Wandering AFR at idle, bad transition on throttle tip in

[60sIron]

I am trying to tune this Holley 750 converted 4150 with a vacuum secondary and I'm totally lost.

Engine is 449 (428 .035over + 4.125" stroke).  Hydraulic roller cam .556" lift  236° duration @ .050".  Single plane Port-o-sonic intake.  C6 Automatic trans

Warmed up it idles at 700 RPM with about 11 inches of vacuum with 12-12.5 AFR.  Setting the idle mixture screws can take it down to 11 AFR and up to 13 AFR without affecting the idle or the vacuum too much.

On the highway it was running about 10.5 AFR @ 2200 RPM and about 20 inches of vacuum.  Stinky and fat, so I went from a #72 primary jet to all the way down to a #68 primary jet to get the cruise AFR up in the 12-13 range. 

Now it dies at low speed when tipping in the throttle to go up a hill.  At 1200 RPM I approach a hill at 30 MPH, then tip in the gas, I see AFR drop to about 10 when the pump shot hits, then in a second or two that burns off and AFR shoots up to 15-16 and it starts to die.  I can nurse it up the hill going on and off the gas to keep pumping it, or I can floor it and downshift which bumps the RPM and draws more fuel out of the primaries.  With either one of these techniques I can keep the AFR below 14 and the engine running.

This is where I'm stuck.  I want to add more fuel at low RPM under load without adding more fuel at higher RPM which is what happens when I increase the primary jet size.  From what I've read, this 'transition' from the idle circuit to the primary circuit is controlled by "air bleeds", but how do I change those?  I could slow down the pump shot, but I don't think I could make that last all the way up the hill.  That doesn't seem like the right way to do this.

Another thing I've noticed is the idle AFR drifts upward as the engine gets hot.  I tune it fully warmed up to 12.5 AFR.  Then I get it really hot after say 10 miles on the highway and the AFR has dropped to the low 11s.  Is this a float problem, or a fuel pressure issue or a defective carb?

Please help.  Carbs are a mystery to me.  What are my options?

[blykins]

Quick question first....where's your initial timing set?

[60sIron]

With the vacuum advance disconnected idling at 700 RPM it is at 13° BTDC

[60sIron]

The fuel pump is a stock mechanical unit with the integrated fuel filter

[blykins]

Total mechanical?

[Drew Pojedinec]

Brent has her on the run..... without appropriate timing, it'll never run correctly.
A/F ratio is massively impacted by ignition timing.

Not saying your carb is perfect, just saying you shouldn't touch it until you have the timing in a reasonable place.... 13 degrees initial for your engine/cam/etc is kinda 5-6degrees retarded by my thinking.

[GerryP]

Is this a 3310 or variant?  Did you buy it new?  Other than the secondary metering block and the primary jet, have you done any more fiddling with it?  If it's a used or older carburetor, is it really clean?  Does it have the proper gaskets?

A 68 primary jet is too small for a 750.  Go back to the 72.  Put the AFR meter away.  It's not helping you.  Tune the carburetor to make the engine run good regardless of what the meter says.  Once everything is well, hook up your meter to see what the engine likes, write that down, and put the meter back in the box.

Forget this being a bleed problem.  Unless they are replaceable and someone fiddled with the carburetor before you got it, the bleeds should be fine to keep you on the road.  Not perfect, but not the source of your problem.

Check your primary and secondary throttle blades to make sure you are not over-exposing the transfer slot.  What can happen in achieving your 700rpm idle is that the secondary throttle could be completely closed, and the primary open too far and exposing too much of the transfer slot.  That detail will make sure your idle and tip-in response is working correctly.

Just so you know, if you have a really stinky exhaust...you know, eyes watering, runs you out of the garage...that is raw fuel -hydrocarbons.  You get that from a lean condition where the fuel doesn't fully combust or you get a lean misfire.  It can also be an ignition misfire, but in your case, you are running too lean.  A rich mixture doesn't have that smell.  And other than really poor fuel economy and short plug life, with a rich mixture the engine will start fine and run okay.  Not perfect, but it won't be giving you a lot of problems.  This is why it's better to have an engine on the rich side, than lean.  Particularly at idle.

[1964Fastback]

Would the power valve be something to look at, for the A/F spike going up hill?  Not saying it is, just asking.

Pat

[blykins]

Yep.  I was trying to see how far he could advance it before needing a recurve, but that's generally where I head too.  About 18 degrees initial.  Increasing the timing will allow the idle speed to be lowered, which will make more vacuum, then you can start tuning.

[60sIron]

I can't really see the degree marks with the alternator bracket in the way so I guessed wrong with the 13° BTDC.  When I did the mechanical advance measurement and I saw where 20° came in, I'd say at idle it is around 16° or 17°.  I had retarded it a little from the initial setting because sometimes it doesn't like to crank over when it is hot, but I could probably put those degrees back into it.

At 3000 RPM total advance is 42° with the vacuum advance disconnected.  I've never curved this distributor.  It is a FoMoCo piece with a Pertronix.  The mechanical advance is not stock, but I'm not familiar with it.

[60sIron]

The carb is a Holley 1367 I bought from Summit a couple years ago.  When I started with the tuning I added the secondary metering block and put a removable top on the vacuum secondary pot.

Based on the idle vacuum I changed the power valve to a standard flow 5.5.

The throttle tip in problem is happening before the secondaries kick in or the PV opens up.

[60sIron]

Sorry that is the date code, I guess it was made in 2017

The model 80508-6

[pbf777]

Put the AFR meter away.  It's not helping you.  Tune the carburetor to make the engine run good regardless of what the meter says.  Once everything is well, hook up your meter to see what the engine likes, write that down, and put the meter back in the box.

    +1       

     Scott.

[60sIron]

If I knew how to tune the carb to make the engine run right I wouldn't need the oxygen sensor.

[Joe-JDC]

What is the carb number on the choke horn?  Body number?  On Holley carbs, everyone thinks adjusting the idle screws or the throttle blade screw is adjusting the idle.  WRONG.  Start by emptying the fuel, turn the carb over and release the throttle blades adjusting screws so that they close completely.  Then open the secondary butterflies until you see the idle transfer slot uncovered ~.015-.020" with the little screw that controls the opening.  Turn the screws on the side of the metering block in gently until they seat, then back them out 1 complete turn to 1 1/4 turn.  The accelerator pump arm should have .015" clearance with the throttle blades at WOT(wide open throttle).  This should get you a clean idle if all the carb circuits are back to stock and clean.  Idle speed is adjusted by that slot in the SECONDARY throttle blades, if the engine is revving too fast, then lower the secondary butterflies a couple of thousandths.  Don't open the primary until you have the idle set.  Just open the primary enough so that it doesn't stick when the throttle is released by letting your foot up off the pedal.  A .028" primary shooter should work for a longer shot on acceleration, but if it backfires, then a .031" shooter can be installed.  Now you can check the O2 for jet changes.  Joe-JDC

[60sIron]

The number on the air horn is 80508-6

The number on the body is 8051B

I set the initial timing to 18° BTDC, the idle increased but no change in vacuum.  More advance degrades the idle.

Then I pulled off the carb to follow the advice of Joe-JDC.

1) I turned the set screw on the secondary butterfly stop until the first slots I saw were uncovered

Will this work if the secondary metering block I have does not have any idle set screws?

2) When measuring the accelerator pump arm clearance at WOT, the arm is spring loaded and is in constant contact with the throttle mechanism.  Should I depress the accel. pump arm to the full stroke and then measure the gap with the throttle mechanism?

3) The shooter on the carb is .035".  I will order an 0.28" and .031" so I can try them.

[blykins]

The number on the air horn is 80508-6

The number on the body is 8051B

I set the initial timing to 18° BTDC, the idle increased but no change in vacuum.  More advance degrades the idle.

Then I pulled off the carb to follow the advice of Joe-JDC.

1) I turned the set screw on the secondary butterfly stop until the first slots I saw were uncovered

Will this work if the secondary metering block I have does not have any idle set screws?

2) When measuring the accelerator pump arm clearance at WOT, the arm is spring loaded and is in constant contact with the throttle mechanism.  Should I depress the accel. pump arm to the full stroke and then measure the gap with the throttle mechanism?

3) The shooter on the carb is .035".  I will order an 0.28" and .031" so I can try them.

The vacuum should go up when you lower the idle speed back down to where it was before you advanced the timing.  What you’re doing is raising the idle with the timing but closing the throttle blade so that you get more vacuum.

[CaptCobrajet]

Is this a 3310 or variant?  Did you buy it new?  Other than the secondary metering block and the primary jet, have you done any more fiddling with it?  If it's a used or older carburetor, is it really clean?  Does it have the proper gaskets?

A 68 primary jet is too small for a 750.  Go back to the 72.  Put the AFR meter away.  It's not helping you.  Tune the carburetor to make the engine run good regardless of what the meter says.  Once everything is well, hook up your meter to see what the engine likes, write that down, and put the meter back in the box.

Forget this being a bleed problem.  Unless they are replaceable and someone fiddled with the carburetor before you got it, the bleeds should be fine to keep you on the road.  Not perfect, but not the source of your problem.

Check your primary and secondary throttle blades to make sure you are not over-exposing the transfer slot.  What can happen in achieving your 700rpm idle is that the secondary throttle could be completely closed, and the primary open too far and exposing too much of the transfer slot.  That detail will make sure your idle and tip-in response is working correctly.

Just so you know, if you have a really stinky exhaust...you know, eyes watering, runs you out of the garage...that is raw fuel -hydrocarbons.  You get that from a lean condition where the fuel doesn't fully combust or you get a lean misfire.  It can also be an ignition misfire, but in your case, you are running too lean.  A rich mixture doesn't have that smell.  And other than really poor fuel economy and short plug life, with a rich mixture the engine will start fine and run okay.  Not perfect, but it won't be giving you a lot of problems.  This is why it's better to have an engine on the rich side, than lean.  Particularly at idle.

+2.........You have to give it what it wants.  Chasing "optimal" numbers can lead you to a place where the engine doesn't really want to be.  Gerry is right-on.  Make it run right, then if you want to put the meter on it and document the numbers for future info, it could help you later.

Often a tweak of the rear blades will allow you to get the idle up without exposing too much primary slot.

Sounds like you could use a tighter advance curve also.  Maybe a 16-18° mechanical and then add your vac pod for keeping it cool and clean at no and part throttle.

[Barry_R]

x3 (or 4 or 5..)

The O2 meter is a useful tool for WOT and part throttle cruise, but it's a trouble maker for idle and tip in tuning.  An O2 sensor reads O2 - NOT an air fuel ratio.  An air/fuel ratio that is too far off the scale for any particular reason will cause a misfire, which in turn will toss unburned fuel into the exhaust, and the O2 sensor will give you garbage numbers.  I have had a couple of the self learning EFI systems that would tune themselves into oblivion because the basic tuning maps were too far away from a usable package for the engine.

In addition, some older o2 systems did not have enough sample rate resolution to get a useful reading at low exhaust speeds - you might be reading pulses or reversion exhaust.

For idle and low speed you feed the beast what it wants to eat.

[60sIron]

The engine didn't "feel" right, like it was "unhappy" before I installed the wideband O2.

It has a lot of vacuum oscillation at idle, like 5 inches peak to trough, which is a lot considering that the peak is about 13 inches and the trough is like 8.  The needle bounces all over the place and so far, setting initial timing doesn't seem to help.  While the car ran OK without backfires or stalling, fuel economy was less than I thought it should be, it certainly smelled very rich, and throttle response seemed lethargic.  At the time I was just running an out-of-the-box Holley 4160, so I thought there was room for improvement.

Perhaps a combination of this cam and a single plane intake manifold is just always going to run like this.  I hope not.

On the dyno, with a different carb and a different distributor AFRs ranged from 12.2 - 13.6 during the full pull.  I'm not saying this is the magic range I'm trying to achieve, but it seems like it is at least possible to control the fuel curve on this engine.  Also, I'm not trying to wring out every last bit of power from this engine, I think it was actually running pretty bad and there were significant improvements possible.

So, this year I decided to try and tune it.  I installed the AFR and just drove it around for a while and see what was going on.  It was consistently rich, running from 10-12.5.  On the highway at 2200 RPM cruising at steady speed on level ground AFR was about 10.1 according to the gauge.   From what I understand about the process, after setting the idle mixture for peak vacuum, the next step is to tune the primaries by running at steady cruise just under the RPM where the secondaries start kicking in.  The first sign of trouble was it took a lot of jetting to get steady state cruise in the window, from #72 down to #67.  Then instead of predictable AFRs, they are erratic; swinging between 12 and 16 with very small changes in throttle and RPM.  So I've gone from sub-optimal to very unhappy.  I did not expect to encounter this kind of trouble.  I haven't even started trying to fiddle with the power valve or the secondaries.

Right now I am following Joe's advice.  I've cracked open the secondary throttle set point and closed down the primaries below the transition slots.  Right now it won't stay running without me on the throttle.  It might idle at 650, but it will definitely not idle at 500.  I am going to pull the carb back off the engine and open the secondaries some more.  Is there any fuel coming through the secondaries at idle?  I do not have idle screws in my secondary metering block.

Yes, maybe a reduction in mechanical advance curve would help.  Right now I'm still trying to tune drivability between 1000-2000 RPM, so total advance isn't a factor yet.  Also distributors are as much a mystery to me as carbs.  I wish I was just changing numbers on a table, but this is what I'm working with.

[My427stang]

The odd thing to me is the bouncing needle, that is generally a valve or in some cases an intake leak

However, it seems like you are closing in. One thing, in my experience an 805xx series of carbs generally have a high IFR and some of those have hot metering problems, potentially due to the IFR but also from goofy main circuit emulsion and kill bleeds. I don’t like them because transition can be sloppy

I would get a timing curve right first though, if wrong, the engine will be happier, then you can finish tuning

[Falcon67]

Vacuum oscillation at idle is a sign something else is out of whack. 

Some hints - https://www.onallcylinders.com/2015/05/08/quick-tech-how-to-read-a-vacuum-gauge-to-pinpoint-engine-problems/

Shooters - order 3 or 4 .028 shooters and get a pin drill set.  Saves $$$ buying a bunch of shooter numbers, just  make your own.  Once you get the car to idle better, try your .031.  If it needs more shot, take one of the .028s and drill it to .033.  Then .035, and so on.  Keep going until you have it right.  You can always use the pin drills to measure when you forget where you ended up LOL. 

"Smells rich" - that's lean idle misfire, not rich.  You smell fuel because the engine isn't burning it correctly, not necessarily because it has too much. 

JoeDs little dissertation ought to be screen printed on every Holley choke plate LOL. 

The more bast timing you can run, the better you can control idle and drive-off performance.  On a stocker igntion, there will be some limits due to the engine pushing back against the starter.  Good time to move to a PMGR start if you don't use one already.  But anywhere in the 16~20 degree range should be doable - BUT - you need to be sure of true TDC vs your balancer marks so you accurately know what the timing is.  Worry about total later, get initial where it runs best. 

[GerryP]

You are probably not aware of this, but single plane intakes almost always want more jet.  Again, you are moving in the wrong direction.  My opinion.  Put a 74 or 76 primary jet in it and see how that does.  At least put the stock 72 jet back in it.

So, is the vacuum gauge needle bouncing real fast, like a regular pulse, or is it more of a drift where it moves from one peak to the next over a period of a second or more?  A fast bounce is usually cylinder pressure moving in the intake from a bad valve seal.  A drift is from a maladjusted carburetor.

[60sIron]

The needle on the vacuum gauge bounces really fast, like it is literally a blur.  This is a brand new engine with maybe 1000 miles on it since it was built. New heads, new valves, new springs, new lifters. What would cause it to have a bad valve seal?  If that were the case, how would I find it?  A cylinder leakdown test?  These are hydraulic lifters, could they be preventing an intake valve from closing all the way?

[60sIron]

The thing is, if there is something fundamentally flawed with the engine in the ignition system or the valve train that is causing it to run rich (or misfire) with #72 jets, then #74 jets are not going to fix it right?  Ignoring the O2 data, given the symptoms, where would you start to try and get this thing running right?  Would it be jetting or something else?

[GerryP]

I think you have found your problem.  The valves are not fully seating.  You will never get proper carburetor metering when you have a bouncing vacuum gauge.  The carburetor is greatly confused.  Do a compression test first.  If the results are poor, then you can do a leakdown if you need a number for data accumulation, but you can also just put air in the cylinder to see if the air leaks into the intake or exhaust or into the crankcase.

As to what is causing the issue, that's hard to say.  It could be the guides are too tight, the seat margins are too narrow, the valve job isn't concentric, the lifter preload is incorrect.

[My427stang]

I think a leakdown test would be beneficial.  You could have a bad valve job, leakdown test would show if one or two aren't sealing well.

If you find one leaking, back off the rockers to make sure not a tight valve.  I think that needs to be done before anything else

I lost track of the jet discussion though. 

- Jets and power valve fuel metering (let and PCVR) have nothing at all for idle, and little for transition themselves if just talking jet sizes, the circuit is dead because there is no airflow over the boosters, at the tail end of transition, the combination of fuel and timing from the various bleeds get the booster started.  However, a change from a 72 to a 76 wouldn't change much other than richen when it got going

- Idle a/f screws adjust a frothy air/fuel mix out of the hole beneath the butterflies in front, if no adjusters, the rear still has the hole, but they are not adjustable.  Both feed idle fuel and air mixed from the low speed air bleed

- Transition fuel is from a similar source as the idle fuel, the IFR, but bypasses the a/f screws and feeds the vertical slot at the butterflies, those need to show a little to put a load on them so when the plates open, aerated fuel comes out.  That fuel along with a pump shot if moving fast, provides fuel until the boosters (mains and PV) can get their aerosol going

If you have a transition problem after the plates are set up correctly, the low speed air bleeds adjust timing of the transition fuel and the IFR provides quantity.   However, I haven't found any of the 805xx carbs to be any good without a metering block swap.  The good thing is, going to an earlier low IFR block, drilled to stock IFR, Jet, and PCVR numbers for the newer carb, is usually a bolt on fix.  Although I would expect someone like Drew could match the main circuit on the existing metering block

[frnkeore]

Does your vacuum fluctuation get better at 2000-2500, no load?

Another quick test you can do, is set the idle to what ever rpm that you can pull a plug wire w/o dying, then pull them one at a time, to see if it's one cylinder related. Note the rpm drop, per cyl.

If it has back fired, the power valve may be blowned, too.

[60sIron]

I don't have a leakdown tester, but I pulled the plugs and did a compression test.

#1  185
#2  190
#3  185
#4  180
#5  185
#6  190
#7  190
#8  190

Based on these numbers, I think I can rule out a leaky intake valve as the source of the oscillating vacuum, unless it is possible for the lifter to hold the valve open, but nobody has said that.

All of the plugs looked pretty much the same, black and sooty.  Some were a little cleaner, #4 seemed to have a bit more hard carbon.  It does not have a steady miss on one cylinder, I know what that sounds like, so I don't know what the painful process of pulling a plug wire on a running engine would prove.

[Falcon67]

Blown power valves are talked about like an everyday thing, in 40 years I can't recall ever holding one.  Yes, I have put the back side of plenty of PV to my lips and gave 'em a "kiss", yep that one's OK.  Leaking PV gaskets - yes.  All the friggen time.  Worse with 10% pump gas.  Also, if the carb is anything "new", like last 10~15~20 years, it will have a check valve in the base plate to protect the PV from a backfire. 

If you want to check the power valve, simple test - run the engine, stop.  Carefully remove carb without tripping the pump shooters,wipe the base plate with a rag and set the carb on a coffee can or similar - quart sauce pan, etc.  Let it sit for a while and check the base plate around the vacuum hole that leads to the PV chamber.  Fuel wetness or drips into the pan = leak from something, lots of time the gasket can be a bit off center and not sealing correctly.  This does not always condemn the PV, the metering block to main body gasket can cause all kinds of trouble if not sealing well.

Also, on a normal engine if you continue to open the idle mix screws, you will see a steady drop in the vacuum reading, not a bunch of jumping around.  So a leaking PV will cause a low vacuum reading, relative to whatever is the normal level.  Same with lean until you get to some point where its too lean to idle well. 

Quick vibration of the vacuum needle screams valve seats, guide issues,  etc.  Yes, the carb can't "think" with all that going on.  If it smooths out at increased RPM, for sure it's in the valve train.  Maybe a cam lobe, that might be a stretch.  As an example, on ly dual 600 t-rammed 357C with the good size hydro roller, 1200 RPM idle, 37 degrees of timing runs right at 12 in/hg steady when everything is close to right on the idle mix, plate positions and such.   

Here's another good list of suspects I ran across on the web:
WHAT TO LOOK FOR,
(paraphrased from a rebuild manual)
1. A steady reading between 16 and 22 in.Hg. at idle...
This is normal for most vehicles.
Radically cammed engines have lower, less steady readings.

2. Normal range at idle, with sporadic drops below normal,
This could indicate a sticking valve.

3. Normal range at idle with needle vibration of about 2 in.Hg.,
This could indicate an ignition problem. Check plug gap, dwell (yes, electronic ignitions have dwell too), cap, rotor, and plug wires.

4. A steady reading slightly higher than normal.
This can be caused by a dirty air filter, or overly advanced ignition timing.

5. A steady reading 3 to 12 in.Hg. lower than normal.
This could indicate one or more of the following conditions,
Intake or carburetor vacuum leak, Late ignition or cam timing, Worn piston rings.

6. Gauge needle drifts slowly over a range of 4 to 5 in.Hg. at idle.
This could indicate an idle mixture that is too rich or too lean.

7. Gauge needle fluctuates rapidly between 10 and 21 in.Hg. at idle.
This could occur when one or more valve springs are weak or broken.

8. Gauge reads normal at idle, but drops slowly as engine speed is increased to 2,500 RPM.
This could indicate a restricted exhaust system.

9. Gauge reads below normal and fluctuates rapidly over a range of about 3 in.Hg. at idle, then the needle becomes steady as engine speed is increased.
Worn intake guides usually cause this reading.

[Falcon67]

I don't have a leakdown tester, but I pulled the plugs and did a compression test.

#1  185
#2  190
#3  185
#4  180
#5  185
#6  190
#7  190
#8  190

Based on these numbers, I think I can rule out a leaky intake valve as the source of the oscillating vacuum, unless it is possible for the lifter to hold the valve open, but nobody has said that.

All of the plugs looked pretty much the same, black and sooty.  Some were a little cleaner, #4 seemed to have a bit more hard carbon.  It does not have a steady miss on one cylinder, I know what that sounds like, so I don't know what the painful process of pulling a plug wire on a running engine would prove.

Hmmm - I just fixed the 393C over the winter for a leaking head gasket.  Leaked enough that once over 160F the engine was pushing water out like crazy.  Like a old camp coffee pot percolation.  All my compression readings were within 15 PSI and leak down didn't show anything out of ordinary.  #5 plug looked clean, even for methanol.  So at least you have really good readings. 

Black and sooty could be from the carb having fits from the crazy vacuum readings.  Or some combination of fuel metering issues

[frnkeore]

The reason for pulling plug wires, is to see if a particular cylinder is effecting how it runs, not if it has compression loss.

It's not painful to do. You just need a plug wire puller, they are well insulated and you get no shock. Burned maybe if you touch the head or exhaust. Wear gloves

If it is a particular valve guild, it may show. You can also pull a valve cover and while running, if you have PC type valve seals, a squirt oil on the intake guilds may show a change in idle if, it's guild related.

[60sIron]

Blown power valves are talked about like an everyday thing, in 40 years I can't recall ever holding one.  Yes, I have put the back side of plenty of PV to my lips and gave 'em a "kiss", yep that one's OK.  Leaking PV gaskets - yes.  All the friggen time.  Worse with 10% pump gas.  Also, if the carb is anything "new", like last 10~15~20 years, it will have a check valve in the base plate to protect the PV from a backfire. 

If you want to check the power valve, simple test - run the engine, stop.  Carefully remove carb without tripping the pump shooters,wipe the base plate with a rag and set the carb on a coffee can or similar - quart sauce pan, etc.  Let it sit for a while and check the base plate around the vacuum hole that leads to the PV chamber.  Fuel wetness or drips into the pan = leak from something, lots of time the gasket can be a bit off center and not sealing correctly.  This does not always condemn the PV, the metering block to main body gasket can cause all kinds of trouble if not sealing well.

Also, on a normal engine if you continue to open the idle mix screws, you will see a steady drop in the vacuum reading, not a bunch of jumping around.  So a leaking PV will cause a low vacuum reading, relative to whatever is the normal level.  Same with lean until you get to some point where its too lean to idle well. 

Quick vibration of the vacuum needle screams valve seats, guide issues,  etc.  Yes, the carb can't "think" with all that going on.  If it smooths out at increased RPM, for sure it's in the valve train.  Maybe a cam lobe, that might be a stretch.  As an example, on ly dual 600 t-rammed 357C with the good size hydro roller, 1200 RPM idle, 37 degrees of timing runs right at 12 in/hg steady when everything is close to right on the idle mix, plate positions and such.   

Here's another good list of suspects I ran across on the web:
WHAT TO LOOK FOR,
(paraphrased from a rebuild manual)
1. A steady reading between 16 and 22 in.Hg. at idle...
This is normal for most vehicles.
Radically cammed engines have lower, less steady readings.

2. Normal range at idle, with sporadic drops below normal,
This could indicate a sticking valve.

3. Normal range at idle with needle vibration of about 2 in.Hg.,
This could indicate an ignition problem. Check plug gap, dwell (yes, electronic ignitions have dwell too), cap, rotor, and plug wires.

4. A steady reading slightly higher than normal.
This can be caused by a dirty air filter, or overly advanced ignition timing.

5. A steady reading 3 to 12 in.Hg. lower than normal.
This could indicate one or more of the following conditions,
Intake or carburetor vacuum leak, Late ignition or cam timing, Worn piston rings.

6. Gauge needle drifts slowly over a range of 4 to 5 in.Hg. at idle.
This could indicate an idle mixture that is too rich or too lean.

7. Gauge needle fluctuates rapidly between 10 and 21 in.Hg. at idle.
This could occur when one or more valve springs are weak or broken.

8. Gauge reads normal at idle, but drops slowly as engine speed is increased to 2,500 RPM.
This could indicate a restricted exhaust system.

9. Gauge reads below normal and fluctuates rapidly over a range of about 3 in.Hg. at idle, then the needle becomes steady as engine speed is increased.
Worn intake guides usually cause this reading.

#7 Is the closest to my condition.  I can pull the valve covers and look.  I know what a broken valve spring looks like, but how do I find a weak valve spring?

[cjshaker]

The reason for pulling plug wires, is to see if a particular cylinder is effecting how it runs, not if it has compression loss.

It's not painful to do. You just need a plug wire puller, they are well insulated and you get no shock. Burned maybe if you touch the head or exhaust. Wear gloves

Just a suggestion, but you can generally eliminate getting shocked by pulling the wire at the cap. When pulling it at the plug, all that energy at the end of the wire is looking for an escape, and the holder is that escape. Pull it at the cap and it has nowhere to go.

[Falcon67]

If this is a more standard FOMOCO with a Pertronix points eliminator, it won't bite that bad at all - if any.  Try my MSD 7AL-3 which hits the coil with 525V before firing.  That'll defib your whole being. 

[BigBlueIron]

What idle RPM are you targeting? I have ran into when trying to idle an engine very low the crazy bouncing/vibrating vacuum needle can appear at idle, even though it runs ok. 

[60sIron]

I'm not married to any particular idle RPM.  I've got a higher stall converter, so 700 works fine.  Lower would be better, but I'm not that concerned about it.

[60sIron]

Update:

I pulled the valve covers and did not find any broken springs.  These are triple coil so it is progressively harder to see if the inner coils are broken, but I'm pretty sure they are good.

I did find evidence of five of the pushrods rubbing on the intake.  I've attached a picture of the worst one.  So this is something that needs to be fixed.  These are oil through pushrods for the T&D rockers.  Are there more slender pushrods I could use, or do I need to pull the intake and grind out the offending holes?

[cammerfe]

Blown power valves are talked about like an everyday thing, in 40 years I can't recall ever holding one.  Yes, I have put the back side of plenty of PV to my lips and gave 'em a "kiss", yep that one's OK.  Leaking PV gaskets - yes.  All the friggen time.  Worse with 10% pump gas.  Also, if the carb is anything "new", like last 10~15~20 years, it will have a check valve in the base plate to protect the PV from a backfire.

Just a 'by-the-way', I suffered a blown 'rubber' in the power valve on the primary carb on my just-delivered '64 427 Custom. The car was two or three days old when I took it back to the dealer because it wasn't acting right. That's more than fifty years ago, but it DID happen.
 

KS

[1968galaxie]

Update:

I pulled the valve covers and did not find any broken springs.  These are triple coil so it is progressively harder to see if the inner coils are broken, but I'm pretty sure they are good.

I did find evidence of five of the pushrods rubbing on the intake.  I've attached a picture of the worst one.  So this is something that needs to be fixed.  These are oil through pushrods for the T&D rockers.  Are there more slender pushrods I could use, or do I need to pull the intake and grind out the offending holes?

What valve springs are you running?
I understood from your first post that you have a small hydraulic roller camshaft (236 duration and 0.556" lift)
I like the single conical springs for hydro rollers.
Perhaps the valve spring pressures are far too high?

Juts a thought I had while reading the thread.

Cheers!