Crankshaft worn

[Puffs]

I'd polish a cam like that using an oil stone.

And finish with some 1200 --> 1600 grit sandpaper, also wetted with oil.

But if you tried another cam (known to work well) in the past, it's probably not the cam.

Assembly issue?

[Blurredman]

Okay,

First thing I did was check for light coming from bulb whilst very very slowly turning the wheel, I got nothing out of the ordinary.

I removed the cam washer and put the cam back on, still ran backwards. The cam is a little grooved, sure but we move on.

But decided to change points and back adjuster plate the time, still the same.

Then I put the washer back on with the cam and something changed- the timing was way off, the small recess which is a bit of bent down metal (presumably with a hammer and tiny chisel) that coincides with the key of the cam broke off- so... i take a different rotor from one I have spare...., time it all up, and it times up the same as before - everything is exactly the same! Still it runs backwards half the time.

I then put original back adjuster plate and points on but you guessed it- Backwards running.

I also put total gap as 0.4mm but same thing.

The only thing I did do for experimentation was have the points gap about 1mm- that seemed to stop it running backwards, but that isn't a solution.



Hmmm


At the minute, I think the only thing I can do is swap the electronic ignition from my 251 onto this ETZ250, and equally put the contact breaker system onto the 251. And see what the results will be- for example, a question in my mind is would the 251 run backwards?

[Puffs]

The outcome of the 1mm lift height test is very interesting.

If you increase the points' lift height, they stay open longer, so that the dwell time / coil charge time is shorter. If that has an impact, it might point at the coil? Maybe we're looking at this from completely the wrong angle, it might not be in the points system where a bounce-induced extra contact is made, maybe there is a loose contact in the coil, which makes & breaks the primary circuit due to vibrations? Try a different coil.

The other thing that happens when you increase the points' lift height, is that the points close & open at a different place on the cam. The rate of rising & the rate of closing will be different, probably slower & more gentle. This would cause less bouncing, so less chance on chatter & an extra spark.
One of the reasons I replaced the points system on my Jawa was for the shape of that cam: it had (speaking in the normal direction) a very gradual rise, then after the contact was broken, the points lifted high, and at the end of the open time they closed sharply (presumably to make proper contact). The downside of the gradual rise is that the timing is less accurate, and needs more frequent adjustment. But it never ran backwards.
I don't know, but if you can solve your problem with a higher lift height, I don't see much downside in that. Surely longer open time, lower coil charge time (hence easier on the charge system too), a bit more wear on the cam follower (which is another reason to polish the cam), but that's all less important. Of course, you still have to get the timing right.

Some EI systems happily produce a spark running backwards too, so replacing the points by an EI is not necessarily the solution. To test the system on your 251, take the plug out & rotate the rear wheel backwards, with the plug in the cap & on earth, to see how that EI system works. If it breaks the connection for the width of a magnet you may get a perfectly timed spark going backwards too.

This backwards-running issue reappeared after you fitted a new piston. It might confirm that you have a good compression - maybe even too much? The distance between the gudgeon pin and the piston's crown is not necessarily identical for all pistons, and may differ between brands. How does the compression feel on the kickstarter? Have you ever measured the compression pressure? At some stage I found close to 9 bar on my 251, but of course it strongly depends on the squish you set.

Edit: I just came across a computer translation of Lothar's words in that document, which are relevant:
"
Which factors lead now to an increased bounce inclination and thus to a possible reverse run?
Generally it happens with worn contact surfaces, so with increased roughness or even existing craters. But also engine vibrations can contribute to the fact that the contact flutters briefly. Setting the contact spacing too low, in combination with unevenness on the cam surface, can lead to uncontrolled switching that causes sparking near TDC.
"
Surely try new points then, set at a higher lift heigth.

[Blurredman]

That's a good point concerning the EI on the 251. I never have done that test on the particular EI I have. And I can do it without touching anything but the spark plug.. This will tell me whether it's even worth attaching it to the Orange ETZ250.


You're also right to consider the point on the location of the gudeon pin and thus potentially influencing compression. But the bike never ran backwards with old piston when I had the compression @ about 12:1 for a brief period. But, I wasn't looking for it then- however, it was running nicely at idle. Typically the backwards running right now, and back in '17/'18 coincided with a lumpy idle when the engine was warm.

[Blurredman]

Okay,

So I have found something interesting. I have a maximum contact breaker gap size @ TDC exactly, and the points do not continue to open, or stay open for what is normally considered (on a single cylinder) roughly half the rotation of the crankshaft. Infact the contact breakers on the bike start to rise (as timed by me) at 2.5mm BTDC, and then as they just swipe past TDC they are already closing and are closed at about 2.5-3mm AFTER TDC. Therefore I think this is where the issue is. Total points opening is no potentially more than 50 degrees of rotation!


But: The way around it? I suppose the only thing I can think about if I want to retain points is to see whether I can open up the gap of the points considerably to see whether I can prolong the amount of time before they close.... But then whether the timing could be adjustable to compensate for that I really am not sure...

Another possible thought is that I could implement some sort of tapered ring or indeed a slight shim between the cam and the rotor opposite the highest lifting point of the cam in order to hopefully influence a longer opening time. If that is a possible workaround.

What are people's thoughts?

[Puffs]

Well you already said so in your 29/08 post, but somehow it didn't sink in and I didn't realise the consequences of what you said... But again, that's not a normal lift profile.

But yes indeed, if the points close some 2.5mm aTDC when going the normal way, they'll open around 2.5mm bTDC going backwards and you do not need any convoluted explanation why your engine can run backwards. It will just want to do that very naturally, and just about equally well as forward.

I checked in my ETZ251, and there they close near BDC, typically open half the time & closed the rest. This is fully in line with Lothar's PDF. Yours are open 14% of the time, and closed the rest (that might cause a hot coil & more power use). Open for half the time is more or less normal for a battery/coil/points ignition like this, I thought, but for other types of ignition it need not be. For instance here an older coil/points ignition (probably from a 60's MX bike):

Cam ignition without battery.jpg (26.59 KiB) Viewed 1093 times

which lifts for some 120°. And I have others which lift even for a shorter period. But those are all ignitions that function without battery (a temporary current is generated by a coil & permanent magnets inside the rotor - no battery, so the points can stay closed to reduce wear). I do not think yours is a normal lift profile for an ETZ... Maybe a cam for an ISDT machine without battery?

For the record a pic of a normal ETZ cam:

Easy to see where it lifts, not so much where it closes, which goes much more gradual.

The solution? A standard cam.

[Puffs]

As the timing & duration of the points lifting appears at the heart of the matter, I made a little film of it, see attached zip. I wired the witness bulb such that when it burns, the points have lifted, and v/v. I regret not having taken a LED bulb, as for the film, the filament takes a bit long to heat up & cool down.

I had also marked the location where the points open & close, but that's pretty much invisible on the film, hence this picture:

Points lift for about 240°.jpg (40.54 KiB) Viewed 1089 times

The black dots indicate the position of the cam follower on the cam at the moment where the points either start to lift, or close. See the points are lifted for about 240°, and closed for some 120°. On the pic they are closed.
Edit: so in this engine there is 240° between those trigger points (= where the spark occurs), when comparing normal & backwards running. When going backwards, I get a spark some 40° bBDC = 9mm bBDC, which will never cause combustion. (See later improved estimate.)

The film covers just 1 cycle, but if you play it in continuous repeat you can see that the bolt seems to wiggle. Specifically: towards the side of the end of the lifting period. I don't think the bolt is bent or anything, and AFAIK original. This offset may be in the crank, or in the rotor. As said, this has never given me any hassles.

The lot is set so that the points lift to a max of 0.32mm, and start to lift 2.75mm bTDC, as set with a clock in the plug hole. Previously I had it higher & earlier, and then it revved more easier, but like this (pretty much standard) it's slightly stronger in the lows. It runs fine & never backwards.

Attachments

Timing film (1 single cycle).zip
(655.54 KiB) Downloaded 20 times

[Blurredman]

Here is video as requested.

As you can see, the points lift for about 90 degrees but spent the rest of that time closed.

If I adjust the gap much wider, then the cam seems to be doing it's work and the points are open for 3/5 to 3/4 of rotation. These cams are original to the bikes I do definitely believe, and seem correct as evidenced by the witness marks on them indicating that a good 3/4 of the cam's rotational surface has been in contact with the points actuating heel.

[Puffs]

If you say the points open for 90° I'm not going to contest that.

To get a clear indication of the lifting period, I used the lamp to assess where the points start to lift, put a dot on the cam there (felt pen), and where the points close (next dot). Now, rather than eyeballed, I 'measured' the angle, and it looks like 107° between the dots is a better estimate. Hence open for 253°. (There is a minor parallax error in this new estimate.)

Points closed for about 107°.jpg (27.85 KiB) Viewed 1087 times

So, in my case, going forward, the points open at 2.75mm = 21.3° bTDC, so they close at 51.7° = 14.8mm aBDC. So going backward, I get a spark at 51.7° = 14.8mm bBDC, which can still never cause combustion.

If your points lift for 90°, and you set them to lift (in normal rotation) at some 21.5° = 2.8mm bTDC, that would mean in backwards rotation they would lift at 90°-21.5° = 68.5°= 24.1mm bTDC. I don't think that will run; the exhaust port has only just closed (for some 12.5mm) so there'll be insufficient pressure. So with this 90° opening period of the contact breakers, we again need another explanation for why it can run backwards: then the spark you get when the points lift comes too early. So we're back to the bouncing...? Or does the point where these breakers lift depend on the direction of the cam's rotation?

One thing I notice is that your points seem to look quite different from mine: while mine have a clear & narrow cam follower, shaped like a circle around the rotation point. I cannot see that on yours. Could that have any consequence?

And also, you say the cam bears witness marks indicating that a good 3/4 of the cam's rotational surface has been in contact with the cam follower. If the cam follower rests on the cam, the points are lifted, right? So that means the points were lifting for some 270° (=360° x 3/4), which is reasonably close to my 253°, yet quite different from the 90° you said it does now, in your film. So once again: could a higher lift height solve your problem?

I looked into the wiggle seen in the head of my bolt, and removed the bolt, as well as the cam. The wiggle is doubtlessly due to torquing it up with a spanner, which pushes the head a bit in the direction you push; it's a thin & long M7 x 90. The cam has a wider hole, the wiggle of the bolt's head has no consequence for the position of the cam. The cam fits very snugly into the rotor, with very little room to move when seated properly in it's key (Haynes fig. 35.1d); after re-fitting it I had quite the same lift height & timing.

[Blurredman]

Thank you Puffs for your extensive investigation, but It comes to me to realise finally that this is all self inflicted and the route cause of the problem is an out of true crankshaft. For anyone reading, I urge you not to use hammer force when inserting the crankshafts. I have been part lucky, and part unlucky in this process (lucky in the case of my Trophy for example) in the past, evidently, but ultimately doing so will cause the crankshaft to be unbalanced and out of round, and it doesn't take much of a wack to do that (after all, that is how you balance them in the first instance). In the short term, being unable to to use points effectively as I have found out, and if you use electronic ignition to get around it then in the long term the issue with main bearings may become apparent over the course of time and many thousands of miles.

My only real option is to disasemble the engine again, and get the crankshaft trued and balanced (I don't have the equipment to do it myself), and then reinstall THE PROPER WAY. Of course hindsight is wonderful but it took me a lot of hours, and pondering, and also failed attempts to simply plaster a solution to this- but ultimately, I am back to square one, and I will have to re-build the engine. My fault- and never again - lesson finally learnt. I'll update this thread with the result of getting the crankshaft trued (it doesn't need to be rebuilt), and will update with what the outcome is.

[Blurredman]

So I bought and have had delivered a new crankshaft. It looks good.

Came from Germany directly, so I wonder if I should stop by a machinist to double check the balancing is right. You never know how these packages have been treated in transit.

[Puffs]

That looks very nice & shiny!

They're not made of cheese, so if it was packaged properly and without damage to the packaging, you'll have to assume it'll be fine.
At any rate, if you cannot measure it properly yourself, or have a workshop nearby, assessing imbalance will be difficult. You can send it to a workshop, and they can measure it (and, if required, re-balance), and after you pay them, they send it back. And in that transit it can be damaged again. So still no certainty...

Sounds like the equivalent of Heisenberg's uncertainty principle for the modern age.