U20 EFI Stroker build (Scarlett's new heart)

[Daryl Smith]

"...from what I've seen lot of the discussion around header design is theoretical and speculative and not backed by hard data comparing apples to apples."
Not theory or speculative. Proven many times over by racers and OEM's. IF your primary source of comparative data is enthusiast forums, then, there, yes it is theoretical and speculative. I have done a ton of research and thoroughly believe (I guess that makes it theoretical? ) a 4-2-1 or 6-2-1, properly sized and properly built, will give you the best all around performance under ~8000 rpm.

"While I agree that a 4-2-1 configuration is generally better for mid-range performance in a street car, It's also my opinion that header design, exhaust system, compression ratio, and camshaft tuning need to be considered holistically. From my personal experience with 4-1 and 6-1 setups, I’ve observed how these components can work together to deliver comparable top-end power on 4-2-1 vs 4-1 and surprisingly similar mid-range performance in a 4-1vs 4-2-1"
Absolutely agree with most of your points re: consideration of the whole system. However, still think the dividing line between 4-2-1 vs 4-1, for performance, is the operating rpm above all else.

All that being said...as I was mentioning to Greg a few days ago....on a well though out and built exhaust system of both styles, the actual difference between the two may not be much overall. Tho, honestly, I don't think much of the currently available header. It works, it is cost effective, and it is much better than the stock manifold. I know Dean has put some time and money into some improvements (which I have not seen), which is commendable, but it could still be improved upon.....but, would it be cost effective? (really, how old is this header design, and what were it's design parameters?)
Making those improvements and still having it fit, without other modifications to the car is a real challenge.
Sound and appearance are pretty much irrelevant to me, performance and weight reduction are my primary objectives.

A cost effective way to give yourself a long tube header would be to cut the collector off the current header, add some pipe and a new (perhaps 'better'?) collector.......https://www.coneeng.com/collectors.html

I'm gonna quit messing up your thread and enjoy the build......Thanks for the thoughful discussion!

[Alvin]

My focus is to match or exceed those numbers while enhancing midrange torque and usability in the 2500–5500 RPM range, which is where I spend most of my driving time.

Sounds like you need to build an R16 stroker!



Just kidding, I'm sure the U20 will easily surpass the R16 numbers, while revving freely past 7500rpm and enjoying the benefits of ITB EFI.
This is gonna be fun to watch!

[Gregs672000]

Well, not necessarily! Those are truly excellent numbers... nice torque... nearly 150ftlbs as early as 4000rpm! That would eat my lunch I think. What's the displacement of this engine? 2.2L? Robello knows these engines well. How did zcar pull another 8hp out of it with the same torque (its not rpm)? I wonder what causes the drop in the torque band from 4200-4800rpm (reversion? Ignition timing?). I think we had seen the crank dyno numbers but these wheel numbers are more impressive to me.

[Pjackb]

Just kidding, I'm sure the U20 will easily surpass the R16 numbers,

Honestly Alvin I'm not so sure it will, Erik and Peter's engine made around 190hp with 180lb/ft peaking at around 6000 rpm
I spoke at length to Dave about these builds and options to build mine and his comment was that when he build these street 2.1l regardless if based on the U20,R16 or H20 he always ends up with about the same numbers, he says the U20 outshines on high RPM big cam engine
I expect mine to have a little less torque and similar hp at higher rpm due to cam selection.

while revving freely past 7500rpm and enjoying the benefits of ITB EFI.
This is gonna be fun to watch!

Definitely , I'm still not 100% decided what cam will end up in it but all of the selections have powerbands in the 3000-7000 + rpm

Well, not necessarily! Those are truly excellent numbers... nice torque... nearly 150ftlbs as early as 4000rpm! That would eat my lunch I think. What's the displacement of this engine? 2.2L?

Agreed , strong numbers
2.1L (2116cc) is the size if I remember correctly (mine will be 2.2L 2165cc)

[Pjackb]

Cylinder Head

As everyone knows for these Datsun engine power lies in the cylinder head, and getting it right can make or break the engine.

Over the years, I’ve consulted with several experts, including Lou Mondello, Jim Tyler, Dave Rebello, and David Levinson, all of whom have extensive experience building U20 heads. Collectively, they’ve built hundreds of these heads, and their insights have been invaluable in helping me navigate through the noise and making intelligent selection.

It's common knowledge that the U20 head flows well and has nice well made large ports, but what does that really mean? My personal research revealed that a stock U20 flows approximately 185cfm at .400” lift, 200cfm at .450” lift, and peaks at 205 cfm at .500” lift. For comparison, most L-series heads peak at under 170cfm. This indicates that the U20 head has strong flow from .400 lift and even in its stock form has the potential to support up to approximately 210hp. Given this flow capability I decided against any porting Instead focussing on cleaning up the ports to remove casting marks and deposits ensuring smooth airflow without altering the original geometry.

For the valve and spring setup, I selected DRP high-performance oversized stainless steel valves. While I chose oversized valves before fully understanding their implications, I’m still glad I did. They will complement the engine’s increased charge capacity and compression ratio, promising excellent performance I believe . That said, cam selection will be critical to maximizing their potential. In a previous L6 build, we used these same performance valves and achieved a 7% improvement in peak flow on the bench from the valves alone and more importantly, the flow curve between .350” and .450” lift increased significantly, and I’m optimistic about seeing similar gains with the U20.






The valves were paired with DRP performance springs and bronze guides, providing flexibility for cam selection and ensuring durability. Of course, new seats were installed to complete the setup.




The head I started with had some valve damage caused by pre-ignition, requiring us to remove about 1mm from the surface and cleaning up the combustion chambers, my builder did a great job of this I think.




After assembling the head , we measured the combustion chamber volume with a stock gasket and got 48cc.




This equated to a compression ratio of approximately 11.35:1, which was higher than my target.
To address this I consulted with Jim at DRP and selected one of his .088” head gasket he looked trough his stock I found one that measured slightly less at .080” uncompressed which should worked perfectly.

This adjustment will bring the compression ratio to a more manageable 10.4:1 to 10.6:1—right on target.

Right now most of the block and head are assembled and missing only a few parts to start assembly in the next few weeks ,

Next up cam selection

[Pjackb]

IF your primary source of comparative data is enthusiast forums,

From my research and experience there’s more than one way to achieve great results! My insights are primarily based on my own dyno testing .

So far, I’ve only completed one session with the U20, but I’ve done around 6-7 sessions with my L28, with the shortest session lasting 3 hours. During those sessions, we tested various configurations, including different cams and intake setups. We managed to achieve above-average horsepower for the configuration, along with exceptionally flat torque curves from 2500–6000 RPM using 6-1 long-tube headers.

The 2.2L will also go through at least 2–3 sessions on the same dyno I use for all my cars

[Gregs672000]

Thanks for sharing, enjoying the ride along! Previously I was at 11.7 to 1 compression... too much for the street and I just couldn't control A/F ratios well enough. On the last build I had the (Arias) pistons machined a bit more and dropped it to about 11 by our estimates (didn't cc etc). Cold cranking compression is 210-215lbs so it's still pretty high. Yours sounds like it will work well!

[Pjackb]

Thanks for sharing, enjoying the ride along! Previously I was at 11.7 to 1 compression... too much for the street and I just couldn't control A/F ratios well enough. On the last build I had the (Arias) pistons machined a bit more and dropped it to about 11 by our estimates (didn't cc etc). Cold cranking compression is 210-215lbs so it's still pretty high. Yours sounds like it will work well!

I think with EFI and the right cam, I could make the 11.x compression ratio work fairly well. However, as a renowned builder once told me, you can make safer power with slightly less compression while running full timing, compared to being on the edge of detonation and having to pull timing.

So, a 10.5:1 compression ratio will give me more margin.

And speaking of cams, read up

[Pjackb]

Cams

Or the heart of the engine.

Right off the bat, spoiler alert—I’m not 100% decided yet on which cam I will be using but have a strong leaning.

If you’ve read everything I wrote so far, it should be clear by now that I do very extensive research on everything before making a final decision, and cams are no different. I’ve obviously talked to the same builders, but I’ve also had numerous and very informative discussions with Ron at Isky, Geoff at Colt, Kevin at Schneider, and Jim at Racer Brown. I’ve also read numerous papers shared by these experts, including Racer Brown’s excellent article on Datsun Camshaft and Valve Timing: https://www.datsport.com/blogs/news/dat ... acer-brown.

The next thing I did was look at an unordinary high number of dyno sheets and results. A few patterns and insights emerged—I'll spare you the boring details—but the main takeaway for me is that U20s seem to like lift and compression and respond very well to it.

If you look at the cars that made higher power than usual, like Greg's Will's or Brian's, they all have in common either higher than 10.5:1 compression ratios, cams with higher lift and duration or both  along with well-tuned induction setups and exhaust systems. This combination allows the engines to take full advantage of the increased airflow and deliver consistent, high-performance output.

With these insights in mind, I approached the selection of potential cams for my build with care and precision. After thorough research and evaluation of various options, here is the refined shortlist of what I’m considering:

Factory B Cam:
The B is a really good street cam. The specs are .327 cam lift, .459 at the valves, and a measured 225° duration @ .050, with a 109° LSA. It’s so good, in fact, that when Shadbolt Cam (now Colt) wanted to build a strong street cam for the L-series, they looked at it and copied it onto an L-series core, creating the Shadbolt M445/Colt C.542.S. This is an exceptional cam with great midrange torque and power to 7200-7500 RPM. This is what a few friends and I run in our Zs and 510s

Lou shared the following about his usual big-bore builds (89x83): With a B cam, 44 PHH Mikunis, 37 chokes, a decent set of extractors, and 10.25–10.5:1 compression, all engines with this spec have made between 171–175 hp with 155–158 lb-ft at the crank.



Factory C Cam:
The specs are around .340 cam lift, .476 at the valves, and a measured 240° duration @ .050, with a 109° LSA. I like the C cam a lot for my build. The lift is more in line with what I want for my build and compression ratio, and it’s a direct fit without modification. The only dyno I could find is Brian Kippen’s U20, and it did very well with 142hp at the wheels on a slightly overbored U20 with 9.5:1 compression. It was a little low on torque below 3000 RPM, but with my higher compression and stroke, this should be well offset. My only concern is that I measured my regrind a little lower on cam lift (.331”). It won’t make a huge difference in power, but I want to have every inch.



Schneider Custom Grind:
Based on the Schneider 284-92F but with custom specs: In/Ex of .340/.345 cam lift, .476/.483 at the valves, and a measured 225°/235° duration @ .050, with a 109° LSA.

This is my take on the Racer Brown findings and discussions with the cam grinders. It seems the Datsun OHC heads really like asymmetrical cams with longer exhaust durations to help with scavenging and exiting gases. Dual-pattern asymmetrical cams, which feature slightly different intake and exhaust profiles  appear to enhance overall engine efficiency and power delivery. This feels to me like a good improvement on the B cam, with more lift and slightly longer duration on the exhaust. I think it will make a little less peak power than the C but perform better everywhere else and be a better overall cam for a street car.

Honorable Mention:
Jim Tyler at DRP has a Delong 360-10 on hand. That’s the cam I really wanted. Unfortunately, it has a little too much lift at around .500” for my build since my head is cut 1mm (.040”) and it would require the pistons to have fly cuts and It’s a little too late for that.

This is basically the Bonk cam that’s really popular on the L-series and makes tons of power from 2500–7000 RPM. If someone is building a hot U20 right now, I would highly recommend talking to Jim about it.

I will have all three cams on hand and most likely will start with the Schneider if it arrives in time, given their long turnaround (5weeks) . Otherwise, I’ll begin testing with the C cam.

[Gregs672000]

Great stuff, appreciate all the info and specs, they can be hard to come by or confirm. I had always thought that the 280 "B" cam was around 248 degrees @ .050, as my Isky Z-197 has 278 advertised and 246 @ .050, 110 LSA. Lift is .531 total, so not workable in your case and may actually be too much in general, but my motor did not like the Racer Brown grind I had with similar duration and less lift, and you can't just "grind off" a bunch of lift at the cam grinders, so... ya run it! Pretty close to coil bind too, so that's a consideration. The Isky Z-196 has a lot more duration 306/264 @.050, and more lift (.530 but that's with a 3 to 1 rocker, not our 4 to 1, and they don't list the cam lift). Spyder ran that cam in his motor with 50mm Mikuni and some shorter rear end gears so he didn't miss the low end torque loss as much. Too much cam for me. I no longer find the z-197 on Isky's site, but I have the cam card from when they ground it. They were very nice to talk to!

My goal was to have a duration similar to the "B" for the power band, so I thought I was pretty close with the 246/.050. For what it's worth, it comes on really well at 3200rpms and is not a dog down low at all. A very streetable cam even with the light flywheel, no worries starting on hills etc. I have never considered the slight difference in LSA on the Isky grind and how that will change the effects of duration and lift and where the engine makes power. From what I've just read, "a tighter LSA (109) prioritizes high-RPM power while a wider LSA (110) prioritizes mid-range torque and drivability." I really look forward to your testing!

[Pjackb]

Great stuff, appreciate all the info and specs, they can be hard to come by or confirm. I had always thought that the 280 "B" cam was around 248 degrees @ .050, as my Isky Z-197 has 278 advertised and 246 @ .050, 110 LSA. Lift is .531 total, so not workable in your case and may actually be too much in general, but my motor did not like the Racer Brown grind with similar duration and less lift, and you can't just "grind off" a bunch of lift at the cam grinders, so... ya run it! Pretty close to coil bind too, so that's a consideration. The Isky Z-196 has a lot more duration 306/264 @.050, and more lift (.530 but that's with a 3 to 1 rocker, not our 4 to 1, and they don't list the cam lift). Spyder ran that cam in his motor with 50mm Mikuni and some shorter rear end gears so he didn't miss the low end torque loss as much. Too much cam for me.

My goal was to have a duration similar to the "B" for the power band, so I thought I was pretty close with the 246/.050. For what it's worth, it comes on really well at 3200rpms and is not a dog down low at all. A very streetable cam even with the light flywheel, no worries starting on hills etc. I really look forward to your testing!

The thing with cams is that unless you’ve measured them yourself, it’s hard to directly compare specs from different manufacturers. That’s because not everyone measures duration the same way, so you need to know how the numbers were calculated to make a fair comparison.

For instance, companies like Isky and Schneider give their @.050” duration values measured at the lifter. These don’t account for lash, so the numbers aren’t exactly what the valves are seeing. On the other hand, manufacturers like Racer Brown and Colt typically measure duration at the valve, which includes lash and is closer to what your engine will actually experience. That’s why it’s important to factor in these differences or, better yet, measure things yourself to get an accurate picture for your setup.

Take my custom Schneider cam as an example. On the cam card, it lists durations of 236/244, but those are at the lifter. Once I factor in my setup’s lash, the actual duration at the valve drops to around 224/235. That’s a big difference, and it’s why understanding how the numbers are measured is so important—otherwise, it’s easy to overestimate or misunderstand what you’re getting.

This also explains why you might have thought the B cam had way more duration. Incidentally 248 is what the C cam would measure at the lifters, but it would be less once you account for lash. It’s all about how and where the measurements are taken, and without that context, it’s easy to misjudge.

Now, moving on to lift: The stock cylinder head’s flow peaks around 0.500” lift, and after that, it starts to drop off. That means going for higher lift won’t really give you more flow, so it’s not always worth the extra stress on the valvetrain. A cam with a lift closer to 0.500”, combined with a quick intake ramp to get the valve open fast and long duration to keep it open longer, would likely work better for this setup. It’s about working with the head’s natural flow characteristics and making sure the cam complements them, rather than just chasing big numbers.

In the end, matching duration and lift to your specific engine and goals is what really makes the difference. It’s not about having the “biggest” cam, but about making the whole setup work together for the best performance.

[spyder]

The Isky Z-196 has a lot more duration 306/264 @.050, and more lift (.530 but that's with a 3 to 1 rocker, not our 4 to 1, and they don't list the cam lift). Spyder ran that cam in his motor with 50mm Mikuni and some shorter rear end gears so he didn't miss the low end torque loss as much.

My ears were itching. Buy the way I have two of those cams and thicker lash pads for them. The one that was in my car and a spare unran one, just saying!

[Pjackb]

The Isky Z-196 has a lot more duration 306/264 @.050, and more lift (.530 but that's with a 3 to 1 rocker, not our 4 to 1, and they don't list the cam lift).

The Isky Z-196 has a lot more duration 306/264 @.050, and more lift (.530 but that's with a 3 to 1 rocker, not our 4 to 1, and they don't list the cam lift). Spyder ran that cam in his motor with 50mm Mikuni and some shorter rear end gears so he didn't miss the low end torque loss as much.

My ears were itching. Buy the way I have two of those cams and thicker lash pads for them. The one that was in my car and a spare unran one, just saying!

The Isky Z-196 is actually their version of the Delong 360-08 baby brother to the Delong 360-09 , it has 353 lift at the cam but more duration
Isky list it as a .530 but they're using 1.5 ratio instead of the actual 1.48 of a Lseries rocker so i should be .525
That grind in a Roadster gives .495 lift

but it does have a ton of duration so would work well on you guys 11.5+ CR engines

[Gregs672000]

Spyder, do you have the cam card for the Z-196?

On the Isky website it only provides total lift and doesn't give rocker ratio. On my Z-197 card it specifies 1.3 rocker and has a total lift of .499, but when using the Roadster 1.4 it's. 537. Mine was measured at Delta Cam after it was polished due to a rocker issue and it was .531 total. For all this time I assumed an L-series was 1.3 based on the assumption that the 196 was for a L-series so maybe the 197 was too... guess wrong! As I think about it now, I do recall reading somewhere that this grind was also for a Lotus, but which one I've not heard.

Interesting, so it sounds like the 197 is a little different idea... a bit more lobe separation, not as much duration as a 196 or a "C" cam (though we seem to have some questions regarding these specs being apple to apple... any thought on what my duration might be? Peak torque has been at 4800, suggesting "B" cam territory) and high (highest? Too high?) lift. Well, we're covering the bases eh?

[david premo]

So the Nissan L series rocker has a 1.45 ratio and the U20 has a 1.4 ratio per the Nissan comp catalog. However it is important to remember that the lift that is advertised is at zero valve lash, so depending on how tight or loose you set the lash will determine the actual lift. Duration is also a bit tricky as it too is measured with zero lash and depending on what standard is followed you can see different durations. In the United States you measure duration starting at and ending at.050 of valve lift. In most of the rest of the world they use .040 lift or 1 mm, so depending on the standard you use you could see a different duration. R16 engines also have a ratio, so again you would need to calculate your actual lift by measuring the cam and applying the ratio. So the math is simple take the cam lift off .353 and multiply it by a 1.4 ratio and you then had valve lift of .494 at zero lash. The same lift with the L series ratio of 1.45 and you would have a .512 lift. All of this said now I will totally mess with this, the ratio is only truly valid when the rocker has been set up correctly and wipes through the center of the wipe pad.
Dave