Holset Turbos, PART 3

[IPT]

The stocker 9cm is really bulky. Spidey seemed to make it fit ok. The BEP is a much smaller housing.

 

[selmerguy]

After reading again I dont think I will mess with the bep housing. the 9cm housing might be alittle amall for my goals. Even the 12cm housing might be to small. What I dont underdstand is why the holset housings are "small" but they seem to take up more space than the bep housing.

 

[Spidey]

I don't even know *which* turbine housing I have, but its quite large compared to any bolt-on housing.

 

[dsm-onster]

The hx35 has at the smallest a 12cm^2 twinscroll turbine housing. The 9cm^2 housing is attached by a v-band to the bearing housing. And only comes on the hy35, from what I've seen. The hy35 supposedly has a smaller turbine wheel. So if you have a 9cm^2 turbine holset then you have an hy35 and can't run a BEP housing anyway because of the v-band attachment and the "non-hx35" turbine wheel.

The stock holset turbine housings are bulky. They are casted with thicker walls. The twin scroll housings are especially large. But that is because of basically melding two turbine housings into one. Looking down the chute of my 18cm^2 housing, I see that the volutes have a bit smaller area vs. my 7cm^2 MHI housing. Each exhaust pulse sees 7cm^2 with my 16g. Each pulse will see only half of the holset twin scroll housing. Twin scroll housings have to be bigger to flow well. Add better cylinder evacuation and less backflow, nd you start to see why a twin scroll design flows so well and still spool so quick.

 

[selmerguy]

Yeah I havent bought a holset turbo yet. I was aware of the hy with the 9cm housing. There seems to be some conflicting information in these holset threads about the conversion from cm2 to a/r. I found these numbers on some honda forms:
6 cm2 = 0.41 A/R
7 cm2 = 0.49 A/R
8 cm2 = 0.57 A/R
9 cm2 = 0.65 A/R
10 cm2 = 0.73 A/R
11 cm2 = 0.81 A/R
12 cm2 = 0.89 A/R
14 cm2 = 0.97 A/R
15 cm2 = 1.05 A/R
16 cm2 = 1.13 A/R
17 cm2 = 1.29 A/R
19 cm2 = 1.37 A/R

I remember reading someones post that said the 12 cm2 housing would choke the hx35 and that they should go with the .70 a/r bep housing. If the afore mentioned numbers are correct how could the 12 cm2 housing choke it when it is bigger that the .70 a/r?

 

[Steve93Talon]

I remember reading someones post that said the 12 cm2 housing would choke the hx35 and that they should go with the .70 a/r bep housing. If the afore mentioned numbers are correct how could the 12 cm2 housing choke it when it is bigger that the .70 a/r?

AR is only relevant when comparing housings within the same family. Saying that a Mitsu 7cm housing is ".49AR" doesn't automatically mean that it will outflow a Garrett T3 .48 housing. People throw around things like AR and trim like they actually mean something, when in reality it doesn't tell you a thing unless you're debating whether to go with a .xy or .xz housing within the same brand and family of turbos.

This is why FP and BEP were so reluctant to give an AR number for their proprietary bolt-on housings, people are going to just throw it around and compare it to Garrett/BW/Holset/Mitsu numbers and it's not even close to apples to apples.

To further illustrate the point, take a GT28R with a .86 hotside and sit it next to a T67 in a .68 T4 hotside. Take a guess which one's bigger.

 

[dsm-onster]

Exactly, Steve. I'm not saying that those numbers are correct. . . But since a 12cm^2 turbine housing is .89 a/r, having each pulse pass through half of a .89 a/r housing would be much more restrictive than having eah pulse pass through a whole .70 a/r turbine housing. IF those numbers compare dimensions. It all depend on whether the numbers compare overall physical dimensions. Or compare flow/spool performance,which I doubt. If they compare dimensions. Two different brand housings at .89 a/r can flow much differently

. . . This is why we shouldn't compare turbine housing numbers between manufacturers .

 

[Spidey]

My housing is a twin scroll one, but it bolts to the center section. I like it a lot, I honestly would not go with a bolt-on housing with anything larger than a 16g(if considering a 20g, there are better Garrett alternatives IMO), they're just very restrictive, even the ".70" a/r BEP housings.

 

[ceedawg]

Yeah I havent bought a holset turbo yet. I was aware of the hy with the 9cm housing. There seems to be some conflicting information in these holset threads about the conversion from cm2 to a/r. I found these numbers on some honda forms:
6 cm2 = 0.41 A/R
7 cm2 = 0.49 A/R
8 cm2 = 0.57 A/R
9 cm2 = 0.65 A/R
10 cm2 = 0.73 A/R
11 cm2 = 0.81 A/R
12 cm2 = 0.89 A/R
14 cm2 = 0.97 A/R
15 cm2 = 1.05 A/R
16 cm2 = 1.13 A/R
17 cm2 = 1.29 A/R
19 cm2 = 1.37 A/R



I remember reading someones post that said the 12 cm2 housing would choke the hx35 and that they should go with the .70 a/r bep housing. If the afore mentioned numbers are correct how could the 12 cm2 housing choke it when it is bigger that the .70 a/r?

Innaccurate don't follow those 's

The #'s you see here at the bottom of my sig are the #'s quoted by a representative from Holset in England. Look at my sig. These #'s may be a lil off as well but alot more closer to the ones you posted.

12cm-.58-.60
14cm-.67-.70
16cm-.77-.80
18cm-.87-.90

 

[ceedawg]

The stock holset turbine housings are bulky. They are casted with thicker walls. The twin scroll housings are especially large. But that is because of basically melding two turbine housings into one. Looking down the chute of my 18cm^2 housing, I see that the volutes have a bit smaller area vs. my 7cm^2 MHI housing.

Matt I have an 18.5^cm3 housing and it definitely has more area than this ported bored out new 14^cm3 TS housing. I've also notice that a 3 instead of a 2 is used on the TS holsets, I could not tell you why. This 14cm was bored out to work on the HX40 turbochargers.
I can't speak for the 18cm hsing but the 18.5 seems to have a niced sized area in the volutes. I would love to compare my 18.5 to a TS divided garrett A/R.78.

Here are some pics 14cm vs 18.5 cm(rusty for now) holset divided TS hsing:

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

I will be selling the 14cm which was done up pretty nice. The 14 will spool up nice and have good top end on the HX40 but I'm more concerned with topend more so than spool up. If I could find another size bigger than the 18.5 cm I'm going to buy it and try it. For now I think the 18.5 cm will be perfect.

 

[ceedawg]

Oh and here is the HX40 with ported, bored TS holset 14cm hsing vs the HX40/35 hybrid with the puny BEP hsing that I sold :

You must be logged in to view this image or video.

You must be logged in to view this image or video.

 

[gixrman]

Finally got my manifold welded up today. Have to finish the WG pipes I was so happy I had a design that fit I slapped it on! But then I noticed its hitting the upper rad hose. Not bad for my first manifold. Hopefully all this hard work will payoff with some quick spool.

 

[IPT]

Swap in a 1G radiator and all will be good. I had to do that on my Divided T4 manifold.

 

[IPT]

they're just very restrictive, even the ".70" a/r BEP housings.

The BEP .70 a/r housings have made 700whp with some S300 turbos. IT is much bigger than the .55 a/r bolt on housings they have.

 

[Steve93Talon]

The BEP .70 a/r housings have made 700whp with some S300 turbos. IT is much bigger than the .55 a/r bolt on housings they have.

I was gonna say, the .70 T3 is anything but restrictive. It's at least as big as a .82 T3 Garrett housing.

You must be logged in to view this image or video.

 

[gixrman]

Swap in a 1G radiator and all will be good. I had to do that on my Divided T4 manifold.

Really, is the outlet farther to the left? Thanks!

 

[IPT]

Really, is the outlet farther to the left? Thanks!

Yeap.

 

[dsm-onster]

Holy he!! that bep .70 a/r is a big turbine housing!

Gixrman, that look good. Are the runners equal length? The don't appear to be. Not that it matters much.

My housing is a twin scroll one, but it bolts to the center section. I like it a lot, I honestly would not go with a bolt-on housing with anything larger than a 16g(if considering a 20g, there are better Garrett alternatives IMO), they're just very restrictive, even the ".70" a/r BEP housings.

You have at least a 12cm^2 housing then. Not 9c^2.

A .48 a/r garrett turbine housing with a gt30r turbine flows WAY more than a .63 a/r t3 stage1 turbine. Turbine housings are not independently restrictive. They influence the power band. They must be coupled to turbine wheel to compare flow potential and determine if the whole turbine side is restrictive to the compressor. Apparently (based on several dyno graphs) the hx35 turbine wheel in the small bep housing flows plenty well for the edge of the hx35 compressor map. 480-500whp has been achieved more than once.

Ceedawg, yea, mine actually states 18.5 cm on the housing. I've been just referring to it as an 18cm housing. And excuse my airheadedness. I have my t3 garrett, my 16g and my 18cm holset housing on my desk. And I didn't realize I was comparing the .63 t3 housing to the 18cm holset. The t3 .63 housing actually seams to have a larger volute than each of the single volutes of the 18cm holset housing. The MHI 7cm housing is still the smallest around here .

Do you guys see what I'm saying about volute size with a twin scroll turbine? Look at the volutes individually to better guage spool potential. Exhaust gasses enter the turbine housing in pulses. Using holset housings. . . Each pulse "sees" the 9cm^2 volute of the undivided 9cm^2 hy35 turbine housing. Each pulse sees HALF of the 12cm^2 total of the divided 12cm^2 turbine housing. But with a twinscroll, the pulses have twice as long to get through the turbine if the exhaust runners are divided properly. The exhaust gases can hit the turbine blades at a more aggresive angle because of the smaller individual volutes yielding more work done earlier-- or quicker spool--, and still have plenty of time to evacuate before the next pulse which yields more flow since the time neccesary to flow a certain mass has increased. The individual volutes can be nearly the same size as a non-divided housing volute, spool quicker, and flow at least the same. Size not being important , the quicker spool at the same or better flow is the goal of a ts housing.

Add the fact that the cylinders see much less to zero pressure from another exhaust pulse because of the runners being divided as they are, and you get much better cylinder evacuation/filling, which means more VE. Higher VE earlier in the rev range also increases spool speed. Higher VE up top yields better peak hp. So the better exhaust gas angle without a loss in flow, plus the better VE greatly increases spool AND flow.

But there is nothing wrong with the bep housings. They are physically much smaller which is better for clearance issues with manifolds that put the turbo in the stock location. And they flow plenty with each turbine wheel for their matching compressor wheel, while still spooling GREAT! They cannot be as efficient as TS housings. But the .55 bolts on. And the .70 is better than running a divided holset housing with an undivided exhaust manifold. And you get a more convenient o2housing bolt pattern with both.

 

[Spidey]

So wait, thats the .70a/r T3 housing on the right up there? I'll retract my statement, if so. I must have seen a picture of the wrong housing when looking previously, as it looked like the standard BEP housing on the left.

 

[dsm-onster]

Well, I have some logs. But I can't copy from my floppy to my online pc. Because of "java.io.IOException: Data error (cyclic redudancy check)". WTF !

Looking at the log on my other pc, I logged 31.5 lbs/min and 2200-2300hz up to 5200rpms. After that, the 1g maf went nuts (maf overrun). At 5200rpms with the small 16g, I was seeing 28.6 lbs/min and 1915hz. Both of these were done at 20psi. . . Well, the small 16g spikes a little at 5200rpms to 22psi. The h1c flows quite a bit more at lower boost with nearly the same spool as an evo3 16g.

I'll try again to recopy the files to the floppy. . .

 

[dsmconvict]

So wait, thats the .70a/r T3 housing on the right up there? I'll retract my statement, if so. I must have seen a picture of the wrong housing when looking previously, as it looked like the standard BEP housing on the left.

If I'm not mistaken BEP also made a .55a/r t3 housing too. Thats what you may have seen.

 

[IPT]

If I'm not mistaken BEP also made a .55a/r t3 housing too. Thats what you may have seen.

They USED to make that T3 inlet .55 there are no more left.

 

[dsm-onster]

Wow. I just blew my headgasket. . .

There's a miss and when I pull the number 1 fuel injector pigtail, the idle doesn't change I guess it's better than holing a piston. I pushed it a little. And I shouldn't have with the 1g maf. The lm-1 read lean WAY too fast for me to get out of her in time. Just a small amount of knock was actually recorded.

This turbo flows WAY more than any 16g at 20 psi. And spools quick enough to deliver tht familiar uncontrollable torquesteer but at only 3000-3500rpms in 4th gear. This is when my headgasket started leaking. The 16g hits harder. Or better to say: the spool isn't linear. It starts spooling later but rushes to 20 psi between 2500rpm and 3K. The h1c start spooling showing positive pressure at 2K an smoothly rises to 20+ psi by 3300rpms. I took the bait and pushed it a little too hard. I didn't really think there would be that much flow difference at 20 psi. About 15 rednecks were throwing their fists and cheering when I got into her pulling from the beachfront (no traffic or intersections). Most were my neighbors, who new what I had in her ("not much" to them ). I've been running the small 16g all day long at 20psi with the hacked 1g maf. I'm an idoit. There goes $50 and 2hours of my time . But the 2g maf will cure this.

I still can't deliver any good logs. It seams that the windows98 on my in car pc (not a lap top) is corrupting my floppy drives to the windowsxp my online pc is using. I can format floppies with my online pc and look at the empty contents. Then open each floppy with the pc in the car and download the files to those floppies. But, when I put any of those floppies into the online pc, the disk appears unformatted so I can't upload the logs to my online pc.

Nevertheless. The 16g wasnt maxed out and flowed 300hz less in volume flow than the h1c at 2 less psi. That's about 4 lbs/min difference to me, if the boost were the same. It appears that any td05h 7cm^2 turbo will flow that much less at 20psi with 272s and stock everything but a 3" exhaust and a 3" intake and medium-sized ebay FMIC. A headgasket swap, the 2g maf, and my online pc put in the car should show some more.

The h1c is fun and feels like a 50-trim.

 

[GORBS88]

Wow. I just blew my headgasket. . .

There's a miss and when I pull the number 1 fuel injector pigtail, the idle doesn't change I guess it's better than holing a piston. I pushed it a little. And I shouldn't have with the 1g maf. The lm-1 read lean WAY too fast for me to get out of her in time. Just a small amount of knock was actually recorded.

This turbo flows WAY more than any 16g at 20 psi. And spools quick enough to deliver tht familiar uncontrollable torquesteer but at only 3000-3500rpms in 4th gear. This is when my headgasket started leaking. The 16g hits harder. Or better to say: the spool isn't linear. It starts spooling later but rushes to 20 psi between 2500rpm and 3K. The h1c start spooling showing positive pressure at 2K an smoothly rises to 20+ psi by 3300rpms. I took the bait and pushed it a little too hard. I didn't really think there would be that much flow difference at 20 psi. About 15 rednecks were throwing their fists and cheering when I got into her pulling from the beachfront (no traffic or intersections). Most were my neighbors, who new what I had in her ("not much" to them ). I've been running the small 16g all day long at 20psi with the hacked 1g maf. I'm an idoit. There goes $50 and 2hours of my time . But the 2g maf will cure this.

I still can't deliver any good logs. It seams that the windows98 on my in car pc (not a lap top) is corrupting my floppy drives to the windowsxp my online pc is using. I can format floppies with my online pc and look at the empty contents. Then open each floppy with the pc in the car and download the files to those floppies. But, when I put any of those floppies into the online pc, the disk appears unformatted so I can't upload the logs to my online pc.

Nevertheless. The 16g wasnt maxed out and flowed 300hz less in volume flow than the h1c at 2 less psi. That's about 4 lbs/min difference to me, if the boost were the same. It appears that any td05h 7cm^2 turbo will flow that much less at 20psi with 272s and stock everything but a 3" exhaust and a 3" intake and medium-sized ebay FMIC. A headgasket swap, the 2g maf, and my online pc put in the car should show some more.

The h1c is fun and feels like a 50-trim.

What are you using to tune?

 

[ceedawg]

I didn't realize I was comparing the .63 t3 housing to the 18cm holset. The t3 .63 housing actually seams to have a larger volute than each of the single volutes of the 18cm holset housing. The MHI 7cm housing is still the smallest around here

Are we talking about the two volutes in the 18.5 cm housing combined?

If you were to eliminate the divider portion in the 18.5cm housing would it look bigger inside,volute wise than the T3 .63 single opened housing? If not then it looks like I'll have to find an HX50 or a 22cm holset divided turbine hsing. The 18.5cm Holset maybe = in size to a Garrett A/R .78 ,.83 or even smaller if that's the case.

I would love to get my hands on a TS T3 A/R .83 divided housing and compare it to this 18.5 cm holset hsing.