A short study on ignition timing and combustion

Makes me think about the old mag system ,dishing out sparks when the advance-retard springs deceide :roll: a system that was good enough for 60 years, or go one better manual advance....bank comming up..better retard a bit :lol: and a bit of Fag paper and a stick to time the engine. :!:

john robert bould said:

Jim, Are you saying my Pazon unit advances instantly at tick over, to take account of the low fuel/air? Only strobing it there is no advance at tickover, it appears static untill reving high?

Click to expand...

Of course not. A Pazon is not a load sensing ignition.

To have a load sensing ignition you must have the input from either a MAP sensor or a TPS. Even then they are normally configured so they do not advance with vacuum at idle to keep the idle quality good. Jim

Great thread, thanks Jim.

And, great explanation and thanks to Dances with Shrapnel for a concise statement of the big twin head's design advantages.

I'm still struggling with the trigonometry, however I've downloaded a copy of Thinbasic. Years ago this would not have been a problem to me as I had access to a computer which ran a decent high level interpreter language. I will be struggling with the language format for a while, however I will stick with it - trying to calculate the shape of a curve which gives the same finishing time for the fixed duration combustion event as the revs rise between 3000RPM and 7000 RPM.

Which EI fires on the 1st revolution?

jseng1 said:

Which EI fires on the 1st revolution?

Click to expand...

Pretty much any analogue system that I know of. Most digital systems take two. Jim

My Boyer fires as I connect the power to it.

Most references say they are shooting for highest brake torque pressure to hit ~15-17' ATDC as best compromise of necessary lead time pressure rise vs what's left on for power stroke. Analog boyah's firing on each new application of power is helpful for diagnosis by key on and plinking at various conductors till a snap or flash sensed then home in from there. Seems like there should be a sonic way to montior best spark curve but would be a treat to get rocket science level mapping of spark lead to rpm pressure peaks.

Btw Fred Old Brits and the Powerarc creator say with 3 close hits low down no need of a special up/dn change in curve for idle stabilization because there's so many miss fires at idle on the single spark units. I of course want to believe its so.

acotrel said:

I will be struggling with the language format for a while, however I will stick with it - trying to calculate the shape of a curve which gives the same finishing time for the fixed duration combustion event as the revs rise between 3000RPM and 7000 RPM.

Click to expand...

You won't have fixed duration combustion. It will be almost that at lower rpm, and a lot of engines start off using about 1 degree advance per 100 rpm until 2000 or 2500 rpm. After that, at higher rpm, the increased turbulence speeds up the burn time. Advance increases at a lesser rate.

You'll reach a stage where you can run full advance, and the increase in turbulence speeds up the burn time enough that you don't need any further increase in ignition timing.

X-file said:

acotrel said:

I will be struggling with the language format for a while, however I will stick with it - trying to calculate the shape of a curve which gives the same finishing time for the fixed duration combustion event as the revs rise between 3000RPM and 7000 RPM.

Click to expand...

You won't have fixed duration combustion. It will be almost that at lower rpm, and a lot of engines start off using about 1 degree advance per 100 rpm until 2000 or 2500 rpm. After that, at higher rpm, the increased turbulence speeds up the burn time. Advance increases at a lesser rate.

You'll reach a stage where you can run full advance, and the increase in turbulence speeds up the burn time enough that you don't need any further increase in ignition timing.

Click to expand...

And in many cases- especially true on a Norton -you can build more power by retarding the spark a few degrees over 5500. About 2 degrees per 1000 rpm. Jim

And in many cases- especially true on a Norton -you can build more power by retarding the spark a few degrees over 5500. About 2 degrees per 1000 rpm. Jim

Click to expand...

Shoot Nortons don't thrill that much below 5500. Will have to have a word with Powerarc about that.

acotrel said:

I'm still struggling with the trigonometry, however I've downloaded a copy of Thinbasic. Years ago this would not have been a problem to me as I had access to a computer which ran a decent high level interpreter language. I will be struggling with the language format for a while, however I will stick with it - trying to calculate the shape of a curve which gives the same finishing time for the fixed duration combustion event as the revs rise between 3000RPM and 7000 RPM.

Click to expand...

That sounds simple but it is not. To be able to create that curve you would need to know the burn rate at all RPMs. The burn rate is going to change as the RPM rises due to mixture motion. It's going to depend on a lot of little variables like compression ratio, port shape, cam timing and squish area so it is going to vary from engine to engine. One size does not fit all. Jim

Jim, in doing the calculation a few assumptions would be needed - standard comp. ratio, standard shaped combustion chamber, fixed acceleration in the burn rate of the fuel. I started thinking about this stuff and memory of my involvement in closed vessel testing of gun propellant came back to me. For different explosive materials there is such a thing as 'brisance'. From memory it is the pressure rise rate divided by the time to reach maximum pressure. This might be the effect of changing octane ratings of fuel:

https://www.princeton.edu/~achaney/tmve ... sance.html

What I am suggesting is that we should do the calculation of the advance needed over the usable rev range of a commando engine which must be applied to the start of the fixed time combustion event - with the purpose of simply compensating for the increase in revs, so that max. pressure arrives at a fixed point in the timing.
The first assumption should probably be that the 28 degrees static advance is correct for a commando engine on petrol, at about 3,000 RPM (i.e. after the cam and exhaust starts working properly). Then assume that max combustion pressure occurs 10 degrees ATDC. And work from there in 500 RPM steps, and jet to that curve. For different fuels, or detonation occurring - move the whole curve by adjusting the static advance. I think that some Norton twins had centrifugal advances with spring loaded bob weights, - we can probably do better than that.
I cannot get my head around jetting first then adjusting the timing curve to suit. How do you get the jetting right without a fixed timing regime ? Particularly with two stroke engines, changing the timing can be deadly.

How do you get the jetting right without a fixed timing regime ? Particularly with two stroke engines, changing the timing can be deadly.

Click to expand...

I use a gas analyzer and 02 sensors to get the mixture right. To get good readings of the burned mixture the timing needs to be about right first .
Applying full load to test fuel mixture on an engine who's timing is not fairly close can cause serious damage. Jim

comnoz said:

How do you get the jetting right without a fixed timing regime ? Particularly with two stroke engines, changing the timing can be deadly.

Click to expand...

I use a gas analyzer and 02 sensors to get the mixture right. To get good readings of the burned mixture the timing needs to be about right first .
Applying full load to test fuel mixture on an engine who's timing is not fairly close can cause serious damage. Jim

Click to expand...

When you said that to get good readings of the burned readings the timing has to be about right first, surely that is a chicken and egg situation ? What I am suggesting that it is the SHAPE of the advance curve which is important and not the static advance. If you have the shape near right then you can balance the mixture against it by using different carb needles etc. I think it would be interesting to use the calculated corrections to the advance over the rev range which compensate for the change in acceleration of the piston as the revs rise, and then jet to that. If you then wanted to seek max torque by playing with adjusting the curve, you would have a decent starting point.
About O2 sensors and gas analysers - I'm an industrial chemist, engineers use our black boxes often without regard for calibration and response times. I suggest you need to think about what the readings you get from them actually mean. The O2 content from different fuels must vary with the composition and all petrol is a mixture of a very large number of components.
So what do you do - calibrate the O2 sensor from an ideal situation using the fuel of your choice ? It is again a chicken and egg situation.

The O2 readings are the amount of oxygen left in the exhaust after combustion. It just tells you how much of the oxygen was consumed. Exhaust temp and HC readings will tell why -if there was excess oxygen left.

The oxygen sensor is calibrated with the standard amount of oxygen in ambient air.

Timing vs. mixture is a bit of a chicken and egg thing. That is why you lather, rinse and repeat until you get it right. Jim

I notice that ignition systems such as the Dynatek 2000 offer a range of different curves (8) which you can switch between. Does that mean that the shape of any one of them is ever the optimum to suit the jetting ? One thing that I've noticed is that with methanol, getting the midrange jetting correct is critical as the change in performance is very noticeable if you know what your bike usually does in certain situations. And if a change in gearing is achievable the opportunity for it is very deceptive - not easily picked up by the rider, especially with the heavy crank. With petrol it is twice as critical to get it right. In my own case I use fixed timing with static advance added to allow for the properties of the methanol. Because I am still using low compression, the jets are still small and I can use Mikuni petrol needles in home-made needle jets. There is still a lot of margin for error, and I believe I am not getting anywhere near the best (potential) performance right across the rev range and at all throttle openings as I accelerate out of corners.

One aspect which interests me is the relationship of the high pressure of the combustion event with the opening of the exhaust valve. I use a two into one exhaust system with a fairly large diameter tail pipe, and I've advanced the standard 850 cam to compensate. I think the exhaust valve now opens at about 92 degrees BBDC. If the high pressure point of the combustion event occurs as late as 10 degrees after TDC, the piston only gets full thrust for 78 degrees of crank rotation. The torque characteristics with that type of exhaust are obviously affected by kadency effects just as they are in a two stroke. Perhaps 'what you pick up on the merry-go-round, you lose on the hurdy-gurdy' ?

More details on monitoring spark and mixture via plug chops, if ya believe a Honda expert can advise us usefully.
http://honda-tech.com/forced-induction- ... *-3063102/

I only use plug chops for setting the main jet size. On a race circuit, getting the best acceleration of corners is the primary issue. I usually jet down on the needle and needle jet until I get the cough in corners, then raise the needles one notch. While doing this I usually have main jets fitted which are slightly too large. If you think about how often you have the throttle wide open blasting down a long straight on a race circuit it doesn't really happen so much. If the mains are grossly undersize you can do damage, however most of the time you are using the mid-range jets. With petrol, if the mains are too big the bike will probably slow substantially when ridden flat out - with methanol the problem is not so great. When setting the mains, it is wise to fit one heat range hotter plugs then replace them with racing plugs after you've leaned off the jetting. The plug readings you take will then usually keep you in the safe jetting zone.
The shape of the needle is important - I use Mikuni petrol needles in Mk2 Amal carbs on methanol, the opportunity for fine adjustment over the whole range from 1/4 to 3/4 throttle, is better. It is extremely important to get the jetting right at 3/4 throttle, and that is where my primary activities usually are. If you need to, it might help if you put a spot of paint on the twist grip to indicate that point. And with petrol the changes make a much bigger difference than with methanol due to the fact that every methanol fuel jet is about 1.6 times the diameter of a petrol jet. Mid range jetting only has to be the slightest bit too rich in either case and you can lose a lot of performance. Increasing the main jet size does not usually affect the mid-range jetting. However once you have set the midrange it is important to fit larger main jets and do a plug chop to check that the transition from mid-range to the mains actually happens. It is possible to have the situation - meter off the tip of the needle and the mains never operate - you won't normally get that if you are using petrol.
About reading plugs, if you see beads of aluminium on the plug that might mean the jetting is far too lean or you timing has moved, and you will probably need more money. With a commando engine the problem is never as disastrous as it is with a two stroke. Keep the black ring on the porcelain.