That good old hanging rpm topic again

[tunnelrider]

Woops I should've read Snowbums notes on carb syncing first, looks like you don't set the idle mix by the shorting method, but adjust the idle speed first. Make it quick between shorting sides, 3 secs initially, then shorter.  I'll have to revisit it.  First time I did that I did both sides fairly individually, 30 secs to a min each side.

How fast should I expect a return to idle from 3K rpm, hot engine, please?  Would be a good reference point.

[Barry]

It should return quickly to idle and you are not expecting too much that it does. I'm going for a ride this morning so I'll note how long it takes but I'd guess 1 or 2 secs at most.

[Barry]

After 20 odd miles mine dropped to idle in 2 seconds.

[Justin B.]

I'll have to pay attention but I think all of ours return to normal idle immediately after a "normal" high speed circuit.

[tunnelrider]

OK thanks Barry and Justin for the info on what I should be getting for a return to idle.  Rather than do another carb sync session I'm going to overhaul the carbs first, I (embarrassingly) remembered last night, from a thread a couple of years ago, that I bought the bike with an R80 carb on the left side and I have all the gaskets, OEM diaphragms, needles, new main and idle jets and new venturi to make them the same.  Better do this first I think    I'll post again with the results.

[tunnelrider]

Overhauling the carbs has not given me a quick fix to the slow return to idle, it's just the same.  Granted I still need to spend some time syncing the carbs again, but after a 40min run with a few stops to give a basic set up, the revs are still taking a long time to return to idle, still approx 4-5secs from 4000 rpm.

So if I can't fix the slow idle return by syncing the freshly overhauled carbs I guess I'll be looking at the bean can.  But there's one thing that confuses me if this is the problem - how can a couple of small weights hold up the revs?  I don't believe the inertia of them would over power the friction and compression of the cam and crankshaft.  So would the bean can cause hanging idle because the firing, if it's advanced, causes the engine to speed up?

Or could my problem of slow return to idle be caused by poor compression?  'Cause I have to admit I was pretty surprised to have only 100 psi each side, with new rings and a fairly new head recondition. Hmmm... 

[Barry]

how can a couple of small weights hold up the revs? 

If the bean can is at fault then the obvious part of the answer is the weights are sticking and allowing too much ignition advance.

The more subtle part of the answer requires an understanding of the impact of ignition advance on idle revs.  Ignition timing at idle is 6 Deg before TDC  and at full advance 32 Deg before TDC.

The key to the phenomena is to realise that 6 deg BTDC at idle is a compromise between easy starting and and a smooth idle. The most efficient amount of ignition advance at idle in terms of the power produced when there is no load on the engine is quite a lot more more than 6 Deg BTDC. What this means in practice is if you apply more ignition advance at idle the revs will rise without touching the throttle. 

So if the centrifugal advance mechanism is not doing it's job and allowing too much advance as the revs fall then the engine is going to either hang up at higher revs or the revs will fall slowly.




The impact of ignition advance on idle speed was utilised by some electronic ignition manufacturers to provide an idle speed stabilisation effect.
If you look at this graph the amount of ignition advance actually increases if the revs for some reason fall below the desired idle speed range. This will tend to stabilise the idle speed by producing a little more power to push the revs back up again. Note that the advance curve is flat between 900 RPM and 1100 RPM which would be the normal range of idle speed settings. If the idle speed was set above 1100 RPM then there would be a potential for idle hang up.
 

[tunnelrider]

The more subtle part of the answer requires an understanding of the impact of ignition advance on idle revs.  Ignition timing at idle is 6 Deg before TDC  and at full advance 32 Deg before TDC.

The key to the phenomena is to realise that 6 deg BTDC at idle is a compromise between easy starting and and a smooth idle. The most efficient amount of ignition advance at idle in terms of the power produced when there is no load on the engine is quite a lot more more than 6 Deg BTDC. What this means in practice is if you apply more ignition advance at idle the revs will rise without touching the throttle.

So if the centrifugal advance mechanism is not doing it's job and allowing too much advance as the revs fall then then the engine is going to either hang up at higher revs or the revs will fall slowly
                   

Thanks Barry for your concise explanation of the subtleties of idle and ignition advance. Yes I've suspected this.  I read on an old thread I found by searching the topic, posted by an ex member 'airhead' that he reconditions bean cans and is finding that a lot are coming in, suggesting they have found their service life.  This was a post from 2008!

At least my carbs are clean!

[tunnelrider]

The impact of ignition advance on idle speed was utilised by some electronic ignition manufacturers to provide an idle speed stabilisation effect.
If you look at this graph the amount of ignition advance actually increases if the revs for some reason fall below the desired idle speed range. This will tend to stabilise the idle speed by producing a little more power to push the revs back up again. Note that the advance curve is flat between 900 RPM and 1100 RPM which would be the normal range of idle speed settings. If the idle speed was set above 1100 RPM then there would be a potential for idle hang up.

Yeah that's interesting and the graph, thanks for finding that.  Is the electronic tachometer (rev counter) usually pretty accurate in high km BMW's?

I think I can borrow an external tacho, and test if it's still accurate.  If I recall correctly it works by counting a spinning mark when aimed at it, like a timing light.  Only problem is that I've got three marks on the flywheel (clutch carrier thing) that are painted.  Have you or anyone out there done this and can give me a couple of tips?  I know this won't fix the problem though but I've been a bit curious to how accurate it is 5K+ rpm.  It got knocked once and the upper revs dropped 500rpm on the tacho for a while, has since returned to how it was.

[Barry]

I've not tested a tachometer but it should be reasonably accurate certainly more so than a speedo.   

On the basis of observing road side radar speed readings, if I had to guess I'd say mine is modestly optimistic, perhaps over reading by 2 or 3%.  It will be interesting to see what you find.  How about a spot on the end of the alternator rotor or is the diameter too small for that to work.

[georgesgiralt]

It could be easy to test a tacho if you have a small power outlet giving you 12 V in alternating current. (no so widespread because they are often used to charge batteries, hence direct current).
Your alternating current is say 50 periods per seconds. So it is 50x60 periods per minutes. So 3000 periods per minutes. If your tacho react on negative signal the tacho should read1500 RPM...
If on 60 Hz mains, make the math yourself ;-)
My tacho was suspected to be reading low. So I opened it and renewed the capacitors in it and it returned to a more plausible reading... 35 years of life in a device exposed to hot, cold humid or excessively dry atmosphere is a harsh environment...

[mrclubike]

If you are going to take your bean can completely apart
Make sure you get some new advance springs
they are cheap and you will more than likely  over stretch the old ones getting it apart

But you can just knock the pin out of the drive and pull the guts out and not dissemble any further
clean it up with some electronic parts spray
And then lube the advance mechanism and bushing with some light oil
Put it back together
It still would not hurt to replace the springs

You also need  to check and see what your   advance is
It should be 26 deg (6 minus 32 equals 26)
If it is more than that you should bend the stops in just slightly so it doesn't over advance   and cause spark knock

[Justin B.]

The last one I did had the weights rusted and I had to soak it in penetrating oil for a while...

[tunnelrider]

It still would not hurt to replace the springs

You also need  to check and see what your   advance is
It should be 26 deg (6 minus 32 equals 26)
If it is more than that you should bend the stops in just slightly so it doesn't over advance   and cause spark knock

Thanks Mrclubike yeah I'm just about to order new springs.  I presume checking your advance involves looking at where the advance line is in relation to the centre of the timing hole, once timing has been set at idle?  If so mine is ok.  I remember reading about this in one of Snowbum's fantastically involved pages, which I'll probably trawl through again once the springs turn up and I do the 'bean can job' but my advance was in the centre, as it was at idle.  I'm going to use this link as a guide for dissembling, which should be enough info.
http://gunsmoke.com/motorcycling/r100gs/auto_advance/index.html
Once again, I'll post with the results!  It takes about 10 days for Motobins parts to arrive here.
Thanks all you've been great with your information.

[tunnelrider]

I had time today to pull apart the bean can.  The springs were obviously stretched compared to the new ones (see the photo below comparing a new spring and the old one). Apart from replacing these all I did was take off the weight arms and lube the pivots. I also lubed the main shaft that the Hall sensor rides on. All back together successfully but unfortunately didn't go for a test ride due to the weather!  I set the timing and noticed full advance is not achieved until now 4000rpm, which seems a bit high to me (it was about 3000rpm before).