think many are. I know of one other person who bought it and exchanged for another one once they saw one of my posts..

Jeromy was tuned by another tuner, from what he told me he was around 21-22 PSI IIRC, and he saw the same exact results (spiking during WOT up shifting). He then switched to Rob, turbo kept spiking. He got sick of it and switched to a BNR S4 and the GTX is sitting in his closet.
The guy who just installed his, this is what his base map looks like ( Base map is from Cobb, notice spool).


I can't comment much about that spool issue, but to me it sounds like this turbo/hotside/wg whatever is wacky. He's getting tuned by Cobb soon as far as I know, and from what he told me, he doesn't plan on pushing it and just wants a conservative 91 tune.

I understand that manufacturers have issues, no one is perfect and that is not the reason of this thread. I have been talking (emailing) with ATP since iirc Sept of last year. You can see in the email thread we did nothing but provide data, logs, tuner logic etc... Now they went completely silent.
I'm sure we will make this turbo work one way or another, and that isn't my main concern. I just want to document and share my experience with people who are looking to buy a turbo kit to know where they should/shouldn't invest their hard earned money.... and when shit hits the fan, which manufacture is willing to stand behind their product.

 

Ill buy the turbo in the closet. Lol

 

Here's me! Tuned by Chris Haag at Cobb Tuning Surgeline in Portland OR, I worked with them before so I got my tune half off

Here's the 20, 18 and 16 psi tunes on their awd mustang dyno. It was just under 60 degrees today and he stopped as he became knock limited, on 92 btw.

Oh and night before solved my issue before that Ge3k mentioned. Turns out the hotside on this turbo is just too shallow for the CS DP's divider, it just barely (didn't) fit when I bolted the turbo and DP together, so the final torquing down caused the divorced DP's divider to bend in a way that it blocked the IWG flapper just slightly from closing all the way, that was enough to give me that awfully slow lag.

I cut most of the divider off of the CS DP before my tuning appointment and all was good, although I'm sure the design is no longer as effective as it would be with on a compatible turbo.


[doublepost=1456887919][/doublepost]Forgot to mention, it's raining and hailing so I won't be able to do logs or play around much just yet. But at least on the dyno, the boost control issue characteristic of this turbo did not appear. However I wonder if I'm just outside of that power/boost range where that happens so far.

 

Good stuff!
I can't wait to see your shifting @ WOT logs. That's where the spike shows, especially when pushing 21-22psi +. You can see in the attached logs/email convo with ATP. 4th gear logs/dyno pulls, this thing is an absolute beast. Here a 4th gear log from my car on 93 pump, and I'm having some boost leak issues which causing the fuel curve to not hit targets:

 

I love my GT3071. The 30 series will take you farther on 93 than the smaller snails and won't be screaming for help at 400 on a E-85 tune.

 

Atp is shit. I will never buy any products from them again. My 2871r didnt last a year turbo was installed correctly and tuned freek but still failed should of went bnr.

 

I'm not sure if Smelson is on here, but he had a honey badger 2871r. He was putting down like 350whp on 93 (Virginia fuel is apparently good) with his 2871 at like only 22psi. He was the second owner, and it was sold to him with porting done on the IWG. He had a Russell EM too. The setup flowed so well his tuner asked if he had headwork done. Zero issues. Here's him running his ethanol tune (also only 22psi, just more timing) at the strip:



His tuned numbers are in the OP here:

http://mazdaspeeds.org/index.php?threads/dfw-big-turbo-club.924/

 

Hey, if you're 100% serious I'll make a FS post.


Sent from my iPhone using Tapatalk

 

Hi
it's not the turbo problem definitely, you have to properly set WG duty cycle and WG dynamic configuration tables. I had the same issue and solved it by updating these.

Sent from my Nexus 5 using Tapatalk

 

It most certainly is the turbo. Changing WGDC did nothing for us. We ended up having to port it, and it still spiked, albeit not as much. Ask me how I know when I put at least 10 hours into running all sorts of different stuff with this turbo trying to control it. It's any GTX series with a 71 hotside that has this problem.

Sent from my SM-G955U using Tapatalk

 

there are more tables involved: WGDC, WG Duty Boost Error Comp and WG Duty Boost Error Ramp Limiter High and Low limits.
Its a typical problem with controlling a device with feedback loop. One of most frequently used techniques here is having PID controller and this is what Mazda ECU has. It's very common that such circuit can spike/oscillate if wrong PID settings are used. You can google "pid controler oscilates" or "pid controler spikes" in google. Unfortunately we don't have access to all the parameters of such controller.
Turbo is especially hard to control because of its non linearity(air flow vs boost pressure). Please have a quick look at wiki how PID works https://en.wikipedia.org/wiki/PID_controller. It may be not so easy to understand without a bit of math background. The important thing is that you can get rid of oscillations/spikes when you set WG Duty Boost Error Comp to 0. Its not ideal solution because then you just disable PID at all. As the consequence ECU looks only on WGDC and cannot adjust boost if it is not equal to what it was set to in boost conf table. There still can be a "long lasting spike" but no oscillations - this one is caused by turbo non linearity which can be easily adjusted by WGDC.
This is just few sentences and there are books on PID control. If you would like to put effort in installing this turbo back and sending me the COBB logs with parameters I will list then I'm 99.9% sure that after analyzing these I will be able to provide you the detailed settings at which the turbo will not spike/oscillate (with PID enabled) Of course there may be few iterations needed to get to the final settings.

 

@togozdz; sadly I sold the turbo off to a friend of mine who plans on fixing it.

Sent from my SM-G955U using Tapatalk