107 Cam Maligned?

[drdoom]

Greetings,   For both amusement and engine build planning I muck about in a rather well rated engine sim program, "Engine Analyzer v3.4". Eventually this led to something perhaps of interest, particularly to owners whose cars are powered by 907. The modeling indicates the 107 to be clearly superior to the C type cams if the baseline data I'm using are valid. While the 104 MOP cam timing has broadly been hailed as the way forward, another option is to clock 107 cams at 110 MOP. The attached screen grab compares a Federal 907 bone-stock ( "last results" curves in red/turquoise ) versus the same but with 107 cams at 110 MOP ( "current" in green/purple ). Results remain consistent with a well tuned upgraded 907 as well. Offered in support of our winter   Cheers Steve

. . . Offered in support of our winter amusement.

 

[EXCEL V8]

Very interesting Steve!  I'm planning my supercharged build using a pair of 107s - what would the "current" data look like at 104 MOP instead of 110 (I've got green dot pulleys 🙂)?  Just for more winter entertainment!

Pete

[drdoom]

Here's the 107's timed at 104 MOP, graphed as "current", versus the same OEM Federal as in prior sim.

Cheers

 

[EXCEL V8]

Thanks very much!  Looks like I need some red-dot pulleys 🙂.  I assume you've used 110 MOP for intake and exhaust on a turbo engine?  Of course this begs the question why the 107s are run at 104 MOP in factory cars - at least here in the UK?  Emissions?

Pete

[drdoom]

Basic theory holds that inlet cam MOP ( AKA centerline in 'Murica ) is central to where the torque/HP peaks will be. With the turbo driving power output on those Esprit it was wise of Lotus to go 104 MOP for the sake of picking up lower end power if nothing else. Boosted engines must always be built mindful of the effects of positive inlet side pressures, as overlap will factor in on the matter of charge retention more so than with N/A engines. The 900 series engines appear to be overly large on port sizes for a street N/A application, IMO rather well accounting for why the 907 with its lesser displacement needs to rev before output is respectable. Emissions are another matter entirely, one I'm happy to treat as secondary at this stage.

I've gathered up a pair of 107's having salvaged a 910 dry sump from a written-off early Turbo, no experience with that engine otherwise. Lotus seem to have sought improved lower range power on the 907 soon after commencing production, moving on from D profile to C early in the program, then on to the 107's when they'd been developed for the Turbo. That's my thesis, at least.

[drdoom]

Peter, I've just run through several sims of your build once again and it's quite clear that the blower makes your lump quite indifferent to the cam choices discussed. It is clear though that you'd be wise to switch the exhaust MOP over to red dot 110 wheel, this giving earlier ex valve opening in aid of clearing the increased combustion volume and limiting flow-through at TDC during overlap via earlier closing. The green dot looks fine for the inlet side.

Cheers

You will want to run the highest octane fuel that can be found, perhaps water/methanol injection in some measure as well. The low RPM volumetric efficiency numbers are in excess of 120% at very modest boost levels.

[EXCEL V8]

Thanks very much Steve - really grateful.  I remember you saying that about cam choice.  I like the idea of running the 110 MOP on the exhaust for the reasons you've mentioned.  I'll be running a chargecooler and monitoring intake the temps closely.  Very modest boost levels is probably all I'll get! 😄.  "Super unleaded" - 98/99 RON is widely available.  I believe Esso's 99 is alcohol-free as well.

I'm just about to start building the engine in the next few weeks!  Best bit 🙂.

Pete

[snowrx]

 My 910 ('85 LC but with 8:1 pistons) is now running 107's on 104/104 MOP. My EFI intake runners are 1.4" longer than the original carbs, and I'm running the Alunox exhaust into a small twin-scroll turbo. 

Do you think the 110 exhaust pully is for me? I have one on hand if a 104in/110ex curve looks better than the 104/104. Thanks for posting up your results!

[drdoom]

Hi Thomas,

I'd be happy to work up data for your engine. The Sim requires a bit of input modelling for the turbo itself, eg: Air Research TO4B, S-3 Trim. Some example comparable to what you've got if not precisely "it". Tables include Rajay, Rotomaster, Schwitzer.

[MrDangerUS]

On 15/01/2022 at 21:32, drdoom said:

Greetings,   For both amusement and engine build planning I muck about in a rather well rated engine sim program, "Engine Analyzer v3.4". Eventually this led to something perhaps of interest, particularly to owners whose cars are powered by 907. The modeling indicates the 107 to be clearly superior to the C type cams if the baseline data I'm using are valid. While the 104 MOP cam timing has broadly been hailed as the way forward, another option is to clock 107 cams at 110 MOP. The attached screen grab compares a Federal 907 bone-stock ( "last results" curves in red/turquoise ) versus the same but with 107 cams at 110 MOP ( "current" in green/purple ). Results remain consistent with a well tuned upgraded 907 as well. Offered in support of our winter   Cheers Steve

Steve,

Very exciting! 

Your findings confirm Garry Kemp's observations. He said to me that 4cyl engines with 107 cams "like" 105° MOP intake/109° MOP exhaust timing combination. My "paper and pen" modeling confirmed advantages of running "uneven" MOP-s in this range, which reduces overlap and increases lobe separation angle (LSA). BTW, according to David Vizard, our engines SLA should be larger.

Could you please run 107/107 cams combo with 106° MOP intake and two different exhaust positions: 108° and 110° ?

Thanks

John

 

 

[drdoom]

Thanks, John, we are wise to mind whatever the sagely David Vizard has offered.

[snowrx]

I'm running the Borg Warner EFR 6258-G turbocharger. Info attached. If memory serves my 45mm intake runners are about 360mm from valve to bell mouth, into an open plenum of about 2 liter. I haven't measured the primaries on the Alunox manifold, can do if you need that info.

EFR 6258-G.pdf

[MrDangerUS]

I'm running two "107" cams at:
108° MOP Exhaust
and
106° MOP intake with SLA=107°, 38° overlap and +1 intake cam advance
Car runs like a scalded cat!

Thinking about "104"/"107" cam combo if/when the head (eventually) has to come off.

[drdoom]

Getting there, Thomas, is it 10psi max boost ? Intercooled?

[drdoom]

Sim menu offers Air Research TO4B V-1 trim, using 50% efficient intercooler and alternative reduced turbine nozzle spec achieves full boost by 4500 RPM, peak HP at 280 by 6500. Sound close?

[drdoom]

Results in hand after adjusting 1 error in the set-up: peak of 275BHP on 107 inlet cam, 285 BHP on 104 inlet. Best torque curve for either with exhaust cam at 110 MOP, inlet at 104 MOP. The 104 cam a bit weaker below 3200 RPM, gaining a notable margin over the 107 soon after 3200, all the way up. Precise figures impossible owing to vagaries of the input data however it looks to be that the Sim results are quite valid for comparative purposes in terms of the inlet cams and of the timing MOP's.

Results graphed for 107 inlet cam in green/purple, 104 inlet cam in red/turquoise. 

Cheers

[snowrx]

6 hours ago, drdoom said:

Sim menu offers Air Research TO4B V-1 trim, using 50% efficient intercooler and alternative reduced turbine nozzle spec achieves full boost by 4500 RPM, peak HP at 280 by 6500. Sound close?

Slow reply, I've been to town and will have to look at the above later, My twinscroll gives me my 10 psi by 26-2700 rpm at full throttle. Although its a T4 flange with a .85 A/r, the small light wheels on ceramic bearings seem to spin up well. So probably not a relevant match to the T04B.  I'm running an ebay 4.5"x 4.5"x 10 air to water intercooler, efficiency unknown, but it keeps the IAT down considerably.

Thanks for the interest, hope I don't find myself custom broaching some oddball MOP pully!

[drdoom]

It seemed late getting full boost as I saw it. I've done very little work modelling turbo set-ups but now recalling that I'd had a go at figuring the original 910 I pulled that out of the library, ran the sim and then tweaked the set-up to emulate Thomas'. Attached graph is more representative, Thomas's in green/purple, Lotus 910 in red/turquoise. Looks better to me altogether, the earlier curves were too erratic.

[snowrx]

Well it's been 5 years since I took these charts, which don't reflect full power runs or plenty of tweaking since then, but here's what my datalogging showed for the calculated torque/power at the wheels. (rpm rise vs weight & gearing ) Looking at the top points gives a sense of output, but bears some skepticism given road grade is not accounted for and weight is from published road tests, not my car exactly. Lower points are cruise/decel etc.  107 cams on 104 pullies

[MrDangerUS]

LSA  (LCA) Lotus Esprit cam spacing, according to David Vizard

From the diagram on page 107 of DV's book "How to Build Horsepower":
LCA=106° is the best for Esprit ‘107’ cams with 35.5mm inlet valves.
The same engine, but w/36.8mm (S4s) inlets engine needs LCA=107°

Comparing to any two-valver, four valve heads exhibit a huge low-lift flow.
Rule ONE:
*The greater the low-lift flow is in the relation of the displacement of the cylinder being fed, the wider LCA needs to be.
*If you bore and stroke your engine, the LCA has to be tighter.
*Increasing the CR requires the optimal LCA to spread, and decreasing CR tightens it. For every CR above 10.5 add 0.75° to LCA
*The faster the intake valve is accelerated off its seat (high lift lobes), the wider the LCA needs to be.

David's decades of dyno testing reveal two factors that are evident:
1. There is a narrow range of LCA-s over which the engine delivers its best.
2. It is better to err on the tight side rather than the wide side. Lower LCA [-1° or -2° from optimal] delivers better torque at the lower rpm. TOO TIGHT LCA produces overall negative characteristics, like choppy idle and diminished low speed output. Just how much of a low-speed manners have been lost ultimately depends on the overlap used.
~~~~
Turbo engines don't like big overlap values. It is difficult to build boost while the back door is open (I.E. overlap, when both valves remain open).
Before the boost onset, the engine acts as N/A and large overlap is beneficial for scavenging.
When the turbo comes in, the large overlap hurts your performance. This is an excellent candidate for VVT !

Ideally, it would take two LSA settings to cover the whole rpm range. On DHOC engines, one can achieve it by phasing the intake cam.
VVT technology was developed for this purpose.


~~~~~~~Simple cam tuning rules for BOOSTED engines:
Advance intake and exhaust => more low-RPM power, less high-RPM power
Retard intake and exhaust => more high-RPM power, less low-RPM power
Less overlap => lower EGTs, faster turbo spool, less fuel consumption
More overlap => higher EGTs, slower turbo spool, more fuel consumption


*The more retarded in relation to the crank the whole event, the more top end,
*The more advanced = the more bottom and mid range power

 

[snowrx]

With due respect to Mr Vizard, the chart shows the general relationship only. With the different compression ratio, valve axis, number of valves, number of cams etc, we would need a new chart to apply to our 9xx engines.  That said, cam specs and timing are not intuitive to me at all so I say optimize the 2-3000 range for good off boost efficiency and cruise behavior, and let the turbo offer what higher rpm power it can. I'll be happy with whatever bolt-on assembly of commonly available parts gets me there, no interest in getting cams ground, searching out a 104 cam etc to get more power that I can't put through the Citroen box.

So my interest was modeling the difference for a turbo 910 with 110 exhaust pully vs a 104 exhaust pully, assuming the 104 intake pully was acceptable. If I get a stronger transaxle (Anyone, how's the Boxster 6spd adapter project going?) It might be worth $pending more time optimizing the valvetrain and upping the boost.

A two-piece cam pully with a centrifugal advance mechanism like a distributor's could give you a very basic VVT if you had the dyno time to establish optimums to shoot for. A clever design could be flipped over to retard on one cam and advance on the other.

[drdoom]

On track Thomas! Boosted engines are less fussy about the sorts of resonance tuning we N/A guys fiddle with. In terms of cam timing turbo's I'd say it's about maximising exhaust energy for quicker spool up, and managing overlap judiciously. You're correct I'm sure to focus on what the turbo can do for maximum output. From what I've read the 104 pulley on inlet, 110 pulley on exhaust is fine though let's not forget Lotus could have easily gone 110 on Ex yet chose to go 104. Not a world of difference in this case it seems.

[MrDangerUS]

IMO, Lotus wanted to achieve a brisk "street" performance at the low end.

104°/110° MOP delivers reduced overlap (38°), decent SLA=107° , but +3° intake cam advance shifts both curves toward high rpm region.

[drdoom]

Convention has long been established that intake closing point is the prime determinant of where power peaks. 104 MOP on intake is an earlier closing point, thereby peaks at lower revs than 110, as I understand it. Difference in outcome between 104 and 110 on exhaust side is quite minimal on the Turbo, perhaps meaningful on a tuned N/A engine with free flowing exhaust. Agree Lotus have stated all along their intention that the Turbo be a non-peaky engine.  

[tkathulevej]

Very interesting thread 🙂

I just discovered that my 2.0L 907 N/A actually has '777' marked camshafts (ie. probably 107's after-installed then). I haven't verified, but it SEEMS to be still the 102.5 MOP pulleys used....

Would be interesting to see theoretic torque/power curves for this setup compared to if switching to 104 MOP or 110 MOP on IN and EX.

...BTW my engine management book (Haynes, written by Dave Walker) says somewhere: "Take a power curve with a known cam timing and then retard the cam slightly. Theory says that this will give you and increase in top-end power at the expense of mid-range. This seldom, if ever, happens in real world. Often there is just a sweet spot that the engine likes, and moving he cam either way just just reduces power everywhere".    - I know that that moving 102.5 > 110 is not simply retarding (but retarding IN plus lowering overlap). However would very be interesting to know how close theory and reality matches for the 907(...anyone tried?)