Timing strobe jitter

Check the rotor is solid on its steel insert. 

Loose timing chain?

Just looked at the Pazon web-site and max advance (at 3000 rpm) is quoted as 30º, so variation of five to eight degrees is a significant fraction of this.

If you don't have the necessary spares in stock, I'd be inclined to order them right away in order to check the cam chain a.s.a.p.

What type of strobe and what type of external power supply ? My old Gunson definitely likes a separate battery.

Some cheaper Strobe timing lights have an arrow on the HT lead inductive pick up clamp, if fitted the wrong way they can give an erratic reading.

But I'm sure you've already sussed that out?

Hi Richard ,

Had the same issue on a Triumph Daytona which had a very old EI fitted. In that case swapped it out for a Pazon Altair job sorted.

If you have access to another black box then substituting it for yours might be quicker than a timing chain check.

Dont know if an iffy battery, poor ignition switch or poor electrical/ earth connections to the box would cause flutter

Might be tempted to ask Pazon what they think

Coincidently, I actually strobed mine only a couple of days ago with similar results.

850MkIII with Pazon Altair, brand new strobe gun.

This issue was extensively investigated a couple of years ago by a fellow member of the Thames Valley Branch. He found that the cam chain drive idler gear was eccentric by 0.0085", this was found to be because the idler gear itself was poorly machined and the pressed in bush simply perpetuated the situation.
He punched the bush out, manufactured and fitted an undersized bush and machined it correctly, the runout was now less than 0.001"

He now has a steady timing line.

When I had the timing cover off a few years ago, I checked the chain and it appeared to be worn as it varied it tightness as the engine was rotated. I fitted a new chain and had the same issue.

I then did some calculations based on the 0.0085" eccentricity measured above. This gave a potential variation in slack of 10mm based on a tight to a loose chain.
Calculations attached.

I never managed to work out what variation this could give on the angle, anyone else fancy it?

Note:- A worn or slack chain in itself is unlikely to cause the jitter at 3000 RPM, as the top run is always in tension and the slack will be on the lower run.
However the eccentricity of the idler gear could cause the issue.

I have not done the mod to my bike yet, this will have to wait until I get back from Sweden, I do not went to start taking it apart just before going away.

So, take the timing cover off, (or an 850 remove the inspection cover), rotated the engine backwards ( so the slack is now on the top run) and see if there is ANY variation in slack as the engine is rotated, there should be zero. If not, you most likely have an eccentric idler gear.

I may pop out to the garage now and check it on mine, will update later.
 

Just realised, an easy way to tell if the jitter is down to the strobe / ignition / noise etc. is simply to point the strobe light at the ignition pickup magnet thing itself. If that is rock solid, then the problem is the connection between the alternator rotor and the ignition pickup, and the only variable there is the cam chain part and hence the idler gear eccentricity as described above.

If you do have jitter when pointed at the magnet, then you know it is related to the ignition system or strobe in some way.

Give it a try.

While I take Tony's point about the load on the cam chain at constant 3000 rpm, I'm not entirely convinced.

It's true that at constant engine speed it must be the case that the power needed to drive the cam will be constant — when averaged over a number of revolutions (at least two, recall).

But some of the energy used to compress the valve springs will be returned as the followers bear on the closing side of the cam lobes?

At the very least there must be some variation in the tension in the upper run of the cam chain, even if it never actually goes to zero.

I agree, and it is very difficult to prove either way. I was thinking it was unlikely to go from completely tight to completely loose, but I think a high speed camera may be required and that is one step too far!

But if the chain IS varying in slack as you gently rotate the engine (best way is going backwards by actually turning the camshaft) then either the chain is worn in a rather strange way, or the idler gear is eccentric.

Again, if someone would like to do the calculations as to how many degrees jitter you could get with the chain varying in slack by whatever you find on your bike, it would be interesting to know if the numbers work out.

Just done the calculations. If we assume the maximum slack is 4.8mm as per the manual. If the chain is tight on the top run, the slack is on the bottom, if it is tight on the tensioner, it will be slack at the top. So the 'tight' distance between the timing marks on the intermediate gear sprocket and the camshaft sprocket will be the 10 rollers, so 3.75 inches. (95.25mm).

If we now have all the slack at the top (the valve springs are now driving the camshaft) , the actual distance between the timing marks will be less, I have worked this out to be 0.244mm

So on the basis of the effective circumference of the gear is roughly 170mm. The actual angular change is only just over half a degree.

I have no idea how much backlash there is in the gear between the crankshaft and the intermediate gear but I would not expect it to be any more than another half a degree.

So if we say it is a total of 1 degree, but we have a 2/1 ratio between the crankshaft and the intermediate gear, this means the jitter on the crankshaft could be 2 degrees. And that is if everything is optimum.

So, the things to check are:-

Strobe the magnets and make sure this is rock solid.

Check the chain tension and make sure you have an even or maximum 3/16 (4.8mm). By the way, I can think of no reason why the chain would wear unevenly.

If it is not even, check the eccentricity of the intermediate gear.

See if you can detect any backlash in the crankshaft to intermediate gear drive. Nothing you can do about it if bad, but at least you would know.

Other than the obvious of making sure the alternator rotor is tight on the shaft, I cannot think of anything else.

Check the air gap between the rotor magnets and stator pickups is not too wide.

What is too wide? is there a spec?

The spec should be available in the set up info. Form memory I think the TriSpark gap is no wider than 1.5mm and there's a gauge supplied in the kit. I had a problem with a new Boyer years ago and they sent me a rotor with extended magnets to cure it.

I've got a Pazon Altair fitted to my 850, there's no mention of rotor gap adjustment on the fitting instructions.

But it states that a poorly seated rotor (on the camshaft internal taper) may cause timing variations.

Better to check that before pulling your timing cover off...

You dont mention the make and model of the timing light. (sorry, just read through it again and you do) Some of the inexpensive ones are only specified to sense up to 5000 sparks/min, but they don't even seem to do that. With the wasted spark you are only measuring 3000 sparks/min but....stange things happen on cheap mass produced circuits. 

You also don't mention if both Commandos have electronic ignition (with wasted spark), which could be why the gun works better on one than the other.  Also, be sure to keep both HT leads away from each other if you have a poweful spark discharge set up, as the one you are not clamped too can still interfere with the signal of the one you are clamped to. 

P.S Its really hard to find the maximum spark rating of most timing lights. Typical values are between 5000 to 8000 untill you spend over £100. Its hard not to buy chaep tools when you only expect to use them one though. 

Thinking on this original problem, if the actual timing was jumping around 5 or more degrees, wouldn't the engine be 'hunting'. Assuming you are holding a steady throttle. Also when used on the road a 5deg change in timing surly would cause jerky running?? My thoughts are that the ignition and strobe are interfering with one another.  The ignition will upset cheap digital voltmeters and the cheap strobes we tend to use can be triggered randomly. I continue to watch with interest.

Today I strobed the magnets, it it was rock solid. Smoothly advancing up to 3000 RPM

So the jitter I see is definitely somewhere between the rotor and the crankshaft. Not electrical interference or dodgy strobe light 

I am off to Sweden on the 7th so am not about to take it apart, will look when I get back.

Richard had his strobed with a higher quality gun (maybe a Snap On) and the problem went away, so it looks more like his gun. His other commando has a different electronic ignition and the gun works fine, so it looks more like the ignition.  Tony has a different gun but the same electronic ignition and has the same problem so it looks more like the ignition. If he could get his strobed with an expesive gun would the problem go away too?  There are no consistent results to draw conclusions on yet. 

Are you strobing from a different lead when you strobe the magnets on the right side of the bike compared to the timing marks on the left side of the bike? (My strobe, which looks identical in every way to Richards but is not a Gunson, only works on the left lead, the right lead does nothing. Can't figure that out at all. Thats with TriSpark and a single TriSpark coil.... The bike works fine!)

If it would just stop raining for half an hour we would all be out riding and not figuring out problems. (Those of us in the UK that is) 

Doing research on single coils with twin leads, it appears that the current flow in one lead goes out to the plug - the other lead has the current flowing back from the plug. When using a strobe, you have to put the pickup on the other way round for the lead where the current flows back from plug or else the strobe will not work. There are articles where you can test this current flow yourself.

Another potentail problem with single coils is that spark plus are designed for the the current flow in the normal direction. Where the flow is back to the coil, the spark plug may erode more as the spark is jumping in the 'wrong' direction and people have reported having to replace that plug earlier.

2 off individual coils are supposedly wired correct to + - on them, this gives the same polarity of spark. A double ended coil has one spark of reverse polarity at the plug, hence the strobe fails to flash. And you have a differently worn plug (this was all in the 2CV manual years ago) Harley Davidson, Honda, and a lot more continue with practice, it does work, but I have seen double ended coils fail/cause grief more often than normal coils. But we still haven't reported on engine 'hunting'. So I am of the opinion that it is an ignition/strobe issue-not an actual engine issue, but what do I know?

I did think that the only other difference between looking at the alternator marks and the timing magnets was that I was looking at the other HT lead, but it is the same on both, as you would expect. 

So I do not believe it is anything electrical but is mechanical.

Given that the jitter we see is when the bike is on no load (tricky to do on load!), I would not expect the engine to complain in any way, so hunting is unlikely. 

As regards double ended coils,  see:-

https://www.nortonownersclub.org/html/archive/rh/324/index.html#15

Personally, I would always avoid them if possible.

Ahh very intersting indeed. I wasn't aware that is how a single dual output coil works. That expalins the timing strobe anomaly. Thanks for providing the link too. I couldn't find anything on the physics of it on-line myself. I'll keep a close eye on the plugs and make sure I never swap them from side to side just so I can monitor the difference. At least I know which side is which now. So far I haven't had any problems with the single coil at all, but I havent done many miles with it either.

It's for reasons such as this that I imported a Tri Spark coil all the way from Aus instead of using an easier available alternative. This way Tri Spark have the entire onus for my ignition system. 

Anyway sorry for the hijack of this thread. Back to timing strobe jitter .... 

Keeping a close eye on the plugs will show after reasonable millage (a few hundred) at least that they will wear different. The centre electrode will be more rounded on one plug. The 2CV manual says to swap the plugs to even the wear. It shouldn't make any difference to our engines provided all other factors (carburation, oil burning) are in order. But be warned if you run a twin output coil and you have an issue on one cylinder, then swapping the HT leads could alter the problem IF the problem is exacerbated by the spark goodness.

If you are seeing jitter on the EI trigger, then that obviously needs sorting out before anything else along the line.

If the ignition is doing what it should, then there should be no observable jitter. Things to confirm / check. 

Use a separate and well charged battery for the strobe light.

Keep the pickup for the strobe as far away as possible from any other wires.

If there is no spec for the gap between magnets and pickup, then check that the gap is equal either side.

Make sure the magnet / rotor is not loose on the taper.

If possible, try a different strobe light.

It is very unlikely to be dodgy wiring but check that anyway.

If all else fails, replace the EI unit.

Best of luck ...

Have a look at the attached picture. This is my plug / coils / ignition test rig. Consists of a fairly ancient Boyer driven by the Dremel, two old coils, HT leads, new suppressor caps and old plugs.

The bottom trace on the scope is the pickup from the rotating magnet part. The top trace is a pickup on the HT lead. I expected the two to be rock solid, but I found the HT trace was sometime jumping slightly early. I calculated this to be about 12 degrees.

I then used the strobe light to look at the magnets as and I could observe the same jitter.

It may be there is noise in the system, it may be I have a sick Boyer (it is over 30 years old), it may be that the slight jitter is not a problem. But I do know that my new-ish Pazon Altair is absolutely rock solid when doing a similar test.

So, I would suggest is anyone is trying to time their bike with a strobe, the first thing to do is to strobe the pickup mechanism itself before you do anything else. If there is jitter, it is NOT mechanical, and remember a degree of jitter on the pickup is 2 degrees of jitter on the crankshaft.

If you see jitter, talk to the supplier / manufacturer, tell them what you have done and get their opinion (and post it here). 

hello,

i wonder if the air gap between the sensor and the rotor is critical.

barry