OHV Cam Theory

[Ben_Flierman]

Hi guys,

I had some thoughts about the cams and stuff on the pushrod B18/20s. I started a new thread on Brickboard.com , but figured I might aswell throw it in here:

As a follow up of <a href="http://www.brickboard.com/RWD/index.htm?id=920672&show_all=1">this thread ion the brickboard about cams</a> I would very much like to hear all your opinions on cams, profiles, valve lift, cam wear, lifter shape, etc on the B18/20 pushrod engines.

I made some drawings on the way a lifter responds to cam shape:

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/1.jpg">

The blue curved line shows the cam profile (as if measured with a micro meter). The red line show the actual profile the lifter would follow. It's clear that they differ.


When plotting graphs of lobe profile and actual valve lift you can again see that they differ. The blue and red line again correspond to lobe shape and valve action. They are basically the same lines as in the figure above, only plotted in a graph.

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/2.jpg">

You can see that the actual valve action starts earlier then you would expect based on cam duration numbers. Naturally, with a large vale lash this effect gets smaller. With an extrem valve lash it would theoretically look like this:

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/3.jpg">


What one would want for a better valve action is more lift at low duration. You want the area under the curve to be as large as possible:

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/4.jpg">

So to get to this valve action, would would the lobe shape look like? You would a 'fatter' start/bottom of the lobe, so that the valves open further, more quickly, like the cam in Case B below. However, when you would use the same CAM-duration, this would yield in a <b>very</b> large increase in actual VALVE duration. A lot more then allready was the case with the stock cam in Case A.

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/5.jpg">

To eliminate this effect of strongly increased Valve duration, one could create a lobe with less overall duration, and create something like the cam below, in Case C. This lobe would be quite good imo. One could maintain a low overall duration, but the valves would open quicker so the area under the curve would be increased. However, the side load on the lifter would be relative high, which would result in increased cam wear.

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/6.jpg">

IMO, the only way to solve this is to create a new lifter shape: a lifter with a smaller bottom. That way you could use the cam of Case B, but not have the effect of increased valve duration.

<img src="http://home.wanadoo.nl/j.w.flierman/camtheory/7.jpg">



So, what do you guys think of this?

Cheers, Ben

[740Weapon]

do you have matlab? these are the equations that govern cam design and shape. you can run this program and play with it if you want. i suggest you get a copy of chapter 8 from "design of machinery" by robert L norton 3rd edition. also the program "dynacam" and read about things like pressure angles and undercutting and why your pictures need larger flat taped lifters or perhaps roller lifters. because what you have depicted just isnt how it happens. seriously. your lobe shapes would create impossible pressure angles unless you had lifters so large that the dent in the lobe was never followed. a pressure angle is the amount the cam pushes on the side of the lifter compared to pushing the lifter up. and this is just to get the cam to physically work. never mind what is best for the gas flowing in and out of your engine. in which case having the largest area under the curve as you put it probably isnt the best scenario at all. atleast for a street motor. maybe at high rpms thats the goal. but not on the street.

This is a matlab program that uses a modified trapezoidal acceleration (minimizes force required to cause lift and is also continuous through displacement, velocity, and acceleration. and that is what goes into cam shape design. i attached images of a cam at 3500 rpms (7000 crank) with a .5" lift. and a lift, dwell, and fall of 30 degrees each (probably around the hottest concieveable cam under the most severe operating conditions)

i dont know if this helps in any way but its so yall understand what really goes into cam design and what is and isnt acceptable. besides that i would say a good cam is something that somebody else already designed that works well. i.e. oem or production aftermarket. a custom cut cam and discussions of such things make me very very nervous. unless its for a engine in which thousands and thousands of people are hopping them up. because their is a reason why a larger lift cam hurts parts of the powerband at the expense of other parts. and the biggest lift and duration will probably not work well with anything else under your hood. including your engine internals and your intake and exhaust shape and size.


% omega is angular velocity in radians
%h1 is maximum lift
%theta1 is the angle in radians allowed for lift
%dwell1 is the angle in radians that the cam dwells at lift
% theta2 is the angle allowed for the fall of the cam

omega=input('omega=')
h1=input('h1=')
theta1=input('theta1inradians=')
dwell1=input('dwell1inradians=')
theta2=input('theta2inradians=')
dwell2=2*pi-theta1-dwell1-theta2;
beta1=theta1
beta2=theta2

%constants specific to a modified trapezoidal function

b=0.25;
c=0.5;
d=0.25;
Cv=2;
Ca=4.8881;
Cj=61.426;

%rise

x1=[0:0.0001:(b/2)];
s1=h1*Ca.*[b.*x1./pi-(b./pi)^2.*sin(pi.*x1./b)];
v1=h1/beta1*Ca.*[b./pi-(b./pi)*cos(pi.*x1./b)];
a1=h1/beta1^2*Ca*sin(pi.*x1./b);
j1=h1/beta1^3*Ca.*pi./b.*cos(pi.*x1./b);


x2=[(b/2):0.0001:((1-d)/2)];
s2=h1*Ca.*[x2.^2./2+b.*(1./pi-1./2).*x2+b.^2.*(1/8-1./pi^2)];
v2=h1/beta1*Ca.*[x2+b.*(1./pi-1./2)];
a2=h1/beta1^2*Ca;
j2=h1/beta1^3*0;

x3=[((1-d)/2):0.0001:((1+d)/2)];
s3=h1*Ca*[(b/pi+c/2)*x3+(d/pi)^2+b^2*(1/8-1/pi^2)-(1-d)^2/8-(d/pi)^2*cos((pi/d)*(x3-(1-d)/2))];
v3=h1/beta1*Ca*[b/pi+c/2+d/pi*sin((pi/d)*(x3-(1-d)/2))];
a3=h1/beta1^2*Ca*cos((pi/d)*(x3-(1-d)/2));
j3=h1/beta1^3*-Ca*(pi/d)*sin((pi/d)*(x3-(1-d)/2));


x4=[((1+d)/2):0.0001:(1-b/2)];
s4=h1*Ca*[-1.*((x4.^2)./2)+(b/pi+1-b/2)*x4+(2*d^2-b^2)*(1/pi^2-1/8)-1/4];
v4=h1/beta1*Ca*(-1.*x4+b/pi+1-b/2);
a4=-1.*h1/beta1^2*(Ca);
j4=h1/beta1^3*0;

x5=[1-b/2:0.0001:1];
s5=h1*Ca*[b/pi*x5+2*(d^2-b^2)/pi^2+((1-b)^2-d^2)/4-(b/pi)^2*sin(pi/b*(x5-1))];
v5=h1/beta1*Ca*[b/pi-b/pi*cos((pi/b)*(x5-1))];
a5=h1/beta1^2*Ca*sin((pi/b)*(x5-1));
j5=h1/beta1^3*Ca*pi/b*cos((pi/b)*(x5-1));

%dwell

x6=[0:0.0001:1];
s6=h1;
v6=0;
a6=0;
j6=0;

%fall

x7=[0:0.0001:(b/2)];
s7=h1-h1*Ca.*[b.*x1./pi-(b./pi)^2.*sin(pi.*x1./b)];
v7=-h1/beta2*Ca.*[b./pi-(b./pi)*cos(pi.*x1./b)];
a7=-h1/beta2^2*Ca*sin(pi.*x1./b);
j7=-h1/beta2^3*Ca.*pi./b.*cos(pi.*x1./b);


x8=[(b/2):0.0001:((1-d)/2)];
s8=h1-h1*Ca.*[x2.^2./2+b.*(1./pi-1./2).*x2+b.^2.*(1/8-1./pi^2)];
v8=-h1/beta2*Ca.*[x2+b.*(1./pi-1./2)];
a8=-h1/beta2^2*Ca;
j8=-h1/beta2^3*0;

x9=[((1-d)/2):0.0001:((1+d)/2)];
s9=h1-h1*Ca*[(b/pi+c/2)*x3+(d/pi)^2+b^2*(1/8-1/pi^2)-(1-d)^2/8-(d/pi)^2*cos((pi/d)*(x3-(1-d)/2))];
v9=-h1/beta2*Ca*[b/pi+c/2+d/pi*sin((pi/d)*(x3-(1-d)/2))];
a9=-h1/beta2^2*Ca*cos((pi/d)*(x3-(1-d)/2));
j9=-h1/beta2^3*-Ca*(pi/d)*sin((pi/d)*(x3-(1-d)/2));


x10=[((1+d)/2):0.0001:(1-b/2)];
s10=h1-h1*Ca*[-1.*((x4.^2)./2)+(b/pi+1-b/2)*x4+(2*d^2-b^2)*(1/pi^2-1/8)-1/4];
v10=-h1/beta2*Ca*(-1.*x4+b/pi+1-b/2);
a10=1.*h1/beta2^2*(Ca);
j10=-h1/beta2^3*0;

x11=[1-b/2:0.0001:1];
s11=h1-h1*Ca*[b/pi*x5+2*(d^2-b^2)/pi^2+((1-b)^2-d^2)/4-(b/pi)^2*sin(pi/b*(x5-1))];
v11=-h1/beta2*Ca*[b/pi-b/pi*cos((pi/b)*(x5-1))];
a11=-h1/beta2^2*Ca*sin((pi/b)*(x5-1));
j11=-h1/beta2^3*Ca*pi/b*cos((pi/b)*(x5-1));


%dwell

x12=[0:0.0001:1];
s12=0;
v12=0;
a12=0;
j12=0;


X1=x1*theta1;
X2=x2*theta1;
X3=x3*theta1;
X4=x4*theta1;
X5=x5*theta1;
X6=theta1+x6*dwell1;
X7=theta1+dwell1+x7*theta2;
X8=theta1+dwell1+x8*theta2;
X9=theta1+dwell1+x9*theta2;
X10=theta1+dwell1+x10*theta2;
X11=theta1+dwell1+x11*theta2;
X12=theta1+dwell1+theta2+x12*dwell2;

S1=s1;
S2=s2;
S3=s3;
S4=s4;
S5=s5;
S6=s6;
S7=s7;
S8=s8;
S9=s9;
S10=s10;
S11=s11;
S12=s12;
figure

plot(X1,S1,'r',X2,S2,'r',X3,S3,'r',X4,S4,'r',X5,S5 ,'r',X6,S6,'r',X7,S7,'r',X8,S8,'r',X9,S9,'r',X10,S 10,'r',X11,S11,'r',X12,S12,'r')

title('Displacement vs. Angle')
xlabel('Angle (rad)')
ylabel('Displacement')

V1=v1*omega;
V2=v2*omega;
V3=v3*omega;
V4=v4*omega;
V5=v5*omega;
V6=v6*omega;
V7=v7*omega;
V8=v8*omega;
V9=v9*omega;
V10=v10*omega;
V11=v11*omega;
V12=v12*omega;


figure

plot(X1,V1,'r',X2,V2,'r',X3,V3,'r',X4,V4,'r',X5,V5 ,'r',X6,V6,'r',X7,V7,'r',X8,V8,'r',X9,V9,'r',X10,V 10,'r',X11,V11,'r',X12,V12,'r')

title('Velocity vs. Angle')
xlabel('Angle (rad)')
ylabel('Velocity')

A1=a1*omega^2;
A2=a2*omega^2;
A3=a3*omega^2;
A4=a4*omega^2;
A5=a5*omega^2;
A6=a6*omega^2;
A7=a7*omega^2;
A8=a8*omega^2;
A9=a9*omega^2;
A10=a10*omega^2;
A11=a11*omega^2;
A12=a12*omega^2;

figure

plot(X1,A1,'r',X2,A2,'r',X3,A3,'r',X4,A4,'r',X5,A5 ,'r',X6,A6,'r',X7,A7,'r',X8,A8,'r',X9,A9,'r',X10,A 10,'r',X11,A11,'r',X12,A12,'r')

title('Acceleration vs. Angle')
xlabel('Angle (rad)')
ylabel('Acceleration')


J1=j1*omega^3;
J2=j2*omega^3;
J3=j3*omega^3;
J4=j4*omega^3;
J5=j5*omega^3;
J6=j6*omega^3;
J7=j7*omega^3;
J8=j8*omega^3;
J9=j9*omega^3;
J10=j10*omega^3;
J11=j11*omega^3;
J12=j12*omega^3;

figure

plot(X1,J1,'r',X2,J2,'r',X3,J3,'r',X4,J4,'r',X5,J5 ,'r',X6,J6,'r',X7,J7,'r',X8,J8,'r',X9,J9,'r',X10,J 10,'r',X11,J11,'r',X12,J12,'r')

title('Jerk vs. Angle')
xlabel('Angle (rad)')
ylabel('Jerk')


man this post is nonsense.

graphs in inches, inches/s, inches/s^2 and inches/s^3.

edit: and by the way the cam in the graphs is impossible to make. (the dwell and lobe shape are all wrong for like a car...and is pretty much shaped like the your last picture. engines use more duration and much less if any dwell.

[gsellstr]

Judging by the first run of pic's, I'd say also a roller lifter setup would be much better than the flat lifter design (aside from the lower friction due to the rollers) although I don't know of any such lifter off-hand for our engines.

[JohnMc]

Actually, finding roller lifters would be the easy part. The B18/20 shares a lifter bore size with the SBC, and you can probably find roller lifters for those just about anywhere. Buy a set for an Iron Duke (1/2 of a SBC) to get 8.

The hard part would be machining the block to accept them. They need a keyway (AFAIK) to keep them rolling straight. A machine shop could probably do that.

And then, of course, you'd need a different cam profile to match, but Isky et al could probably whip you up ne for not much more than regular cost.

As for the topic of the original post - you are assuming that the cam profile was designed without taking into consideration the interaction of the lobe against the lifter, but I'd bet large amounts that that isn't the case. The 'optimum' line as described by the center line of the lobe isn't what they were after designing the cam, they wanted the end result. They take the rocker ratio into consideration as well.

[mikep]

The newest roller lifter guides look like this
http://www.international-engines.com.br/manual_i/powerStroke/cap_g5_2.ASP

[Ben_Flierman]

your lobe shapes would create impossible pressure angles unless you had lifters so large that the dent in the lobe was never followed.Well thats exactly my point. You could create an improved(?) lobe shape but the wear on the cam and side load on the lifter will be enormous (what I said in 1st post). So it might work ok, but will last for about 30 sec...
Thats why I figured, maybe there was a way of getting the vale action youd want AND decrease cam/lifter wear.

your pictures need larger flat taped lifters or perhaps roller lifters.Wider lifter would be great, but there is no room for them in the B18/20. Roller lifters would be the best imo, but like JohnMc allready pointed out, you need to keep them lined up with the cam, so there will be ($$) machining neccesary.

the biggest lift and duration will probably not work well with anything else under your hood.Thats why I kept lift and duration the same. I just wanted to create more area under the curve. But still, these are no exact graphs or what ever, just to visualise the concept.

man this post is nonsense.Well, I'm a relative newbie, so I might be completely wrong, but I just ponder about engines and stuff, and am eager to learn more about them. I do have a CD of matlab, although only slightly familiar with it. I will install it and see what happens when pasting your routine in there. Thanks.

As for the topic of the original post - you are assuming that the cam profile was designed without taking into consideration the interaction of the lobe against the lifter.No no, not al all, but IMO the design on the OHV cam/lifters is just crap. The way they interact does work, but not without relative much wear. Which is why the cams/lifters in the B18/20 need to replaced a few times in the engine's life.

[stylngle2003]

well, i enjoyed reading your thread, Ben.

it is a very interesting concept. i wonder what the costs of machining the engine to accept roller lifters....the gains could be substantial if the actual graph is anywhere near to the theoretical-to-prove-a-point graphs you have presented. perhaps by messing with matlab you can get some more realistic information. keep at it!

[Ben_Flierman]

well, i enjoyed reading your thread, Ben.

it is a very interesting concept. i wonder what the costs of machining the engine to accept roller lifters....the gains could be substantial if the actual graph is anywhere near to the you have presented. perhaps by messing with matlab you can get some more realistic information. keep at it!
Thanks! And like you said, what I presented are just theoretical-to-prove-a-point graphs, no exact figures what so ever. Perhaps I should have made that a bit clearer. Just theoretic bla bla so to speak...
I'll have to create some free space on my hard dsic to be able to instal matlab, but I wil certainly have a go at that. Also the possiblity for roller lifter would perhaps be worthwhile to look into a bit further.
In the mean time, keep those replys coming!

Cheers, Ben

[JohnMc]

The newest roller lifter guides look like this
http://www.international-engines.com.br/manual_i/powerStroke/cap_g5_2.ASP
Veddy interesting, I could se adapting something like that to a B20. You'd need the lifter guides with the proper center-to-center for the B20. And the lifter guide retainers would probably need to be in 4 pieces as well given the restricted access we have to the lifter gallery. If made to handle just a pair of lifters each one could be lowered down to hold a single lifter guide and fastened to the side of the block - only block modification being the bolt holes for that.

Actually, the more I think about it, given my lack of fabrication tools, I'd still probably be better off just taking a block to a machine shop and having them carve out a keyway for the older style of roller lifter.

{EDIT}But wait! How about these:
http://www.wickedcases.com/charger/lifters.jpg
They would need to have the proper center to center measurement, but that linkage has to allow some side to side movement already, so it wouldn't have to be *exact*, just close enough. And you'd need shorter push rods, but no block modifications at all. Anyone have access to a SBC that can measure the center to center of a pair of lifter bores?

[740Weapon]

"Well, I'm a relative newbie, so I might be completely wrong, but I just ponder about engines and stuff, and am eager to learn more about them. I do have a CD of matlab, although only slightly familiar with it. I will install it and see what happens when pasting your routine in there. Thanks."


i was refering to my post being nonsense. not yours. mine is filled with like 150 lines of code and graphs that dont represent a buildable cam... their graphs for a cam that looks exactly like the cam you made in the last part of your post. i wasn't trying to tool on you in any way. just my own half assed atempt at conveying information. and my code has to be put into a m-file... incase you have never used matlab before. and their is some way to make matlab post the displacement graph radially and if you find out the base circcle of a B20 cam and add that to every "S" in the program you could get matlab to actually show you a cam lobe. and then you could figure out the physical limits of cam shape... and maybe somewhere on the internet you can find pressure angle limits for a automotive cam.

edit: and their is some way to make matlab calculate pressure angles. but i cant visualize it right now.

[Ben_Flierman]

Actually, the more I think about it, given my lack of fabrication tools, I'd still probably be better off just taking a block to a machine shop and having them carve out a keyway for the older style of roller lifter.
What do these 'older style roller lifters' look like? Do you have a pic?

[V8BRICK]

I can easily find out the lifter bore spacing on a sbc block if someone can measure their torn apart/blown b20, I know there are a few out there ;)

I like the idea of the linked bar solid roller lifter if it's close to the same bore spacing as a sbc. Though it wouldn't be hard to cut/weld the link bar to add or remove material.

[JohnMc]

I have a bare B20 block in the garage. I'm not sure what luck I'll have in getting my calipers in to measure them, but I'll give it a try. I could also probably measure the side-to-side measurement on a cam between a pair, although those are rough case and the figure might not be precise. And the cam lobes are offset to the side (to make the conventional lifters spin) and maybe they all aren't offset the same way.

By old style roller lifters I mean the ones that just have a pin poking out of the side that rides in a groove in the lifter bore. Maybe those aren't old, maybe they aren't even the most common, but I was under the impression (quite possibly false) that that was how roller lifters are more commonly installed if from the manufacturer, not added in later. (PS - I just tried to find some that look like that and I couldn't - perhaps I was totally off base in that thought)

PS - Some different roller lifters - showing the slot and how it scoots back and forth, and that the connector apparently doesn't need to be very heavy - that looks like sheet metal:
http://fireandfuel.com/crower%20roller%20lifters.jpg

[gsellstr]

You guys SUCK. My motor build keeps getting more involved and more expensive seeing things I should do. ;) I'm liking this roller lifter idea even more now assuming the details can be worked out. Less drag, better following of the cam profile, can't really see a down side yet. Definately something to pursue and get part numbers out there for the few of us planning a B20 build!

[JohnMc]

Better following of the cam profile - is a bogus goal. The cam profile *itself* is not what the desired movement of valve is, unless you have an OHC cam that pushes right on the valve. Otherwise the cam profile is a 'reverse engineered' shape designed to give the proper movement at the valve, taking into consideration the lifter/cam interaction and the rocker arms and their changing ratio as they, umm, rock.

So when the roller rocker interacts differently with the cam lobe then you need a different shaped lobe to end up with the proper valve movement. And over on the twin Brickboard thread John Parker is saying that he doesn't know of anyplace to get roller cams for B20's, and that apparently Isky won't/can't just whip up a roller cam using a Volvo blank. Perhaps it is because of the way the lobes are offset of the lifter center to make flat lifters spin as the cam turns. I'm guessing you would want a lobe on a roller cam to be centered and wider too. Perhaps they just plain don't make blanks like that.

As for some P/N's - the SBC lifters that fit right into the B20 (with longer pushrods) are P/N 202-H - so I would have every reason to believe that the 202-RH roller liftes would have the correct diameter. ( http://www.iskycams.com/productdisplay.php?sku=1429&hdwt=31101&loc=101&dealer=no ) - or possibly these (http://www.iskycams.com/makesearch.php?catid=3495&engine_id=21) since they also list 202-H as the solid lifter option. They might possibly need to have that little connector arm length changed.
They certainly look taller, so you'd need shorter push rods, and you'd still need to track down a cam with a lobe shaped for the roller rockers instead of a flat lifter. I can swear I've seen those somewhere. not on Isky, not KgTrimning. Maybe it was Paeco, I'll try to find their catalog at home tonight.

[Ben_Flierman]

over on the twin Brickboard thread John Parker is saying that he doesn't know of anyplace to get roller cams for B20's, and that apparently Isky won't/can't just whip up a roller cam using a Volvo blank. Perhaps it is because of the way the lobes are offset of the lifter center to make flat lifters spin as the cam turns. I'm guessing you would want a lobe on a roller cam to be centered and wider too. Perhaps they just plain don't make blanks like that.
That might indeed be the case, although do remember that when using a roller lifter, you'll need a lot more 'flesh' on the lobe, since a roller lifter will will move in a way very simular to the actual lobe shape. There is no mushroom effect, so you need largers or 'fatter' lobes. (height unchanged offcourse). Just an other thought on what could be the problem though.

[JohnMc]

I can easily find out the lifter bore spacing on a sbc block if someone can measure their torn apart/blown b20, I know there are a few out there ;)
I managed to wiggle in a caliper and got a rough inner edge to inner edge measurement of 1.2 inches. I guess you'd have to add a single lifter diameter (2 halves) to get the center to center. I think the are something like .842?

[Volvo Heretic]

It's four years later, has anything more been figured out on the subject of roller lifters?

[JohnMc]

No roller cams available, so that makes the use or roller lifters pretty moot. But probably not that hard to do.

The point remains - you can't get a roller cam. The lobes need to be wider and not offset slightly to the lifter bore centers. And the shape of the lobe needs to be different as well.

[JE Jr]

Ok, you wonät understand anything, but there's some guys in Sweden working on some roller cams now. Not yet finished, but there are some pics here:

http://www.agap.se/Produkter.htm

[fidel]

No roller cams available, so that makes the use or roller lifters pretty moot. But probably not that hard to do.

The point remains - you can't get a roller cam. The lobes need to be wider and not offset slightly to the lifter bore centers. And the shape of the lobe needs to be different as well.

i have been talking with my cam grinder about this, the cams are doable but on custom billet cores. they would cost twice as much as any cam on the market now.

the other issues we would run into are that there isn't much room in the lifter area to put bridged roller tappets in so we would most likely have to go with keyed lifter and either cut groves in the lifter bores or bore them out and install slotted guides. second we would have to keep the cam from having as much movement front to back. the bronze trust plate is less than ideal. the timing gear would have to be modified like i do on my motors so we could run roller bearing .

[JE Jr]

An old Timo / TRS rollercam form the 70s or 80s
They used some kind of ball bearing as the axial bearing.
http://web.telia.com/~u67800607/V-B20-RollerCam.jpg

[JohnMc]

second we would have to keep the cam from having as much movement front to back. the bronze trust plate is less than ideal. the timing gear would have to be modified like i do on my motors so we could run roller bearing .
Most of the newer replacement cam gears (fiber) seem to have been designed by people that didn't realize the back side of the center section was actually a bearing surface that rides against that thrust plate.

[mikep]

No roller cams available, so that makes the use or roller lifters pretty moot. But probably not that hard to do.

The point remains - you can't get a roller cam. The lobes need to be wider and not offset slightly to the lifter bore centers. And the shape of the lobe needs to be different as well.

There's a cam grinder on Black Dog Alley in Easton MD who used to grind B30 race boat and B18 race boat cams. I can stop by and talk to him.

[Volvo Heretic]

Well, if it's going to be that much work and expence, I might as well just dream of this B304 with two B204 heads;

http://i143.photobucket.com/albums/r123/VolvoHeretic/Volvo%201800%206%20Cylinder/B204-6CylinderEngine.jpg

[dbh86]

Well, if it's going to be that much work and expence, I might as well just dream of this B304 with two B204 heads;

http://i143.photobucket.com/albums/r123/VolvoHeretic/Volvo%201800%206%20Cylinder/B204-6CylinderEngine.jpg

Mismatched cam gears and all?? :-P

[Volvo Heretic]

Is that your picture I molested? Too bad I am not a mechanic, machinist, or know anything about these motors, or maybe I could get it done. Just a dreamer...