New Miata Junk 2 Intake
#67
The issues with A/B testing go beyond the simple "it's expensive" issue. The A/B test results on setup A is going to vary drastically from the A/B test results on setup B. This part is going to see different gains on a stock-ECU BP05 vs. a built VVT head vs. a low-boost Rotrex vs. an SR20 T25 at 12psi vs. a built E85-fueled EFR6758. Add in the variable of which of four different OEM manifolds you are testing against, plus the variables of which header, how much power, what fuel, etc, and there are easily hundreds of different combinations, all of which will see slightly different effects. So you can spend hundreds of dollars generating an A/B test that is only relevant for maybe 5% of the market that you're trying to provide that data to. Of the other 95%, maybe 30% of them will end up misconstruing that data. At the end of the day, you're out of pocket for dyno time and labor and you have a net loss in actual accurate info being disseminated.
This creates a serious issue for a company like Skunk2, which is why I don't necessarily fault them for not publishing charts. They could publish 100% legitimate charts, and the Internet Hive Mind would still run amok with them, claiming that the tiny gain seen on an NA car "isn't worthwhile for a 400whp turbo car", and so on and so forth. There's virtually no upside for them to publish that data.
So yeah, I won't bother doing a true A/B for that reason. No manifold swaps on the dyno for me. I will get my current setup dialed in on the squaretop, and I will take the car back to the dyno with the Skunk2 setup on a different day. I'll try to control for as much as I can (same/similar ambient conditions, SAE correction, same Dynojet, no other changes), but I'm not going to try to generate scientific data that will be irrelevant for 90% of users.
This creates a serious issue for a company like Skunk2, which is why I don't necessarily fault them for not publishing charts. They could publish 100% legitimate charts, and the Internet Hive Mind would still run amok with them, claiming that the tiny gain seen on an NA car "isn't worthwhile for a 400whp turbo car", and so on and so forth. There's virtually no upside for them to publish that data.
So yeah, I won't bother doing a true A/B for that reason. No manifold swaps on the dyno for me. I will get my current setup dialed in on the squaretop, and I will take the car back to the dyno with the Skunk2 setup on a different day. I'll try to control for as much as I can (same/similar ambient conditions, SAE correction, same Dynojet, no other changes), but I'm not going to try to generate scientific data that will be irrelevant for 90% of users.
totally get that, but seeing some A/B plots on a few different cars wouldn't hurt for us to make informed decisions.
Hell Corky himself kept saying his cast IM did little benefit below 300HP and we are disagree and wanted them to release it -- the setups I've personally seen using it gained tons of spool AND top end.
#69
The issues with A/B testing go beyond the simple "it's expensive" issue. The A/B test results on setup A is going to vary drastically from the A/B test results on setup B. This part is going to see different gains on a stock-ECU BP05 vs. a built VVT head vs. a low-boost Rotrex vs. an SR20 T25 at 12psi vs. a built E85-fueled EFR6758. Add in the variable of which of four different OEM manifolds you are testing against, plus the variables of which header, how much power, what fuel, etc, and there are easily hundreds of different combinations, all of which will see slightly different effects. So you can spend hundreds of dollars generating an A/B test that is only relevant for maybe 5% of the market that you're trying to provide that data to. Of the other 95%, maybe 30% of them will end up misconstruing that data. At the end of the day, you're out of pocket for dyno time and labor and you have a net loss in actual accurate info being disseminated.
This creates a serious issue for a company like Skunk2, which is why I don't necessarily fault them for not publishing charts. They could publish 100% legitimate charts, and the Internet Hive Mind would still run amok with them, claiming that the tiny gain seen on an NA car "isn't worthwhile for a 400whp turbo car", and so on and so forth. There's virtually no upside for them to publish that data.
So yeah, I won't bother doing a true A/B for that reason. No manifold swaps on the dyno for me. I will get my current setup dialed in on the squaretop, and I will take the car back to the dyno with the Skunk2 setup on a different day. I'll try to control for as much as I can (same/similar ambient conditions, SAE correction, same Dynojet, no other changes), but I'm not going to try to generate scientific data that will be irrelevant for 90% of users.
This creates a serious issue for a company like Skunk2, which is why I don't necessarily fault them for not publishing charts. They could publish 100% legitimate charts, and the Internet Hive Mind would still run amok with them, claiming that the tiny gain seen on an NA car "isn't worthwhile for a 400whp turbo car", and so on and so forth. There's virtually no upside for them to publish that data.
So yeah, I won't bother doing a true A/B for that reason. No manifold swaps on the dyno for me. I will get my current setup dialed in on the squaretop, and I will take the car back to the dyno with the Skunk2 setup on a different day. I'll try to control for as much as I can (same/similar ambient conditions, SAE correction, same Dynojet, no other changes), but I'm not going to try to generate scientific data that will be irrelevant for 90% of users.
Ceteris paribus and all.
#70
If a vendor dynos it, you'll see two things. First, people will whine about the test conditions. "Why didn't you test it with a Rotrex?" "Why didn't you test it naturally aspirated?" Repeat for every other possible power combination you can think, cams, head types, different types and sizes or turbos, different boost levels, with/without the S2 throttle body, etc.
Second, no matter what the results are and how many disclaimers the vendor posts, people will take that dyno sheet as a promise, buy the part, and then whine when they can't replicate the numbers.
I totally understand why vendors don't want to go there. Why spend a thousand bucks on dyno rental and at least that much again of their own lost shop time to produce a bunch of numbers that will just generate hassle, whining, badmouthing of the vendor's reptutation, and probably not even generate any more sales?
Look at the discussion in this thread about S2's own dyno results, or at the discussion resulting from Jeremy's efforts in testing the squaretop manifold at FM.
--Ian
#71
Spec Miata guys do it because they're spending their own money to gain a competitive advantage on their own car.
If a vendor dynos it, you'll see two things. First, people will whine about the test conditions. "Why didn't you test it with a Rotrex?" "Why didn't you test it naturally aspirated?" Repeat for every other possible power combination you can think, cams, head types, different types and sizes or turbos, different boost levels, with/without the S2 throttle body, etc.
Second, no matter what the results are and how many disclaimers the vendor posts, people will take that dyno sheet as a promise, buy the part, and then whine when they can't replicate the numbers.
I totally understand why vendors don't want to go there. Why spend a thousand bucks on dyno rental and at least that much again of their own lost shop time to produce a bunch of numbers that will just generate hassle, whining, badmouthing of the vendor's reptutation, and probably not even generate any more sales?
Look at the discussion in this thread about S2's own dyno results, or at the discussion resulting from Jeremy's efforts in testing the squaretop manifold at FM.
--Ian
If a vendor dynos it, you'll see two things. First, people will whine about the test conditions. "Why didn't you test it with a Rotrex?" "Why didn't you test it naturally aspirated?" Repeat for every other possible power combination you can think, cams, head types, different types and sizes or turbos, different boost levels, with/without the S2 throttle body, etc.
Second, no matter what the results are and how many disclaimers the vendor posts, people will take that dyno sheet as a promise, buy the part, and then whine when they can't replicate the numbers.
I totally understand why vendors don't want to go there. Why spend a thousand bucks on dyno rental and at least that much again of their own lost shop time to produce a bunch of numbers that will just generate hassle, whining, badmouthing of the vendor's reptutation, and probably not even generate any more sales?
Look at the discussion in this thread about S2's own dyno results, or at the discussion resulting from Jeremy's efforts in testing the squaretop manifold at FM.
--Ian
#76
All of the notes in the first picture are played with the slide in first position (fully retracted). Multiple resonant frequencies are available with no change in length of piping.
#80
Compared to a lot of the custom manifold I've seen made where the user builds their own plenum/runners, this looks pretty well designed regarding equal flow to all 4 cylinders. I don't know if the runner length they choose is ideal, but it looks like it could be right for improving top end power. I'm interested to see how this works for folks. I've run VICS and now a squaretop, but if this new manifold outperforms both, I'll be upgrading at some point.