General Diesel Injection Question

There is no obstruction on the return line.  When I pump the manual lift pump it pumps freely and I can hear the fuel splashing into the tank.  Pumping the manual lift pump has no affect on how the engine is running when it is running normally or when it is acting like it is starving.  The engine keeps running normally at working speed, but once it idles down it starves fairly quickly.

When I start the engine and leave it at idle, never revving it up, the centrifugal advance does not advance.  I can verify this by watching it, but the engine will still starve after about 5 minutes of idling.  I can also see it smoothly advance and retard when I rev and slow down the engine.  One test I have not done, but can, would be to rev the engine while watching the centrifugal advance then, after running a few minutes at higher RPM, let it slow to an idle while watching the centrifugal advance.  Since all other indications are that this is not the problem I'm not expecting any surprises with this test.  

The smoke is black, but not heavily so.  I still haven't thought of a definitive way to figure out if the engine has gotten flooded with excess fuel at this point or if it is starving for fuel.  It does seem more like it is flooded.  If I crack an injector line in this state it seems like the amount of diesel squirting out is about the same as it is when running normally, but maybe it doesn't have enough pressure behind it to force it  through the injector.  Is there some guage I could put on an injector line that would solve this mystery?

The injectors are working fine.  The engine starts well enough, although it usually needs a little ether unless it has been running recently enough to be warm (It doesn't start as well as it used to, so the timing is probably at least a little off.).  It runs, responds to throttle and pulls well until the RPM are allowed to drop.  

Your understanding is correct, but I've always started with the hand throttle open.  If left idling it takes about 5 minutes before it starts to sound different.  It is sublte at first, but within another minute or two it is barely able to run and has no response to throttle position (or very little and delayed).  If pulling it will pull well as long as it is not allowed to idle down.  If it idles down it starves.  I test run it in my driveway, which is uphill.  I pull up the hill and wherever I stop and disengage the engine the engine starts to starve.  I put it in reverse and the hill allows me to pick up RPM on the engine and it seems like it is pulling normally going downhill, but when I stop at the bottom of the hill it can barely move the tractor in low gear at full throttle.

The engine was running fine, but I had a mishap with a seized injector pump that broke the key in the timing gear.  I replaced the injector pump and have tried many, many, many timing positions.  The only possibilities I can think of are:
1) The centrifugal advance mechanism was damaged when the pump seized.   If this is the problem I will get another timing gear that does not have a centrifugal advance.
  2) The new injector pump is defective, but why would it work fine until it idles?  Might there be something in the governor that does this?  There is an injector shop about 75 miles away, but when I talked to them on the phone they said it is very unlikely that they would be able to test this injection pump.  If it is a problem with the pump, it is probably a problem of starving for fuel.
3) The timing is wrong.  This seems the most likely problem, but I've changed the timing and the start position of the centrifugal advance so many times and have not seen any clear pattern to help me zero in on the correct timing.  This engine was, I believe, the first  “EPA” engine with centrifugal advance.  I think they changed the timing gear advance from the one I have, but I believe  all the timing gears and the one injector pump used with them are interchangeable.  I suspect the timing mark on the  front pulley was not changed to accommodate the centrifugal advance timing gear, but I  would guess that the spill test timing would be the same with  or without centrifugal advance.  I have bent the injector lines so many times to make major changes that I'm reluctant to bend them much more.  If I  don't find out some test for the injection pump my next move will probably to get a new timing gear without the centrifugal advance.  One big advantage of that will be that I can make timing changes without rotating the injector pump.

So, to sumarize, I think I have either
1) a timing problem resulting in too much fuel collecting in the cylinders due to it being injected out of time, requiring over throttle to run normally until it is so “wet”, or flooded, that it can't ignite properly at lower RPM.  If this is even possible it seems like it would be experienced commonly by others  when engines aren't timed quite right.
or
2) A defective injection pump.  If this is the problem it would no doubt be a starvation problem (no smoke?).

Jim,

This is very interesting.  I'm not aware of any adjustments or how they are made.  I'm told it is a Bosch injector pump, but don't know where I can find out about any possible adjustments. 

Did yours run OK at higher RPM until it was allowed to idle down?  I suppose it's possible that a limiter would not kick in and limit until the RPM drops or the fuel delivery rate drops.  I would very much like to learn more about this.  It is an EPA engine.

/Jon