Betty is dieseling

[Lebro]

So, looking at Peter's graphs - should we be setting our ignition timing at high revs to 30° BTDC rather than the static / tickover figure of 4°

 

Bob.

[graeme]

I had a TR4A in the seventies, it had similar mods to my current 4A, 10.5:1 compression, gas flowed head, big pistons, camshaft, manifold etc. It ran on/dieseled exactly as my current car does. Even with five star full lead fuel.

 

Cheers

 

Graeme

[Peter Cobbold]

So, looking at Peter's graphs - should we be setting our ignition timing at high revs to 30° BTDC rather than the static / tickover figure of 4°

 

Bob.

 

Bob,

Yes - but not on the static. That's exactly why we have the centrifugal advance. It advances the spark to about 30BTDC at 3000rpm to ensure the flame speed and the piston motion are synchronised to bring peak pressure at around 15 ATDC. Above about 3000rpm the turbulence in the cylinder keeps pace with revs and the flame keeps pace with the turbulence so no more spark lead is needed.

All the above is for wot operation.

When the engine is throttled back to cruise, the mixture is compressed less, heats up less, is usually made leaner. So it burns slower. That's why we have vacuum advance. Typical cruise advance 45deg BTDC. Lots of combustion as the piston is rising, but still more pressure is pushing it down. That explains the much worse thermal effciciency and rise in BSFC.

 

I never understood the reason for the spark advance curve - for decades. Classic car books / magazines rarely get to expalin the subject of combustion.

That graph was taken from Heywood's textbook. It is full of stuff that makes the design of engines understandable. Its a book that needs work - far from an easy read. And I have to ignore all the equations. Its really a book for undergraduates specialisng in engines. But highly reommended for understanding the 'why's of engines. I read the graphs mostly!! About £37 on Amazon.

http://www.prioritytextbook.com/product_images/o/166/9780071004992__87287_zoom.jpg

 

Peter

Edited July 18, 2014 by Peter Cobbold

[Peter Cobbold]

I had a TR4A in the seventies, it had similar mods to my current 4A, 10.5:1 compression, gas flowed head, big pistons, camshaft, manifold etc. It ran on/dieseled exactly as my current car does. Even with five star full lead fuel.

 

Cheers

 

Graeme

Graeme

That convinces me its not hotspot igniting the mixture but a spark. If youd had hotpsots causing the running-on the pistons would have succumbed to preigntion. Its very difficult impossible to envisage a hotspot that only appears upon switch-off.

None of the 8 causes of preigntion listed here would happen only after switch off:

https://www.hastingsmfg.com/ServiceTips/detonation_and_preignition.htm

 

You had points/coil ignition back then? and now?

So how do some sparks keep firing after the key has been switched off?

If thats impossible then attending to the source of the hotspot might be a good idea. Try the plug heat grade first.

Most web sites suggest its glowing carbon or wrong plug grade etc. But they dont say why that source of hotspot ignition is not present during driving.

 

What dos the ingition tell-tale do during 'dieseling'?

 

Peter

Edited July 18, 2014 by Peter Cobbold

[graeme]

I went out in the TR4A today and did some tests. Not sure they are conclusive. I have a switch which isolates the coil from the dissy. When switched off rather than the ign, the engine still diesels. I tried switching the ign off 'slowly' as per Roger but this didn't stop the dieseling. If I let the engine idle for a minute it stops OK.

 

I have an alternator and Magnetronic ign module on this car, but as far as I remember, the car in the seventies had a generator and points ign.

 

I didn't notice what the Ign light did when dieselling occurs.

 

The head was off about two years ago and no obvious signs of anything to cause a hot spot.

 

Incidentally, about 5 years ago I ran 40 DCOE webers for a short time and the engine never once dieselled. When the ign was switched off the engine stopped dead so I always assumed it was the SUs to blame for allowing fuel to flow, however something must be igniting the excess fuel.

 

Cheers

 

Graeme

[Peter Cobbold]

I've been reading the textbooks, but info on 'running-on'. or 'after-running' in USA, is sparse to say the least. But Heywood mentions its as being a compression ignition process, not surface ignition. But without any further discussion.

http://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0CCAQFjAA&url=http%3A%2F%2Fpersonal.mecheng.adelaide.edu.au%2Fmarcus.boyd%2FACT%2Fscans%2Fpg450-479.pdf&ei=qyrKU8XPHof3O8SWgZAO&usg=AFQjCNFh7fygBJiZVM2M8N6QdJbaU3KEMQ&bvm=bv.71198958,d.ZWU

( p452)

So it is indeed 'dieseling', not hotspot ignition.

How could this occur? I had been thinking that compression behind a closed throttle must be low - the cylinder is starved of air by closed butterlflies. But that is not true if the revs drop very low, say from normal tickover at 800rpm to 100rpm. On my old Aston the running-on happened not when the key was first turned but when the engine had very nearly stopped- then it would cough and knock and rattle the timing chain briefly. Do TRs run on like that - the rpm drop well below tickover then at very low rpm the run-on happens? At 100rpm the inlet valve is open for 8 times as long duration as at tick-over, but the air leak around the butterlfy stays the same and the air consumption by the engine is 8 times lower. So the pressure in the manifold is higher at the point of running-on than at 800rpm tickover and the cylinders fill much better. So the compression and heating might bring the mixture up to the point of compression ignition.

 

This compression igntion would be disastrous in loaded petrol engine so something must be different in running-on ignition - but Heywood does not elaborate. However there is the phenomenon of cool flames which ignite at low pressures and temperatures, 200C or so, due to a chemical reaction. Long chain normal paraffiins are prone to this.( see fig9-66 in pdf above) . In the dying rpm of a switched off engine. maybe 'cool flames' provide the ignition for the rest of the mixture. In which case higher octane fuels where aromatics replace paraffins ( which are low octane) will help avoid running-on. In normal engine operation cool flame ignition might be too slow to influence combustion, as suggest by this wiki:

http://en.wikipedia.org/wiki/Cool_flame

""A cool flame does not start instantaneously after the threshold pressure and temperature are applied, but has an induction time. The induction time shortens and the glow intensity increases with increasing pressure.^""

 

If this is the mechansim then practical solutions involve ensuring the engine is not running hot ( correct mixture and spark timing **) , allowing the engine to idle to remove heat from the chambers, and to try different fuels.

On a positive note, there's no reason to suspect that running-on involves hotspots that are likely to lead to piston-melting preignition. Unless the engine is old and tired and clagged up with carbon.

 

Peter

 

** running retarded heats up the exhaust valve abnormally. Most engines use vacuum advance to advance the spark at cruise If the v-a capsule fails then the engine will be running retarded at cruise. Its difficult to check correct operation of this, apart from checking the diaphragm is not punctured.

Edited July 19, 2014 by Peter Cobbold

[Peter Cobbold]

I went out in the TR4A today and did some tests. Not sure they are conclusive. I have a switch which isolates the coil from the dissy. When switched off rather than the ign, the engine still diesels. I tried switching the ign off 'slowly' as per Roger but this didn't stop the dieseling. If I let the engine idle for a minute it stops OK.

 

I have an alternator and Magnetronic ign module on this car, but as far as I remember, the car in the seventies had a generator and points ign.

 

I didn't notice what the Ign light did when dieselling occurs.

 

The head was off about two years ago and no obvious signs of anything to cause a hot spot.

 

Incidentally, about 5 years ago I ran 40 DCOE webers for a short time and the engine never once dieselled. When the ign was switched off the engine stopped dead so I always assumed it was the SUs to blame for allowing fuel to flow, however something must be igniting the excess fuel.

 

Cheers

 

Graeme

Graeme,

My apologies for raising red herrings. It turns out it is dieseling -see #31. Your test is definitive - proves its not sparks.

Yes, there's too many reports of fresh engines running-on for carbon deposits and hotspots to be serious contenders.

The Weber - SU difference is interesting. Do Webers have a fuel cutoff solenoid in the idle circuit? The SU float chambers woud I guess provide fuel for a minute of running-on!!

 

Can I ask, do the rpm drop well below tickover rpm before the running-on starts?

 

cheers

Peter

[Lebro]

It's quite easy to check the vacuum advance mechanism. Disconnect the pipe from carb end, remove distributer cap, and while watching the points base plate suck hard on the pipe. The base plate should move a short distance. Place tongue over the end of pipe to seal it, & watch for the plate slowly moving back to it's start position. if is does not then the diaphragm is OK, if it does then you have a leak somewhere - diaphragm, pipe, or tongue !.

 

Peter.

Next time I have the bonnet open, I will re-time ignition to 30° BTDC at 3000+ RPM, vacuum pipe off, then see what the timing is at tickover - should be interesting.

 

SU's will continue to get a supply of fuel as long as the engine is turning - mechanical fuel pump.

 

Bob.

Edited July 19, 2014 by Lebro

[Peter Cobbold]

It's quite easy to check the vacuum advance mechanism. Disconnect the pipe from carb end, remove distributer cap, and while watching the points base plate suck hard on the pipe. The base plate should move a short distance. Place tongue over the end of pipe to seal it, & watch for the plate slowly moving back to it's start position. if is does not then the diaphragm is OK, if it does then you have a leak somewhere - diaphragm, pipe, or tongue !.

 

Peter.

Next time I have the bonnet open, I will re-time ignition to 30° BTDC at 3000+ RPM, vacuum pipe off, then see what the timing is at tickover - should be interesting.

 

SU's will continue to get a supply of fuel as long as the engine is turning - mechanical fuel pump.

 

Bob.

Bob,

Yes that's a good v-a check to ensure its not knackered. But knowing the actual degrees advanced is tricky to measure on an operating engine. It will depend upon the spring retraction being correct and the vacuum being taken from the correct position on the throttle edge- tricky if the carbs have been changed. And the engine has to be at cruise load.

 

 

 

The centrifugal swing should give you 30BTDC - or near that - at 3000rpm without re-timing. But without load on the engine the butterflies wont be open enough to create the venturi depression that feeds the capsule. Even at fast tickover there's very little power needed to spin the engine - around 1-2hp - compared with cruise ( say 20-30hp). So the air flow is '10 times' less and the butterflies far less open. Thats why its tricky to measure- engine has to be swallowing air for the cruise load, as well as at cruise rpm.

 

Peter

[Guest ntc]

That's what a rolling road is for.

[Peter Cobbold]

£10 USB inspection camera/endoscope , 5m cable to laptop. And strobe lamp. Both pointing at the crank timing marks.

Worth a try

Edited July 19, 2014 by Peter Cobbold

[Guest ntc]

http://www.gunson.co.uk/item.aspx?item=3442

[BrianC]

That's what a rolling road is for.

I think that sums it up nicely. No matter how much you spend on diagnostic tools and time, it's only worth doing if you really enjoy it. For a quick, efficient and ultimately cheaper solution, a rolling road with an experienced operator (especially one who is familiar with your engine and fuel system) is the way to go. I couldn't believe the change in my car after a RR session. It still runs on but wasn't my concern and most 4 cylinder TRs seem to have the same problem. It's a sports car so as long as it starts and runs well, switching off is a minor issue. [Just my opinion of course]

[Peter Cobbold]

http://www.gunson.co.uk/item.aspx?item=3442

Mine is much simpler - and very much older. It merely flashes when the HT lead fires.

Its the Vampire model - disappears in daylight.

But under a closed bonnet it should be bright enough.

I'll tape the camera to it and clamp it to the radiator stay.

http://www.amazon.co.uk/s?ie=UTF8&field-keywords=usb%20endoscope&index=blended&link_code=qs&sourceid=Mozilla-search&tag=firefox-uk-21

Peter

Edited July 19, 2014 by Peter Cobbold

[peejay4A]

I think Neil is pointing out that a strobe with an advance feature will be needed to check advance accurately at 3000rpm. Does the Vampire do that?

Edited July 19, 2014 by peejay4A

[Peter Cobbold]

I think Neil is pointing out that a strobe with an advance feature will be needed to check advance accurately at 3000rpm. Does the Vampire do that?

Yes, the Vampire will. If the pulley is marked with true TDC and a scale stuck on the timing case with TDC 10 20 30 40 50 deg advance marked.

All in white paint.

So when #1 HT lead fires it lights up the scale and the pulley mark is opposite the degree number. It'll work at all rpm.

But only after sunset.

I keep a selection in the garage, never know when they come in useful:

http://www.retrosellers.com/images/vamps.jpg

 

 

I use it without the camera to check the centrifugal advance -it works fine. Neil's device gets round the lack of a scale- see pdf. It'll show the pointer aligned with the pulley mark if the advance matches the desired value thats dialled into the device.

 

Peter

Edited July 19, 2014 by Peter Cobbold

[peejay4A]

Not that accurate then. With a dial you bring the TDC marks together and read off the advance. Alternatively, set the required degrees advance and check that the the TDC marks align.

Edited July 19, 2014 by peejay4A

[Guest ntc]

Not that accurate then. With a dial you bring the TDC marks together and read off the advance. Alternatively, set the required degrees advance and check that the the TDC marks align.

Correct Pete

[mike3md]

Interesting debate.

 

I get occasional running on with a mildly modified engine (on SU's) set up on a rolling road earlier this year, with DD distributor. The timing is now 10deg BTDC @ idle (800 rpm) and 30 deg BTDC @ 3000rpm.

 

I know Ian Cornish has an anti run on valve fitted to 4VC - has anybody else done this?

 

Mike

[Peter Cobbold]

Not that accurate then. With a dial you bring the TDC marks together and read off the advance. Alternatively, set the required degrees advance and check that the the TDC marks align.

Pete, Neil,

That's nonsense. Its the other way round.

For the Vampire method degrees can be calculated from the circumference with great precision and marked off. No meaningfull errors there.

And the strobe fires at exactly the time the HT fires, on every cycle.

 

By contrast, the trick device has a chip in it that averages many cycles to determine rpm and hence the delay as to when to fire the strobe. In doing that it removes any evidence of timing scatter due to cam walking, end-float in the disy etc. All that information is averaged out.

Gunson give us no details of the avaraging period: 10 cycles? 100? It will spark on every cycle yes,but in timing its flash it has to calculate the millisecs' time delay from an ensemble of preceeding cycles. Ideally it will be a moving average but we are not told.That way the ca milliseconds delay between the actual spark and TDC is determined

The trick device may give an impressively stable reading but in doing that it conceals useful info on spark scatter and hence mechanical play in the crank to disy drive.

 

 

 

Peter

Edited July 19, 2014 by Peter Cobbold

[Lebro]

Have to agree with Peter here, I have a "fancy" zenon strobe timing light (Halfords) which can do advance, & dwell angle as well as measuring RPM,

but I only use it set to 0° advance, & then read the actual advance off the crank pulley.

This is a narrow belt damped type which prior to fitting I stamped with markings:

5, & 10 ° ATDC, & 5, 10, 15, 20, 25, 30, 35° BTDC.

With the Zennon strobe these marks show up well in full daylight.

 

Bob.

Edited July 19, 2014 by Lebro

[Guest ntc]

Checked against a Snap On and there was nowt in it.Just learn how to use it.

[Peter Cobbold]

Neil,

For me its no use.

By recording the timing by simple strobe+camera while I drive I can choose the speed, gear etc, on the flat, up hills etc and read off the spark timing.

To do that on a RR requires me to record manifold depression rpm gear etc for each load condition. Now thats a real hassle.

 

 

Anyone wanting to check the operation of their v-a capsule will find it easy and cheap as chips with 'my' method. No RR££, no expensive kit.

Less than £50, plus their laptop.

 

Peter

[peejay4A]

To be honest I can't be bothered to fart around checking the advance curve. It seems futile without a rolling road as seems to be the consensus. I'll stick with my Snap-On strobe and enjoy the car I think. I don't think I could mark the pulley accurately enough either.

[Peter Cobbold]

To be honest I can't be bothered to fart around checking the advance curve. It seems futile without a rolling road as seems to be the consensus. I'll stick with my Snap-On strobe and enjoy the car I think. I don't think I could mark the pulley accurately enough either.

Pete,

 

Measuring the centrifugal advance is really easy with the car stationery stationary. Just get a driver to set the rpm and you read the timing .No need at all for RR.

 

On a RR you're not getting the best tune for yout TR .Almost all tune ups jsut sue wot. But most road driving is at cruise -only around 5-10% is at wot, even for TRs, less for MGs obviously. And those cruise conditions are not replicated, even remotely, on most RR tune-ups. The va capsule plays an important role in cruise timing and it never operates at wot on the RR.

 

No need to mark the pulley- put the marks on the timing case. Google 'timing disc'.

 

Peter