[addo]

Good points. One of the standard checks I always use is how does a non-starter go with ether spray? If it coughs and nearly catches, or starts and runs, your problems are likely in fuel delivery.

Some hard-to-start Pugs I’ve seen are casualties of a “paint pen dot” timing belt job, without confirmation of mechanical timing - they can be out most of a cam tooth. Usually though, they present a cam sensor fault so the workshop change the sensor without need. If problems persist, confirming mechanical timing might be a good step.

 

[addo]

If it is injector spray, a dose of ether based start spray into the inlet manifold/plenum chamber prior to 1st start may help identify any fault there.

I suggested that in Post 55, it must have got lost in the static.

 

[addo]

Buy the Supercrap equivalent ether spray; cheaper but same strength.

Ideally try to introduce the spray past the throttle plate, a vacuum port is often suitable. It‘s mostly captive on that side of the throttle, so if you do the recommended few seconds, then let it stand briefly - optimum distribution is achieved.

 

[addo]

I honestly doubt fuel quality is an issue for you. As per @Fruity, I think fuel delivery is somehow the crux of your woes.

Thinking of it like a diesel, my first thought is weak lift pump or its check valve not working. This would result in at least the mechanical pump having to cycle longer to build up a pressure that permits spark (assuming it doesn’t fire the coils if rail pressure is below threshold). A strong HP pump could also draw air through slightly tired O-rings on the “suction” side.

Mechanically logging the low pressure supply with a gauge teed in, over several hours would show if lift pressure was bleeding off. The “short and snappy” alternative would be to jump power the lift pump for 15 seconds before keying on.

Digressing to diagnostics, I have an OBDLink SX cable from Mike at Gendan. It seems to work fine with the licenced version of MES. As a bonus it also plays nicely with the eastern bloc diagnostic software for Fords.

 

[addo]

The lift pump only has one outlet, it’s internally regulated and fitted with a check valve to hold pressure over many hours. At least overnight, as a vernacular guideline.

Teeing in first at the tank end is probably simplest if you can bear the residual smell of fuel. You can butcher a dead fuel line for the female coupling and peel off the plastic to expose its barbs for use with regular SAE J30 R9 rubber hose. The male spigot is a rarer beast, not sure where my cut pieces came from but worst case try someone like EFI Hardware for a brass one.

If you get good pressure at the tank and it stands over time, close up there; move to between filter cartridge and HP pump. There will be a quick disconnect at the pump. Be really fussy about cleanliness, like a diesel.

 

[addo]

Normally the in-tank pump runs for only a few seconds prior to cranking, and won’t resume until it receives a discernible timing pulse. That’s why temporarily jumping it with 12V just before keying on, can build up a head of pressure to show if check valve is at fault. While you shouldn’t “dead head” it for a long time, a minute or so to get a stable gauge reading is not a problem. If pressure doesn’t approach stable and spec when wiring is jumped, then either pump is weak or regulator faulty.

I had a faulty 18 month old aftermarket pump in a Citroën two years ago, the regulator was extruding itself and line pressure stayed so low it would start with ether but was gutless on the open road due to low supply volume. As the owner is a fellow after my own heart (inclined to thrift) we repaired it by drilling holes and strapping the regulator in place with galvanised tie-wire. It runs well to this day!

Note that once the JTS starts, like a CRD, it will likely have sufficient suction at the injection pump to draw fuel from the tank.

Edit: We have one of these where I sometimes lurk, have added a few more DIY and modded fittings but it’s a good piece of kit. A trick is to tuck the gauge under the wiper arm so you can see it while stress testing.
Brand new ABV fuel injector tester kit | Power Tools | Gumtree Australia Wollongong Area - Wollongong | 1288932768

 

[addo]

I can’t comment on your gauge other than to say it looks like it possibly tries to cover too many bases, to give a dependable result for all.

It‘s unlikely there is a pressure sender in the low pressure line, MES will be conjecturing. So you’re stuck with relying on gauge readings. The few seconds of pump run when keying on, is only to top up the standing pressure in the lines. I note the nominal pressure in eLearn is 4.5 Bar, just over 65PSI. Off the top of my head, free (zero head) fuel delivery with a pump of this nature should be over 4 litres/minute, post-filter.

So, if you’ve got a dependable gauge plumbed in, jump the lift pump and run it (ignition off) until the pressure stabilises and is within spec - then leave for 8+ hours and review gauge thereafter. I can’t offer a number but it should absolutely not be near zero.

Not sure if you’ve seen this flowchart: B13 ENGINE STARTS WITH DIFFICULTY' - Alfa - 156 - eLearn - 4CarData

The last step is blackly humorous.

 

[addo]

Out of curiosity, when you replaced the lift pump was it because of cracking in the output spigot like so many do?

 

[addo]

It’s been my feeling that your “good days” have largely coincided with the warmer temperatures. Is this correct?

 

[addo]

I think the metal threaded bottom of that bush is “keyed” and should engage in a slot on the bracket.

It’s common for the upper part of the mount (the bushing) to fill with standing water and corrode the threads, affecting ease of torquing.

Suggest you investigate the keying possibility, if so make sure it is correctly indexed and use a jack to bring the parts into mutual compression. While stabilised in this way, ease the bolt out, coat the threads with cavity wax or waterproof grease then reinstall.