Electronic Ignition Advance Curves

I realize a lot of this has been hashed out for years, so I'm not really adding much to the conversation. However it was a fun project and I learned a lot. I basically constructed a test harness to spin ignitions and measure the advance. I wrote up the results here: http://www.doovsprojects.com/projects/commando-ignitions. To cut to the chase, here's the summary graph



Note: the advance and rpms are crankshaft units. If you read my write up you'll see why the PowerArc advance is so different. Hopefully this info makes sense. Always open to any constructive feedback.

-- David

There are several that have been done, curious as to how close they track, brand by brand.

No Pazon, no Sparx...

Nice work. Advance curve has been in the past of concern for my combat engine. The Boyer advance is too fast so switched that for a Lucas RITA several years ago. When the RITA died it was replaced with a Tri Spark based on comments that support your conclusion. The downside of the Tri Spark....first unit lost it's inductor function with less than 1,000 miles use. The company shipped a new unit no charge from Australia, so good customer service. So far so good on the second unit. I'm still a bit nervous about the location of the box in the former points position. As a semiconductor veteran, how do you view high temperature and vibration environment for ICs. Sounds like rel testing and that spec is typically 1000 temp cycles.

Just going by this forum in an unscientific way, while the Tri- Spark must work well when it's working, the failure rate seems to be quite high compared to the other EI ignitions?

Glen

I have had what may be non-typical amazing resuts with Sparx ignitions. Two failures, on the same bike, and over 40 sold & installed.

napanorton said:

I realize a lot of this has been hashed out for years, so I'm not really adding much to the conversation. However it was a fun project and I learned a lot. I basically constructed a test harness to spin ignitions and measure the advance. I wrote up the results here: http://www.doovsprojects.com/projects/commando-ignitions. To cut to the chase, here's the summary graph

Note: the advance and rpms are crankshaft units. If you read my write up you'll see why the PowerArc advance is so different. Hopefully this info makes sense. Always open to any constructive feedback.

-- David

Click to expand...

Thanks for the new data Dave - you did a very thorough and well thought out job, especially including the references to other's work.

Thought there would be merit in gathering a couple plots, cited by you and previously posted here by "rick in seattle", to provide more information in one place which allows the reader to visually compare the results of the various undertakings.

http://www.accessnorton.com/commando-timing-advance-curves-compiled-reva-t6488.html



The plot below was updated to include the Altair.

I too had an early Trispark fail due to heat. Trispark replaced it quickly and no problems since. As I understand it, Trispark redesigned the electronics to be more robust (using higher spec components, mechanical redesign). Based on the overall build quality, the performance, and customer service, I'm more than willing to believe the unit is adequately designed and tested. Although the points cavity is a horrible environment for electronics, it's no worse than a lot of industrial processes, some military environments and certainly space. We've managed to get semiconductors to work reliably in those places, so a commando should be easy!

--David

napanorton said:

... Although the points cavity is a horrible environment for electronics, it's no worse than a lot of industrial processes, some military environments and certainly space. We've managed to get semiconductors to work reliably in those places, so a commando should be easy!

Click to expand...

I seem to recall hobot and perhaps at least one other, who produced pencil-traced layouts of engine vibration at certain point(s) on the engine, (perhaps?) to study the effects of different hardness of rubber inserts in the isolastics.

I wonder how harsh the shaking environment is, at the points cavity location?

I read your article and appreciate the difficulty you have in setting up the EIs you tested and graphed.

Can't dispute your conclusion. I used to run a PA and struggled with it for 3 years - wore out a kick start pawl in that time.

Now use Trispark in my bikes and they are 1-2 kicks to start every time.

I suggest that in a petrol engine best power depends on a certain combustion condition and the difficulty lies in maintaining it under all conditions of carburation/vacuum, revs and comp. ratio. A decent engine management system with fuel injection needs inputs including vacuum sensing, throttle position and revs sensing so that it can determine fuel needs. On a commando with normal vacuum dependent carburation, we usually choose an advance curve and comp.ratio then jet our carburation to suit. I suggest that the shape of the taper on the carb. needles and the shape of the advance curve are inter-related. The symptoms of mistuning the carburation are often similar to those which occur when we have an ignition system problem. Basically, the main reason we use an advance curve is to allow for the differences in rock-over time which depend on rod length/stroke ratio and the revs.

when I look at both of these graphs, I see that they both have plotted the boyer microdigital advance curve. I wonder why Napa's graph doesn't show the slight reversing of the advance at the begining of the curve that Dave Comeau's curve does show. Are they 2 different boyer ignitions?? Also, does the fact that the Dave Comeau curve matches up to the trispark curve shape, mean that the boyer microdigital also has an accomodation for idle stabilization built into their advance curve??





I supose I might have to ask Dave Comeau...

Napa's graphs have no points plotted below 500 rpm, Dave's do so they show the retarded timing at kicking speeds.

Not to be too argumentative, but the reversed slope of both curves which show some retarded spark is well beyond 1000 rpms on both graphs, so I would think that napa's graph would show a little reversed curvature prior to showing the advance.... But anyway,...

.... my point was that: I wondered if the boyer microdigital (which I have on my bike btw) has the same feature in it's advance curve that the trispark is noted for. That being the retarding of the spark at kickstart RPM's in order to make starting the bike easier... If it does have that, I didn't know that, and that's why I asked the question...

From reading the details in the link napa provided, it looks like the Boyer in his test is an analog MKIII, not a micro-digital. That would explain why the curve doesn't have a retard feature for low rpm starting. Identifying Boyers can be a bit confusing with the different names, like Boyer-Brandsden, Boyer Micro-Power, Boyer Micro MkIII, Boyer Micro-Digital. Then again, maybe it's just me that's confused.

Ken

What is the best system to use if you want to programme your own curve into it ? I've seen it claimed that it is impossible to mathematically derive the advance curve from the rod length, stroke, time of combustion event etc. And I don't know whether I believe that. I wonder if the curves which are supplied with some ignition systems were derived by first setting the unit up arbitrarily, then adjusting by trial and error ? Seems to me that within reason, an error in the shape of the advance curve can often be compensated for by adjusting the carburation. That approach might not give a performance optimum.
I understand that these days a decent high level computer language for calculating tables from algorithms is Python. Have any of you ever gone down this path ?

I have held off from buying a programmable ignition because it would be expensive to buy a unit, then find it is not versatile enough to be adequately programmed.
When we do this tuning stuff we are playing with 3 variables - ignition advance (curve), mixture and comp.ratio. Usually we fix the comp.ratio and the ignition advance, then tune the carburation. When we use an advance curve we have the opportunity to cope with the way the engine geometry changes the rock-over time - that affects the position before TDC that the combustion event begins and also when it finishes. Angularity and when maximum thrust occurs on the piston, probably has an effect on power output.
I think for most guys this tuning activity is a backwards and forwards process and we end up in the middle of space somewhere which is not necessarily the optimum (maximum potential).
With my own bike, I use fixed ignition timing and I still have not tried all the available Mikuni carb. needle tapers in my Mk2 Amals. And I know there is probably advantage to be had in doing that. However it might be a case of putting the cart before the horse, if a programmable ignition gives more control over the timing of the combustion process.

My Micro digital has the reverse curve idle stabilisation, makes tuning the idle a longer exercise as the boyer will advance the timing if the revs drop and the revs then go back up so you have to factor that in when tuning the pilot mixture.

I did not see a Micro digital on Napa's graphs, only the trispark which has the reverse slope, so not sure what I said that was wrong.

I am currently working on adapting a car ECU to fit Norton Commando's, this has fully programmable curves via bluetooth and as well as varying the curve by revs also will vary them according to inlet depression, so you get a mapped curve that will retard timing with a wide open throttle at low revs more than if you just feather the throttle open, which especially on Triumphs pinking is an issue. There are other inputs it could use as well but I am not suing them but may do later, these include, throttle position, engine temp. Software is open source and runs on PC and a cutdown version on Android. It will do wasted spark or sequential spark and I will first try for sequential, this should work as below 500 rpm it reverts to wasted spark to ensure on each switch on it retimes itself quickly. Coils will be digitally driven and could be coil on plug if I can find one short enough. The injector part of the firmware will be unused. I need to install a crank position sensor for mid cycle monitoring using the engine sprocket teeth as the trigger wheel and a cam sensor looking at a second trigger wheel to show each TDC for firing ie one gap per firing point reference. All electronics will be in a box away from the engine heat.

I have never kick-started a standard Commando. Are they more difficult than other old British bikes ? I can remember trying to start Triumph Bonnevilles and getting the kick-back on every attempt.

kommando said:

My Micro digital has the reverse curve idle stabilisation, makes tuning the idle a longer exercise as the boyer will advance the timing if the revs drop and the revs then go back up so you have to factor that in when tuning the pilot mixture.

I did not see a Micro digital on Napa's graphs, only the trispark which has the reverse slope, so not sure what I said that was wrong.

Click to expand...

I don't think you said anything wrong either. I was responding to oOnortonOo's post, where he wondered why napa's curves didn't show the same retard feature for the Boyer that dynodave's showed. I probably should have quoted him to make that clear.

Ken

acotrel said:

I have never kick-started a standard Commando. Are they more difficult than other old British bikes ? I can remember trying to start Triumph Bonnevilles and getting the kick-back on every attempt.

Click to expand...

My 1972 T120RV always starts first kick hot or cold and it never ever gave me a kick back. Pazon ignition, Amal 30mm concentrics, Morgo 10-1 pistons and jugs.
8)