I forgot to mention the carb spacer on the J-spec manifold had a channel connecting both primaries together. There was communication of both primary runners above and primary to secondary runners below. Either of the secondary ports could pull on either of the primary carb barrels. What were they thinking? Probably designing a carbed 13B for maximum low end some 10 years before the RE-EGI GSL-SE engine came out. The evidence for this is all around. Consider the following:



B6ET three ground points plugs (really hot range)

190°F thermostat

full time coolant bypass for hotter operation (weird thing inside waterpump just below thermostat)

Auto rear counterweight

super small primary ports no larger than an S4 NA (the J-spec intermediate pictured above)

secondary ports the same size as stock 12A

manifold with runners connected at the top (like a '79 RX-7)

carb spacer with primary runners connected (not sure what other models had this)

Hitachi carb with tiny fuel jets and huge air jets for low RPM operation (90-140 fuel, 90-160 air)

exhaust ports with late opening and semi-late closing - same as GSL-SE (better than FC port timing for low RPM)



This engine was not set up for a sports car. It had an automatic tranny and most likely came out of a luxury car meant to be lazed about town. It has several features I like as a potential baja bug engine. I'll keep the auto counterwight for the VW tranaxle adaptor kit, switch to a 12A intermediate plate for a port mismatch and reduction in reversion while keeping the channeled manifold and carb spacer. Upping the fuel jets and reducing the air bleeds for better mid to high RPM operation should minimally affect low RPM power. No porting. All 4 ports will have stock 12A timing. Only the casting flash and harsh edges will be removed and a smooth rough textured surface will be left behind (should help with fuel-wetting issues during quick stabs of the throttle). Exhaust ports will remain stock. 180° thermostat, S4 T2 waterpump for slight reduction in weight and maybe better flow.



No one has ever done a rotary baja. Or at least a successful one as far as I know. I want to be the first, and I think this engine combination will work. It will come close to the GSL-SE in power and torque but will surpass it in reliability (unless the carb doesn't like extreme angles... no problems so far on other vehicles). It ought to be more reliable than a VW engine and should have twice the power with only 40 pounds more weight.

 

I played around with alot of spacers I modded. And with the Holley, a transfer slot across the primaries was the only way to get the vac secondaries to operate properly. Mostly due to the 80* abdc closing on the S4 N/A. Maybe thats the reasoning behind the slot, trying to keep the benefits of an IR manifold, yet usability of a plenum design.

 

bob, the 1st gen that races with dave, uses the weber, 13b, peak power is 9600. interestingly they added length to the exhaust, mostly to let them add more mufflers (it goes around the fuel cell), 19feet long, 16 foot car, the engine actually made more power...

 

is there a simple explanation of why you would have thought this? all of the stock manifolds (Rx-7s) that i've seen thus far (some flowbenched, some not) have put channels where there originally were none. just curious.

 

The part where RONIN saysHolds true for my J-spec manifold. The port closing difference of about 10° between primaries at 30° and secondaries at 40°. That is, the primaries close at NA FC spec and the secondaries close at 12A spec. That is one potential reason for Mazda's cutting of channels in the top of the manifold.



Another example is a really large streetport I made in a 4 port 13B. The primaries close at 55° and the secondaries close at 60° (Racing Beat spec). It's a fresh rebuild with only about an hour of run time in a non-driveable REPU. I used a stock 13B carb with a channeled manifold which I cut myself. It fired up and ran smooth down to 600 RPM within 2 minutes of run time (that's better than my other REPU with a fully broken in engine with smaller ports). The exhaust consisted of a cast iron manifold with a short downpipe and a mostly dead Rotary Engineering glasspack (really loud). I copied T2 exhaust ports and closed them 10mm later (whatever that is in degrees). I may have described this engine earlier in this thread, so sorry if it's repeat.



That large streetported engine with a channeled manifold and different size ports had a strong idle. I drove it forward about 5 feet in my driveway and it handled that just fine. Maybe the manifold channels are the secret?

 

If Jeff20b didnt answer your question already, I wasnt absolutely sure of the benefits of merging the two ports (prim, sec) without speculating. Which is why I was picking his brain to see what he knew about it.

I wasnt bieng facetious, if thats what you took from it...



Or maybe I have you question totally wrong. Care to elaborate what you mean, or want to know?

 

no, didn't think you were being facetious at all. i'm not very well-versed in this stuff and i was basically asking the question to see if there was something concrete out there that opposed what i'd seen thus far in practice, that's all. i was also curious as to why you said you could see merging the secondaries as being better. i'd never seen that in practice before.



my next question would have been asking you to direct me to the source. i'm interested in getting to the basic principles of airflow through these engines without jumping too deep, too fast.

 

I would tend agree with you in theory, but...........

Well I ran some tests using different spacers on an RX-3 mainfold (Indpendent runners). Open plenum spacer made the best power up top (5k up) but lost considerable amount below that. The IR spacer was great till about 5K then felt weak (probably flow limited as the primary ports on a 12A stink for flow). The dual plenum was the best compromise (at least for street use). Very small loss at low to mid range, and decent top end (probably from the good flowing secondaries "borrowing some of the primary intake charge through the channels, at least thats my theory after flowbenching stock irons). I agree that the channels should hurt VE at low rpms (probably responsible for the small power loss), but it seems the best compromise for a street engine. This was all on a 1st gen, 12A, stock port, RB SP exhaust, Sterlingcarb. Stopwatch, g-tech and butt dyno.



Bottom line to me is that whatever works best for the application is what should be used. Theory is all well and good but we're limited by what Mazda has already engineered so we have to work around that, so compromise is the best we can reasonably achieve, unless you want to get into custom fabbed parts, which most of us dont have the resources to do.

 

Very Excellent Carl! Thankyou for taking the time to test three different carb spacers and report about the differences. Your findings mirror mine.

 

Amen.