For engine cooling I'm going to use a VW A2 style radiator with dual fans. I'm looking to run the coolant hoses through the longs or possibly behind the rockers. I know that's not how Renegade does it but I really don't like how they do it.

I mentioned earlier that Japan has higher octane fuel than we do and it is required for this engine (somewhere around 97 - 100 RON). If you run the factory ECU how will you handle the fact that you can only get 91 or maybe 93 octane fuel depending on your area? (Ok, I'll admit I can get 100 RON but only at a racetrack and it's like $6 a gallon).

I don't believe knock sensing is a requirement just yet. I'll use my ear.

For fear of sounding ignorant, I've never heard of AVCS before and I'm not sure it's on my engine. Looking over the engine sensors and timing belt I'm not seeing anything particularly exotic that lead me to believe I have it. There are programmable ECUs that will control AVCS (Motec M4 from what limited reading I've done) and I could go that route if neccessary but I don't think it will be.

Are you checking in on any of the Subaru BBS's out there? This engine has been successfully transplanted into other Subarus here in the US. I'd rather you get your non-914 related information from someone who's done it vs me who will have to do it 3 times to get it right.

Jon

Uh. As far as I can tell my engine doesn't have variable valve timing. I think the AVCS is only on the sti engines, but I could be wrong.

AFAIK we both got the motors from Nippon.

I do have a stock ECU, but I ain't using it because I aint interested in joining up like a million wires, and then not being able to debug it. I am doing megasquirt-n-edis instead, since it is not impossible to understand it.

http://auto.howstuffworks.com/question90.htm

Our lower octane rating basically means we can't compress our fuel as much without it spontaineously combusting - I think this will translate into us not being able to run quite as high boost levels as are possible with the higher octane gas. I think we'd do well to try read the knock sensor for tuning purposes so we can determine what we can get away with.

On the plus side, it means we can crank up the boost whenever we get racetrack fuel :-) And boost is the kind of thing we may be able to mess with more often if we get the right boost controllers etc.

l8r,

Fiid.

If you run stock levels of boost you will likely have to use less ignition advance due to the crappy gas we have. All the more reason to use the knock sensor. By the way, I can tell you from working with the programable engine management I'm running... the engine is knocking LONG BEFORE YOU CAN HEAR IT.

IMHO we would all be wise to play it very conservatively with the boost and advance unless the knock sensor is used.

-Ben M.

Unless you lower the charge air temperature through better intercooling, water injection, NOS, etc.



How? Does your MIC3 have knock detection? (I haven't read up on it in a long time) Anyway, I understand that knock happens before you can hear it, but I don't believe (please tell me if I'm wrong) that a few seconds of knocking will melt a piston. By a few seconds I mean 3 or less. I know my engine will knock some while auto-tuning but I'm not overly concerned. I'm not saying that knock sensing is bunk, it's vital to get max hp and fuel economy out of an engine; but you can run your fuel maps on the rich side as protection from knock. It's something I can worry about later, like when I try to go from 300 hp to 350.

Let's talk about knock detection - is there an open project to get this running? I would be happy with just an indicator light or something that I can use to tune from.

Also - have you guys thought about doing a water - air intercooler instead of the usual air-air? Lotus have done this reasonably successfully and some of the miata supercharge kits seem to take this approach.

Regarding knock... No, my MIC3 has no knock-sensor input. My experience was seat-of-the-pants. There were conditions when the motor sounded fine with the naked ear, but knock could be heard clearly if a stethoscope was used. If you don't want to place a stethoscope to the cyl. head while the engine is screaming under load (Two people and a top-down 914. You can use your imagination to fill in the blanks. Do not try at home.) you can epoxy a pizo microphone directly to a large metal aligator clip and record to a tape or camcorder. Recording to a camporder mic input would be good because you could video the gauge cluster or laptop tuning screen so you could watch the video to SEE what condtions were when knocking occured. Autospeed.com has a howto on using the mic and alligator clip, but the video cam idea is mine.

Knock does more damage than just busting pistons. It is hard on bearings, wrist-pins, rings, head-gaskets, you name it. An engine that is knocking regularly, but not severly may never bust a piston, but it might hammer out a rod bearing to the point that one spins some day. (disclaimer: I have no scientific data to back that up. It's just my guess).

Intercoolers... I'd like to stick with an air-air design for simplicity. I was thinking about sticking it in the fenderwell, duckting air to it from a scoop/scoops. Another thought would be to have it above the engine like stock units, but with air ducted to the underside of it from quarter-panel scoops. Because of the shape of the car, the engine grill area should be a low-pressure area. It should be feasable to get airflow ducted in from side scoops to exhaust well through the IC from the bottom and out the top of the engine bay. To confirm this I do plan to do pressure testing this spring with a homemade manometer.

(sound of power tools eminates from basement)
Mom> Boys, what are you doing down there?
Boys> Nothing Mom.
Mom> You better not be building a two-stage intake charge cooler!
Boys> No Mom, we were just... ah...
Mom> Why is all my dry ice missing? Don't you make me come down there!

-Ben M.

... has an incorrect mixture and will show up on the O2 sensor. Knock is a better and quicker indicator but, as I said, it's not a "requirement".


A top mounted air-air intercooler will get heat soaked by the engine any time you are stopped and will take a while to cool down once you get air moving over it again. Just something to consider.

It can have the correct mixture and still knock.

Andrew

Correct. Air/fuel ratio, compression ratio, intake air temp, fuel composition/octane, humidity, fuel temp, ignition advance angle, and manifold pressure all come into play. (And other obscure things like combustion chamber shape).

I meant that you could drive 50,000 miles with mild knock that you can't hear with your naked ear and never know it. I put 3,000 miles on my TypeIV last year with it knocking under heavy loads between 2,800-3,200rpm. I could not hear it. I only found out by listening with a stethoscope. Backed off the timing 2 degrees in that rpm range (I love laptop tuning) and the knocking went away. AF ratio was spot on before and after. In fact, as an experiment I richened it up in that rev range before backing off the timing and the knock remained.

Heat-soak with the intercooler is a valid concern. I was thinking that I might be able to duct air from the sides into an airbox. The IC would form the top of the airbox, and hopefully the airbox would insulate the IC from the engine heat to a degree.

-Ben M.

TI has a chip for knock sensor interfacing. We might be able to do closed loop knock detection for the megasquirt.

http://focus.ti.com/lit/ds/symlink/tpic8101.pdf


It's entirely possible that with the 6 spare I/O ports on the standard megasquirt processor we could set up knock control input as well as stepper idle air control.

Reading the article you referred to on Autospeed.com, it is not a trivial matter to detect knock electronically. It is hard to separate the sound of interest (detonation) from other engine sounds. The author on autospeed noted that a technician using the microphone technique detected knock way before an aftermarket electronic sensor did and that a sensor design that the author worked on had a hard time detecting high RPM detonation. Even a well sorted out system that works on one engine may (probably) will not work on another.

Andrew

Dave just mentioned this in another thread. The Jap Subaru motors are designed for 100 RON in Japan - so we actually are pretty close to that as long as we use premium.

Yep, the human ear/brain combo has amazing analog filtering ability!

-Ben M.

and you're supposed to use premium with a turbo'd car anyhow

See what I get for relying on my brain. Never again, I say.

I'm done knocking this retarded horse to death. Knock sensing will wait until the car is driving.

Not spot on, but not really far off either. 100 RON automotive gasoline should be roughly equivalent to 94 R+M/2. Best I can get locally at the pump is 92 (R+M/2).

-Ben M.

Talk to sam an others who have run thier water hoses like that and you will find out how awefull it makes trying to get the air out of the system. Give it a try, you may have a way of making it work better.

Huh? I thought we were having fun!

By the way, I've been doing my scooby homework and you've got me convinced that you'll be fine without the knock sensor if you keep the boost at reasonable levels. From what I've been reading the impreza crowd is getting pretty darn agressive with the boost before the knock sensor calls and tells them to turn the music down. I'm sure you'll have the motor running great with the megasquirt.

-Ben M.

That's a possibility. However I've seen 6 cylinder cars run oil cooler lines that way. Wouldn't that setup also be prone to air bubbles? Actually I've never heard of air bubble problems with oil coolers but it's a big system with a thermostat so my guess is it's possible. My radiator setup will most certainly be pretty stupid regardless of how I run the lines

Fiid is running Megasquirt, I'm using Autronic. My Autronic doesn't have knock sensing, now you understand why I'm not so hyped about it. If I had it I'd be singing its praises from on high, but since I don't, eh who cares? I'd rather have knock sensing integrated into the ECU instead of a separate system that will never integrate properly, blah blah blah. I would need a few more reasons to upgrade my ECU, at this point I think it does everything I need.

And I have no intention of keeping boost at reasonable levels

Hey look what the header fairy left on my doorstep yesterday

Right, right. Sorry. I knew that.

The weather suddenly got nice around here so I drug the 914 out of hibernation last night. I gave it a pre-flight, some fresh gas and went for a spin. <-- it was fun.
I was having a little off-idle hesitation when I put the car away last year and it was still there. I had been studying my ECU manual over the winter and had some thoughts about clearing that up. 2 minutes with the laptop and the little stumble at idle transition was gone. What is my point? I was all knotted up about useing a stock Subaru ECU or a pre-programed Link ECU with the Subaru motor, but after the reading I've been doing about the Subaru tuning, and after working with my ECU last night I'm thinking there's absolutely no reason not to use the ECU I've already got in the car. It should do just fine. As will the megasquirt and most definitely your Autronic (best of the three).

-Ben M.

ps - Car is fun to drive even with 80hp. Can't wait to quadruple that #.

Hey Jon,

that looks nice, does it exit towards the front of the motor or the rear???

Hey,

Jon's postman gives head (er). LOL

The header should exit to the rear of the car, but should be rotatable if necessary.

I worry about the air-bubble problem. I suppose we will need bleeder valves just about everywhere. I just don't like the idea of losing my coolant on a speed hump.

I'm not so sure about that. Have you seen the tuning software

They finally "upgraded" it to a windows version as it was still a DOS-only app until recently. The windows version is just like the DOS version, it could definitely use a re-work.

I shouldn't comment any more on running coolant hoses cuz I don't know how it'll go until I mount the radiator. There's a possibility I might run them under the car after all. I like the idea of putting the car on ramps to fill and burp the coolant system.

My tuning software is a DOS app too. As long as it works well, I'd probably prefer a DOS based application. Ever tried to drive a car and use a mouse at the same time? With the DOS software I can do everything from the keyboard and I can do it on an old junker laptop I paid $35 for. <-- that's me with the "tightwad" trophy.

-Ben M.

The MS protocol is open, so I have choice of several tools I can use to work with it.

I am still thinking of building a visualisation UI and replacing all the guages with a flatpanel. :-)

Yeah, a windows program doesn't preclude the use of the keyboard but unfortunately that's how everyone writes windows software. In the past I considered reverse engineering the Autronic protocol (since that's what I do for a living) but I'm too much of a capitalist to do it for free. One of the things I like about the Megasquirt is that there are a few options for tuning programs (including Palm Pilots IIRC). I bought into the last Megasquirt group buy for 1 full kit and 1 extra PCB. Wish I had time to play with it.

Ben, where did you end up mounting your ECU? I bet all our aftermarket controllers are packaged similarly (weather resistent but not water proof) and I haven't decided if it was stay dryer in the engine bay or in the passenger compartment

The megasquirt has a temperature sensitivity problem, that is one of the main reasons they say not to mount it near the engine. More than just weather issues...

I mounted my ECU to the backpad right behind the driver's seat down near the floor. It is most definitely NOT weather-proof. Just about every page of the documentation states that it must be installed inside the cabin.

-Ben M.

You are a rebel.

You could always mount it inside a rubbermaid container, just don't forget to burp it

What are the Megasquirt guys using for sensors? Are you using stock sensors? Which engines are you running MAF engines or speed density engines? If I order an engine from Nippon will it be a 2000+ engine or some other engine?

Brett,

for the MS, you can use almost any brand sensor, I'm using GM parts, others here have used the stock 914 sensors.......I'm sure the subaru sensors can be used with the MS as well..........

Hi Brett, sorry I kinda hijacked your thread. These turbo subaru engines came stock with MAF injection, but everyone I've heard of who is pushing huge power out of them has converted them to speed density to get the intake restriction of the MAF eliminated. Almost all the aftermarket ECU's being run on even the mild street WRX's use speed density.

If I use the MIC3 ECU I'm currently running on my TypeIV I'll have to use the sensors that came with the ECU instead of the stock sensors. I can't recalibrate it for different sensors*. The megasquirt could be recalibrated to use the stock sensors though. Fiid will have to speak up regarding his plans for sensors. I believe Dave Hunt recalibrated his Megasquirt to use the stock 914 cyl. head temp sensor (as opposed to a coolant temp sensor it was orignially calibrated for).

-Ben M.

*Technically one could build hardware between the sensor and ECU to modify the signal, but that could be a major pain.

Yeah I wan't real keen on using MAF system. I want to build a daily driver but I see no reason not to use what ever sensors the injection unit was designed for. The thread hasn't been jacked, All this is related to the WRX conversion.

Subaru switched to Speed density in 2001 I think. Just FYI.

My megasquirt on the 1.7 engine lives underneith the battery. I try not to run my car in rain, but it has been run in at least 2 quite severe rainstorms (one on the way back from laguna seca) with no problem whatsoever - and I have no raintray.

I'm going to use all subaru sensors. Subaru's VR sensors work with the edis unit and teh ford crankwheel too.

If you have brass coding balls (assembler kind) you could rewrite the megasquirt to read the stock subaru crank and cam sensors and get it to drive coil on plug. I do have brass balls, but I don't have the sport mounts, and more importantly I don't have the time to get it working.

right up front, no pictures !!!!!


I saw Jon's WRX motor today while bolted up to the transmission for mocking up for the engine mounts and I've got to say is:

"damn, that is one tight and compact engine"


it looks sooooo much better with all of the extra cr@p off of it...with only the intake manifold on, it's a very simple and clean looking engine. The intake manifold could use some detailing such as removing the unnecessary bosses and brackets that won't be needed...it looks like it's cast aluminum which should clean up nice as well..........

Why thanks Mikey. I'm not used to getting compliments on how small and compact my junk is.

Hopefully the radiator will keep with this design philosophy. Not to mention the "Powered by MasterCard" philosophy I've adopted to get this project done by the WCC. Be sure to come check it out for if it really is done by then I'm pretty sure monkeys will start flying out of my butt soon after.

* "Powered by MasterCard" is a registered trademark and is by no means an excuse for vendors to rape me blind.

Fiid,
I just had an idea regarding your power steering pump/idler-converion... What if you keep the pump and used it to circulate coolant (hydraulic fluid in this case) for a air/water intercooler (well, air/oil actually)?

Anybody have thoughts on how well this might work? I don't know anything about the thermal properties of hydraulic fluid. I'm sure it would have lower efficiency than water, but it might still be the best intercooler solution for these conversions.

Speak.

-Ben M.

ps - a few days ago there was a cooling system discussion going on and someone quoted size/model/price/supplier of a radiator that was the same or virtually the same as the Renegade unit. For the life of me I can not find that thread/post despite desperite searching. Anybody know the info? Thanks.

by the way, how are you mounting that thing into the car??

check out some of the Vanagon web sites regarding the Subaru conversion. It is somewhat popular with them, lots more power than stock Vanagon. i have been considering this for some time for my 86 Vanagon. The main drawback for me is the need for my 86 Vanagon to pass Nazi california smog laws.

the second drawback is the ground clearance issue, i believe it is several inches lower pan, however i understand that some converter (I think it was Kennedy???) is working on a low profile oil pan for the Vanagon, which may work on the already low 914 conversion. You do loose oil capacity doing this, and I would imagine that hard cornering with the low profile pan might cause you to loose oil at the pick up tube (this is just a guess here) It probably aint ever an issue in a Vanagon, but the way 914's hang the curves, it is an issue to consider.

other than the oil pan issue, the subaru seems like a nice choice. ecspecially since our 914's are pretty much smog exempt now days in California (you 76 owners will be next year)

One thing that scares me however is the proposals to have road side automatic smog checks, and if the state finds you modified your car, when they tag you for a roadside inspection you will be in trouble, even if the car is old enough to be exempt from the bi-yearly tests, technically you can not modify your car unless it is like pre 1963 or something like that, even if exempted form the tests. engine swaps are a big no no out here, unless you jump thru all kinds of hoops with a referee. to be legal you must add all the smog equipment for the year engine you install (ie cats, vapour cans, OBD systems, engine check lights on dash, and what ever other crap was required on the car that the engine you swap came from. pain in the royal ass.

i will be spending over $500 for my 1980 spitfire (bone stock) in the coming months to get it thru the new "enhanced area" smog test, hell I only drive the car less than 1500 miles per year. Yet my nieghbors burn wood in their stinky fireplace all winter long, making me gag every time I go out side (smell even gets in my house)

God damn stupid laws, and politicians, off with their heads (sorry I got off topic here!)

good luck with the subaru, let us all know what you find in researching this.

dave

Ben, I considered using the PS pump for an oil/air intercooler (no really, I did, I'm not just saying that). There are arguments on both sides as to which type of intercooler is best. Water cooling will increase intercooler efficiency but once the intercooler becomes heat soaked it is much harder for it to recover compared to an air/air setup. This has always been a problem area for subaru's due to placing the intercooler on top of the engine; and you can see it's a bigger problem area for the 914 where a hood scoop isn't available.

I'm going to start with a traditional WRX air/air intercooler to get the car running. I figure I'll revisit stuff like air/water intercooling or water injection when I have trouble making more horsepower.

I believe you. I've never had an original thought. Well, okay, maybe a few of the really dumb ones...

-Ben M.

Some notes:

P/S pumps run at several hundred PSI. Intercoolers like 5 PSI. I don't think running water instead of P/S fluid is going to drop it enough to be useful. It's also a higher volume that recommended.

I'm going to run a water/air cooler if I ever get there. Older Mercedes use a seperate Bosch water pump for the heater. A small oil cooler or heater core w/ fan should work for cooling.

This is the last time I'm going to say this... run rubber radiator hoses under the car. They will not tear on round speedbumps, rocks, cones, commercial driveways, etc.! The rubber is a respectable conductor of heat, I know this cause my floorboards get hot. If I ever pull mine off I'm going to run an insulator. There is a 10-15 degree cooling effect from airflow across the hoses.

VW Golf3 radiators are a nice fit and cheap. I might even make a kit out of it now that I have the facilities... or maybe just a Spal fan package w/shrouding. (shameless plug)

OK, I'm done. Get back in the garage

Chris

I would rephrase that as "capable of running at several hundred psi." Nothing says they have to be run at those pressures.
I think running water would ruin the pump. Hydraulic pump needs to be lubricated by the fluid it's pumping.
Now that's cool. Keep talking.
Do they have enough cooling capacity for a 300hp motor though?

-Ben M.

Hey Ben,

Your "twin" brother Jon emailed me to harass me about deleting that post as well

I'll send you a copy of it that I saved.

The check valve in the pump could be removed allowing it to flow freely instead of putting out hundreds of PSI. Not sure about the volume requirements.



Yeah I might end up doing this, but I want it to look better than the V8 cars I've seen done this way. Saggy hoses and electrical conduit clamps under the car?



You would use this radiator to cool what engine? A basic rule of thumb is to have the same surface area as the factory used to cool a particular engine. You can try getting around this with a double pass radiator but they are less efficient than just having the right number of square inches. I'm not trying to tell you what to do, you obviously have fabrication skillz, I'm just throwing the the info into this thread for everyone to see.

Whoa, there's an echo in this thread.

I just realized what I'm going to do for intercooling. I'm going to run an air/air intercooler in the airstream above the car. Picture John Cusack holding that boom box in "Say Anything"