The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support Fuel Injector Clinic
Please Support Rix Racing

Rotary Valve Cylinder Head Prototype

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

The way I envision this concept in my head is basically cutting a 90° port through each valve and having both the inlet and outlet in a teardrop shape so you can use the overlap working with the rotation to get the most air in and out of the cylinder. A setup like this seems like it'd work well for a turbo application because it will allow the air to compress in the cylinder before the compression stroke even happens. I wish I could provide a visual representation of it but that's a no-go on my phone.
 
Definitely a superior design to what I saw from those other guys. I see exactly what you did there and that should work pretty well with a turbine setup! Great work man, I am happy to see some ingenuity and engineering going into these platforms. Keep us posted!
 
So if anyone else is as confused as i was the first 3x times looking at this thread, here is an animation that calms the mind. :)





Op im interested in learning more about the results of this 4G head.
I have always been interested in making this engine non-interference, and all of the other positive gains that come with the change...

Great work !
Stuff like this is what keeps me in Eng School.
 
Last edited:
I think this is a truly amazing idea, I remember reading about rotary valve heads, way back when Coates was kicking the idea around, I'd love to see a working design too, op your efforts are much appreciated and thanks for sharing the info, some really cool stuff here.
 
So I bought a talon for an engine to use for this project. The guy I got it from was fed up with it because it had so many problems. The guy he got it from had really screwed it up. His words were "it was like he took scissors to it". So it had a new turbo on it but unfortunately it probably never seen any oil. The oil feed line going to the turbo (an-4 fitting) was stripped out. I went a bought brake line to make a new line for it only to find out the turbo does not have an-4 fitting rather a plain M12-1.25 hole. I just got done disassembling the turbo so I could clean up all the debris after chasing the screwed up threads. No crush washers on the bottom side of the coolant banjo bolt either smh...

What is the usual connection for the oil feed line to the turbo? I'm not seeing any M12-1.25 (male) to an-4 (female) adapters.

15123009_1301250196584058_8190891815376808554_o.jpg



Here is the car. Notice the super awesome headlights ;)

13310438_1128120983896981_8814718759914238666_n.jpg
 
AdR, great observations. As far as timing goes I have not got there yet. My initial design has the valves open/close exactly at TDC/BDC without any overlap for the proof of concept prototype to be changed later. The valve ports are in as cast condition so the tolorance are not great and I have yet to measure what the timing is at now that its built... When/if I can get the engine to run reliably I hope to do extensive dyno tuning on the valve timing. I still need to finish the water brake I'm building....

I expect you are right and optimal timing would be close to that of a traditional poppet valve setup. That being said there could be some physics going on with a rotary valve that I have yet to discover that would require different timing.

I lost what you are saying midway thru your third paragraph. Some of the Coates designs have a shared intake/exhaust valve. My design has completely separated intake/exhaust valves and porting like a conventional head. So in theory from a turbocharging perspective I don't think it will be too much different than a conventional head.. If there is something I'm missing come back at me?

I still need to get this head on a flow bench and see what she does... Anybody know of someone that has one in the Prescott, AZ area?

There is a flow bench in Prescott Valley At Mountain Valley Performance.
5792 fulton dr.
Prescott Valley, AZ
(928) 632-5470
Talk to Bobby Marcum
 
What is the usual connection for the oil feed line to the turbo? I'm not seeing any M12-1.25 (male) to an-4 (female) adapters.

You'll be hard pressed to find an m12x1.25 to -4an female. I would get the banjo adapter to male -4, or the direct male fitting. the banjo would allow you to drill/tap a set screw into the piece easier for restrictors (what I did).
the factory oil feed for a 2g should be from the OFH, which you can put a BSPT to -4an male fitting in and run a straight -4an feed line to the turbo.

Id love to see some flow bench numbers.

@goldenfab - I was thinking more about the effects of not having a seal on the SV inlet port, and i was wondering if you have the bearings sealed off on each side of the valve. my concern is boost pressure, when the CC is sealed, reaching other valve areas and possibly bleeding pressure over. if they are sealed, i think your assumption is right that the effect will be minimal.

Also, I need to read it again, but was there concern of the individual valves inlet-to-outlet overlap for port flow? I will re-read when I get a chance, as I'd be interested in the timing specs you have currently, to help in the future.
 
I called that shop and talked to the wife and am waiting for a call back. Else I'm going to TX for turkey day where there is a shop I used to work for that has a flow bench.

Thanks for the ideas on the oil line. Mine had a 1/4" brake line which the ends have a 7/16-24 threads which I thought was an-4 but an-4 has 7/16-20 threads. And I think the angles of the flair might be slightly different too. Anyways I got some adapters yesterday that I think will sort it out...

The bearings are not sealed off from one another in their respective intake/exhaust valve areas which I originally thought I was going to need to seal but I think it's okay since each port is at the same pressure since they all feed from a common source and pressure (they all link together in the manifolds). The intake and exhaust areas in the head are sealed off from one another so no issues for making boost.

No I don't think there is concern about inlet to outlet overlap flow per individual valve. I was just explaining my design - that there is 100% overlap rather than some intermediate chamber in the valve. Was clearing up a misconception which someone else thought the design was like. Kinda hard to understand what I'm doung I realize as I have not explained it very well.. The videos miliman13 shared do a good job as my design is very similar.
 
All good info, cant wait to see what the flow is like! what is the CC inlet size roughly? it looks about the same or slightly smaller than a sohc 2.4 heads exhaust valve, but without the poppet valve in the way, my hope is the "early opening" point will take advantage of that and keep velocity higher, but still have big "full open" flow.
 
so... how exactly are you sealing the rotary valve to the head?
are your "cam shafts" "rotary valve shafts" whatever we are calling them, cut by yourself also?
are they machined cast iron or something easier to work with for the prototyping?

anyway, always glad to see new products.
 
I would be interested to see a 16 rotary valve version. This is so awesome though, any YouTube video of the whole process? Yes as above did you machine your own rotary valves? How and where did you cast the aluminum? Did you melt down some old heads? Very curious. love to see the flow this head make. will increase with two Rotary valves per side? Intake and Exhaust. I bet you can't wait till you see the results.
 
The cross section of the valve ports at the seal is 1" ID, exhaust and intake are the same size to make it easier to build for now, it will likely change. I forget what the ID of a stock valve seat is, maybe slightly smaller. But my prototype has half the number of valves so I'm with you - hoping I get the same or better flow and higher velocity charge.

The valves are cast from aluminum and hard cost anodized. Yes the exhaust valves will likely melt, you guys can start taking bets ;) and grenade the turbo, it's just a starting point for the prototype since I do not yet have experience casting ferrous metals. The valves are fixed to a solid steel shaft. The seals are a special metalized graphite material, I have many more materials to test when I get it running
 
Yes a 16 valve version could be good but easier to do 8 for now until the design is proven.

I have some video I need to get up on youtube. Don't know when I'll get around to editing and posting it up though..

The valves are cast from aluminum and then machined. I was able to do all the casting and machining for the had in my shop. It took a LOT of work to get there. I used re-melted old heads so far but have ingot to use in the future.
 
i would think 8 'valves' would be fine but they could be a little wider than what you have and the inner area of the ports could be designed like a ying-yang shape or 2 tear drops like someone said.

yes getting it running before doing efficiency tuning is in order, just ideas for the future.

i like the graphite sealing material especially for prototyping. but it may prove to be the best cost effective long term solution also.

as far as casting, iron and aluminum have a lot of similarities. especially if you are doing sand casting. i've designed and built a lot of ATV parts over the years and the worst part is producing a nice cost effective product that the market is all over during prototyping and then coming to market and having dismal sales. a few times i resorted to giving stuff away trying to generate word of mouth, it only generated people asking if they could have free stuff.
 
So I got the head on a flowbench yesterday. A little disappointed in the results but now I have a baseline to work from. I have room to re-design the ports for more flow. Of course static flown on a bench is only part of the picture but the numbers were still really low. I only tested with the port 100% open. I need to do some more testing to see how it flows at different degrees compared to a stock head.

The results at 28" of water were 85cfm for the intake and 130 cfm on the exhaust.



 
I think the quick port transitions from horizontal to vertical, as well as 90* turns and large cavity volume changes may be impeding flow on the intake side. the exhaust is almost there, so my thought is larger valve inlets, smoother transitions into the RV itself and larger CC inlets/outlets.

still pretty cool! just gotta keep tweaking till you find the sweet spot.
 
I agree I really need to tone down the sharp turns and abrupt changes. One of the contributing factors to the crappy design was I was not able to draft the shape I wanted in the CAD software I was using. Lowesy excuse I know but now I have another CAD program that should make short work of it so no more excuses.

Back to work today after a bit of a vacation during the holiday. I'm still getting everything configured on the engine test stand I am building. Hopefully this head, although it does not flow great give me some usable data to using for the next re-designed prototype. Also I should be able to work on and get closer to perfecting my seal design with it. For the next head design I am going to get serious about the port geometry and work on the flow numbers and flow quality. I'm thinking a wet flowbench is in order. Also, I'm thinking about how I can fabricate the ports out of a clear material (maybe cold cast out of a clear plastic resin) so I can visualize the flow in it with a wet flow bench. I would love to hear any ideas or resources I should read up on for this type of thing.

I started making some youtube videos. Hopefully it gives a little better insight to my design and some entertainment or maybe its just as cryptic as my other posts ;)

https://www.youtube.com/playlist?list=PLfl9sMSPZMIXFePAnZnn9nTqobBVVkN9j
 
Status as of today:
-I got the high pressure oil line to the turbo sorted out and finished.
-Head is torqued down, all belts and pulleys are installed and the valves are timed.
-Exhaust, intake, and cooling plumbing is finished.
-I have a 2.5 gal fuel cell on order.

Today I am working on the wiring. I stripped all the relevant wiring harness out of the car and am salvaging what I need to make the engine run.

The way things are now, all the oil going to the head is blocked off. Does anyone know if this is going to be an issues (resulting to overflow to other bearings)?

I have a VCDS scanner (my daily driver is a VW). So I'll be using it for initial troubleshooting and parameter monitoring for generic OBD2 stuff. I have been looking into ECU tuning solutions for the stock ECU (E2T61674). Else I am considering using a megasquirt II I have laying around. If anyone knows of any options I should consider please share.

15492414_1331587993550278_1179313716211082613_n.jpg
 
Awesome.
Look into a tactrix cable and "ecu flash" for tuning the stock ecu.

Grab an oil pressure sensor off the oil filter housing. If the pressure indeed gets too high, you can port the housing until you get appropriate pressure.
Keep er up!
 
Thanks for the feedback Sp1kE. So it sounds like you are saying releiving excess pressure (if neeed) at the filter housing would be sufficient. If someone knows otherwise (would need to do it at the head instead) please chime in.

If anyone knows of a write-up for taking a stock wiring harness and stripping it down the the bare engine control essentials please share. I know I can buy a harnesses for this but I'm looking to use what I have since the car this all came out of is being parted out. I found the ecu pin schematics so I'm just going to work it from that as I go.

15380844_1331624123546665_459724165456176423_n.jpg
 
Last edited:
Unfortunately wiring the ECU is really taking a lot if time. ECU pins 13, 26, 92 are all connected to ground. I have ECU pin 80 wired to constant power (+12V) via a 5 amp fuse. The two black and red MPI wires are getting fed constant +12V power via a 20 amp fuse. I have ECU pin 82 getting switched +12V power which also energises the MPI relay which in turn gives ECU Pins 12, 25 (+12V) power. I'm thinking with the above done the ECU should boot up and turn on the check engine light and communicate with my OBD2 scanner. Instead the CEL (ECU pin 36) has no voltage and the OBD2 port is not talking with my scan tool. I'm all ears if anyone has any pointers. Here is a video I made describing the problem.

Good news is the fuel cell came in today but the bad is I could not find any -10 AN fittings in town. I ordered some online that should be here Monday. Amazon prime to the rescue! I also need a new CAS and knock sensor that also were not in stock but should be here next week.
 
I have the complete manual books for the 95, I can go step by step to explain but it will take for ever, I can take pics and send you all the required pages to get it running so you can figure it out.

You are almos ther with the project, I'm anctious to see it running.
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top