Won't start when hot

[mickdavies]

It could be that, when you shut it off, the heat spreads and seizes it up until it cools (common symptom when they fail).

If you mean the fuel pump, this is a second new one. I can't imagine 2 new pumps are bad.

[Rich]

The only way to rule out fuel pressure leakage (loss of rest pressure) is to put fuel pressure gauges on and confirm it.

+1

You mentioned you'd be putting your gauge on it. Plumb it as per the DWM diagram if you can.

Poor rest pressure retention is the root cause of most all hot-start issues with K-Jet cars as mentioned in the Knowledge Base article about hot starting problems. See if the system meets the specs in the rest pressure test (hot engine off, FP off) as mentioned above in post #3 or on DWM page D:02:01.

Your shim mod shouldn't affect that test much.

[Helirich]

Are you saying that when it fails to start hot and you then do the plug swap, it will start and run, but the engine will die if you undo the swap?*.

Hate to hyjack before the OP solves the problem, but can anyone explain this? Is this a valid test? What is it telling us?

[Ron]

Hate to hyjack before the OP solves the problem, but can anyone explain this? Is this a valid test? What is it telling us?

It tells us that there's a fuel problem (v/s ignition, etc). It's starving for fuel when 'hot', so suspect the CPR/WUR (It should run, once it starts and the plug is swapped back...not the 'normal' hard start problem...).
And ultimately,that he needs to throw a CIS fuel gauge set up on it.

[mickdavies]

Solved it!

It dawned on me that the new fuel pump had a fitting but no end "nut". This is the non return valve. I had removed it and put the fuel pipe banjo on with the existing banjo bolt (no non return valve).
I found the removed NRV in the spares box and re-fitted it. Didn't have an an end nut but robbed it off another car. Now starts fine when hot.
I don't see why, with lots of cranking, the pump can't repressurise when hot? But maybe keeping the pressure up stops fuel vaporising and causing air locks.
Anyway, it's working and I'm taking it for it's first run on UK roads tomorrow.

[Rich]

(a) I don't see why, with lots of cranking, the pump can't repressurise when hot? (b) But maybe keeping the pressure up stops fuel vaporising and causing air locks.

That's good news. A missing check valve kills the hot starts because there will be zero fuel rest pressure after shutdown. In a hot engine the fuel in the poorly-pressurized lines on and around the engine will then vaporize. It's called vapor lock. It stops the engine because the injectors start metering vaporized (boiled) gas instead of liquid gas during restart attempts causing a very lean mixture condition.

Answers to your "why" questions. I added reference letters to your points above:

(a) The pump CAN re-pressurize the lines when hot. What this can't do is eliminate the vapor that's still in the hot system. That can only be done by cooling the lines/FD/engine first and then pressurizing or by purging the vapor from the hot lines with lots of hot cranking time and hope to get the vapor out faster then you're making more of it with the fuel from the tank. A theoretical option is to pressurize the lines with enough extra pressure to compress the vapor to liquid in the hot areas of the system.
To speed up the vapor purging an old hack is to floor the accelerator during the hot cranking cycles. This admits more air past the metering plate which in turn increases the injector flows thus purging the vapor past them sooner which is its only way out. This hack works on carbureted engines suffering vapor lock episodes, too. If the battery and starter don't give up too soon.

(b) Exactly. That's what the check valve and the accumulator are primarily there for. To limit the post shutoff rest pressure loss.

Glad to know what you found and how you fixed it. Thanks for the report.

Bonus fact: Vapor lock can cause an injected engine to even die while it's running if the engine and fuel are hot enough (like pulling a very long, steep grade at high speed) on a very hot day. It happened to me (not in our D at the time). Half an hour by the side of the road with the hood up, then restarting with throttle wide open got her going again.

[mickdavies]

It'll start when cold and hot but not warm now.
I though I'd look at the dwell.
I noticed the car starts easily and idles well when cold.
I connected an automotive multimeter which has duty cycle and from cold read 50.2%. Perfect. Then, as the engine warms, there's a noticeable change to the idle being a bit rough and the duty cycle goes down to a whopping 4.5%
I changed the lambda sensor but the exact same thing happens.

[Bitsyncmaster]

It'll start when cold and hot but not warm now.
I though I'd look at the dwell.
I noticed the car starts easily and idles well when cold.
I connected an automotive multimeter which has duty cycle and from cold read 50.2%. Perfect. Then, as the engine warms, there's a noticeable change to the idle being a bit rough and the duty cycle goes down to a whopping 4.5%
I changed the lambda sensor but the exact same thing happens.

That is normal as the engine is warming up the engine runs very rich and the lambda is trying to run it leaner. The cold pressure regulator is what makes it run rich until the engine has warmed up. Check your duty cycle on a hot engine and you should be running around 44 % with it swinging around that value.

[David T]

That is normal as the engine is warming up the engine runs very rich and the lambda is trying to run it leaner. The cold pressure regulator is what makes it run rich until the engine has warmed up. Check your duty cycle on a hot engine and you should be running around 44 % with it swinging around that value.

Check your pressures against the chart in the Workshop Manual during warm-up. The Control Pressure Regulator is what controls the mixture till the motor warms up enough for the Lamda system to go into "Closed Loop". Your CPR may need to be recalibrated. Make sure the vacuum hoses are good going to it and it is fastened down.

[Helirich]

It tells us that there's a fuel problem (v/s ignition, etc). It's starving for fuel when 'hot', so suspect the CPR/WUR (It should run, once it starts and the plug is swapped back...not the 'normal' hard start problem...).
And ultimately,that he needs to throw a CIS fuel gauge set up on it.

Based on his current post, he was running on the startup fuel only? Can it rev up like that? Could it be driven like that? So if he didn?t swap the plugs back immediately, would the problem sort itself out after a few minutes? If so, any guess on how many minutes?