If the car is idling in park and I slowly and gradually increase the throttle, the engine is smooth and fires evenly until around 900 rpm all the way up to something like 1400 rpm.  In that range, it misses, shakes, and is generally unhappy.  It seems to improve at higher rpms.  If I accelerate through that same trouble range with the vacuum advance disconnected, it does not miss or shake.  The vacuum advance is giving far too much advance at those speeds for an unloaded engine.

I have never known an engine in good tune to miss and shake during any reasonable range or load of operation, or, specifically, when slowly acclerated at idle in park.  So I conclude that something is wrong.  But it does not demonstrate the same problems when driving.  It is generally smooth and responsive.  At high road and engine speeds, say 70mph (3k rpm) it feels like something is off though.

The vacuum pickup on the carb (it's original, autolite 2100) is above the idle screw hole but below the transition holes, so it sees manifold vacuum almost the moment you ease off of the idle stop.  What is so puzzling to me is that all of the pieces of the system seem to be set according to their intended design, yet the actual performance of the system is flawed. 

Any thoughts about what, if anything, is wrong here?

Where have you set the initial advance? Your new distributor will not have come with the same internal settings and springs.

Everyone who suggested that the spark is over-advanced is definitely correct.  I'm just not sure why it's over-advancing.  Let me give some additional detail.

First, this is how I have calibrated my distributor.  I looked up the centrifugal specs in the manual.  It calls for x-y* advance at z rpm.  I have taken these values to be DIST degrees at DIST rpms, so I doubled x,y, and z for each data point.  With the vacuum advance line plugged, I then would rev the engine to the spec'd speed and use my timing light to check for the spec'd advance increase over initial.  I did this mainly for the low-speed spring, because my tach only goes to 2k rpm so I just sort of assumed the high-speed spring was where it needed to be

Then, to set the vacuum advance, I idled the engine as slowly as I could to prevent the increased engine speeds due to forcibly advancing the spark from bringing the centrifugla mechanism into play.  The manual specifies x* advance at y "hg vacuum, and again I doubled x.  Using my hand vacuum pump, I supplied the spec'd vacuum, then looked at the advance.  My advance can is NOT the allen key kind, and as it turns out the spring rate of the supplied spring was too soft and would never produce the spec'd curve, so I obtained several other springs from the hardware store and found one that had almost exactly the desired rate.  The washers move the curve up and down, and ultmately I got all the data points comfortably within their ranges.

I have the initial advance set at 8* by the damper.  Who can say whether it has moved?  I chose the 8* mark partially on assuming the damper hadn't moved and partially on the way the engine ran.  It would run faster at idle if I advanced it more, but I did not want to max out initial advance.

I should also note that at hot idle, there is not enough vacuum to the advance can to move it, but just a touch off of the hot idle stop i get 15"hg.  So, around 1000 rpm, the timing light show over 30* advance--waaaaay too much. 

The way I see it, either my calibration method is wrong or this thing was designed this way--and I'm more inclined to believe the former than the latter.

Oldcarmark, the centrifugal advance is set to an 18/36* maximum, and the vacuum advance maxes out at about 12/24* (dist/crank degrees).

Thanks so much for your help so far everyone!

With a little disassembly - and you've been down to that level already - you can swap the advance mechanism over to the lower setting of the advance, is it a "13L"? That's 13° distributor/26° crank. That plus an initial of 10°-12° will net a maximum mechanical of 36°/38° which is close to correct for one of these stock engines. Mark is right - use the proper springs to get all of the 36°/38° at between 2500/3000 rpm (with vacuum advance disconnected and plugged). The engine should be both run-able and tune-able with these spec's and no vacuum advance.

To get economy you will want some vacuum advance added to the engine at periods of low load - cruise. You will find that with the quicker mechanical curve it works best to restrict the vacuum advance at the crank to something like 7°/8° so the total available at cruise / low load over 3000 rpm won't exceed 43°/46°.

I have found that putting an orfice in the vacuum line to slow the speed at which the vacuum pot makes its additions and subtractions of advance - is extremely beneficial to driveability.  

The 18L/36° slot is provided for use with an engine set up at 0° initial advance. You won't like the way the engine runs with that sort of initial advance - it will have a very hot idle and drink fuel.

I have been studying the specs listed in my manual and in the manual for my sister's 1970 Dodge Polara with a bone-stock, original 318.  Both manufacturers call for a very similar advance curve, and one in which if you add the max advance limits at the crank of each component of total timing, you get a number in excess of 60 degrees at the crank.  When I observed the advance behavior of the Polara, though, I noticed that I was able to increase engine speed a couple hundred rpms before a vacuum signal to the vacuum advance ever occurred.

Further study of that manual showed conclusively that the vacuum advance pickup port is ABOVE the throttle plates when they are resting at idle position.

Is this supposed to be the case for the autolite 2100 as well?  Because I have a low vacuum signal even with the engine idling at 600 rpm in park.  The mixture screws are happiest 3/4 turns from lightly seated.  I am wondering whether the throttle plates might be misaligned in the throttle shafts.  I have messed with them so it is certainly possible, though I'm sure I tried to seat them properly before.

The way things are now, the vacuum pickup port in the carb throat is approximately even with the throttle blades at idle.  Looking into the carb, I can clearly see the uppermost of the two idle transition holes above the throttle blades, but I cannot see the lower transition hole nor the advance pickup port.  Yet, as is, idle speed is just 600 rpm.  I can't shut the throttle plates much more without stalling the engine.

To anyone with an autolite 2100: where are your mixture screws set?  Are both idle transition holes visible above the throttle blades, engine idling?

Mark, can you give me the link for those springs you got .

Steve, Can you explain the orifice? can it be used to lower the vacuum?

AT this point I have set mine at [ with no vac] 13* initial , 28.5* @ 2750 RPM, 35* @ 3300RPM.

with vac 51* @ 2930RPM, 53.5 @ 3280 RPM . I have no pinging. seems I need a bit more work

Greenbird.. CAn you give us the rundown on the orifice ...? How what and where...

After you unscrew the hex cover from the can -there are shims that preload the spring. Adding or removing shims changes the advance travel position versus vacuum applied. If the spring preload gets up to a point where the vacuum can no longer overcome the spring sufficiently to reach the mechanical stop - then it would also have the effect of limiting the total vacuum advance available.

The fiber tube is factory sized to limit the travel of the unit at a particular amount. By installing a longer tube for instance, the total travel would be limited. Both of these items are subject to experimentation.

To run "hard lines" on my outfit, I used brake tubing formed to fit on the manifold - roughly run like original. Its attached to the carb and the distributor with rubber tubing at both ends. At the distributor end - I inserted an electrical "butt splice" (red plastic cover) - and crimped it until the system behaved the way I wanted it to. Others have used a solid piece of brass/plastic/steel and drilled holes to get it to behave properly - try a 1/16 hole and see what that does for you. When mine was working to suit me - I pushed the rubber over the butt splice enough to make it invisible.

Anyway, I was wondering, why do we want all of our mechanical advance to come in by 3500 rpm?  If one desired to take the engine on up to 4 or 5k rpm, wouldn't the optimum spark time at those speeds be earlier than that of 3500 rpm?

Is there a maximum limit to advance after which more is detrimental, even as engine speeds continue to increase?

The "early" advance curves we have described are used to take advantage of the additional torque available - low in the rpm band - when the spark advance is optimized toward building maximum cylinder pressure. It takes about one millisecond to light the fire - as the revs go up the spark initiation point has to get further and further ahead - but you are also moving toward less and less compression of the mixture (away from top dead center on the front end, the piston is rising up the bore). A limit is reached where "enough is enough" - and you can't get any more torque advantage out of the quicker spark curve. The old "Y" cylinder head combustion chamber design is not the best - flame travel isn't like a GM "fastburn" by any means - its 50 year old technology.

This low end torque advantage - you will feel it in the seat of your pants - and it could very well increase your fuel economy because the engine will accelerate the vehicle without as much throttle opening. Thunderbirds like mine are pre-programed from the factory to take off in second gear of the Ford-O-Matic (low is basically a manual selection in ordinary driving). Your two speed might have a better low than I've got - in second. One of these T-birds in poor tune stumbles away from a traffic light trying to transition from idle under heavy load - it's tuned right "or else". When I changed from the "Load-O-Matic" to a '59 distributor (using a manifold vacuum source) - it was like taking 500 pounds out of the trunk. Uses way less fuel too.

Vacuum advance has the effect of leaning out the engine when vacuum is high - low load, slight throttle opening - no call for power. Even in this state - there is a limit to how effective additional advance might be. If the engine has been fitted with a more optimal mechanical advance curve (a higher number of degrees at a slower speed) then less vacuum advance is permissible - to keep the total within bounds. This leaning effect gives better economy - and when the throttle opens a little or a lot you are going back down toward the mechanical limit. About the only way you can get the maximum total spark advance in most vehicles - everything summed (initial+centrifugal+vacuum) - is a high speed downhill in the mountains - with the throttle closed - using the engine as brake torque. You aren't making power in that case - you are using it.

Hope this helps out a bit with the "logic" of the systems as they are applied to our street engines.

STEVE, again your knowledge of all things electrical was explained so well in this quote, very interesting and informative. Best regards bill.

With the distributor set as is, vacuum advance disconnected, it gets unhappy around 1500 rpm but clears up as engine speed increases.

When attempting to set the mechanical side, will road tests with vacuum advance plugged show me whether I'm getting closer to where I want?  I mean, should I make the car run and drive the best I can with the vacuum advance capped off first, then deal with the vacuum side when that's done?

I bought the Mr. Gasket kit 925d.  It contained two identical light springs, a bit lighter than the primary spring I was using to begin with.  I ended up using one of these new springs in the primary position and the original secondary spring.  Any other combo brought all the advance in too early, resulting in spark knock.

I have my vacuum advance limited to 6 degrees.  Even that much is enough to produce some irregularity revving the engine in park in the 950-1400 rpm range, but it is better than it used to be and I notice no negative side effects driving.  The car is quite responsive even without the vacuum advance, and WOT does not incur any spark knock.  It's still not what I'd call a fast car, but it's faster than it was, which is nice when you need it.  I don't know how fuel economy has been affected yet.

Anyway, I'm gonna take it as it is!  Thanks once more for all the help and wisdom along the way.