So I recently completed my fully divided manifold with twin wastegates and 4" exhaust. Initially I was happy since I get 10 PSi by 3400 RPM (5th gear highway) and full boost by about 3800-4000 RPM versus the old 4400-4500 RPM (manifold had cracks).

So I was out "tuning" last night just to get it to 16 PSi so I can try it out on the track for Monday. I ended up not touching the map as it was close enough and I forgot to add the voltage specs in FC-Edit on my new laptop so I couldn't log AFR's so I just eyeballed it. Now the weird thing is even though I get full boost by 4k RPM, the torque hasn't fully kicked in yet. At about just 5k RPM or so its like a second stage of boost kicks in and I feel a dramatic increase in torque even though boost is unchanged, almost as if boost just kicked in at that point.

I have tuned quite a few cars now and I have never experienced this. Normally boost kicks in with the power and then power increases linearly with RPM. My previous turbo was a 6765 with the SP-cover and the power kickin was normal. So I'm wondering if this is a characteristic of the H-cover or theres something up with my setup atm.

Basically, what was changed from last year is...

- New Precision billet 6765 /w H-cover vs. the old SP-cover. Both are 1.0 A/R but the old one was using the turbonetics housing and the new one is with the Precision housing
- New fully divided manifold with twin wastegates and the runners are maybe double the length of the old manifold.
- Complete 4" exhaust using a Vibrant cone for the 3" -> 4" right from the turbine housing
- New Intercooler that is 2" wider and 0.5" thicker with 3" outlets instead of 2.75"
- 3" IC piping pre-intercooler which matches H-cover and IC outlet vs the old 2.75"
- post-intercooler remains 2.75" since the RW 3" elbow isn't bolt-on and there was no time for modifications so I just expanded the old IC pipe to match the 3" hose coupling
- no blow-off valve (i doubt this has any effect but I figure I'll include it)

Anyway, I don't understand why torque would spontaneously appear at 5k+ RPM even though boost and AFR's are unchanged. The only thing I can think of is I'm out of some kind of efficiency which then becomes efficient at 5k RPM. Does anyone have any idea?

thewird

 

ive had that happen with stock/hybrid turbo and stock ecu setups. the best example was a cosmo 13B/s5 hybrid/FMIC/stock s5 ecu+fcon. it would hit 12psi around 3k, but in 3500-4500 range it would stumble/lack power. AFR on the wideband would go from 13's as boost hit to mid 12's in the stumble and then over about 4500-5k it would sort of gather up and take off

after some playing around with it, what actually was happening, was that the flat spot/hesitation was because it was so lean it would actually misfire/stop running. richening it up took the AFRs from mid 12's to mid 12's.... actually i think we had to lower the boost to get it to have enough fuel

you may have something else going on, but its easy to add fuel and see

 

You'd have to put it on a dyno to really verify what is happening at that RPM, but my only guess would be that since you are getting boost relatively early in the RPM band on that size of turbo, you could be running near the left-hand surge line of the (non-existent) compressor map, and only once you get over 5K RPM do you get into the efficiency range of the turbo. Since Precision doesn't give any maps for these compressor wheels, you couldn't even plot it and find out.

This is entirely possible because if you compare the GTX line from Garrett to the standard cast compressor wheels, the new aero shifts the surge line further right. If we had compressor maps from Precision we could be certain. Also, compare the two compressor maps from Garrett for the 3582R, one with a ported shroud and one without. There is a slight difference in the low-flow low-PR area of the map that could have been the difference in your new setup. Are you getting any weird noises from the turbo before 5K?

All I can say is that my 6768 with the H-cover didn't experience that, but it was un-divided with a larger turbine wheel, and thus only reached full boost by about 4800-5000 RPM; it definitely hit hard though. I also didn't get a chance to get my tune down perfect with that setup.

 

my initial plan would be to hook up the AFR logger.

good luck,

hc

 

Not enough information yet to give advice.

 

This sounds like my FC when it was at low boost.

At 12psi the turbo was at full boost by 3500 or so but the kick was from 4500 or so.

Basically the powerband was still N/A like at the lower boost so it was when the ports really started to be effective that I could feel.

Once at 23psi it was boost that gave the torque kick and you didnt really noticed when the ports started to come in to their own as it was already flying.

Ive had the same on turbo piston cars with low boost and big cams, they feel like a N/A engine and its when the engine comes on cam you feel the kick. And once the boost is high enough then its boost you feel more than when it comes on cam.

On some setups even with big boost its the same, but you get the big boost punch, then a crazy cam/port one later. One of my old piston engines running 25psi had that, 3500 then again at 6500, and a friends FD drag car running 35psi has the same, full boost torque kick by ~6500 and second kick at about 9500 that pretty much threw us off the road one day on road tyres.

 

i have a billet 6765 h cover also. issue with the h cover turbo is that is does not come with an o ring like the s cover. you need to put a small bead of silicone around the compressor cover to make a full seal. i learned of this on the supraforums as they were having inconsistant boost and loss of power. if the boost leak doesnt help its probably an afr issue somwhere in the map.

 

Something similar to the plots below but to more of an extreme due to the Aero Design of the billet wheels (if it's similar to the GTX) that shifts the surge line as mentioned.

I would say the GT4092 and the 6762 compressor maps are prolly pretty close as they both 67mm wheels with the 6762 having a much smaller turbine wheel than the 77mm on the GT40.

 

the GT4094 has a 68mm exducer turbine, so it would be more comparable to a 6768, plus the Garrett turbo has a much larger exducer on the compressor side (94 vs ~84) so it's still only a best guess. But you're absolutely right on with the plotting, and I see it as a possible scenario that we wouldn't be able to prove without the right map. My only difference is I use a little more conservative VE's to calculate the CFM of the engine. At 95% VE at 5,000 RPM, I only get ~32 Lbs/min which using this map would be just right of the surge line.

And yes, the H-cover has a ported shroud which generally broadens the map, however doesn't do much for the surge line.

 

I typically use 90% for stock port and sub 4k rpm (such as first plotted point) and a max of 1.15 for a full SP with 3/4in exhaust and all the bells and whistles.

Given your 95 VE that would spot on and makes sense since i was probably a little ambitious with the 115%.

I don't know the trims/rest of the wheel diameters on the precision line so I only had the inducer of the compressor to go off of - it was my best guess but thanks for adding the rest of the picture. I thought the last digit was the inducer of the turbine - but it makes more sense they use the exducer as it seemed like a really small wheel.

Never actaully looked at their line of products and didn't find what i was looking for in a few minutes which kind of annoys me....... LOL

 

the billet 67 has a 67mm minor and an 88mm major. the billet 67 has a bigger blade surface area than the 4094 due to the smaller hub of the billet.

 

I tested it again last night after putting in the voltage specs for the wideband on my new laptop. It came out as I expected, the AFR's were stable. I also tightened all the hose clamps in case there was a leak but that didn't help either. Also, raised the boost to 1.3 bar, as well as turn on my water injection and didn't change anything either.

Also, my H-cover is NOT ported. Sean sent me the wrong one even though I asked for the ported one but I decided to use it anyway.

I have also listened on the highway with 5th gear pulls from low RPM and windows down and it doesn't seem like its surging. I can hear it wooshing as it builds boost but its definitely not surging.

That sounds like a possibility. When I was assembling it, I noticed there was some kind of sealant on it from the factory. I just put it back on with neither cleaning the sealant off, nor applying anything new. I did tighten the bolts as tight as I felt comfortable though. I guess any silicone would do or is there a special sealant? Not looking forward to taking it off again as it will take me a whole afternoon/evening as there is a lot that needs to be removed before I get to it.

Another possibility I thought of is the Synapse gates are cracking open early more then they should. I only have one line running to the wastegates as Synapse is very clear that one line with the boost controller to the bottom port (C) should be enough unless your looking to run very high boost 30+ PSi. It seems to be working as its making boost and holding it early but maybe its not ideal? I have the Set Boost to 14 PSi so there is no pressure on the gate until 14 PSi but I wonder if the spring can hold it closed by itself with our rotary back pressure. I do have the stiffer spring in there as well as the smaller valve seat.

thewird

 

Ugh. Leave it to Synapse to make an unnecessarily complex design and then sell it with a bunch of marketing hype

 

You know I have always advised people not to get Synapse gate if they could avoid it after a few experiences tuning with them. But I had a brand new one sitting in a box so against better judgement, I bought another one and used them for my manifold figuring it must work right if you set it up right. I really hope this isn't the cause though...

thewird

 

I also had this powerband that the OP describes of full 14psi boost ~3,500rpm but power coming on hard at 5,000rpm. In addition it gave very near full power and boost no matter how little throttle I gave it above 5,000rpm.

I switched to a smaller street port that still opens as early as possible, but closes much closer to stock and just clean up in the runners with no other changes and the car has flat torque from 3,400rpm (full boost) to 7,000rpm where it gradually declines. It also gives me partial boost and power when I give it partial throttle- much easier to drive.

I believe it is from a lack of intake velocity below 5,000rpm so that there is much port reversion.

In your case it could be the new large outlet diameter compressor cover and intercooler inlet piping killing the velocity below 5,000rpm.

 

thewird, yea when i did a boost test on my car it would fall off at 10psi. i then fixed the compressor cover to get 18psi. i did find some other leaks as well after that i need to take care of. you should be able to get 20-22psi in there berfore it steadys out.

 

Firstly your prediction is too far to the left on the compressor map. Secondly you won't feel the added efficiency because

#1 air temperature out of the turbo will still be lower in the less efficient region to the left of the peak efficiency island than to the right. Don't dream for a second it will be colder at 78% efficiency and lots of flow, yet hot at 55% efficiency and less flow.

#2 Any change in air temperature that was hot initially and then cooled (as you came into the peak efficiency island) would not be observed as more mass flow because of the temperature change of the air charge temperature. The nature of engine RPM being considerably higher for more flow is what you feel.

To comment on the rush at 5000RPM. My car does it to and its about 4 times as responsive as a Z sized turbo.

My car power band is shaped more like log(x) than the x^2 that you will get with a larger single. As I said its all in before 4000rpm, it goes very well at 4000rpm. Overtaking power at 2500 in incredible, yet it still goes mental at 5000rpm.

 

Firstly, why would you say my prediction is way too far to the left? You’re implying that VE is 100+ below 4k on this theoretical scenario I used as an example to explain a phenomenon? (this is a mute point as we don't even have the compressor map)

Secondly, we are not talking about efficiency at all - it's the surge line we are talking about and if the turbo is even capable of delivering the amount of air the engine is trying to ingest creating a restriction.

 

I'll back this up with proof. This is a Mazda research paper from the 80's testing VE vs. port timing on an N/A PP. Notice how VE drops when IO is retarded, and this is at 50-100* BTDC vs 25* ATDC for a standard streetport.



@ 4,000 RPM on an optimized PP N/A (which will result in higher base VE than any sideport turbo) and with ideal exhaust routing (I.E. no turbo restriction) the engine made maximum around 100% VE, and in other scenarios much less based on IO timing (which btw is a HELL of a lot more advanced than any streetport timing) So I'd say Sesshoumaru's figures were a little on the optimistic side.

And as mentioned above, it has nothing to do with efficiency as far as charge temperatures are concerned, it has to do with mass flow, or in this case possibly the lack thereof.

But once again, as mentioned without a compressor map we're just pissing in the wind to see who can reach further. Let us know when you get it on a dyno thewird, then you can log the powerband vs intake manifold pressure, AFR, etc. and hopefully get an idea of what's going on.

P.S. I would use 80-85% as a more realistic VE at that RPM (unless someone else has more experience or facts on the topic) and as far as calculations go, I use Sean from A-Spec's formula:

CFM=(4-stroke equivalent displacement in L x RPM x VE x PR)/5660

(2.616 x 4000 x 80 x 2.09) / 5660 = 309.11 CFM, which equals roughly 21.65 lbs/min of air (give or take some for air density)

If we transpose this info on a T04R/PT67 compressor map (which is as accurate reflection to the 6765 as anything else, since the Precision-style smaller billet center hub doesn't seem to play much of a role until higher PR's) we get this: https://www.rx7club.com/attachment.p...1&d=1306261795

Now to say that surge isn't a possibility without any other proof is as stupid as saying it's anything else. Not enough info!

 

I am running the same turbo as your weird and i have noticed the same with mine. I have yet yo get out and tune for this year with the new turbo but one thing to think about is the timing. General rule is to keep a eye on timing for peak torque its possible that you dont have enough timing where your engine produces max torque or before that. Some get confused with fuel knock and timing knock when tuning. mine made max torque at about 5200rpms with mild street port and the non billet 6765 last year. this year i am running the billet version so i will have to play around a little. If i remember right we have about the same set ups.

 

It wouldn't be surprising for the torque to bump up where the natural VE/ torque peak is, especially if timing isn't retarded through that area as it is effectively like running more timing as the charge will be denser, so the effect of VE through that range will be more than just the proportionate change in chamber filling as it increases flame speed as well, so twofold power effect, unless you have pulled timing in that range. Alternatively it is telling you that you can safely run more timing before & after the peak at the same boost pressure.

 

the Rx8 SAE paper has the VE chart of the Rx8 engine and the S5 non turbo engine. the peak on the S5 engine is about 92% ve, and this actually is right around 4500, it drops off on either side of that, @3k its more like 85-86%. the Rx8 engine is in the mid 90's around 4k, and peaks around 105% from 5500-8k

i'm not really sure how the turbo affects ve though, or i'm not sure how to calculate that change

 

The thing about mazda/SAE volumetric efficiencies is that for a factory engine they will be using factory parts that less people use.

If you take the factory system and put a 3" or larger cold air snorkel on it with a massive air filter, and you take off the factory exhaust and replace it with a 3" or larger mandrel bent exhaust with loose mufflers you have changed mass flow by a large amount. The primary variables that change are intake manifold pressure (increased) and exhaust manifold pressure (decreased).

I'm telling you if you achieve 100% chamber filling to your intake pressure you have filled more than 654cc into that cycle. The volume at bottom dead centre is more like 720cc.