CA18DET Secondary Butterflies Control System

Discuss topics related to the CA18DE and CA18DET series engines.
User avatar
biosehnsucht
Posts: 1839
Joined: Thu Feb 26, 2004 10:20 pm
Contact:

Post

Updated 04/04/2008 - See Method 3

Basic Theory of Operation The amount of fuel and air sucked/forced into a cylinder is dictated by several factors, one of which is charge velocity. Higher velocity means more gets in. At low RPMs, you usually aren't generating much boost, and since you don't have much displacement, you're not sucking much either. A larger opening lets the charge in slower, a smaller one faster. Think of a garden hose: just holding the hose out, the water comes out at a decent rate, but it's in no real hurry. Press your thumb over the opening to restrict the exit area, and flow velocity must increase, for flow to be maintained. At some point, you are restricting total theoretical flow, but at lower flow levels, this will cause a higher exit velocity. The same is true with the CA18DET's secondary butterfly system. At lower airflow points (i.e. RPM), the butterflies are closed, forcing nearly all intake charge to go through half the space, increasing its velocity into the cylinder and therefore the rate at which the cylinder fills. More charge = more power, all else being equal.

CA18DET Secondary Butterflies Control System The secondary butterflies are controlled via a vacuum operated actuator, that sits on the back end of the intake manifold. When there is a vacuum, the butterflies are closed, and when there is no vacuum, are allowed to open. When correctly hooked up, this is controlled via the ECU for (presumably) optimal operation. The stock system uses a fairly complicated arrangement of gizmos, which I will cover towards the end. There are many ways to hook up this system. I'll start with the simplest.

Method 0: Do Nothing (WRONG way) This is the worst method, i.e. not hooking it up at all. You get no benefits at all, and potentially some losses vs. not having the butterflies at all.

Method 1: The Simple way This is the bare minimum anyone should do, and is very straightforward. Basically, you connect it to the intake manifold such that it sees vacuum and boost directly. As described above, when there's vacuum it'll pull the butterflies closed, and when there's no vacuum or boost, it'll let them open. This is not necessarily optimal, and from multiple accounts, is nowhere near as good as the correct way of hooking it up. You can either dedicate one of the vacuum ports to this, or T it in with something else (i.e. the FPR).

Method 2: The Stock ("Correct") Way This is how it should be done, if you're running a stock ECU. By stock I mean OEM type engine control, retuned ECUs inclusive. This is quite complicated. The ECU controls the butterflies by turning a solenoid on and off. When on, they close, and when off, they open. This is achieved by connecting (when on) the butterflies to a vacuum source, and then when off, connecting them to a vent to ambient air pressure.

In detail, the vacuum source comes off the intake manifold, goes through a vacuum delay valve (possibly check valve), into a vacuum tank, then to the solenoid, then another vacuum delay valve (they’re to smooth the on/off transition) and finally into the actuator. The solenoid also has a connection running around to the intake side of the engine, behind the MAF, before the turbo, where it "dumps" the vacuum when the butterflies are switched off. The first valve may be a check valve (makes more sense), but the parts catalog in all regions has it listed "delay", whilst the FSM says "check" - and "check" also for the other valve, which can't be right or it would never release the actuator! Use the OEM parts and you won't have to care.

What happens is that due to the vacuum check valve built into the vacuum tank, the engine is always pulling a vacuum, or at least, not "venting" the vacuum, from the vacuum tank. This keeps it ready to go, all the time, in theory. After the tank, we have the solenoid. This is the one the ECU controls. When the solenoid is on, the butterflies are continuously held in vacuum by the vacuum tank. When off, the solenoid shunts the vacuum in the actuator and it's vacuum hose to the ambient air pressure by the intake, which releases the vacuum and opens the butterflies. Between the manifold and the vacuum tank, and also between the actuator and the solenoid, are vacuum delay valves. You could live without the valves, they’re just there to smooth things out. They merely restrict airflow to make the pressure changes in the system less sudden..

Like I said, it's complicated. It also takes up a lot of space in the engine bay...

Here are some images showing the layout/connections, and how the solenoid operates. You can find the entire section 223 image in original form here, however it is not nicely colorized and quite hard to read. I have also blanked out all the unrelated bits in the section for the purposes of this discussion. Diagram of the stock layout and connections (if your browser (ex. Firefox) squishes it to fit the forum, right click and 'View Image' to make it legible):
Image
A note on above image: The uncolorized stuff in the bottom right with another solenoid and a few hoses I left in as it is connected to the same ambient source, but not highlighted as it is unrelated. This is the stock boost control system, when ECU sees more airflow than it thinks is ok it switches the solenoid to allow wastegate to see real boost, otherwise it sees the ambient pressure in that little cap.

Solenoid energized (on), secondary butterflies closed:
Image
As you can see, it is connecting the butterfly system to the vacuum tank, which provides a vacuum source to pull them closed.

Solenoid de-energized (off), secondary butterflies open:
Image
And here they are instead connected to the ambient air source, allowing them to open as they are no longer pulled closed.

Technical Details: - Part number for solenoid is 14956-35F10 (JDM CA18DET) or 14956-27M00 (USDM CA18DE); partcodes of 14956VA and 14956U respectively - Part number for vacuum tank is 22372-V6700; in the diagrams partcode is 27085Y (both JDM CA18DET and USDM CA18DE) - Part number for delay valve, to manifold, is 14958-H9910; in the diagrams partcode is 14958QA (JDM CA18DET) or 14958M (USDM CA18DE) - Part number for delay valve, to actuator, is 14958-V6700; in the diagrams partcode is 14958Q (both JDM CA18DET and USDM CA18DE) - Solenoid is ground switched by ECU, pin 8 on the ECU. This means one wire to pin 8 on ECU and one to power (12V). - Vacuum tank and both delay valves can be had from CA18DE equipped Pulsars. - Both the above listed JDM CA18DET and US CA18DE solenoids are functionally identical, but are slightly different due to dimensions and mounting, hence different part numbers. Any solenoid will work as long as you connect it properly and it has 3 ports, A, B, Common, in a configuration such that when de-energized ports A and Common are connected, and when energized B and Common are connected. physical positioning of ports does not matter.

*NEW* 04/04/2008Method 3: Cutting out the middle men

Quite a few Volvo, Volkswagon, etc cars from the mid to early '90's had a vacuum operated cruise control system. With this system, they had a vacuum pump with built in release solenoid. I have just slapped one of these in, and it awesomely simplifies things without sacrificing anything.

They look like this:
Image
I pulled this one off a Volvo wagon of some sort, didn't think to check what model. If you see that it has cruise control, follow the second throttle cable to the diaphragm, and from there follow the vacuum tube and you'll find this puppy.

As I have noted in the image above, there's three wires on their harness connection, which are for 12V, ground switched release solenoid, and ground switched pump engagement.

Connecting this in lieu of the standard junk isn't too difficult. You'll need to slap a SPDT relay inbetween the ECU and the pump as follows:
Image
SPDT Relay:Terminal 85 (or 86) - Switched 12V from engine harnessTerminal 86 (or 85) - ECU butterfly control switched groundTerminal 30 - GroundTerminal 87 - Release SolenoidTerminal 87a - Pump On

Vacuum Pump:12V input - Switched 12V from engine harnessRelease Solenoid



Note that for relay terminals 85 & 86, the orientation 12V vs ECU doesn't matter unless your relay wiring includes a protection diode, which you don't really need anyways since the ECU is built to drive a solenoid with that wire anyways. If you do have a diode (for example, bought a relay boss with it built in) then make sure whichever side of the coil / diode that has the line on it marking the cathode is connected to the switched 12V, and the other side is connected to the ECU.

Finally, run your vacuum line from your pump (white connection, black is the atmospheric vent) to your butterfly actuator (via delay valve if you wish to keep the OEM smoothness). Now cap off or remove all the vestigial bits from however you had it setup before.

I'd have to say now method 3 is the way to go, it is both very simple to setup and very functional.

(NOTE: SEE PAGE 3 W/A POST BY BUDDY WORM FOR UPDATES ON WIRING FOR THIS METHOD)

Alternative Methods If using some form of standalone or other gizmo letting you use MAP instead of MAF sensing, you could modify the "correct" method where it would normally dump the vacuum through the solenoid to the intake post-MAF, pre-turbo piping. You could just let this vent out (might want to put a small filter on it) instead, since the MAP will cover for any variances due to the "leak" effect. This will at least remove one run of vacuum connections.

Another possibility (one which I've toyed with "on paper" but haven't had the time/energy to pursue) is replacing the actuator with a servo or stepper motor, which is then controlled via a microcontroller of some kind. Many variations on this theme could be done, such as merely full open/close in response to an on/off switch a la stock ECU, or precise degree open/close controlled via a standalone. One would presumably install limit switches and something that rotates with the rod driving the butterflies to trigger them, so that you can shut down the motor when you hit full open/close, rather than keep sinking current into it for no purpose at all.

Keywords: butterflies butterfly secondary secondaries vacuum check delay boost valve actuator
Modified by biosehnsucht at 6:30 PM 4/4/2008


bentvalves
Posts: 1435
Joined: Sun Sep 03, 2006 8:58 am
Car: 89 Silvia K's

Post

awsome write up man, and it couldnt have come at a better time!

LongGrain
Posts: 52
Joined: Fri Oct 27, 2006 2:14 pm
Car: a piece of junk

Post

great writeup, i was thinking about not hooking mine up at all, so they were just always open, but now i know not to do that, thanks :D

User avatar
float_6969
Moderator
Posts: 19857
Joined: Mon Aug 26, 2002 1:55 pm
Car: CA18DET swapped 1995 Nissan 240sx (too many mods to list)
2015 SV Leaf w/QC & Bose (daily)
Location: Topeka, Kansas
Contact:

Post

Thanks Soooo much for the writeup! I'll get this linked in the stickies and see if we can't get you a little compensation for your troubles. Keep an eye out for an email from one of the administrators/super moderators/article guys.

niscort
Posts: 177
Joined: Sat Jan 03, 2004 2:47 am

Post

- Part number for delay valve, to manifold, is 14958-H9910; in the diagrams partcode is 14958QA (JDM CA18DET) or 14958M (USDM CA18DE)

this is actually a one way valve.. not a delay valve

User avatar
Bwana
Posts: 7320
Joined: Thu Sep 16, 2004 9:38 am
Car: Denali
Location: Escondido, CA

Post

Writeup has been articlized! There's a missing link in the middle of it, if you can get that to me I'll add it into the article, if not I'll just remove it. Let me know what you want for a custom title and I'll get one of the admins on it. Great job btw!

User avatar
biosehnsucht
Posts: 1839
Joined: Thu Feb 26, 2004 10:20 pm
Contact:

Post

niscort wrote: - Part number for delay valve, to manifold, is 14958-H9910; in the diagrams partcode is 14958QA (JDM CA18DET) or 14958M (USDM CA18DE)

this is actually a one way valve.. not a delay valve
Well, thats the funny thing. FAST in all regions says its a delay valve .. and the US Pulsar FSM says check valve for both valves, which if that was the case you'd never release the vacuum on the actuator and you'd be up a creek as you'd always run with 4 runners only. And the parts between the two (USDM Pulsar and JDM S13) are the same, so obviously one or both (most likely) are wrong. I have seen items listed as 'check valves' in FAST however, so its not like they call all 'check valves' 'delay valves'.

it would make more sense than dueling delay valves.. however I'm pretty sure my vacuum tank holds a vacuum without it, so.. it may be redundant. At first I had two of the same delay valves as the one after the solenoid, and replacing it with a "correct" part didn't make any discernable difference.

The point of the correct way was using the OEM parts, which if you obtained from a junked Pulsar or bought from Nissan would work fine regardless. I am going to go ahead and put a comment in there that it may in fact be a check valve..

User avatar
biosehnsucht
Posts: 1839
Joined: Thu Feb 26, 2004 10:20 pm
Contact:

Post

Bwana wrote:Writeup has been articlized! There's a missing link in the middle of it, if you can get that to me I'll add it into the article, if not I'll just remove it. Let me know what you want for a custom title and I'll get one of the admins on it. Great job btw!
wups.. fixing that now.

-edit- fixed.

You can delete our two posts about the link if you want, since they're not technical posts?

User avatar
iliketocrash
Posts: 1109
Joined: Sat May 31, 2003 5:58 am

Post

probably a delay valve and a check valve. cause i've tried blowing through them and you can definitely only blow one way through them but it is definitely harder to blow through than the actual vaccum lines that they connect to. so it does seem that they restrict some flow

niscort
Posts: 177
Joined: Sat Jan 03, 2004 2:47 am

Post

biosehnsucht wrote:Well, thats the funny thing. FAST in all regions says its a delay valve .. and the US Pulsar FSM says check valve for both valves, which if that was the case you'd never release the vacuum on the actuator and you'd be up a creek as you'd always run with 4 runners only. And the parts between the two (USDM Pulsar and JDM S13) are the same, so obviously one or both (most likely) are wrong. I have seen items listed as 'check valves' in FAST however, so its not like they call all 'check valves' 'delay valves'.

it would make more sense than dueling delay valves.. however I'm pretty sure my vacuum tank holds a vacuum without it, so.. it may be redundant. At first I had two of the same delay valves as the one after the solenoid, and replacing it with a "correct" part didn't make any discernable difference.
It may be that both are delay valves just for the DE version. The DET definately has a one way valve, at a guess so that the vacuum tank only ever recieves vacuum. Anyhow, this is the image from the JDM FSM which labels it as a one way. Hope it helps


User avatar
biosehnsucht
Posts: 1839
Joined: Thu Feb 26, 2004 10:20 pm
Contact:

Post

That would make more sense. tho that still means the JDM parts catalog is mislabeling it (not the first time). I had a scan of the JDM FSM tho that said the same thing, but maybe that was just a misprint that was corrected.

User avatar
themadscientist
Posts: 29308
Joined: Tue Nov 12, 2002 3:30 pm
Car: R32 GTR, DR30 RS Turbo, BRZ, Lunchbox, NSR50 Sportster 883 Iron
Location: Staring down at you with disdain from the spooky mountaintop castle.

Post

Nice writeup!

User avatar
float_6969
Moderator
Posts: 19857
Joined: Mon Aug 26, 2002 1:55 pm
Car: CA18DET swapped 1995 Nissan 240sx (too many mods to list)
2015 SV Leaf w/QC & Bose (daily)
Location: Topeka, Kansas
Contact:

Post

I tried the part number listed and it's definitely a check valve, not a delay valve.

And as stated, it allows the valves to close, but not open.

The way that I have mine setup is that on the outlet of the solenoid, I have a vac cap with a VERY small hole in it. It took me quite some time to get the opening size right.

The valves still shut quickly, but they open slowly now. It makes the transition from closed to open at the 3750rpm changeover much smoother.

bentvalves
Posts: 1435
Joined: Sun Sep 03, 2006 8:58 am
Car: 89 Silvia K's

Post

wasnt this going in the stickies?

the colored portions of the diagram make reading and understanding much, much easier.

User avatar
float_6969
Moderator
Posts: 19857
Joined: Mon Aug 26, 2002 1:55 pm
Car: CA18DET swapped 1995 Nissan 240sx (too many mods to list)
2015 SV Leaf w/QC & Bose (daily)
Location: Topeka, Kansas
Contact:

Post

It is in the stickies. He wrote this because I asked him to so that I could put it in the stickies.

zerothread/136061

User avatar
biosehnsucht
Posts: 1839
Joined: Thu Feb 26, 2004 10:20 pm
Contact:

Post

I rewrote it from scratch, my original even I almost couldn't understand, and I wrote it

Slidejenkins
Posts: 5
Joined: Thu Nov 16, 2006 12:36 pm
Car: 90 Silvia

Post

I've been reading thorough this thing at work , pretty brain dead at this point , so please forgive this retarded question... but I dont understand what the set-up offers us ?! getting rid off the valves ?!

or do we want solenoid energized (on), secondary butterflies closed ?!

or (off . butterflies open?)


User avatar
float_6969
Moderator
Posts: 19857
Joined: Mon Aug 26, 2002 1:55 pm
Car: CA18DET swapped 1995 Nissan 240sx (too many mods to list)
2015 SV Leaf w/QC & Bose (daily)
Location: Topeka, Kansas
Contact:

Post

The point of the valves is to increase low end torque.

solenoid on, valves open.

User avatar
Darius2250
Posts: 91
Joined: Sat Jun 21, 2003 7:31 am
Contact:

Post

Is there a way we can get a actual picture of this set-up I have had 2 CA's and nether one had this set-up hooked up. I can assume that there was some japanese looting of parts on both my clips.. well I can think that at least.

User avatar
superJoy
Posts: 202
Joined: Tue Jun 20, 2006 9:56 am
Car: 1989 RS13

Post

I don't know how much help this actually is, but here's a look at all the components brand new from Nissan:



It mounts on the bracket under the windshield washer motor between the firewall and the passenger side strut tower.

User avatar
Darius2250
Posts: 91
Joined: Sat Jun 21, 2003 7:31 am
Contact:

Post

non of that came on my clip, thanks

nocwage
Posts: 280
Joined: Tue Aug 10, 2004 9:01 am

Post

Super Joy, How much did all of that cost? Do you have a price breakdown on those parts??

missionsix43
Posts: 168
Joined: Sun Dec 03, 2006 12:44 pm
Car: 1989 240sx hatch ca18det

Post

The holes in the manifold for the butterfly shaft are small enough to be drilled out with a 1/8" NPT tap drill bit, and tapped to 1/8" npt and use 1/8" npt plugs.

User avatar
superJoy
Posts: 202
Joined: Tue Jun 20, 2006 9:56 am
Car: 1989 RS13

Post

nocwage wrote:Super Joy, How much did all of that cost? Do you have a price breakdown on those parts??
If I recall correctly... solenoid = $65, valves = $10 each, tank = $20. Well worth it IMO.

User avatar
The_Chosen_One
Posts: 1664
Joined: Tue Oct 17, 2006 2:05 pm
Car: 2009 Subaru Impreza 2.5i
Contact:

Post

where did you get it? Are you still able to get more? If I gave you the cash could you get it for me?

User avatar
superJoy
Posts: 202
Joined: Tue Jun 20, 2006 9:56 am
Car: 1989 RS13

Post

http://www.onlyfactoryparts.com

Just enter the part numbers in there and you're golden. They even have the JDM solenoid listed so you don't have to use the Pulsar one.

dattodude
Posts: 430
Joined: Wed Jun 29, 2005 2:51 am
Car: CA18DET Datsun 1200 B110 Sedan
Contact:

Post

FYI, after some more dyno testing today. I'm using 270/9.25mm cams for both intake and exhaust. I'm running the cams at 7deg in, 2deg ex at the camshaft (14/4 at the crank).

I did a power run with the butterflies 'closed always', and one with them 'open always'. And then figured where they stopped being a benefit. The point was 4800rpm for my configuration. So I set my Autronic SMC to open them at 4800rpm instead of the factory 4000rpm.

My CA18 is pulling 188kW (250hp) at the wheels now at 16psi, I'm running a big turbo, and should be running 20-25psi boost. But the engine is still young.

Hope this adds to the body of knowledge in regards to these much maligned CA18 accessories.

nismoplsr
Posts: 406
Joined: Sun Oct 19, 2003 6:45 pm
Location: CT

Post

Sweet dude!

Do you have some dyno graphs of each test so we can see the difference? That would be amazing.

User avatar
float_6969
Moderator
Posts: 19857
Joined: Mon Aug 26, 2002 1:55 pm
Car: CA18DET swapped 1995 Nissan 240sx (too many mods to list)
2015 SV Leaf w/QC & Bose (daily)
Location: Topeka, Kansas
Contact:

Post

IIRC somebody claimed a 20% increase in torque down low. They were using big cams and a big turbo as well.

I think the more "race" the setup becomes, the more noticeable the effect of the butterflies.

nocwage
Posts: 280
Joined: Tue Aug 10, 2004 9:01 am

Post

float_6969 wrote:IIRC somebody claimed a 20% increase in torque down low. They were using big cams and a big turbo as well.

I think the more "race" the setup becomes, the more noticeable the effect of the butterflies.
I've got proof..Stock engine, but T3 super-60 turbo

I had the problem where the solenoid was never opening them and discovered it when i dyno'd my car and found the power curve was like an american OHV engine

CLOSED butteflies..

OPEN butterflies..

I lost 20lb-ft of torque pretty much from 2500 all the way to 4000rpm..

That's a BIG loss, and I'm not running anything fancy in the engine itself. If I lost 20lb-ft of torque than anyone with the stock cams lost it too no matter what boost or turbo..


Return to “CA18DE / CA18DET Forum”