equal length headers, and why you should get em.

[Rhitter]

Lemmings off a cliff. Lol.

 

I have a bridge to sell if any of the subscribed are interested.

 

I am confused by this. I am the only one who asked about cheap knock-off headers from a brand whose downpipe you used to recommend (my understanding is the CAT on the CNT downpipe doesn't last as long and you no longer recommend it). All the comments said - no name brands kind of worked but they all broke really quickly...which seems to align with your comments.

 

I was guessing / hoping that if the welds were good enough on the CNT downpipe they would be fine for the headers. But like I said there is very little information about them or really about headers in general out there.

 

All the brands mentioned in the original post - you sell on your website!

 

There isn't enough information in this post for it to become gospel... just one tuner saying he has had good luck with them on 6 cars - sharing his experience.

 

And then you come in and just make snide comments without actually saying anything to educate people where these comments are coming from.

 

Usually you spend the time to educate people based off of your experience - which we all know is significant and very valuable to this community.

 

I am just confused why you have posted in this thread three times to "bash" it with cryptic responses instead of just stating what you mean.

[adosdrummer]

I also am confused with the lack of direction on this subject besides the non educational posts from mspank. I have posted my finding from different cars, different tunes, different headers. Granted I have not tuned a car using a aftermarket UEL header. But there Is apparent gains, and raised knock threshholds exp in cyl 2 and 4. If you have had different findings then please chim in.

[adosdrummer]

Maybe in your finding the results are not so apparent. But have you seen actual negative affects?

[covertrussian]

To me getting similar results even on two cars would mean that it would probably yield similar results universally. But I haven't tuned 6 cars..., I only own 5

 

I also am confused with the lack of direction on this subject besides the non educational posts from mspank. I have posted my finding from different cars, different tunes, different headers. Granted I have not tuned a car using a aftermarket UEL header. But there Is apparent gains, and raised knock threshholds exp in cyl 2 and 4. If you have had different findings then please chim in.

 

I haven't really looked into the aftermarket UEL headers, but the cylinders are separated for a good enough of a length they should still be much better then stock. The further the joining is from the head the better they will work.

 

Btw are you disabling the timing compensation that pre 07 LGT's add to cylinders 2 and 4?

[m sprank]

Lets clear some air.

 

CNT down pipes... Manufacturing plant changed approx 2 years ago. Pipes were never the same from the "new" factory. Bellmouths are rough, welds are poor, steel is low grade, cats are worthless and they now have fitment issuess on some cars (not all the pipes seem to fit the same). So, I stopped selling them and stopped recommending them.

 

"EL" headers are rarely truly EL. The boxer rumble is the sound of inefficiency. It is the sound of exhaust gases out of sync. Get the pulses in sync and you can make more power. But... there is more to it than that.

 

Most EL headers are a 4 into 2 into 1 design. The larger primary piping leading into the larger runners tends to slow exhaust gas and decrease spool (there is too much volume in the headers piping). Then the primaries are a different length and have different bends. So, the exhaust gas pulse velocities are different and power is lost. The metal is also generally thin and the energy from the exhaust air is transferred into the piping until it gets hot enough to not absorb the heat (energy) from the air. So, spool is decreased even more until the headers are good and HOT (decreased turbine efficiency).

 

Because the metal is thin and at best the welds are machine tig, meaning a poor tig weld over a poor mig weld yielding poor penetration, you get a brittle metal and a tendency to crack/leak from heat cycles. Turbo cars can easily exceed the temp rating of 304 on headers and up pipes. On some the quality is so poor the flanges are not even flat and they wont seal without being machined.

 

There is still more to it. I am tired and the knowledge I have acquired did not come free so I dont give it all away for free.

 

I am not hating. I am laughing at the masses huddling at the feet of the soothsayer who has tuned 6 Subies. He gave no data, no repeatable scientific studies nothing to refute the claim that EL headers are good and you should get them, but here we are and people are already jumping to install eBay crap that might do more harm than good.

 

EL headers are good. When you buy the good ones. KillerB and FullRace are the two best we have tested. KillerB is cheaper and (honestly) makes more power for most users. Perrin is Ok, but far from perfect. Inividia is ok, same as the Perrin. You get what you pay for. Want headers that work as advertised and make 20whp more than Perrin or Invidia? Want them to last? Buy KillerB.

[adosdrummer]

I would love to tune a Perrin header and then on the same car a Killer B. you are correct, I did not post before and after dyne plots, or screen shots of the differences in the timing and vvt tables and data logs. Frankly I didn't see the need to do so. I stated my universal findings from a handful of different cars and three different headers. Do you have data to show that my claims are false or not withstanding? And if it makes any difference I have tuned 13 subies. And another that makes 14 was a n/a wrx. This is not a discussion based on my tuning or the number of subarus I have tuned.

 

I stand by my opinion that a good quality (be it a true equal length or not) elh is a worthwhile mod. Do your own research, talk to different tuners, and make an informed decision. Yes. I have fitment issues with my Perrin. It takes a little muscle to get it to go down on the header studs. Is it the best quality metal? No. Is there a chance for metal fatigue? Yes. But the gains and improvments on my car and those I have seen on others justify my purchase. And I would do it all again. In a perfect world I would have gotten the killer b myself. But I'm not looking for every HP I could on a car that won't be more than a stg2 with a vf52

[510Finn]

I agree with Mike, Killer B is going to be my choice, the 321 over the 304 and quality that's top notch, but I want to chime in and mention that the Tomei headers are a probably along the lines of Perrin and Invidia, at least that would be my pick if I were trying to save a little money and get good performance. If you're using the wrap they provide, I would still spray coat them after wrapping with the the Thermo Tec spray or equivalent. It's interesting they did mention the EL vs UEL, and from their testing with the EJ25 they picked the UEL over the EL. They mentioned EL had a decrease in turbo response, they tested it. So the data is there, and if you believe EL is better than UEL without testing you may be in for a surprise. http://www.tomeiusa.com/news/Report-EJ25-V1.html#C3.1

[m sprank]

We have tested the GTSpec and Tomei variants (same header/up pipe). They work equivalent to the Perrin and Invidia. We actually run the GTSpec on max LGT track car. The price was right for the gains. We have over 2 years and 35 track days on those headers and they are still hanging in (wrapped too).

 

Are they as good as the KillerB. No. But, they were under $700 with the up pipe. At the time it was about money, not about absolute power and we already knew the headers were good enough to last.

 

I agree a decent set of EL headers makes a nice gain. an AP OTS stage1 map makes more of a gain than headers do and when you do a cost to horespower comparison the headers do not equate to value for some customers. Also, Perrin, Invidia, GTSpec, Tomei, etc. all delay peak boost until the headers are super hot. Some customer dont get the headers hot enough to over come this and then they are not happy.

 

EL headers are not for everyone. Removing timing compensations is easy enough to decrease low load knock (which is actually not a major issue and not why cylinder 4 fails first). I personally like a good set of EL headers on my Subies. But, I have built far too many to still think that what I like is best for everyone.

[heiche]

As far as poor cylinders 2 and 4 (especially 4)...

 

• Better exhaust header (which is the subject of this thread) helps
  
• Timing compensation: 2.1 degrees is a fair amount of extra timing advance for cylinders 2 and 4 in the stock tune. Removing it evens out both banks and alleviates 2/4 a bit
  
• Other reasons: cylinder 4 probably also doesn't appreciate having a Dog Leg to compound the above issues with more restrictive flow
  

[m sprank]

Ok, off topic but here we go...Cylinder four is considered the sacrificial lamb by Subaru engineers.

 

The single knock sensor is located directly over cylinder #4. So, Subaru engineers felt it was wisest to put in timing compensations on cylinder #4. These compensations and the thought process behind them are at least two fold. First, cylinder #4 is closest to the knock sensor, so the sensor should pick up any knock in cylinder #4 first. The timing compensations are there in order to retard cylinders #1, #2 and #3 enough to stop the knock event from happening in those cylinders (as was detected in cylinder #4). Second, cylinder #2 and #4 hold the most heat. Their EGT's are higher than cylinder #1 and #3 due to the design of the OEM exhaust manifolds. Hot exhaust gas is scavenged out of Cylinder #1 and #3 faster than #2 and #4 (longer primary tubes in #2 and #4). This effect causes the Boxer "rumble" but also leaves hot exhaust gas in cylinder #2 and #4 longer which increases temp and can be detrimental.

 

This then leads us to the "why did Subaru engineers feel it necessary to run SCARY lean for over 2 seconds as the car transitions from closed loop to open loop? Everyone all at once, US DOT and CARB regulations!!!

 

Some food for thought.

[covertrussian]

I've seen people talk about leaving some timing in cyl #4 so that it would knock first and knock sensor would start retarding the timing. I checked against other turbo Subaru's and none of them had as much timing. Do you know why in 07+ LGT's they zeroed them out?

 

As for the closed loop to open loop delay, I actually got better gas mileage post removing it, but to the EPA mainly cares about emissions. I do think it's Subaru's fault, they could have handled it in another way instead leaving the car filthy lean.

[dga]

To slightly detract, but on topic of headers, how many melted bottom plastic engine shields do you see? I'm getting pretty close to getting a header, but would not wrap it and want to keep the stock plastic shield. Maybe glue some heat shield to the cover, if I can find the right stuff.

[m sprank]

^^ Subaru has adjusted the tunes slightly for differences in turbos, efficiency of heads, etc. B heads and a VF-40 on 05-06 and D heads and VF-46 on 07-09. Also, possibly due to complaints of issues because of the amount of compensation in cylinder #4.

 

As for the lean condition... Not Subarus fault (afaik). We have been informed that DOT/CARB forced it in order to pass emissions.

 

^ lots. Many headers will melt the plastic splash tray.

[Nrw]

Why are you against wrapping/coating the header?

[Nightmaresmk]

Becuase it traps heat the more heat the more likely your are to fatigue the metal. turbo cars can produce enough heat to wreck 304 stainless which 95% of all Subaru headers are made out of.

All subaru Turbo headers should be 321 in reality. thick wall to keep the heat in.

[dga]

...

^ lots. Many headers will melt the plastic splash tray.

 

That was what I was figuring. Thanks.

 

Why are you against wrapping/coating the header?

 

From my experience on motorcycles and what I've seen on older wrapped headers on muscle cars, the wraped header, through the cycles of heating, cooling, and getting wet looked like garbage shortly afterward. A friends wrapped motorcycle header actually caught fire after some oil dripped on it. Not sure for how long was it catching oil, but it did start to burn. Don't want to go through that on the Subaru.

 

Becuase it traps heat the more heat the more likely your are to fatigue the metal. turbo cars can produce enough heat to wreck 304 stainless which 95% of all Subaru headers are made out of.

All subaru Turbo headers should be 321 in reality. thick wall to keep the heat in.

 

This was the second reason. I want to really get the Killer-B header, but the price is tough to swallow for me and I do belive it's worth all of that. I have a MIG at home and I can weld stainless if it cracks, so I'm not too worried about cracking 304.

 

I really like the GTSpec for it's price point. Mike any reason you don't carry it any more?

[m sprank]

GTSpec stopped production.

 

Wrapping headers works when done right. Rarely do I see it done right. To do it right takes far more time, effort and material than most are willing to put in (as well as a sealant to stop oil impregnation and fires). The returns are also not all that worth the time, effort and expense (imho). I am a big fan of coating.

[Nightmaresmk]

Yeah Killer b is always out of stock. Full race is nice but i'm not running a turbo big enough. I'm picking up a maddad ceramic coated setup. UP

[BarManBean]

Yeah Killer b is always out of stock. Full race is nice but i'm not running a turbo big enough. I'm picking up a maddad ceramic coated setup. UP

 

Nah, there's lots of vendors that have killerb stuff in stock at different times. Maybe a pretty good lull going on right at this moment with the redesign...?

[Nightmaresmk]

Yeah i tried to get mine second hand. unfortunately some one beat me to the punch on the killer b's

[Code]

FWIW, Moore Performance will build you a set of headers with custom specs for runner diameter and materials. They offer multiple UP configs including EWG, twin-scroll, -.5 scroll, and different turbo styles (Subie,Mitsubishi,t3/4, vband, rotated)

[dga]

GTSpec stopped production.

 

Wrapping headers works when done right. Rarely do I see it done right. To do it right takes far more time, effort and material than most are willing to put in (as well as a sealant to stop oil impregnation and fires). The returns are also not all that worth the time, effort and expense (imho). I am a big fan of coating.

 

God. Damn. It. Should have picked em up last time I was getting stuff from you.

 

Honestly, never wrapped a set and not sure what a good job versus a bad one is, so I'd just avoid it all together.

 

Would coating do the same job as a wrap? Lower the plastic melting temperatures?

[m sprank]

Coating does the same "job" as wrap. Even coated or wrapped many headers will melt teh plastic splash tray.

 

I have 1 set of V1 KillerB headers stashed. I have 6 sets of V2 on order... delivery delayed again.

[BarManBean]

Mike, what's the latest on what will be changed for the v2 headers?

[m sprank]

Top Secret.