TR6 PI (Imported and converted to PI/RHD) - misfire on 5... Bubbles in the lines.

[OKibblewhite]

So. I've got a misfire on 5.

First pass, plugs, number 5 is black (but dry) and the rest are brown, new plugs and gap checked.

The second pass, I've got bubbles in the number 5 injector pipe.

https://photos.app.goo.gl/313YERvYq4J36WWY7

https://photos.app.goo.gl/UrAG2DhHCtqiAJaX7

Videos Here ^^

 

So seals in the union to the metering unit is the next question, as they're only happening on number 5 and not on the others... 

But no apparent fuel leak.

 

NEXT, I discovered that the injector on 5 wasn't firing properly, so swapped it with 6 and it started working fine on 6. The injector that was moved TO 5 started dribbling. Probably because the line is full of bubbles and not getting the pulse properly.

 

I then pawed through my Haynes manual, this diagram seems to sugged that the REAR most port on the metering unit TOP, is actually for cylinder 6... NOT 5, which is what it's plumbed to currently.

Leading me to ask a whole load of other questions about how exactly the hell it WAS working before (firing on all 6 and smooth-revving).

Thoughts anyone?

 

A new injector is on the way from TRGB and some other bits for the clutch slave job.

[NonMember]

I do remember from when I owned a PI estate that nos 2 & 5 are the ones with banjo bolts. And the setting up procedure for the correct injection timing is complex. However, engines are mostly a lot more tolerant of injection timing than ignition timing - remember that the Bosch K-jetronic didn't even bother timing, it just dribbled constantly, like a carburettor, and worked well enough to outlast the Lucas PI by over a decade! Remember, too, that people put engines together with the distributor drive miles out and "fix" it by rotating the plug leads... and the engine runs fine. The same is perfectly possible with the PI unit!

[OKibblewhite]

Update. After prodding and poking the injector on no5 and replacing the plug the misfire has gone.

My TR Garage said they had never seen clear injector lines on a 6 in their 30 years of business and suspected that given it's running normally, they may all be like that, just unseen inside the normal solid black injector lines.

[JohnD]

OK,

If you don't have a manual, here's the online one: http://www.lucasinjection.com/Lucas Mk2 manual page 2.htm

Pis are VERY sensitive to air in the lines, and getting it out if the injectors have been off the car can take ages.

If you have the type of injector with a 'needle' (the pintle) poking out of the tip, then pull on that with your fingernails, not not use pliers!    The bubbles prevent the pressure rising high enough to compress the spring inside and release the fuel/air.

JOhn

[OKibblewhite]

Done that, the has been line cleared of bubbles, bubbles reappeared from the unit. 

https://photos.app.goo.gl/5v16m1ZhEaFMLm48A - See here for a video of the bubbles emerging from the metering unit.

This raises all sorts of questions about where the air is coming from, as conceivably anywhere where air could get in... fuel at some level of PSI would be able to get out. 

But no smells or drips are evident. And the bubbles are not interfering with the actuation of the injector (pulled out, in a glass jar etc)...

 

ETA: - Thanks for the manual link, I have a few but not one for the PI system!

[thescrapman]

If air is appearing it is coming from somewhere, and I would have thought fuel at 105psi would come out of any leak rather than air sucked in, so it could possibly be getting in on low pressure side of pump?

worth fitting a new one-way banjo valve, or swap 2 and 5 over and see if it moves, just in case

 

[JohnD]

As above, when the system is pressurised to 100psi+ from the pump onwards, that limits the possibilities for air to get in!

I note the title - perhaps some pics of the upstream?    The pump/tank relationships and connections.

[GT6M]

This is the problemo wid the PI system,

some thing I found oot yonks ago, wid the use of clear lines,

but did the TR Brigade listen, no, thats why their cars are constantly conk,n oot.

 

then they tell folk to leave it for 1/2 hour

fill up wid  new petrol, which is cold

fit a bosch pump and wrong filters, which are selt in a wrong kit, or right kit, wrongly fitted.

fit a cooling coil,  { which meks it even wuss after a wee while,

as fuel is heated by pump, an the MU, as its fitted to engine, so gets v v hot fuel returning t,tank.

or, this is best ,n,    .. you have an air leak..., really,     air leaking in, at atmosheric pressure,

but fuel at 100 PSI+ dont get oot side the lines. !!!

 

Air is coming frae cavitating fuel, this frae pump, PRV,an MU, an also tank inlet .

it then goes aroond an aroond the system, mek,n zillions of bubbles in the tank,

the tank is litterally glistening wid bubbles whenst ye look in thea, as I used t,doo t,fin oot what was going on.

which is then sucked into the pump, which then starts t,get hot

which then heats fuel even moer, which then boils fuel mek,n even moer bubbles in system

 

the longer its run, the wuss it gets, untill yer car wont start

leave it for 1/2 hour it,ll start wid a misfire as the frothed fuel int tank es settled, BUT the stuff int lines, as ye have seen, leaves big air pockets

 

its thee,s air pokets that stops yer injectors frae work,n

 

think of yer system as basicly an older jerk pump diesel set up,

its pressure fed an opened, an if any air is in that type of system, then the engine will no run at all.

as pump compresses the fuel t,open a valve

but if airs thea, then it compresses the air, which can compress, so the  valve / injector dont open,

 

Which is what yee,v got on yer PI

the MU shuttle compresses the fuel, which then compresses the air int line, which compresses, BUT dont open yer injector, follow,!!

 

 

What you need t,get rid of the air int system is some filters mounted verticly,

the OE CAV is correct, but pipe work is too small for inlet,  so too is ootlet frae tank,

it can, an does cause cavitation if a bigger bosche pump fitted

 

the ones that really doo mek a differecns is one int return frae PRV an MU

thee,s can be inline filter on their sides, as fuel will then go into a bigg bubble, and then enter tank, an dissapear v v fast to top.

 

but biggest gains will be got frae a bigg filter fitted ..verticly.. right beside the MU

wid the inlet to MU at bottom, and inlet frae PRV at top

this way, all frothed up fuel an air will go to the top, leaving good ole solid fuel at bottom

the longer the filter, the better it,ll wuk

 

IF ye dont fit the filters for a while, then it,ll keep doo,n it

butt, remedy is t,get some revs up when car is on the road,

lift off throttle, an pull the choke full oot,

doo this a few times, it,ll clear, till lines get full agen

 

Note, listen t,the TR Brigade, then it wont work. { some ev it sussed, but vast majority ev,nt a clue }

just go read the sheite they post evey  week or so aboot the problems ye got.

 

M

[JohnD]

I've sometimes wondered if a PI fuel filter will double as a swirl pot, or more accurately as a debubbler, but it would need a lifter pump ayway to circulate fuel to it and back to the tank.

Not sure if M's solution will work - what happens when the filter is full of bubbles?

It may be excessive for road use, but in competition, as soon as your tank isn't full to the brim, on corners the pump will suck air and a swirl pot is essential.   Otherwise, the engine cuts out on the apex and won't start again until you're past the exit!      If you're going to fit a lifter pump, might as well get a proper swirl pot, that introduces the fuel at a tangent to spin it, and returns it form the true centre of the top.    THEN your feed to main pump, from the bottom, is guarnateed not to have any bubbles in it.  I've shown this before, the pumps and pot set up in my boot, but OK may not have seen it.   Been a few mods since, but the layout is the same.

Neither am I convinced that cavitation bubbles will survive long in a pressurised environment.    They occur where the pump action reduces pressure to less than the vapour pressure of the liquid, and are filled with vapour, so as soon as the pressure is increased, they collapse.    VERY small bubbles, microbubbles, may have a longer life but that'snota problem, unless they accumulate as foam.

I'm risking revealing my ignorance here!   Glad to be corrected educated!

John

[OKibblewhite]

8 hours ago, JohnD said:

As above, when the system is pressurised to 100psi+ from the pump onwards, that limits the possibilities for air to get in!

I note the title - perhaps some pics of the upstream?    The pump/tank relationships and connections.

 

The braided line runs to the Metering Unit

 

[JohnD]

Pump well below tank ( I think)

That right angle elbow in the feed tube puts two unnecessary connections in the line, lose that, and ensure the intengrity of the hose that replaces it.

Else, without surge entraining air, I short of any further suggestions!

John

[GT6M]

see them filters,  wrong, totally wrong for a mechanical system

OK for injectors opened by electikery,

 

And, i see a bosche pump, 

has the inlet been med bigger frae tank, and also the lines too

they should bve 10mm min, no the 8 mm / 5/16th of an OE

which could easily be only 7 mm, as olives reduced the dia.

 

Some home work for them wid some time ,spare, or the OE Poster want,n t,fin oot aboot His PI probs,

Have a ratch aboot ont tinternet,

and show me an older type Deisel filter fitted on its side

or even a modernsih deisel filter on its side

 

John, the bubbles are thea,  zillions of em

the tank will be shimmering wid em as the cars conk oot

the pump sucks em into it, which then puts em into fuel lines

then thee,s bubbles come back frae MU,PRV an go back t,tank, mekn it even wusser

a vicious circle the longer it runs, thats why PIs conk oot.

 

the original 5 had its MU an PRV fed into filter,tkeep it topped up

it conked oot, soon was changed so it went into tank,

 

butt, they did,nt fit a filter to them lines,so the mod made,

still med PIs conk oot.

[Nick Jones]

Interesting video, not seen bubbles that big before, but it goes a long way to explaining why PI injectors (often #5) sometimes vapour lock for no obvious reason (when your lines aren't see-through at least!)

As already said, if the bubbles are air, they are not getting in after the pump but before it.  This might be drawing through pipe joints on the suction side, through the pump shaft seal (Lucas pump only, not the "canned" Bosch one obviously) or, most likely,  coming down the suction line as a result of fuel surge exposing the outlet or fuel that is is already aerated from a returning flow. 

You also need to consider that maybe your bubbles are not air but vapour, from the fuel and of the fuel.  Again, these would form on the suction side of the pump, most likely at the entry point to the pumping elements themselves where pressures are lowest and velocities highest.  They are the result of some of the more volatile fractions of the fuel boiling and forming bubbles as the first part of the process of cavitation.  In normal cavitation, these then collapse back into to liquid, (or rather, implode, releasing a fair bit of energy) as they pass through the pump and the pressure goes up.  This creates noise and batters the pump. However, modern fuel is very strange stuff with the ethanol content playing merry hell with vapour pressures and it is actually quite likely that at least some of the vapour produced stays as vapour and merely gets compressed, to expand again at the MU and get sorted into the lines at the higher points.

Anyhow, the trick is to stop the vapour bubbles forming in the first place.  Basically you need to maximise the pressure at the pump inlet and keep the fuel as cool as possible.

Do:

- Mount the pump as low as possible to maximise the static head.
- Use large line sizes with minimum possible bends and fittings
- Use generously sized filters/strainers 

Don't:

- Use a true filter on the suction side of a Bosch pump, they cause too much loss.  Use a generously sized gauze strainer - or even nothing - a real Bosch pump has the strainer built in.
- Use a pump with far higher flow than you'll ever need.  The extra flow rate increases velocity & pressure drop for a given line size and increases the ratio of returned fuel through the PRV, doing nothing but turning the energy from the pump effort into heat in the fuel.  Bosch pumps typically used are often too big.
- Use a "cooling coil" on your Lucas pump motor.  What it really is is a heating coil for your fuel and hot fuel is BAD because it boils (even) easier under suction.  If your Lucas pump motor is getting hot, put some effort into running some decent, heavy gauge wire to supply it so it sees true system voltage.

Nick