Hey I know how we all love any and all info about 4 port 13Bs, so I aims to please.

gearhead-42 and I both got our hands on a couple of special R5 engines recently. What makes them special? They are believed to have come out of the rotary powered bus in Japan back in the '70s. Tons of low end torque, but appearantly not much up high. I got the nearly complete engine and studied the carb. It was set of speicifically for low RPM operation, as were the spark plugs' very high heat range, the 190° thermostat, and there was even a funny water bypass valve that would open under certain conditions allowing hot coolant to recirculate regardless of the valve at the bottom of the thermostat. Interesting stuff. No to mention the lack of a blind plug at the rear of the eccentric shaft, and the oil return hole at the base of the rear plate; designed for an engine oil filled torque convertor with a clutch and manual transmission! Talk about smooth shifts!

Carl and Sterling will appreciate that the Hitachi carb had 90 primary air bleeds, 160 secondary bleeds, and 90 primary fuel jets with 140 scondaries. Good low end but forget about much above 4500RPM. I tested this carb on my Cosmo and it was slow. When the secondaries opened, they were so uh weak, for lack of a better term, that I could hardly feel them. Again perfect for a bus that needs to keep its passengers happy. Don't want to spill their coffee when the secondaries kick in.

Oh man, I could keep going but I should get to the point.

The goal for my engine is to build a torque monster that will have a rev range in excess of 4500 RPM. We're talking closer to at least 6000. Is that possible? Well, ask yourself is it possible that a GSL-SE can have power above 4500 without the aux ports functioning? The answer is no, not really. I bring up the comparison to a GSL-Se because the bus engine had the same size intermediate ports. The end plate low RPM ports on an -SE are about the same size as any '76-'85 12A ports, as are the bus ports. The only major difference between them worth noting is the progressive EFI throttle body vs a carb with channels at the top of the manifold. While the bus engine would have great low RPM power, and a ton of midrange, it still can not approach the significant low end of the GSL-SE.

I suppose I should mention at this point that another one of my goals is to come close or even match/surpass the low end torque of a GSL-SE, but do it with a carb for simplicity. Why do I want to keep it simple? Why not simply go with EFI like a GSL-SE core engine or something? Because I'm stupid!

If you believe that, maybe you should stop reading now.

Still with me? Good. I want to keep it simple because the engine is going into a VW Baja. Simple, carbed, less to go wrong. Now the need for ultra low end tire twisting torque becomes appearant. Actually the nice thing about a baja is it's so light that even a gutless 1600 with a tall geared (stock) bug tranny can take you to many places. Of course being the 13B 4 port guy that I am, you know I've gotta throw one in the back of my baja at some point. Why not sooner rather than later? Gotta build the engine first. Then get the adaptor. Then stick it in there and fab up an adequate cooling system and exhaust. This thread is to shed some light on the build for all you guys wanting to do a 4 port 13B some day.

Oh, one more thing to think about. Another reason for the comparison to the GSL--SE above, the exhaust port timing is exactly the same. Read this Paul Yaw article and especially pay close attention to the last setion near the bottom entitled Low RPM Operation. . http://www.yawpower.com/febtech.html

Ther later opening exhaust ports of the bus and GSL-SE are key to getting awesome low RPM torque. I will take full advantage of this wonderful little feature.

There's lots more to type and lots of pictures to post but I've gotta go do some car stuff this morning. Yay!

gearhead-42, if you'd like to post some pictures and a link or two to your engine's history, that'd be great. Also be sure to let us know what its intended purpose will be, and other significant details. Thanks.

 

its nice to see people doing alot of research and development on these engines ... seems like alot of you guys are doing more research on them then Mazda. We should really come up with a commitee of sort, and try to have Mazda somewhat sponsor them for doing so, and by sponsor them, i mean supplying information from their own engineering depts, maybe some blank housings and sideplates, and just stuff like that. I just hope one of you guys can read Japanese

 

I don't need anything like that. I think I've done a good enough job piecing the engine's purpose together adequately from all the clues present when I received it for the rebuild. Yep, an engine purpose built for low RPM operation and an oil filled torque convertor. Can't get much stranger than that.

Hey you guys are in luck. I snapped a couple progress pics of a waterpump mod I'm doing to the baja engine for weight reduction. This one shows an aftermarket aluminum (actually, I think it's pot metal) GMB waterpump impeller section sitting on top of my last spare 2nd gen water pump housing. Nice weight savings. The gasket sitting on top shows the bolt hole offset.


Do I drill a hole completely through the GMB?

This pic shows three different alt brackets. Bottom is FC, middle is FB and top is '74-'75 designed for a really big alt housing. It has an offset which normally interferes with the taller electronic dizzy caps, but would allow it to bolt semi easily to the stud position in the waterpump just to the left of the regular bracket position. It wouldn't be perfect, and the slot length is shorter, allowing less total alt movement up and down for tightening, but it's a thought.


The other option is to drill a hole right through the GMB at the FC position and secure it with a helicoil and a bolt. Or just simply cut some 10mm x 1.5 threads, like I did in the FC housing on my REPU engine, or possibly 11mm x 1.5 threads in the housing, for extra security, and use a stud or a bolt. Note the FC alt bracket must go on my REPU engine so I'll have to fab one from the FB or '74-'75 bracket sitting there. If I choose to drill a hole through the GMB, of course. Note the bottom part after the bend must be lengthed a little.

I'm not sure what to do at this point. All I know is I want to use the GMB and the FC housing together as any weight savings, especially at the extreme edge of the engine (in this case the rearmost end) is a plus.

Oh and before any of you start dreaming of your own lightweight waterpump solution without having to shell out the big bucks for those aluminum racing units, this pot metal on aluminum thing probably won't support a clutch fan. Mazda used a ribed cast iron impeller housing for a reason.

 

Jeff, this thread might help you with the GMB pump. It is aluminum.

https://www.rx7club.com/1st-generation-specific-1979-1985-18/racing-water-pump-what-do-i-have-367550/

 

Yeah, that's the thread I was thinking about. Stevan moved his alt bracket to the adjacent stud position.

I spent some time eyeballing it today before dropping the gasket in the FC pump on my REPU engine and getting it all torqued down. What I realised is I can keep things BACKWARDS COMPATIBLE. Hehe.

In other words, there is no need to drill through the GMB, or cut and reweld an alt bracket or any of that complicated catharsis (did I use that word correctly?).

My solution is to cut 10mm x 1.5 threads in the normal alt bracket hole in the GMB and use either a short stud or a bolt. I think I'll use a nice little fairly intact looking stud that I pulled out of the REPU engine. The threads were a little crushed but I attempted to chase them with a die. It galled a little, but this was a few weeks ago. When I looked at it today, it looked potentially good enough to use if I poked the good threaded side into the GMB and let the bad side stick out. Next I needed a nice tall nut in 1.5 like any old exhaust stud nut. You know what I'm talking about. All I had were two complete sets of four. Don't want to break a set just for this. Ah, what's that there? It's an oversized 15mm nut that's also taller than an exhaust nut. I don't know where it came from, but it's perfect. Nice smooth internal threads. Doesn't care about the ratty threads on the stud. Should be long enough to cover them up.

There is a chance that a bolt might infact work better here. Well, I also found a bolt that is only 1mm shorter than perfect length, so I'll see how they both fit and make my decision after cutting the threads in the GMB and seeing which fits best.

Oh, and this mod is backwards compatible, as I mentioned above, because the threads are only going into the alt bracket hole in the GMB so if I ever have to move it to a cast iron waterpump, for which it was intended, the threads will not cause any compatability issues at all. This is just like DLIDFIS in a way where I don't cut a single stock wire in the dizzy. You gotta like that way of thinking.

Hey trochoid, did Stevan ever post any follow-ups to his waterpump mod? Any longevity issues? Did he use an electric fan?

 

OK, now this is starting to get weird

Look at one of the other goodies I just happen to have laying around in prep for this build:



Jeff, are you spying on me?

 

Nope but you've gotta be spying on me. (I've had the GMB since '01 and the bus engine since '05)

Were you going to mount yours on a cast iron or aluminum housing? I can't wait to see how well it holds up on the aluminum housing with a stud in the alt bracket hole. I'm definitely not using a clutch fan so those extra stresses will not exist.

Thanks for posting a pic of the GMB by the way. It looks better than my pic, and clearly shows the alt bracket hole.

 

Jeff, I might have gotten a little lost..

My theory is:

If you are building a normally aspirated, and carburated low end torque monster, but want high end power...

Why don't you just build a 6 port block with working actuators, and a sidedraft carb? Make sure you got the exhaust (or other means) of opening the 6 ports and you're golden.

You can get even more power if you streetport or bridgeport the secondary ports.

That's basically a combination I had over 15 years ago, with a GSL-SE that whipped @$$ on everything short of a Turbo monster back in the day.

 

The goal was to get GSL-SE-like low to midrange without the EFI stuff, and if 6000 is the ceiling, that's fine by me.

No need for power above 7k to be honest, but anything up there is fine as long as I don't keep it there for long. Everyone tells me VW trannies don't like high RPM, unless you have them built to handle it. $$$$

I'll just get by with a 3 rib bus tranny.

Having said all that, I know what you're thinking... that I'm going to somehow wind up crippling the rotary to work in a VW. Well, no. In my opinion, the original bus engine was crippled, as is a GSL-SE with non functioning aux ports. Well I'm not going that route. The engine will not be crippled in any way.

Although I do have a set of GSL-SE (J-spec RE-EGI) front and rear plates, I don't have the manifold. I also don't have a side draft carb and have never tuned one. I'm also affraid of low RPM bogs, which is what the current dual port 1600 ACVW gives me. I want to be able to punch it at low RPM and have it respond - something I think a side draft just won't let me do.

The 1600 has nothing between idle and 1500. From about 2k to 3k is the sweet spot of this engine. Then power drops off the closer you get to 4000. I've only reved it out to 4500 once or twice and all it did was make more noise. The engine only maybe had 50 to 55 HP when new. It's old and tired now, with a recent valve job, so maybe it's close to 50. Maybe... on a good day.

Close to 50? C'mon, I'm into rotaries. It just doesn't feel right to drive something so wanting in the power department. But I gotta be careful about the powerband on whatever I do.

I read something recently about the old carbed 13Bs like what came in the REPU. It said at 3500 it had 117 pound feet of torque or at 4000 it had 120 torque. Oh, and 110 HP at 6000 RPM.

I don't recall the torque figures of a 1600, but I'd venture to say the rotary can pretty much double the numbers by at least 3500 to 4000, and match them at say 1500 to 2000. That might be wishfull thinking, but the engine I'm planning should produce higher numbers than those posted above.

I'm not expecting much from idle to 1500, (it's not a V8 you know) but I have a couple secret weapons up my sleave that may level the playing field or perhaps even give me the advantage. More on them later.

 

Interesting post! Keep us updated!


Don't forget about somender-singh's intake grooves!

http://somender-singh.com/content/view/101/37/

https://www.rx7club.com/showthread.p...ghlight=groove

I believe this slightly helped my low end torque and unquestionably improved idle quality.

 

Hmm... I don't think it would hurt anything to cut grooves in the intake.

I'd like to experiment with grooves in at least one manifold. I think I know just the one. It's a slightly less than perfect '74 manifold. I was going to swap it on my white REPU anyway (because I need its current manifold for something else). Might as well cut grooves before bolting it on. I've also seriously been considering channels in the top, which also have shown to improve idle quality.

I know I shouldn't change too many things at once. The exhaust needs to be dealt with on the white REPU, too. Ok, I'll change one, take it for a test drive or two, then change the other. That works.

I also need to swap manifolds on the Cosmo and add a header kit and RB presilencer. More importantly, I have everything I need to convert it to a 5 speed, so that will be coming up shortly. What do you think about a GSL-SE flywheel instead of a stock 30 pounder in a Cosmo? Should be an improvement, and the car is so heavy and the rear diff geared so tall I don't think a light steel or even worse an aluminum aftermarket flywheel would be very good. I just have to order a pilot bearing and seal. Already have a freshly resurfaced GSL-SE flywheel and a good 225mm disc and pressure plate. Then that auto crap can come out.

I think I'll change all the Cosmo stuff at the same time because it's pretty involved. I won't have the time to change one thing, button it up, drive it around and hope to remember exactly how it drove, then tear it back down to change something else and drive it again. The only thing that will be changed after it's running again will be the Cosmo manifold because I'll need it for the baja engine.

So, after the Cosmo is running, I will drive it around for a while to get a feel for how the car was supposed to drive, without all those power robbers everywhere. It should feel like a new car with a nice header, manual, and the correct manifold with the reversed runners... although I do have some concerns that maybe throttle response won't be quite as instant as it is on the shorter normal runner manifolds, like what is currently on it (the '74 manifold that will go on the white truck). Everything takes a back seat to throttle response and driveability in the baja project. Of course having short secondary runners that flow into the intermediate plate may limit flow up high, but when they open the response should be immediate.

I also have another reversed runner Cosmo manifold that's going on my MG Midget project. I already tried this manifold in this car before. It was on the '73 3B engine that I just rebuilt. I borrowed it out of my red truck to get the MG running. Unfortunately the intermediate plate was an NO casting with the tiny GSL-SE sized ports which are short on the outside. Compared to the Cosmo manifold, which has tall intermediate ports, because they're the seciondaries, you can imagine a severe port mismatch hurting secondary flow, which it kinda did.

I switched to the Cosmo manifold originally to give the REPU a little extra grunt, which did help since most of my driving was on primaries. The engine was still quite gutless in the truck becasue the exhaust was not ideal (too big in places, too restrictive in other places). Then when I installed the engine in the MG for test fitting and got it to run, that Cosmo manifold really made that MG fly. Or maybe it was just the significant difference in weight? Or maybe a less restrictive exhaust? Heh, all three.

The primaries alone were kinda scary. Then when the secondaries would open it only gained a little bit more. I'm sure the cool engine I recently built for it, with '74 spec ported side plates and a tall port intermediate plate, that the primaries will be just as scary as they were before, and the secondaries should be downright terrifying.

Anyway I'll be able to test the throttle reponse in the MG and compare it with the Cosmo since both will have Cosmo reversed runner manifolds. The MG has '74 spec ported Y castings and stock '74 exhaust ports (smallish). The Cosmo has stock '76-'85 carbed 12A and 13B sized intake ports and stock exhaust ports (same small size as '74).

It will be interesting to find out the differences between '74 and stock ports in these two vehicles with the same manifold and exhaust port size. It's my guess the Cosmo's smaller 12A size ports will provide better low end, and if that's the case, the baja engine will stay at stock 12A size, which is easy because it will consist of Y casting side plates. The only difference will be the exhaust ports, which open later and close later than '74 and '76 spec. They're actually the same port timing as a GSL-SE, which had legendary low end torque. We're talking 133lbs at only 2750 RPM! That would be a great engine for the baja, but I don't have access to one and don't care for the EFI stuff for offroading where mud and water will splash onto the engine from time to time. If there's a problem I'd like to diagnose it with simple tools, which one can do much more easily with a carbed engine.

This thread is also for anyone out there who does have a complete GSL-SE engine and ECU/wiring harness laying around and wondering what to do with it. A baja would really move with one of those engines. I would have gone with one but I already had the next best engine; the bus 13B, which is carbed and fits into my comfort zone. Plus with a GSL-SE there is no challange; It's already a complete running package. I didn't want that. I wanted to get my hand dirty, to build this thing from the ground up from spare parts, to have the knowledge base from which to pull ideas and knowing what will work with what, and all that. I realise not everyone is in this position, and that's fine - you're more likely to succeed with a complete running engine package anyway, especially if this is your first time. I helped a friend install an REPU 3B engine in his baja back in '99 and have helped him work on little things over the years to get it to where it is today. Now he wants to pull it back out and redo a bunch of stuff. Even go with a smaller ported engine since his REPU engine was streetported with a Holley. Basically it's gultess and has poor driveability in the baja. I think he'd prefer a stock carb too. I'm trying to avoid these kinds of problems entirely with my setup. Now you see where some of my concerns came from. That's only if you read this far into my boring long winded post.

So there you have it. Driveability and low end torque. Very important. The challange is getting it from a carbed engine. I think I can with a little know-how, decent seals, good side plates, nice chrome, small ports etc. It won't cost nearly as much as a VW engine. Instead of the limited RPM range of the bug or rotary bus engine, it should rev well to 6000 or even beyond. If '74 spec ports are viable, it'll have good power to at least 7000. This engine will not be crippled to choke it down to work with the VW tranny or anything like that. Nope. It started out with good low RPM characteristics, and I will keep them intact while bending the rules to break past the 4500 barrier. I believe the Cosmo manifold is the key in all of this.

So just for the record:

1st choice is a GSL-SE engine with stock ECU and wiring harness. Its only drawback is the complication and my lack of desire to want to fudge with it come troubleshooting time somwhere out in BFE. Plus I don't have access to one. That makes it easy.

2nd choice, which I think is better for someone like me, is the bus engine from Japan. There are a few of these around; gearhead-42 for instance, and I think wackyracer had one. He's got the intermediate plate from one anyway. Maybe it was actually a delivery van engine? They seem at least sort of semi-common and findable, like the 20B and 12A turbo. It was set up by Mazda purely for low RPM operation. It would be very good in a baja as stock, but I'm used to the great midrange and high end of our US-spec rotaries. If I wanted a limited RPM range, I'd have gone with a 1950 ACVW or some other torque monster and skipped all the cooling system issues and extra expense of the adaptor kit etc.

Again, the Cosmo manifold, in my opinion, is the key to great low end and high end without EFI. The holy grail of intake manifolds if you will. I just need to carry out some experiments first. I guess this isn't really a tail of two engines anymore.

 

Pretty impressive you took all that time to go through mixing and matching parts in search of the perfect NA combo.

But I gotta ask, why not just go forced induction? Seems like it would produce twice the power with half the meddling? Take the worst matched intake manifold you can think of and throw some positive pressure to it and you'll still end up out performing the "holy grail" of NA combos.

Still... very interesting stuff!

 

Not a fan of forced induction for a baja. If it was a sand rail, that'd be great since you need acceleration power with paddles, but it's just too much heat production for a baja, where the cooling system is lackluster at best and totally inadequate at worst. I'd like to mount the radiator in the back for simplicity. It can be made to work back there.

 

I got more work done on the GMB.

Here you can see the short stud and the oversized nut.

 

Lynn E. Hanover said something of great importance over on nopistons recently.

He said, and I quote, "Picture is of a bridgeported 12A iron but shows what you will be looking for. Closing line is the same as late closing the intake valve on a piston engine. The later the closing point the poorer the bottom end."

I've kept my eyes peeled for any and all information on how the intake port closing timing affects low RPM as that is the only major difference between '74 spec and '76-'85; my two choices. I've tossed around the idea of '76 spec primaries and '74 spec secondaries, which seems plausible since the bus engine had NO/GSL-SE spec (small) primarys and '76 spec secondaries, with channels in the manifold. The channels technically allow the vacuum signal to be shared, and whatever port closing differences exist, kind of dithers between them (that's just my guess, and it could be back-asswords). Perhaps the end result was a vacuum signal resembling that from an engine with ports closing somewhere between NO and '76 spec? No such engine exists, but Mazda did have a use for one. The solution was to have small primaries and normal secondaries. Again, it's just speculation.

Now let's take what we know from the bus engine and apply it to the proposed baja engine. Let's plan to use a channeled manifold and primary ports that are smaller than secondary ports. Only testing will tell whether the Cosmo reversed manifold will perform the way I want it to, and whether cutting channels in it is a good or bad idea.

Gotta do some testing on two cars here. The Cosmo has stock '76 ports, a thermal reactor and auto trans. Current plan is to swap in a header kit and test run it. Or swap in a 5 speed at the same time (it'll be up in the air, why not). Eventually the reversed runner manifold will find its way onto the engine. Then I can do the testing.

Another test vehicle is the MG Midget project with the freshly rebuilt 4 port 13B conisting of '74 ported Y side plates and stock '74 rotor housings (small exhaust ports, like the Cosmo). It already has a channeled Cosmo manifold (it was the only type that would fit) and the reason why my Cosmo needed another one for testing. Once the tests are complete, if the Cosmo manifold does what I want, it'll go onto the baja and the Cosmo will get either the bus manifold or a '78 RX-4 manifold (both are similar internally). It did well enough on the RX-4, and the current '74 manifold on the Cosmo can come off and go onto some other project.

If the Cosmo manifold sucks, I'll probably recreate the bus engine and just know that my limit is 4500. Actually, with some minor porting, and since the carb has already been rejetted and air bleeded (lol) for higher RPM, I bet 5000 would not be a problem.

Blah, I'll think positive thoughts that the Cosmo manifold will do what I want it to do, and perform well up to and beyond 6000 on the Cosmo. Then it wouldn't be a problem to have '76 primaries and '74 secondaries. The bus engine had staggered ports after all.

Feel free to comment if you want. I know a lot of this isn't your average every day 1st gen section material.

 

Here's an odd thought for you that I discovered recently when tuning the 1/2bp. The DCDs carbs have 4 separate runners to each individual throttle plate. Since each runner is separate, and stops at the butterfly, as long as the secondaries are aren't open, it now drives like a streetport. On the way back from the Iowa rotary meet, after a little tuning and jetting changes thanks to Ians's help and the wideband, I was cruising through a small town at 1500 rpm, which is lower than I had the idle set. When I first had the Mikuni, then the untuned DCDs on, I could not cruise below 3k without surging, even had to drive in 4th gear on the highway many times. Now it runs nearly as smooth as a stock engine with no bucking/surging and smooth acceleration. I do seem to have a bit of ignition breakup, it's still stock.

I had changed to another set of jets before I left, had bad surging at certain rpms, needed to downshift on many of the hills because it wouldn't maintain speed. Mileage was a little over 22 mpg. On the way back, no downshifting, cruising at almost any rpm, except a very narrow band around 3k, and 20.5 mpg. I can now even feel the secondaries open up.

In a nutshell, if one wants to run a 1/2bp on the street, find an intake that has 4 individual runners. The RE intake is the only one I've seen that does have 4 separate runners. The wrap around intakes merge about 2" inside the intake, mucho overlap. All the other merge into a single runner for each rotor and that's where the overlap comes into play and kills the low end torque.

One thing I probably should do is change to a smaller primary venturi. I'm running RE's 13B/SP set up, which is fine for the bridged secondaries, but with the separated runners, I should be running the 12A SP venturi. Easy to do on the Mikuni and other Webers, I think this one takes a special Weber tool to remove them since they are not fixed in place with side screws as the others models are.

As an after thought, I realize the stock intakes are separate runners, this is good for maintaining low rpm torque and eliminating overlap when running on primaries only, where most driving is done. I do think Mazda should have made most of their intakes like the reversed runner one that you have. That may be a good carb to run a 1/2bp, with the eyebrows cut in the primary instead of the usual secondaries.

Just a little extra food for thought in the torque wars, lol.

 

Nope I never did, but it's still on my DD. I took it off once to drill and tap the houseing for the stock temp sender so I could let the car warm up while I tie my shoes in the morning(I get to work just on time every day).
Yes I'm using an e-fan.
I like your idea for the location of the alt bracket, but worry about the additional stress on that section of the water pump without the bolt on the left side.
The JB weld I put in that area shrank down. It would be nice to fill that area with aluminum and grind/machine the top and bottom smooth and drill a hole through it for a bolt.

The impeller is aluminum? the two GMB pumps I have (one from junk, other bought new) have the crappy stamped steel ones.

Great thread.

 

So trochoid, you're saying having 4 seperate runners has been best for low end torque and part throttle smoothness?

Stevan, I'm concerned about the alt bracket too. I won't know how well it works until I try it.

The GMB housing is aluminum. The impeller is stamped steel.

 

Exactly Jeff. The seperate runners eliminate the usual overlap when the secondaries are heavily ported. The 1/2 bp is almost tuned to the point that it runs/drives like an SP when it feed by only the primaries. By the time the secondaries open up the rpms are high enough, i.e. full throttle acceleration, the bridge simply makes more power than a full SP would.

If I were to install smaller primary venturi, it would be even more streetable.

 

Ah, so I guess it's come down to channels vs no channels in the manifold. If I was to recreate the bus engine, the answer is easy - I'd simply use the bus manifold and carb; bus manifold came stock with channels. Since I'd like to bend the rules and use the Cosmo manifold instead, I'm still debating whether to use channels or not.

For some reason I get the impression I'm being too greedy for wanting '74 spec secondaries. Trochoid, I know if left the manifold stock, it should still behave like a stock ported engine on primaries. I'm only worried that the carb may not idle right on a seperate runner manifold, since the idle circuit was sized presumeably for channels.

Also when I was breaking in the RX-4 wagon's engine before the supercharger, it had the bus carb on it and a channeled manifold, and the low end torque was surprisingly good. Mild street ports and T2 ported exhaust ports. All new seals and GSL-SE rotating assembly. It had better low end than my REPU, which had a seperate runner manifold.

Every engine that's received a channeled manfiold, whether stock or cut by me, has always had a good, strong idle, capable of reasching down past 500RPM. Of course they've also all been normal runner manifolds. I have yet to test my channeled Cosmo manifold in the MG.

I was a little concerned that the long runner length of the Cosmo manifold may hinder throttle response, especially in the MG, so upon examining the runners today, I noticed the runner inner diameter wasn't actually all that large. That's a very good thing. Also the brake booster fitting is on the primary runner. What's the significance? I'm not sure. Would cutting channels mess up power brakes? Don't know. Does it matter in the MG? Nope. Baja? Nope.

Part of my continued struggle with my decision has to do with something rotarygod said about channels and low end torque. He said the channels can actually hurt the low end while allowing more power up high. But all this was based on theory and an experience he had helping a friend tune a Holley on a sand rail (can't remember the engine, but probably a rotary, hehe). It was a seperate runner manifold that got some channels which helped the problem it was having, but introduced another set of problems, if I remember correctly. I don't recall which way the channels went (either front to rear or both primaries and both secondaries together).

 

Well, all I have to go on is my personal experience with channels and so far it's been nothing but positive. Take the projects I've done as proof that channels work.

The GLC huge streetport nitrided R5 13B. Mild streetport primaries and extend ported secondaries. Huge exhaust ports that got the approval of rotarygod. '74 manifold with large diameter runners and channels. Rejetted Hitachi carb. Free flow exhaust. DLIDFIS. It fires right up when you crank it and idles surprisingly well for a streetport. Even with different closing lines of the primaries and secondaries, with channels, the idle quality is excellent. It idles better than the same basic engine in the white REPU with mild ported pri and sec, and upside down D ports. The REPU has the same manifold but it does not have a channels. I swapped in a channeled carb spacer to the REPU and the idle quality increased noticeably! to test whether it was a fluke, we swapped carbs and the results were actually improved on both vehicles. Heh, interesting. The GLC still out idles the REPU though, and the low, mid and high RPM power and driveability are better. Not bad for a carb that's honestly too small for the porting level in that engine. It needs an Edelbrock.

The next engine with a channeled manifold was the RX-4 engine, in which it came stock. Man, even with the old clunker I built for it back in '05 out of spare parts, it could still idle down to 400 or was it 350 RPM? The tranny produced a whine that was so high pitched it was inaudible at normal idle speeds, but get it down below 600 and it starts to show up. 500 and you really start to hear it. 400 or below and anyone could hear it. Only possible with a channeled manifold. Specs of the clunker engine were mild street ports and upsidedown D exhaust ports; same exact shape and size as in the white REPU. All the seals were used and upon teardown showed signs of leakage. It never had very much power. Only some mid and high range power after the RB exhaust was installed (RB header, presilencer, powerpulse muffler for '83-'85 modded to fit the wagon). At the time I assumed the high RPM power and kack of low RPM power was a result of using a channeled manifold. Then I rebuilt the engine with all new seals, GSL-SE rotor housings ported to T2 spec (bigger than upsidedown D), GSL-SE rotating assembly but the side plates remained the same with mild street ports. The power difference was truely amazing. The difference just in low end torque was phenominal! We couldn't rev it very high due to the break in process, but this engine really wanted to move! Then we dropped a 5" Camden SC right around the 800 mile mark (we needed 1500 miles due to new bearings) so we kept the RPM low for a while. It's still alive today and reving to 8000.

The next engine was a 12A with '74 spec ports and a channeled RX-3 manifold. It could idle down to 750 within 20 minutes of first run time (only because it took that long before it dawned on me to try lowering the idle). It could rev easily too. It's still driving strong in my '83.

By the way, the GLC engine could idle down to 600 within 2 minutes of run time, only because I didn't think to try it sooner. Fresh rebuild. Only possible with channels.

The next engine was in the black '84 with automatic, about 10k miles on rebuild, 12A turbo rotor housings, header exhaust, and FB manifold which I removed the shutter valve and cut a channel. It could idle so low the tach would read at the lowest tick (whatever that is). I set it to about 1000 though since it's an auto and when in D or R, it pulls the RPM down to 750 which is appearantly correct.

The Cosmo and white REPU so far don't have channels. Only the white REPU has the RX-4's channeled spacer. The Cosmo has the bus' primary-only channeled spacer. Isn't that interesting? The bus not only had channels in the manifold connecting primary to secondary, the carb spacer also had a channel connecting both primaries together. The carb appearantly doesn't care because I tried it on the rebuilt RX-4 engine with the usual pri to sec channels and a seperate runner spacer and the torque was excellent. Or was it with the RX-4's channeled spacer? I don't remember now.

How much longer can I beat this subject? I'm sure you guys really love reading this stuff.

Lastly the Cosmo manifold for the MG has channels already cut. The seperation between runners is thinner so the resulting channels are smaller. It shouldn't be a problem. All the MG needs is an exhaust sytem. Then I can test fire the rebuilt engine. If the MG works out as intended, I'll dive right into cutting channels in the other Cosmo manifold. Wouldn't you with so much overwhelming evidence?

 

So far the baja engine will consist of:

the bus carb; rejetted for bigger ports and high RPM
bus channeled primary carb spacer, for better idle quality I assume
Cosmo manifold (mostly likely channeled)
bus rotor housings with stock spec (GSL-SE size) ports
Y side plates with stock port timing; only smoothed
'74-'85 (heavy) 13B rotating assembly
new springs and new or very good used seals
.001 or less clearanced new side seals for superb idle quality
GMB waterpump impeller on an aluminum T2 waterpump housing for weight reduction
gutted thermostat with bypass blocked
beehive oil heater (lol)
tall style radiator and 16" electric fan
DLIDFIS with weatherproof mods
KEP rotary to VW adaptor
bug tranny for now, but 3 or 5 rib eventually
short header made from RB flange and 1/8" wall pipe
Rotary Engineering glasspacks with still some life in them

Here are the bus rotor housings.


Check out the exhaust port. That's stock!

 

Mazda has some information about the bus this engine came out of on their website that might be of interest.

http://www.mazdausa.com/MusaWeb/disp...azdaMuseumMain

Click on "Collection" at the top, and then "Classics" in the bottom-left corner. The Parkway Rotary 26 is on the far right of the middle row.

Apparently these motors could pull around a 3+ ton bus and 26 passengers. (Though they probably weren't very happy about it!)

 

Hey thanks for posting that! It looks like the bus engine was capable of 146 to 160 HP! And all at LOW RPM. Not bad.

To think that they could pack that much power into 4500 RPM or less is very cool. I think I know what engine I'm going to build for Tom's baja. A clone of the Parkway Rotary 26's engine.

 

Very informative Jeff!
thanks for the effort,and keep the research going!

karis