Open Tuning/Turbo Theory Discussion

[HawkeyeGeoff]

Found this on YT today:

I must say I was a little surprised because FWIH, Forced Performance generally has a great rep. We have a local with an HTA3076 and that sucker is fast and spools quite quickly. I guess the moral of the story is be wary of the FP SS hotsides.

My XR, pretty much the next version of the HTA3076, spools like a 3071. No problem hitting 20 psi by 3400 starting at 3k.

I would believe the HTA version of the 35R would spool faster than the GT series (maybe even the GTX series).

 

[broda]

My XR, pretty much the next version of the HTA3076, spools like a 3071. No problem hitting 20 psi by 3400 starting at 3k.

I would believe the HTA version of the 35R would spool faster than the GT series (maybe even the GTX series).

I wish I'd be able to participate. I have a HTA3576 but I have a higher CR as well so it'd be too apples to oranges.

 

[Redline]

My XR, pretty much the next version of the HTA3076, spools like a 3071. No problem hitting 20 psi by 3400 starting at 3k.

I would believe the HTA version of the 35R would spool faster than the GT series (maybe even the GTX series).

Maybe the reason they spool fast is comparatively smaller hotsides, like was referred to in the vid? That would definitely give a spool advantage, at the expense of some top-end.

Your spool is quite impressive. Though not optimized for fast/early spool (intentionally, becuz stock block/rods), my BNR S3 going WOT @ 3k hits 20psi ~3,250, so basically the same as yours. But seeing a bigger turbo match that is very impressive.
[doublepost=1477152204][/doublepost]

I wish I'd be able to participate. I have a HTA3576 but I have a higher CR as well so it'd be too apples to oranges.

You can jump in with that, if you want. Originally, I created this thread when talking about the elusive "safe" upper whp limit. We all hear about the "safe" upper wtq limit of ~380 or so, but the whp limit is much less commonly discussed in detail. @Realgib3 raised some very good points. Bigger turbos lend to being safe even closer to 500whp because of how/where they make power, that they don't have to be pushed as hard (so much cooler), and that tons of alky gives you a safety threshold. Though it wasn't specifically mentioned, everything else being equal, a turbo with a larger hotside (i.e., 3076/5858 and above, like he mentioned) won't "bottleneck" the heat in the CC nearly as much, AFAIK. So that's a more specific way of saying what he already essentially said.

My whole thing is whereas the 380wtq limit is becuz rods, the 500whp = no-no becuz ringlands, etc., the way I understood the physics of turbo'd engines. Power and heat level are directly proportionate. But he broke it down further, and has a ton of practical experience with this. The short of it: want 500whp on stock block? Run a 3076/equivalent or above and a crapton of alky. Wtq management becomes easier in terms of boost ramp-up being slower and boost threshold being higher naturally (which both help protect the rods). Then, it's all in the tune: capping that wtq to 350-380 or so for longevity.

Jump in with another turbo topic, if you want, mayne.

 

[broda]

Once I'm done/in the process of tuning I'll throw it in. I'm going to be breaking in the motor so won't have any data for a bit.

That's the reason why I got the HTA3576 because it was meant to be on a stock block w/ meth. I assumed I could be at around 450whp but still keep wtq at a reasonable level where my block would be safe. Unfortunately my engine popped on the base tune so I didn't get to do any testing towards that lol

 

[Enki]

Bit late to the party but I'd like to say that engine life isn't set by a flat number, but a curve. Stay stock and it will last a long time. The further you deviate from stock power levels the sooner you can expect issues.

More torque? Rods will weaken and eventually break faster.
More redline? Cylinders will oval and you'll lose compression faster.
Ragged edge tune? How about them ringlands.

You can make good power and torque reliably on stock and built engines, but if you don't do it right the first time, you will have to do it over again later.

Also hashtagDatEFRLyfe

 

[Redline]

Bit late to the party but I'd like to say that engine life isn't set by a flat number, but a curve. Stay stock and it will last a long time. The further you deviate from stock power levels the sooner you can expect issues.

More torque? Rods will weaken and eventually break faster.
More redline? Cylinders will oval and you'll lose compression faster.
Ragged edge tune? How about them ringlands.

You can make good power and torque reliably on stock and built engines, but if you don't do it right the first time, you will have to do it over again later.

Also hashtagDatEFRLyfe

Metallurgy and other science FTW! I'm currently working on the master equation that takes every single variable into account to let you know exactly when your engine will fail (down to the tenth of a mile), and by what cause. It's a work in progress....

#ThatFutureEFRLyfe

 

[Enki]

Good luck with that. lol

 

[Redline]

Good luck with that. lol

Yeah, it may take several lifetimes if I'm unaided. It's like the String Theory of engine failure

 

[Redline]

It's definitely not the easiest thing to do. I've found interpolating the values in the bottom 4 rows (WOT conditions) of my WGDC down to pretty low throttle positions helps a lot. A lot of tuners, Cobb included on their OTS maps, have this big "shelf" of power you can feel the second you dip above 75% throttle or whatever it is. It gets really annoying when you're trying to gradually roll on the power.

I always find it interesting how different people can be in this community even though we all own the same car. There's no way I could give up the low end torque I have in the BNR. It's part of what makes the turbo and the whole car so much fun for me.

@Redline totally forgot about the 4-port option. That would definitely be worth investigating for OP.

Unfortunately, I'm on a locked FREEKTUNE map. When I first went S3, I was fully confident in my ability to self-tune it great, but I chose instead to go pro because at the time Justin was still not locking maps. As a student of tuning, I saw it essentially as 1) paying to have a pro-quality tune, 2) paying to essentially have tune tutoring, watching everything he did revision by revision to learn further/refine my craft more. I was very inspired many times in that he did exactly what I would've; other times, he'd do something so elegant/simple that I was amazed by his skill. Anyways, all of that to say that I could get further refinements for 1st and 2nd, but it'd be a re-tune. BTW, I totally agree with you about loving the "gutsy" feel of the S3. That's why I covet the EFR7163 if/when I finally go 500whp: it's the rarest type of turbo - one with that gutsy low-end, but still can support ~550 or so.
[doublepost=1478280505][/doublepost]

5. 4 port will not work with our ECU, without an external board to modify the Frequency. With that said, 3port is plenty sufficient and if you're trying to run super low power in 1st gear, there's ways to run well below spring pressure anyway, so super soft spring isn't necessary.

Well, electrical engineers in our community, this looks like an awesome project! With most larger turbos, it's a self-resolving issue due to the nature of their boost threshold and spooling time. But with next-level turbos, like the EFR7163, it'll still be an issue. That's why I'd really LOVE to have 4-port ability. I'm also curious about the EBCS system that already comes on the EFR7163 and how low the boost can be set, especially in 1st and 2nd.

In the meanwhile, care to develop your statement about "there's ways to run well below spring pressure anyway"? @Realgib3

 

[Enki]

I'm also curious about the EBCS system that already comes on the EFR7163 and how low the boost can be set, especially in 1st and 2nd.@Realgib3

On my 6758 it took about 40% WGDC to see ~1-2 PSI gain; the curve got aggressive after that with 100% peaking at about 20-22 depending on weather.

I'm also curious as to how to run below spring pressure on a 3 port.

 

[Redline]

On my 6758 it took about 40% WGDC to see ~1-2 PSI gain; the curve got aggressive after that with 100% peaking at about 20-22 depending on weather.

I'm also curious as to how to run below spring pressure on a 3 port.

Definitely. Either that, or I should throw an AWD conversion on the to-do list when going built/EFR7163

 

[alexwlwsn]

Anybody look into the new Garrett GTX Gen2 turbos yet?

Link

Coming from a GT3582R it's cool to know I can use my same hotside and step up to an 850whp turbo on the same basic frame. I don't exactly need/want anywhere near that though that so I might look into the new GTX3576R or GTX3076R, both of which are rated for 750.

 

[Realgib3]

Alright here's a link to the portion of the thread that deals directly with the "below spring pressure" question, but if you have time, the whole thread is pretty interesting and amusing to read through.

http://www.mazdaspeedforums.org/forum/f418/full-control-load-each-gear-46550/index3.html#post406638

With that said, the super short version is correctly setup load limits/targets tables can hold the throttle closed enough to hit extremely low loads/boost.

EDIT: I'm also partial to that thread, because it's where we had the very first "ahah!" moment about what the "wall" could be and began testing for that soon after.

 

[Redline]

Anybody look into the new Garrett GTX Gen2 turbos yet?

Link

Coming from a GT3582R it's cool to know I can use my same hotside and step up to an 850whp turbo on the same basic frame. I don't exactly need/want anywhere near that though that so I might look into the new GTX3576R or GTX3076R, both of which are rated for 750.

Pretty sure that 750 figure is only for the GTX3576r. The 30-frame series doesn't have a big enough hotside, apparently. The rule of thumb is a turbo's horsepower support level is lbs/min x 10. But if you look at the GTX3076r compressor map on this gen 2 version, it only flows ~67 lbs/min. Oddly enough, the compressor map for the GTX3576r looks pretty much identical. Anyways, that's a far cry from a 750hp turbo:

Regardless, cool find. Any links/whitepapers showing what, exactly, the differences are between the Gen 1 and Gen 2 GTX'es?

 

[alexwlwsn]

Pretty sure that 750 figure is only for the GTX3576r. The 30-frame series doesn't have a big enough hotside, apparently. The rule of thumb is a turbo's horsepower support level is lbs/min x 10. But if you look at the GTX3076r compressor map on this gen 2 version, it only flows ~67 lbs/min. Oddly enough, the compressor map for the GTX3576r looks pretty much identical. Anyways, that's a far cry from a 750hp turbo:
View attachment 4277
Regardless, cool find. Any links/whitepapers showing what, exactly, the differences are between the Gen 1 and Gen 2 GTX'es?

That is strange that everywhere they're "rated" as 750hp yet the flow chart says differently. Even on Garrett's website they're rated for 750 yet the charts peak at around 670lb/min. The GTX3582R which is an 850hp turbo however seems to show peak flow around 840lbs which makes me curious why there's such a discrepancy.

It seems like it's an all new compressor housing with a speed port like a Borg and then a more efficient new wheel design.

 

[Realgib3]

Pretty sure that 750 figure is only for the GTX3576r. The 30-frame series doesn't have a big enough hotside, apparently. The rule of thumb is a turbo's horsepower support level is lbs/min x 10. But if you look at the GTX3076r compressor map on this gen 2 version, it only flows ~67 lbs/min. Oddly enough, the compressor map for the GTX3576r looks pretty much identical. Anyways, that's a far cry from a 750hp turbo:
View attachment 4277
Regardless, cool find. Any links/whitepapers showing what, exactly, the differences are between the Gen 1 and Gen 2 GTX'es?

On the website it does actually say 750hp for the GEN2 3076, but I agree it doesn't add up at all. Idk I really don't trust that these turbos will make a ton of power and be responsive too, all the sudden. Same thing with the new 3584RS they're so hot about. I checked it all out and I think they're high, saying that's a 1k hp turbo. Idk, could be totally wrong, but for right now I'm skeptical.

 

[alexwlwsn]

On the website it does actually say 750hp for the GEN2 3076, but I agree it doesn't add up at all. Idk I really don't trust that these turbos will make a ton of power and be responsive too, all the sudden. Same thing with the new 3584RS they're so hot about. I checked it all out and I think they're high, saying that's a 1k hp turbo. Idk, could be totally wrong, but for right now I'm skeptical.

I agree. It's hard to believe you can physically flow 1000hp worth of air through that. It looks to be tapping out hard at 900lb/min so maybe their ricer math for flow to hp is different than the commonly accepted 10:1 ratio lol

 

[Enki]

Well you can go off the map. It's not entirely good for the turbo but it is possible (see: K04).

 

[Redline]

Well you can go off the map. It's not entirely good for the turbo but it is possible (see: K04).

Indeed, but that's an excellent recipe for shortened turbo life. And if I'm going to drop a few grand on the turbo alone, best believe I'm going to stay in the islands, LOL.
[doublepost=1478298024][/doublepost]

Alright here's a link to the portion of the thread that deals directly with the "below spring pressure" question, but if you have time, the whole thread is pretty interesting and amusing to read through.

http://www.mazdaspeedforums.org/forum/f418/full-control-load-each-gear-46550/index3.html#post406638

With that said, the super short version is correctly setup load limits/targets tables can hold the throttle closed enough to hit extremely low loads/boost.

EDIT: I'm also partial to that thread, because it's where we had the very first "ahah!" moment about what the "wall" could be and began testing for that soon after.

This is a game-changer. I'm surprised I never knew about this. Email out to Justin asking his pro opinion/if he'd be able to implement with a re-tune fee.

Ninja edit, @Realgib3 : is this a hybrid map compatible tuning solution, or strictly load-based?

 

[mazdaspeeder]

IMHO, there are 2 primary concerns:
1) too much absolute torque; too much torque too low; or too aggressive of a torque ramp-up = potentially compromised rods.

I just found this forum so I'm going to revive this to share my knowledge on prevention of this from self tuning over the years.
My technique to combat this torque spike on spool up but keep that turbo spooling as fast as possible is to stop the ignition advance from hitting like a sledge hammer on WOT. This also helps with spool up KR.
Here's how:

1) Set your Closed Loop - Exit Delay B to 40
I have found this controls the time it takes after hitting the Max Load D value
until the transition starts from CL -> OL IGN Table
This allows a decent time to ramp up of about a second after Max Load D has been hit.


2) Set your Closed Loop - Max Load D table up to attainable but close to max on a hot day.
I have found that this table is used to determine when to transition from CL -> OL IGN Table


3) Set your Ign Table - Low Throttle/CL tables from 1.63 down to about 3-5 degrees less
than your Ign Table - High Throttle/OL table values
CL IgnTable

OL IgnTable


See my attached log how at 4k it hits Load of 2 and then after about 3 rows it starts the transition up to OL timing.
You may want to adjust the time it takes after or the starting CL Ign based on your mods to allow maybe some heat soak in your inter-cooler/manifold to be brought down after slow cruising or idling.
My logic could also be completely flawed, I'd like to hear your thoughts.