Jeff20B

Three rotary projects need to run by summer. One engine is already assembled (never fired up) but I could tear it down to swap parts pretty easily. The other two are dissassembled and mostly cleaned up. Just waiting for advice before I start stackin' plates.



All three are early 4 port 13Bs with various porting already performed. They'll go in an MG Midget project, REPU, and a VW Baja.



I want good driveability on stock carbs (Since I have three jetted Hitachis here to use). For this reason, the intake ports are no larger than what came stock in 1974: intake closing at 50° ABDC I believe. Two sets of plates are ported to '74 spec. The third set are Y plates from a US model 1st gen at 40° ABDC.



The rotor housings have various porting. Set 1 are '74 with the classic MAZDA logo on the dowel pin area which had some badly done ports that I corrected and ended up about the same size as T2 ports. The chrome is decent.



Set 2 are '74 with stock ports. They are pretty small but good for low end torque. They open and close at US-spec for any 13B from the '70s or any 12A from '76-'85. The chrome is somewhat worn but they ran pretty well before I tore it down. Idle was smooth and power was great in my MG.



Set 3 are J-spec with GSL-SE port timing. They open later than my stock ported '74 housings and close at T2 spec, just like a GSL-SE. Chrome is very good.



Well there you have it. A run down of the three sets of side plate and rotor housings, as well as the vehicles that need engines.



My next post will cover the parts in the currently assembled engine and why that combination of parts may not be the best overall setup.

Jeff20B

I posted a similar thread on the other forum but it got really long winded and I didn't get many responses. That shouldn't happen here.



Ok here's a brief description of the engine that's already assembled. I built it early last year before the other parts described above were available. However now that I have enough parts to rebuild three engines, I can customise each for the vehicle they'll go in.



Y plates from a 1st gen ported to '74 spec

J-spec rotor housings with GSL-SE ports



It's in my MG but suffered an oil pan leak so it has to come back out to reseal it. Why not rebuild it with better parts while it's out? I don't see why I couldn't, other than time and money.



Now I'll briefly describe the previous engine in the MG. It was from my REPU. It had '74 spec intake ports as stock on some early 3B plates. The rotor housings were the somewhat worn ones with the cool MAZDA logo and the small stock exhaust ports. This engine was a serious beast in such a small car but was kinda gutless in the REPU. The exhaust in the MG is less restrictive and I switched to a light steel flywheel instead of the 30 pounder stocker. The reduced weight of the car also probably helped.



I was kinda hoping to build another engine for the MG with these same rotor housings because their smaller ports will be quieter (I can't add any more muffling due to lack of room) and worked well with the old '74 ports. Plus it's scary to drive it too quickly, so if it's NOT streetported, that could only be a good thing. Trust me, this car was scary with the old streetported 12A it had at one time.



This would leave the GSL-SE ported rotor housings available for either the truck or the baja.

Jeff20B

I've thought a lot about how to build a nice high torque engine for the baja. New seals, good chrome, small ports and all that. I didn't want the added complexity of EFI though, so it must remain carbed.



We know the GSL-SE engine had the most low end torque of any 13B Mazda ever made. Didn't it peak at some unruly low RPM? Like 2750 or something? Wouldn't that be ideal for a baja? Well yeah, but maybe I can come close without all those wires and manifolding...



There was a rare delivery van 13B from Japan that used a torque convertor and a stickshift. It was intended for low RPM operation and had tiny primary ports no larger than stock GSL-SE or S4 NA, except it was carbed. The secondaries were normal '76-'85 12A size. The exhaust ports were the same size and timing as US-spec GSL-SE; they opened later and closed later than US-spec 13Bs from the '70s.



The exhaust ports are actually visually larger than US-spec ports from that era. They open a few degrees later than ours, allowing more time for the exhaust gasses to push on the rotors, and close at T2 spec which is a lot later than ours. I assume the longer duration was to more adequately evacuate the exhaust from the housings at low RPM. Hey, it's pretty much what the GSL-SE ports do, and that engine had more torque at a lower RPM than any other 13B as far as I know. Perhaps that port timing is the secret ingredient in creating a torque monster?



Don't believe me? Read this Yaw article. Scroll down to the low RPM operation section. http://www.yawpower.com/febtech.html



Of course general practice suggests the smaller the ports, the better the low end at the cost of high end, and all that. To that 'end' I have the set of US-spec small ported rotor housings that did pretty well in the MG. I could always use them in the baja and I'm sure the engine would perform pretty well, but I want to try the more unusual housings for the low RPM-intended port timing. Plus they have better chrome.



I just happen to have the carb, manifold, rotor housings, waterpump and a few other odd odds and ends from one of those unusual delivery van engines. I was thinking about using it as a starting point for the baja engine. Now this van engine was designed for ultra low RPM use. The primary ports were exceptionally small, the thermostat was set for 190°F, the plugs were B6ET (very hot), the waterpump had a secondary thermostat type valve in the bypass hole, the carb had tiny 90 primary fuel jets and huge 90 primary air bleeds. The secondaries were 140 fuel and 160 air. Basically this thing would stop metering fuel effectively above about 5000RPM. The manifold also had channels connecting primary to secondary runners, which I discovered helps idle quality because two ports are sucking on the idle circuit instead of one.



I was going to use the intermediate plate with its super small primary ports but then gained access to a Cosmo manifold with reversed runners. The primary carb barrels flow into the front and rear plates through long runners while the short secondary runners flow into the intermediate plate. You must use a tall port intermediate plate with this manifold. All 1st gen 12A Y plates are the tall port type, as are all nitrided R5 and some early non-nitrided R5 plates, so I had a fair amount of choices available. Only earlier 3B and some R5 and NO plates had short ports. The rest were EFI and didn't need tall or mismatched ports.



No need for the delivery van's old heavy intermediate plate anymore. Now I can use a complete set of lightweight Y plates from a 1st gen. They will not be ported. Only the harsh 90° edges and casting flash will be smoothed. I'll leave a nice rough finish that should reduce fuel-wetting-out during quick stabs of the throttle.



The long primary runners of the Cosmo manifold have been said to improve low end torque. They should while making up for the loss of the small ported delivery van intermediate plate, and the Cosmo manifold's tall ports will flow correctly into the tall intermediate ports of the Y plate. The engine should have as much low end torque as before, but with far more power from midrange on up. It's a win-win.



By the way Mike, thanks for the encouragement on the other forum. However having had a chance to sleep on the descisions I made in that thread, I think the idea laid out above is far better for the baja.



I'll talk about the truck engine next.

Jeff20B

That old set of 3B side plates came from a '74-'75 REPU engine. I feel they'd best go back into an REPU. What do you guys think? This just leaves the last set of '74 rotor housings with the decent chrome and ports cut out to T2 spec.



I'd ordinarily prefer smaller exhaust ports for better low RPM torque, but the nice thing about early REPUs is the very low geared rearend. At 4.625, you could get away with bridgeports and it would still haul pretty well.



I ultimately chose the ported rotor housings because I planned to do a supercharger in the future and I needed to correct a bad port job that never flowed very well (they were gutless in my bro's RX-4). Plus the heavy stock flywheel can make up for some of the losses from porting. Besides, my bro's RX-4 received another set of rotor housings that we ported to T2 spec and mildly streetported the intakes which worked out surprisingly well with a new set of seals in NA mode (the old seals were leaky which could have lead to the gutlessness as well). Then he got an an SC and the low RPM operation (before it starts boosting) didn't have a detrimental effect in stop 'n go driving. Infact I'm betting the REPU will do even better because the rear diff gearing is so different from the RX-4's mega-tall 3.636.



By the way, the REPU will be used for occasional towing; hence the concern over low RPM operation.



As I've said above, the MG did fine with those stock ported housings with the less than perfect chrome, and what I've read about the later opening exhaust ports of the GSL-SE and its legendary low end torque convinced me to combine the weird delivery van rotor housings with some unported side plates for the baja.



If you guys can think of a better combination for any of those projects, I'd love to hear it.



Thanks.

Jeff

j9fd3s

yeah having slept on it, i think you need to look at what mazda has done over the years with their motors, especially in reguards to the application.



in the repu, with its low gearing, i think you want something thats got a lot of mid range, and top end. on the FD engines, they use a t2 sized exhaust port (everything after 86...), and the primaries, comparaded to a gsl-se, are huge. the secondaries compared to a 12a, are bigger, but not radically.



mg: seems like total power production isnt the issue, a stock 12a, would be fine...



baja: the other low rpm tork motor, is the jc cosmo, its gotta haul a big boat around with an automatic. compared to the fd, the primaries are small, they are maybe gsl-se sized. the secondaries are basically streetported, larger than the fd. i think the exhaust ports are the same timing, but they do have a slightly different shape.

Jeff20B

Thanks for the input, Mike. What you say makes sense.



That automatic REPU I got running in '05 definitely provided good feedback for the effectiveness of '74 spec ports. The torque convertor allows the RPMs to come up before the truck even begins to move. It's perfect and had a ton of power. The owner was pleased.



'74 spec intake ports have excellent midrange and adequate top end for daily driving with a stock carb and manifold. That is to say stock jetted for '74. In other words, really big.



My friend's '83 GSL 12A, also with '74 spec ports, which I ported and rebuilt early last year, moves that car right along although the primaries alone are kinda on the weak side at low RPM. I installed 94 jets because I didn't have anything bigger at the time. To give you an idea how big they should be, '74 13Bs had 106 primary jets. I'd think 100 for a 12A would be ok. What's more, when he opens the secondaries (stock 160) at even a fairly low RPM, it really comes alive so there is a lot more in the engine down low - we just can't access it with the current primary jets.



So for the REPU and its gearing, I agree that midrange and top end are more important than low end. Infact the current engine is mild streetported with crappy upside down D exhaust ports and is kinda down on power, and it can still tow stuff (a rebuild is in order for that engine too, but not right now). The T2 ported rotor housings should not be detrimental here. Plus they'll flow great when I get an SC.



One down, two to go!



You're right about the MG as well. Power is not an issue but driveability is. Why not exchange one for the other? When I installed the REPU's engine it could do a burnout at 2k with the stock truck flywheel. Then I threw on a light steel flywheel and it drove a lot better. Bingo!



So basically the MG will get whatever is left over after the other two or done. Fortunately that means the stock ported rotor housings with some chrome flaking since they're not really good for anything else. The used apex seals that went with them are still above 7mm so they're reuseable. It almost had too much power with this setup and the truck's side plates before. Lighter Y side plates with the same porting as the truck and fully flowing secondaries will make it scream.



Only one left.



Lastly we have the baja. I think since the JC Cosmo was set up for turbos and an automatic, it doesn't have as much super low RPM torque as say a GSL-SE. I could be wrong but I'd think the torque convertor would allow the engine to spin up enough to spool up the primary turbo before it even starts to move. That would be the only way to move that big old boat.



The delivery van also had a torque convertor with a 4 or 5 speed manual tranny. No clutch. Maybe it was so weak at low RPM it needed to spin up a bit before it would move the van? I hadn't thought of that before. I did drive my REPU engine when it had an NO intermediate plate like that. The NO plate has the exact same size ports as the delivery van and a GSL-SE/S4 NA. The problem with the carbed engines is they only flow through primaries at low RPM while the EFI TBs are progressive and flow two then four smallish ports at low RPM until the aux ports open.



The REPU was not fun on the NO plate. Whatever it gained in low RPM power, it gave it right back in poor driveability everywhere in the RPM range. This inspired me to try the Cosmo's reverse runner manifold. The primaries then flowed through the '74 spec front and rear ports while the secondaries flowed through the tiny intermediate ports. Driveability was vastly improved and I didn't even notice the secondaries were any weaker than before, strange to tell. I traded increased driveability for weaker secondaries which worked out just fine in the end.



Anyway just like with that guy's REPU, the torque convertor allows the engine to come up to its stall speed before the wheels begin to turn. Perhaps that delivery van engine was set up specifically to take advantage of this and the stop and go driving conditions using all its special features (for hot running) described above. I guess it's a good thing I have enough parts to convert it to a stickshift engine that will run cooler with more useable power.



Since the baja and the GSL-SE are both manual, I think I can find some common ground. I'd prefer it to start building torque right off idle if at all possible. I'd like best power from 1500 to not much higher than 4500 like a decently built ACVW, to help the tranny live longer (they don't like high revs unless you build them for it). I think this goal is within my reach. What do you guys think?



The other secret weapon will be the Cosmo reversed runner manifold. It was sold on both auto and manual '76 Cosmos and probably other models (so it's not a manual or automatic-only thing). This engine will be virtually identical to a CD Cosmo (RX-5) engine that someone simply swapped the rotor housings to GSL-SE and kept the other parts the same. Has anyone done that? There's gotta be someone out there.



Stock sized intake ports and GSL-SE sized exhaust ports. Anyone?

partly mazda

Boy them wheels are a turnin



My 2 cents

Lighter cars don't need the low end torque, go ahead and open up the ports.

A nice combination is to go conservative on the primarys and big on the secondaries.

This works real well with the stock carbs because those early reactor engines had to

be a little rich to keep that flame going inside the reactor. The engine will run leaner and

crisper without rejetting. Besides, when you open up the secondaries you will know it.

The seconday throttle plates are bigger , the runners are bigger, the ports can be bigger.

Plug off the intake coolant passage in the rotor housings with pipe plugs for the summer so

you can try different manifolds.(repu)

Use nitrited housings for longer oil control ring life if possible. There used to be 2 different part

numbers for nitrited and pre 79. I though,like to use the old stuff sometimes. Whats the

problem with a little ,smoke anyway.

You can do some porting the the intake manifolds at the carb base to get crossflow between

primarys and secondaries(sp).This might give you a little better low end on the pu.

Don't let the teckies get you down with all of the port analizing and flow charting stuff.

Most don't take into accout that the rotor obscures the port opening most of the time.

Exhaust ports, don't get too crazy, cutting up too much and you will loose power. Cutting

down will give you a quicker rev. Don't worry about flow too much, keep in mind these are

big peripheral ports, and not blocked by the rotor like the intake is.

A restrictive exhaust will make your ported engine run like a stock one.

And if you got the *****, build youself a set of hockey brackes to save a little money on gaskets

when you are experimenting. Opps, they said that couldn't be done! I slipped.

john

j9fd3s

your baja is interesting, never had to maximise power UNDER 5000 before, usually the other way....



but yeah the gsl-se's with the 6 ports closed pull hard up to about 4500-5000... i'm wondering about stacking some carb spacers too, or making the intake longer.



i'd also like a long primary header, or maybe true duals. or ideally you want a header thats optimised for a 4000-4500 peak, which has gotta be really long!

Jeff20B

The forum just ate my post. I'll try to remember what I wrote.



partly mazda, I agree that a small light car doesn't really need low end torque, but the MG was kinda unique. It accelerated like the GLC did when it had the 20B in it. Now that I've experienced that kind of quickness, I want to recreate it in the new engine.



I think I can do one better. First I'll keep the primary ports the same size ('74 spec) so there is some porting although it's pretty mild. 2nd I'll keep the reversed runner Cosmo manifold because it worked out so well before, and actually fits the engine bay. 3rd I won't use the NO intermediate plate with its tiny ports on the inside and out. There was a hideous port mismatch on the REPU engine. So the end result will be primaries just as powerful as they were before, which I fell in love with, and now the secondaries will provide the high end which it was lacking before (although you couldn't tell because they still added noticeable power). This car honestly doesn't need ports any larger. I must also not open up the exhaust ports because I can't add any more muffling.



That's basically what I wrote before.



Oh, nice picture. Was that from a nitrided R5 engine? Or non-nitrided? The non nitrided wear the oil seals out quickly, like you said.

Jeff20B

Mike, the funny part about the delivery van engine is the ports are the same size as the four low RPM ports on a GSL-SE. So it technically should produce similar power in similar conditions. That's the common ground I mentioned.



The longer intake will be handled by the Cosmo manifold. A carb spacer sounds like a good idea too.



This engine might peak at 5000 or 6000 though despite my efforts. The Cosmo manifold's long primary runners are good for low end but flow into normal 12A sized ports which aren't as good for ultra low end. So it might actually break even with the van's short primary intermediate ports with normal runner manifold at ultra low RPM, but probably flow a lot more in the midrange, improving driveability, which is a good thing. +1 -1 = +1 for the Cosmo manifold in the baja on primaries.



Next we have the short secondary runners of the Cosmo manifold which flow into the tall ported intermediate plate. The intermediate plates don't flow all that well period. It's because the runners are narrow and have too much surface area. This is ok for primaries, where the flow isn't very high, but for secondaries it's kinda not that good for ultimate power. This should work out ok for high end in the baja, since I'm not looking for much, but ultimately won't flow as much total CFM as the long secondary runners of the normal manifolds (like the one that came with the delivery van) into the rounder end plate ports. So -1 for total flow but +1 for use in the baja since it needs to peak at a lower RPM. Secondaries broke even.



I'll add another +1 for the Cosmo manifold because it allows all four ports to be the same size instead of the smaller primaries and normal secondaries of the delivery van. The ultra low RPM should be very similar and then the Cosmo manifold should start to hold its own from 1500 on up to 4500. The engine will probably peak above 5000 though.



This is just speculation of course. I won't know until I put it together and try it. But at least I know which combination of parts to use on each engine now.