Cam Design (Or lack thereof) for the MZR DISI

Just a little bumpity bump. Got bored last night and found that Kelford now offers cams for us. They're pretty well loved in the evo community and others. I will say, I'm not too sure about how I feel about their grinds. Kind of odd their B grind has more lift on the exhaust side. Looks like they kept overlap to a minimum as well. Fortunately they offer custom grind options!

http://www.kelfordcams.com/global/camshafts/mazda/mzr-2-0l-lf-ve-2-3l-l3-vdt-disi-turbo?engine=3

Edit: Looks like they also offer 62lb and 84lb beehive springs:

http://www.kelfordcams.com/global/kvs17-valve-spring-set.html

http://www.kelfordcams.com/global/kvs17x-valve-spring-set.html
 
Last edited:
Enki said:
Not sure I like their grinds...
Click to expand...

Not a fan either, but likely for different reasons that you. I think the durations are overall "ok", but they lifts are designed to be used with stock or stock style pistons that do not have provisions for larger valve reliefs. I would just never go through the trouble of putting cams in a stock engine, so that's useless to me and if I'm going to build, I would use pistons that could take a real cam, lift wise.
 
I don't think you need huge amounts of lift to make power though; huge duration is all well and good but not if you're just dumping air back out into the intake or opening the exhaust before the piston has fully harnessed energy from the power stroke.

I mean, look at the cam grind difference between stock:
Stock:
Stock cams.PNG
Kelford Stg 3:
index.php


Its bleeding pressure from more of the power stroke and allowing more air to leave the cylinder during the compression stroke. There's also a bunch of overlap which won't be good for reversion depending on the exhaust setup.

Hell, even Piper does that (stage 2):
piper stg 2.PNG

Overlap in the cams can be pretty bad for us, considering direct injection and all.
 
1. You can't really map out a cam profile like you have unless you know your ramp rates.

2. Having a longer blow down period is done to minimize reversion during overlap. So in this case opening up the exhaust valves earlier isn't hurting much.

3. Same thing on the compression stroke, the air is still filling the chamber and closing the intake valve later doesn't necessarily mean you're losing air charge.

There's easily 50whp on a middle of the road cam like the piper S2 for most people who have built engines.
 
I realize that and I know I'm just using a generic ramp rate for everything but all the ramp rate will do is plateau out the lift profile; in some cases a more aggressive ramp rate reduces power output on the sim. As a side note, there's enough information for the piper cams to calculate a close approximation of the ramp rate since they provide both seat to seat and .050 numbers.

Also, just about any non factory cam grind will produce more power; that's not what I'm trying to say. I'm just saying that the grinds that are offered don't appear to be very good in terms of valve event timing for our displacement/headflow/rpm numbers.

It will be very amusing to me if I put down really good numbers at low boost with less intake valve lift than stock.
 
Actually I wanted a little more DCR too, which is part of the reason why I went the route I did. I spent quite a bit of time figuring out what would be good and what wouldn't in the engine sim and came away with a compromise that I'm pretty satisfied with. If it works out well for me, I'll release the grind info and part number.
 
KCi said:
Sorry, not sure if I interpreted it correctly, Kelford 247-B stated IVO 18 ATDC and EVC 0 TDC. Technically, there should be no overlap unless it was IVO 18 BTDC (which I believed currently drawn in Enki's timing diagram). Please correct if I'm wrong.
Click to expand...

Yep, you're right. The sim didn't allow a negative intake valve open number in the mode I was using to fill in data. I've corrected that and the new map/numbers are below.

Nice bump in VE:
Kelford VE.PNG

And power to boot (360 chp 320 ctq @ 15 psi):
Kelford Power.PNG

This is with an otherwise stock engine and head (without factoring in intake manifold restriction). Still room for improvement though; a quick iterative test generated a cam profile that at 15 psi pushes 456 chp and 384 ctq but it falls way outside of what I personally consider acceptable cam timings for our platforms specific fueling requirements.

If Kelford revised this grind to do nothing more than open the intake valve at TDC (all other numbers the same; this increases intake duration to 238 degrees), peak VE would remain just about the same, but the VE curve flattens out slightly and power production jumps to 395 chp @ 6500 and 350 ctq @ 5500 with HP remaining over 350 until right around 9200 rpm. Still not really a big fan of the intake valve closing when the piston is almost halfway up the cylinder (edit: for a daily) but eh, to each their own.
 
Revived from the dead.
I've been talking with Kelford on a custom grind cam set.
I pretty much came up with my own specs after doing some reading and I'm curious how it plots on the simulator you're using.

Intake Cam
224 @ .050, Total Duration 264
Lift .413, 124deg Lobe C/L
@ .050:
IVO 12 ATDC
IVC 56 ABDC

Exhaust Cam
216 @ .050, Total Duration 256
Lift .378, 116deg Lobe C/L (the exhaust lift was selected to work with +2mm exhaust valves)
@ .050:
EVO 44 BBDC
EVC 8 BTDC
 
Gonna need more information:
IVO, IVC, Centerline
EVO, EVC, Centerline
Both for .050" and seat to seat if you want ramp rates factored in.
[doublepost=1557296624][/doublepost]Here's some sim info I just did for another cam grind I already had lined up (if mine work well); starting off, the stock VE for reference (all motor):
Stock VE.PNG


Now, the VE with the grind and headwork:
VE.PNG


The expected engine power output at 15 psi (assuming perfect conditions, as always, it is a sim after all):
Power.PNG


And finally the grind itself:
Cam.PNG


It's pretty close to my current grind, but with a lot more lift.
 
Last edited:
Enki said:
Gonna need more information:
IVO 8 BTDC (0.005) / 12 ATDC (0.050)
IVC 56 ABDC (0.050) / 76 ABDC (0.005)
Centerline 124 deg, Duration 224/264
EVO 64 BBDC (0.005) / 44 BBDC (0.050)
EVC 8 BTDC (0.050) / 12 ATDC (0.005)
Centerline 116deg, Duration 216/256
Both for .050" and seat to seat if you want ramp rates factored in.[/COLOR]
Click to expand...
 
Your grind (EDIT: This is shown in seat to seat timings unlike the others; I'll post the 050 numbers if anyone wants for comparisons sake):
FSTGrind.PNG

Your VE (no boost, with same headwork as other sims)
FSTVE.PNG

Powerband (15psi, like the others):
FSTCurve.PNG
 
ewww. that looks like garbage. not sure how though. it's basically a de-tuned Piper S2, but more aggressive than CS cams.
 
Glad you mentioned Piper.
Stage 1:
Piper Stg 1.PNG

Stage 2:
Piper Stg 2.PNG

Stage 3:
Piper Stg 3.PNG

Honestly I don't like any of those grinds for our platform.
Corksports grind is a lot better (for the platform), but it doesn't look good in the sim without all the information (which I don't have) present:
Corksport.PNG

And finally my personal grind (but on the same engine/settings as above):
EW220.PNG

Keep in mind my lift is less than stock on intake and about the same as stock on exhaust.
 
i wonder if it's the 120deg lobe separation angle that making mine look so off. stock and piper are 115deg, yours is 111deg.
a wider LSA supposedly makes a flatter power band and should make more top end power on boosted engines due to reduced overlap. while my grind does have more overlap than stock, it's not terribly high. Stock overlap @ 0.005 lift is 5 degrees (practically zero), my grind is 20 deg. piper stage 2 is 38 degrees, stage 3 is 52 deg.
 
Positioning of valve events matters too. My cam has some VVT built in, for example, which raises dynamic compression and improves overall power delivery even with the lower valve lift (which I wanted for higher RPM use, less stress and harmonics).
 
ya, obviously adding in VVT would bring up the low-mid range.
my grind with full VVT
IVO 38 BTDC (0.005} / 18 BTDC (0.050)
IVC 26 ABDC (0.050) / 46 ABDC (0.005)
Centerline 106.5
 
Sorry, didn't realize it would make a difference (it shouldn't, by all rights) but having your grind set to seat to seat timing seems to have dropped power. I redid the stuff using 050 timings and it looks a lot better.

Original:
FSTGrind.PNG FSTVE.PNG
FSTCurve.PNG

And with VVT:
FSTVVTGrind.PNG FSTVVTVE.PNG
FSTVVTCurve.PNG

Some of the numbers don't line up quite right but this thing only allows for symmetrical lobes so...*shrug*
BTW that dynamic compression bump with VVT enabled is quite fierce. Lol
 
You must log in or register to reply here.
Back
Top