GalantVR4.org The Mitsubishi Galant VR-4 Forum
Galant VR-4 Forums » Galant VR-4 » Technical Discussions » ECMlink plus Cyclone
Previous thread Next thread

ECMlink plus Cyclone


Hertz Galant VR4.org Administrator
OneTitle to rule them all.
77/2000


Galant VR-4 org Post #: 802811 posted 06/17/09 12:23 AM     Remind Me!  Send Private Message   Edit Post      
Is anyone here running the dual runner Cyclone intake manifold and doing such with DSMLink or ECMlink? Is it able to control this or are you using a pressure switch?

What is the option if you don't have the stock vacuum canister? If I remember correctly any vacuum reservoir would work, right?

Would I have any issues with a cyclone + 2g head? Shouldn't be any different than a USDM intake + 1g head?

Thanks in advance



Ryan Hertz


Edited by Hertz (06/17/09 12:25 AM)

Posts: 13501 | From: Chicago, IL | Member Since: 07/29/02 | IP: (75.22.239.107) | Report this post to a Moderator

jepherz Galant VR4.org Moderator
Too Clean
1881/2000
50/1000


Galant VR-4 org Post #: 802815 posted 06/17/09 12:42 AM     Remind Me!  Send Private Message   Edit Post      
If the Cyclone was meant to match up with a 1g head, I can only imagine that the shorter inlets of the 2g head will purchase you a large step when going from the intake runner to the head, just like a 1g intake on a 2g head.



-Jeff
1881/2000
50/1000

Posts: 7872 | From: KC, Missouri | Member Since: 08/08/04 | IP: (75.87.128.187) | Report this post to a Moderator

Telecaster
Senior Member
1581/2000


Galant VR-4 org Post #: 802822 posted 06/17/09 01:27 AM     Remind Me!  Send Private Message   Edit Post      
You have control of the fpr solenoid with v3. So you can use it to control the cyclone valves.

Posts: 573 | From: San Jose, CA or Philippines | Member Since: 02/26/02 | IP: (119.111.141.88) | Report this post to a Moderator

chris
Unregistered


Galant VR-4 org Post #: 802850 posted 06/17/09 07:32 AM     Remind Me!  Send Private Message   Edit Post      
i used a vacuum reservoir from a 3g v6 eclipse as they have a cyclone style intake manifold also. It works awsome and is oem mitsu. btw it needs to have the ball inside as it acts as a check valve on one side of the canister.


Edited by chris (06/17/09 07:33 AM)

| | | IP: (72.200.148.16) | Report this post to a Moderator

keydiver
Unregistered


Galant VR-4 org Post #: 802872 posted 06/17/09 08:28 AM     Remind Me!  Send Private Message   Edit Post      
You can use the vacuum canister for the Eclipse GTS or Montero 3.8l with variable induction: MD125817. Its only ~$25.

| | | IP: (74.163.255.107) | Report this post to a Moderator

Hertz Galant VR4.org Administrator
OneTitle to rule them all.
77/2000


Galant VR-4 org Post #: 802896 posted 06/17/09 09:58 AM     Remind Me!  Send Private Message   Edit Post      
Quoting jepherz:

If the Cyclone was meant to match up with a 1g head, I can only imagine that the shorter inlets of the 2g head will purchase you a large step when going from the intake runner to the head, just like a 1g intake on a 2g head.




But I should be able to port that out to match, right? Or am I being crazy?

Regarding the JDM ECU control of the runners is it rpm or airflow based? Would it really be necessary to use the ECM to control it if a Hobbs switch would work as well if not better?

I ask because I've read that pressure-triggered water injection may not be as desirable as load or even temperature-triggered delivery. (I've been reading Forced Induction Performance Tuning by A. Bell )

Plus it sounds like I'll only be able to control one "external device" from ECMlink.



Ryan Hertz

Posts: 13501 | From: Chicago, IL | Member Since: 07/29/02 | IP: (69.209.93.194) | Report this post to a Moderator

belize1334
well bread and nobly conceived
1334/2000


Galant VR-4 org Post #: 802950 posted 06/17/09 12:17 PM     Remind Me!  Send Private Message   Edit Post      
Standard wisdom says that it's an RPM based signal for the runners. Remember that higher airflow measured at the MAS means denser air in the runners but the actual volume flow in the runners is gonna depend mainly on RPM and VE so you want to use RPM to trigger the switch.



Roger B. Scott
'91 Belize Green

Posts: 3313 | From: Dundee, Scotland, UK | Member Since: 11/18/03 | IP: (72.175.127.40) | Report this post to a Moderator

1990ggsxnj
Member ++
1578/2000


Galant VR-4 org Post #: 802960 posted 06/17/09 12:34 PM     Remind Me!  Send Private Message   Edit Post      
Yes, but isn't that slightly skewed seeing as the charge from a turbo can be different from one car to the next?

I've seen everywhere that it's RPM based, but at a certain amount of airflow opening the runners would help more idealy than just rpm. A stock engine/turbo etc running 12 lbs of boost will hit the point where the secondary runners will help much later than a highly modified one where the that same airflow will be sooner. Am I wrong on this?

Is this what you were thinking Hertz?



"Is that Spool Drool? EWW!"
90GGSX Blue-Chesapeake
91GVR4 Nile Black-Sidney (Sid)

Posts: 525 | From: Blackwood, NJ | Member Since: 11/21/08 | IP: (12.175.119.20) | Report this post to a Moderator

mj_rosenfeld
Unregistered


Galant VR-4 org Post #: 802975 posted 06/17/09 02:07 PM     Remind Me!  Send Private Message   Edit Post      
Here's a variation on the merits of RPM versus airflow trigger for the Cyclone. Should it open at the same RPM or earlier if it is used with a stroked motor?
Mike R.

| | | IP: (64.0.248.227) | Report this post to a Moderator

1990ggsxnj
Member ++
1578/2000


Galant VR-4 org Post #: 802981 posted 06/17/09 02:27 PM     Remind Me!  Send Private Message   Edit Post      
Watsun? I would think the different Cam overlap would have something to do with it as well...but one thing at a time here!



"Is that Spool Drool? EWW!"
90GGSX Blue-Chesapeake
91GVR4 Nile Black-Sidney (Sid)

Posts: 525 | From: Blackwood, NJ | Member Since: 11/21/08 | IP: (12.175.119.20) | Report this post to a Moderator

belize1334
well bread and nobly conceived
1334/2000


Galant VR-4 org Post #: 803021 posted 06/17/09 04:24 PM     Remind Me!  Send Private Message   Edit Post      
The issue of one runner vs. two is mainly about optimizing flow (which is related to how much air the engine wants to breath compared to the drag that the air experiences along the path) vs. turbulence (which improves atomization of fuel). The drag will be primarily dependent on the velocity of the air and thus the volume of compressed air that the engine is consuming (not the same as the volume of uncompressed air that passes through the MAS). The ACTUAL amount of air will play a role as dense air is more viscous and thus causes more drag but it will be a small affect when compared to the influence that velocity plays. So, if you wanted you could make an airflow correction to the RPM trigger, but you'd still want RPM to be the primary factor since you can relate airflow through the runners directly to RPMs via the volumetric efficiency.

As for the stroker issue, I'd think that, given the same manifold, the higher displacement will consume more air so you'll want to activate at a lower RPM. The way to figure it exactly would be to take your exact configuration (cams, CR, displacement, etc will all play a role) and just do back to back dyno runs with and without the runners activated. Then set the trigger to the point where the power curves intersect.



Roger B. Scott
'91 Belize Green

Posts: 3313 | From: Dundee, Scotland, UK | Member Since: 11/18/03 | IP: (72.175.127.40) | Report this post to a Moderator

Dialcaliper
Flagration Member
1269/2000


Galant VR-4 org Post #: 803120 posted 06/17/09 09:38 PM     Remind Me!  Send Private Message   Edit Post      
As far as the 2.3/2.4 goes, I'd say that the RPM point should remain pretty much the same. It may move up or down a couple hundred RPM, but the intake resonance that's being exploited is purely a function of the pulses created by the intake valves opening and closing. You may see OEMs designing longer runner manifolds (that switch over at lower RPM) for bigger displacement engines, but that's only because they want the engine to peak at a lower point.

The main issue with using load triggered water injection is that unless both your WOT and part throttle tuning are spot on, and the water injection system is working perfectly, *and* your load sensitive trigger is tuned for all conditions, it will be extremely difficult to tune safely (You have to individually consider every point on the fuel/timing map, and account for hot days, humidity, etc). You will spend many hours fine tuning - It's probably not impossible, but it is not something you approach lightly.



1269/2000 Summit White

Posts: 1284 | From: Mountain View, CA | Member Since: 06/22/07 | IP: (158.184.37.55) | Report this post to a Moderator

keydiver
Unregistered


Galant VR-4 org Post #: 803138 posted 06/17/09 10:27 PM     Remind Me!  Send Private Message   Edit Post      
Exactly ^^^^ I agree. The Cyclone runners are tuned for specific rpm's, which doesn't change with displacement, so rpm activation has proven to be the best way to do it. The original JDM ECU's switched at ~4500 if I recall my tests, but I chose 4100 in my chips, based on graphs I have seen of the HP and torque with and without secondary activation.

| | | IP: (74.163.255.107) | Report this post to a Moderator

Dialcaliper
Flagration Member
1269/2000


Galant VR-4 org Post #: 803188 posted 06/18/09 12:59 AM     Remind Me!  Send Private Message   Edit Post      
Really, the best way to tune it, especially if you have DSMLink or a similar datalogger that lets you calculate "torque" and "horsepower", is to tune once with the vacuum disconnected, and once with the vacuum connected and set to high RPM so the runners don't open.

Compare the two plots and the changeover point will become obvious. My suspicion is that you will end up right around the 4000-4200 range for any setup (unless perhaps you have a thick intake manifold spacer or something). The only factor besides runner length that might affect the RPM slightly is boost pressure (sound pulses travel slightly faster in denser air) but it should be pretty minimal.

Obviously the actual torque and horsepower numbers may not be accurate, but multiple plots will give you a relative comparison of which is better.

It's key to note that the pulses we're talking about are essentially sound (pressure) waves, which travel at the same speed down the runner no matter the actual flow velocity of the air moving through (aka, the speed of sound). As far as it matters, the intake air might as well be not moving. The effect of flow speed on cylinder filling is a completely independent effect.

The "Helmholtz" resonance has an effect solely because the sound waves are nothing more than increased or decreased pressure zones in the intake. The faster the opening/closing of the intake valves, the shorter the length of the pulse. When the "wavelength" matches up with the length of the runner going up and down, you end up with an amplified high pressure zone right at the intake valves. Voila

A cyclone runner specifically designed for a larger displacement engine with the same RPM characteristic would be the same length, but slightly larger in diameter (unless you wanted to speed up flow at lower airflow).



1269/2000 Summit White


Edited by Dialcaliper (06/18/09 01:08 AM)

Posts: 1284 | From: Mountain View, CA | Member Since: 06/22/07 | IP: (158.184.37.55) | Report this post to a Moderator

Hertz Galant VR4.org Administrator
OneTitle to rule them all.
77/2000


Galant VR-4 org Post #: 803303 posted 06/18/09 11:53 AM     Remind Me!  Send Private Message   Edit Post      
Great info on the Cyclone guys, thank you.

Quoting Dialcaliper:

The main issue with using load triggered water injection is that unless both your WOT and part throttle tuning are spot on, and the water injection system is working perfectly, *and* your load sensitive trigger is tuned for all conditions, it will be extremely difficult to tune safely (You have to individually consider every point on the fuel/timing map, and account for hot days, humidity, etc). You will spend many hours fine tuning - It's probably not impossible, but it is not something you approach lightly.




So would you advocate taking those variables out by activating water injection with boost pressure?



Ryan Hertz

Posts: 13501 | From: Chicago, IL | Member Since: 07/29/02 | IP: (69.209.93.194) | Report this post to a Moderator

Rausch Galant VR4.org Moderator
Rock Star Status


Galant VR-4 org Post #: 803306 posted 06/18/09 12:01 PM     Remind Me!  Send Private Message   Edit Post      
Quote:

The faster the opening/closing of the intake valves, the shorter the length of the pulse. When the "wavelength" matches up with the length of the runner going up and down, you end up with an amplified high pressure zone right at the intake valves. Voila


Not to go too far O/T, but this is essentially why they 'tune' intake manifolds, etc. to get better than 100% VE on N/A race applications. I do believe...



Rance lives here...
Founder of Sold out to BMW Cru, then sold out there too.


Posts: 12049 | From: Cleveland, OH | Member Since: 12/21/04 | IP: (63.253.227.122) | Report this post to a Moderator

Dialcaliper
Flagration Member
1269/2000


Galant VR-4 org Post #: 803331 posted 06/18/09 01:36 PM     Remind Me!  Send Private Message   Edit Post      
This is way off-topic from cyclone activation, but here's what I'd do (and might someday if the engine ever goes back in the car...)

I would use some of ECMlink's new "nitrous" features - particularly the ability to run dual maps, based on whether the FPR solenoid is activated or not. Tie the FPR solenoid output to a relay, which turns on the WI system. Tune the base map conservatively, without the WI, and tune the secondary map with timing advance for when it's activated.

The key here is that you can tie the FPR solenoid to specific conditions, including load factor if you want. More importantly, you can also set a knock condition (only activate with 0-7 counts of knock for example) for the solenoid to be activated. That way, if *anything* goes wrong (whether it's a WI system failure, or simply an incorrect timing or fuel setting), the system will kill the WI system, *and* revert to a safe timing map until the knock count drops, instead of only retarding timing slowly.

Once that's in place, you can add whatever fancy features you want. If you can live without a two-step, you can even use the clutch switch input to determine whether the FPR solenoid triggers (and also the WI timing maps tied to it), giving you a dual mode setup, or even use a momentary steering wheel button to make the WI function like a nitrous button. You can even tie a dual-stage boost controller to the same relay output as the WI system that will allow you to run higher boost, but *only* while the WI system is running, *and* you have low knock. If knock is detected, the system will also drop the boost to a safe level.

Basically, you have knock sensor control tied completely into the WI system and the advanced timing that goes with it. It's the ultimate failsafe system that doesn't have to rely on flaky things like detecting pressure loss or a "clog" in the WI system, and will allow you to tune the system safely.

I'd been thinking about something similar with DSMlink V2, using the nitrous control to add a static timing advance for WI instead of retard (if that's possible), but the dual maps system on ECMLink would be much more elegant, and give you a better non-WI tune instead of just a static timing retard.

Anyway, it might be pretty ambitious, and would require a lot of tuning time to do the set-up, since you'd have to completely tune the car twice (unless you dialed in the tune already before installing the WI system), but having knock control tied to a timing map seems like the safest way to do it.

Regardless of what you do, you need to at least restrict the WI system to activate only under boost, probably just specifying a minimum load factor unless you want it active all the time under cruise, or you use something like the clutch switch to manually disable it). With this setup, you cannot have a separate control "downstream" of the ECU that disables the WI, or the ECU will be trying to switch to WI maps without the WI running.

So in short, no, I would take the variables out by tying it to knock instead of to boost pressure.

The only problem I see is that unless you're logging, or paying attention, you may not even notice the power loss in a situation like driving on the track, unless you also rig something like an indicator light to the FPR solenoid output.




Quoting Hertz:

Great info on the Cyclone guys, thank you.

Quoting Dialcaliper:

The main issue with using load triggered water injection is that unless both your WOT and part throttle tuning are spot on, and the water injection system is working perfectly, *and* your load sensitive trigger is tuned for all conditions, it will be extremely difficult to tune safely (You have to individually consider every point on the fuel/timing map, and account for hot days, humidity, etc). You will spend many hours fine tuning - It's probably not impossible, but it is not something you approach lightly.




So would you advocate taking those variables out by activating water injection with boost pressure?





1269/2000 Summit White


Edited by Dialcaliper (06/18/09 02:01 PM)

Posts: 1284 | From: Mountain View, CA | Member Since: 06/22/07 | IP: (158.184.37.55) | Report this post to a Moderator

Dialcaliper
Flagration Member
1269/2000


Galant VR-4 org Post #: 803348 posted 06/18/09 02:10 PM     Remind Me!  Send Private Message   Edit Post      
Note to the above - the solenoid activation will only control the system being on or off. Unless you could use the BCS to run duty cycle, or base it off injector duty cycle, you would still only have whatever control (RPM and/or boost pressure based probably) that the system has.

If you have WI control valves that can be based off injector duty cycle (ala Aquamist), you could essentially treat the water injectors like a set of secondary injectors, either using a fixed water/fuel ratio, or influenced by whatever controller is in the loop.

Then you'd actually have "load based" water injection control, instead of just "load activated"



1269/2000 Summit White

Posts: 1284 | From: Mountain View, CA | Member Since: 06/22/07 | IP: (158.184.37.55) | Report this post to a Moderator

keydiver
Unregistered


Galant VR-4 org Post #: 803422 posted 06/18/09 07:00 PM     Remind Me!  Send Private Message   Edit Post      
I designed a system like that for #1886, one of a kind. I used the $150 Aquamist highspeed solenoid. If I recall, it comes on at 15psi with 50% duty, and by 20psi it is 100%.

| | | IP: (74.163.255.107) | Report this post to a Moderator

prove_it
my racist jokes aren't actually funny
715/1000
1284/2000


Galant VR-4 org Post #: 803426 posted 06/18/09 07:11 PM     Remind Me!  Send Private Message   Edit Post      
Going back to the link and cyclone intake, I'm running the combo with 8.5 wisecos and a stock turbo for now. I had problems with my engine showing about 1-2 degrees knock(sometimes 4 degrees) when the rpm's hit the activation point. I tried the switch over at 4500, 4400, 4300, 4200, 4100, and even 3700 and still showed knock. I set link to pull 3 degrees timing when activated and still showed light knock. Finally I just pulled the vacuum off to test it. Runs strong with no knock at all. Also I was running 93 octane. Anyone have any ideas as to why this would be happening? Could the butterflies be causing a knock like vibration when opening?



#1284/2K NB, Totaled
#715/1K KG, Rebuilding
02 Acura TL daily duty unit

12yr+ Vr4 owner.
Honda/Acura master tech.
Family Guy

Posts: 4201 | From: Sioux Falls, SD | Member Since: 07/04/08 | IP: (96.2.103.100) | Report this post to a Moderator

keydiver
Unregistered


Galant VR-4 org Post #: 803591 posted 06/19/09 09:27 AM     Remind Me!  Send Private Message   Edit Post      
My guess is that it is because of your 8.5:1 compression. The 2G maps are much more retarded than only 3 degrees from the 1G maps in the mid-rpm's during spoolup. Are you running DSMLink v2 or v3?

| | | IP: (74.163.255.107) | Report this post to a Moderator

Hertz Galant VR4.org Administrator
OneTitle to rule them all.
77/2000


Galant VR-4 org Post #: 803722 posted 06/19/09 03:36 PM     Remind Me!  Send Private Message   Edit Post      
How does the JDM ECU (or Keydiver mod) operate the butterflies? Solenoid to vac canister? When is it applying vacuum? Low revs?



Ryan Hertz

Posts: 13501 | From: Chicago, IL | Member Since: 07/29/02 | IP: (69.209.93.194) | Report this post to a Moderator

keydiver
Unregistered


Galant VR-4 org Post #: 803765 posted 06/19/09 06:03 PM     Remind Me!  Send Private Message   Edit Post      
Vacuum is applied to the Cyclone diaphragm to keep the secondaries CLOSED. That part doesn't change. The only thing different between the way I activate it and the JDM ECU activates it is the solenoid operation is reversed. I energize the solenoid at 4100 rpm, which causes the FPR solenoid to dump the vacuum in the diaphragm out to atmosphere, opening the secondaries. The JDM ECU energizes the solenoid UNTIL ~4500 rpm, but has the same result.

| | | IP: (74.163.255.107) | Report this post to a Moderator

prove_it
my racist jokes aren't actually funny
715/1000
1284/2000


Galant VR-4 org Post #: 803767 posted 06/19/09 06:11 PM     Remind Me!  Send Private Message   Edit Post      
Quoting keydiver:

My guess is that it is because of your 8.5:1 compression. The 2G maps are much more retarded than only 3 degrees from the 1G maps in the mid-rpm's during spoolup. Are you running DSMLink v2 or v3?




Running V3 with 2G maps thats why I only used 3 degrees, I didn't want EGT's to soar. Also running a 3" GM maf that is maf comped in V3. Changed plugs several times, NGK BPR7ES gapped at .029" and am using a wideband with street tuned fuel ratio holding 11.0 across the rpm range.

Every time the butterflies open, shows knock pulls timing and then it smooths out. If I leave the butterflies open all the time there's no issue.



#1284/2K NB, Totaled
#715/1K KG, Rebuilding
02 Acura TL daily duty unit

12yr+ Vr4 owner.
Honda/Acura master tech.
Family Guy

Posts: 4201 | From: Sioux Falls, SD | Member Since: 07/04/08 | IP: (96.2.103.100) | Report this post to a Moderator

belize1334
well bread and nobly conceived
1334/2000


Galant VR-4 org Post #: 803794 posted 06/19/09 08:24 PM     Remind Me!  Send Private Message   Edit Post   
I noticed that mine used to get two counts of knock right when the butterflies opened. I never disconnected them to be sure but it was consistent with the rpm range and it was the only knock I saw. I decided that it was either the sensor misinterpreting the sound of the butterfly valves slapping or it had something to do with a momentary disturbance in the airflow reading do to the sudden availability of the extra runners. Either way, it was only 2 counts and didn't affect timing so I was never concerned. This was on my OEM long-block with a keydiver 92 octane timing curve.



Roger B. Scott
'91 Belize Green

Posts: 3313 | From: Dundee, Scotland, UK | Member Since: 11/18/03 | IP: (72.175.127.40) | Report this post to a Moderator


Pages: 1 | 2
Previous thread Next thread

Extra information
0 registered and 20 anonymous users are browsing this forum.

Galant VR4.org Moderator:  curtis, steve, atc250r, jcgalntvr4-244, cheekychimp, jepherz, Rausch, toybreaker, iceman69510, pot, FlyingEagle 

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled

Rating:
Thread views: 6560

Rate this thread


News & Events: News | Events
Galant VR-4: Newbies | General VR4 Discussions | Technical Discussions | How To and Info Archive
Marketplace: Parts For Sale | Cars For Sale | Good Guys | Bad Guys
Community: Members' Showcase

Contact Us | Privacy statement GalantVR-4.org

Generated in 0.115 seconds in which 0.045 seconds were spent on a total of 14 queries. Turbo powered.


Hertz's Galant VR-4 Page