mgs
[Top] [All Lists]

Musings on tuning

To: mgs@autox.team.net
Subject: Musings on tuning
From: "A. B. Bonds" <ab@vuse.vanderbilt.edu>
Date: 10 Jun 1996 13:38:57 -0500
I have finally gotten some cooperation from the MGA I have been
battling for several months, and it appears to be running pretty well.
(Note that discussion of British iron in absolute terms is hazardous
to one's credibility.)  In the process a couple of things that have
been bouncing around the net popped up again, so I thought I'd give these
older threads a tug or two.

(1) Timing.  Static timing must be a wonderful thing, since that is
the only spec available for an MGA, but it is still fraught with
indeterminacy.  On this example (a '58 with the Lots-o-miles feature)
the centrifugal advance was sloppy.  Very sloppy.  Even though I dialled
in the timing mark most carefully and used a meter with a beeper to
determine the orientation at which the points opened, the resulting
setting was about 15 degs off _at the distributor_ which translates to
30 degs off at the crank due to gearing.  Car barely ran.  A good wank
on the distributor (ahem, that's grasp it and _rotate_) evened the
idle, but of course all reference to the "7 degs BTDC" was destroyed.

I ended up dialing the idle down as low as possible (about 600 rpm)
and set the timing with a light to 7 degs BTDC (vacuum advanced
disconnected and plugged), under the assumption that the only advance
that should be present at that speed would be due to the slack in the
distributor.  The slack in the dizzie was not really evident in jitter
of the timing marks under the strobe. This yielded about 15 degs at 
1500 rpm, which is not far off from where I run my B.  Seems to run
fine now. This is one vote; I'd be interested in knowing if any 
others have checked their A timing dynamically. 

(2)  Mixture.  There seems to be a paradox here.  Even though we often
hear that "one can't be too careful", I think in this case, one can.
With healthy carbies, the tried-and-true "twist the nut until it pinks
(!!)" is actually a pretty sensitive method.  I found that good idle
was achievable only over a range of about 4 flats (2/3 rotation) for
either carb.  The smooth idle also came when you get a brief increase
of speed, then a decrease as you lift either piston a smidgen (1/32"
or so).  Further lifting caused stumbling.  Car runs very well with
this setting.  Here's the paradox.  By two other methods, it's too
rich.  The colortune shows a healthy yellow at idle, and my new/old
fancy AC tailpipe sniffer (hadda brag about it) shows about a 12:1
ratio, also too rich.  Moreover, the plugs are blackish, although
there is a hint of tan on the electrode.  When I gun up to about 3000
(slowly) the colortune shows deep blue (heading toward lean) and the
CO analyzer says about 14:1.  If I set the idle mixture _correctly_
according to these measures (blue, 14:1 at idle) the engine coughs
more than the audience at a Beethoven concert in Nashville.  I would
also note that setting the idle mixture by the colortune works very
well on my B, which has a Weber DGV.

So what's going on?  It just looks like that this engine wants to run
a bit rich at idle, although it seems lean enough at speed.  One could
argue that they are using the wrong needle (checked, these are correct
according to the spec) but if you set the engine to the "appropriate"
leanness (leanity?) by objective measures, the engine wheezes and
stumbles.  So it's the right needle after all to get the engine to go,
which is, I believe, the objective.

Comments?
                                A. B. Bonds


<Prev in Thread] Current Thread [Next in Thread>
  • Musings on tuning, A. B. Bonds <=