Why does it run better?

[John Roberts]

Bike is an ’82 R100RS. One cylinder not firing at all. Took out the plug and rested it on the cylinder, reconnected the lead and ran the engine, it sometimes had a weak spark, sometimes none. Now, the odd thing (to me) was that when I connected a spark tester in line with the plug after it was screwed back in the cylinder the engine ran perfectly, and with what seemed to be a good healthy spark visible in the tester.

All this spark tester is is a spark gap that is enclosed in a clear plastic tube which, as said, is connected between the lead and the plug; if you can't see a spark in it then there won't be a spark at the spark plug either. So how come the engine runs OK with an extra spark gap in the circuit?

By the way, it was the same with a fresh plug and when I swopped leads and only happened with the right hand coil. I am now looking for a new coil. (Not looking good, not sure if they are still available, surprisingly.)

 

[Beemerboff]

I have a iGunsons Flashtester gnition tester which measures the distance the spark will jump - and according to it a normal ignition should jump a gap between 9 and 15 mm, so jumping a
0 .6mm plug gap is not a conclusive test that things are OK.

My understanding is that OEM coils are not polarity sensitive, and you always one spark jumping the "wrong" way, not a problem with a good system , but one side will usually start playing up first as the system deteriorates.

 

[Franm]

I remember my uncle telling me years ago that in the older cars (way back) they often put a shirt button inside the plug cap.
The result being that a more intense spark would be generated at the plug because of this gap. The downside being that the plug would wear out sooner.
Maybe some electrically inspired person can explain why this happens.
This might also explain why this cylinder runs better when you have the spark viewer/tester in place.
It sounds like a coil issue though given that you've ruled out the leads and plug caps.

 

[Ogmios]

Cant help / but will watch and hopefully learn too.
My Guzzi has an intermittent fault very similar.
But it comes and goes - and is always cured by taking the offending sides plug cap off and holding it just a little away from the plug.
Thus it HAS TO make the jump, and always fires up fine.

Changed; plugs, HT leads, low tension wiring, coils and two (yes two ) complete ignition systems.
WTF?!

 

[John Roberts]

Funny, but I remember dad telling me about this sort of dodge working- getting the plug to fire by making the current jump another (extra) gap, by now I can't remember the explanation but it did seem to make sense at the time. All this is rather annoying because I do have a fair few qualifications in this electrickery malarkey but it still seems strange that if a current cannot jump one spark gap (the one in the plug/cylinder) then how on earth can it suddenly manage to jump TWO effing gaps?

OK, I know nooothing. Sigh.



And I miss you dad.

 

[Ogmios]

i am not alone!

... All this is rather annoying because I do have a fair few qualifications in this electrickery malarkey ...

John,
thank you for the above.
As a Guzzi keeper with no qualifiacations - in fact i'm electrophobic (i.e; crap at it!) you give me hope and renewed strength to keep on trying with my own!

 

[nick.ct]

The spark you see in your spark tester or when the plug is lying on the cylinder is generated in the benign environment of fresh air and no fuel. The effort to produce a spark under compression in a running engine is much greater.
If you are using a 2 output coil, I suspect the HT lead or plug cap
If you are using 2 seperate coils, try swapping the HT lead and plug cap to the other cylinder and see if the fault follows.
Unfortunately, coils can measure correctly for resistance but still not work in situ. However, if the resistance is incorrect, they really are unlikely to work

 

[John Roberts]

The spark you see in your spark tester or when the plug is lying on the cylinder is generated in the benign environment of fresh air and no fuel. The effort to produce a spark under compression in a running engine is much greater...

Agreed, but why does the plug suddenly decide to spark when the current is presented with the extra hurdle of another gap to jump??

 

[Redboots]

When I were a lad and on my apprenticeship, every garage had a spark plug tester.
Ours was a champion. You would screw the plug into the cleaning hole of unit and blast clean it.
When it was clean, you screwed it into a holder and attached the HT lead and pushed the button that would fire a jet of compressed air into the chamber to pressurise it and activate the HT. You could see the resulting spark or lack of it.


Good explanation of spark gap here: HT gap

 

[Myke Rocks]

Maybe this will help

http://ngkntk.co.uk/wp-content/uploads/2011/03/Motorcycle-Cat.pdf
Very good bike orientated plug descriptions.
It is believed that the button lowers the voltage applied at the plug tip.
If the plug insulator is breaking down, then when a very large voltage is applied, the spark "tracks" down the insulator inside the plug to ground instead of jumping the gap.
Decreasing the voltage, conversely to what is expected, stops the discharge "tracking", and thereby jumping the gap as intended.
If a coil is breaking down by tracking internally, then one would expect that reducing the spark plug gap would make it work, since the spark will take the line of least resistance, so the explanation I have just given is obviously total bollocks.
Hope the above enlightens you. (Not)
Myke

 

[ahutcheon]

Agreed, but why does the plug suddenly decide to spark when the current is presented with the extra hurdle of another gap to jump??

I think it's because the extra gap forces the voltage generated in the secondary winding by the collapsing (for most ignitions, which spark when the coil current is turned off) magnetic field to rise to a gap jumping level before it reaches the spark plug. Without that, a fouled plug can leak the energy from the coil away before the voltage builds far enough to spark. My physics is too rust to be sure, never mind explaining clearly...