PI Difficult to start after a couple of weeks.

[EGCD]

2 hours ago, Spit_2.5PI said:

One little thing mine does - if I start it for just a few minutes or less, in other words I don't let the engine get warm. Then stop and leave it for several days, it will start on the button. None of this 20- 30- 40-seconds cranking nonsense! Do yours do this, Steve and Steve?

Cheers, Richard

Richard, no, mine is still a poor starter if I try a start as you suggest,

Steve 

[EGCD]

Investigating further I made up a rig to check how well the NRV union, hose and injector system holds pressure.

Starting at 50psi, the pressure decayed 20psi overnight and had all leaked away within 24 hours.

Using a leak detect spray the only leakage was a small one past the injector outlet needle.

1,3,4 & 6 were pretty much the same, I haven't taken the metering unit off yet to allow 2 & 5 to be removed and checked.

Given the design and manufacture of the injector needle seat I am assuming that this amount of leakage is acceptable, and that they will never be 100% sealed?

 

Edited November 17, 2022 by EGCD

[Waldi]

Good factual testing done.

It does not surprise me at all that the pressure drops.

In industry, a pressure test is normally conducted in a specific time /duration, 1 hr. is common. The low volume and injector tip-design play a role here too.

New O-rings in the injectors could improve it, not sure by how much.

Waldi

[john.r.davies]

Yes, ECGD, I love measurement!     To lose pressure in 24 hours seems not unreasonable, but if the hose stayed full of fuel that would soon pressure up on rotation of the engine and not delay starting.   I suspect that in the car, any pressure loss is backwards, and that the hoses drain as air enters from the injector tips, so that the hoses must be refilled before it will start.     Each injection of fuel is tiny and with compressible air in the line even less!

JOhn

[EGCD]

3 hours ago, Waldi said:

New O-rings in the injectors could improve it, not sure by how much

Thanks Waldi, the injectors were refurbished last year, only 500 miles ago, so I am presuming they are as good as they are going to get.

Steve 

Edited November 17, 2022 by EGCD

[EGCD]

1 hour ago, john.r.davies said:

Yes, ECGD, I love measurement!     To lose pressure in 24 hours seems not unreasonable, but if the hose stayed full of fuel that would soon pressure up on rotation of the engine and not delay starting.   I suspect that in the car, any pressure loss is backwards, and that the hoses drain as air enters from the injector tips, so that the hoses must be refilled before it will start.     Each injection of fuel is tiny and with compressible air in the line even less!

JOhn

Yes John I agree. The NRVs on 1,3,4 & 6 were not leaking at all. I've yet to test 2 & 5. The leak from the injectors is so small, once the pressure in the hose has dissipated, I feel that the fuel would just remain in the hose, and would pressurise quite quickly, even though each spurt from the MU is so small. 

I am pretty sure the poor starting after a week is fuel related, but I have nearly exhausted all possibilities, apart from striping the MU, trying a clear injector hose to check for air locks (as previously suggested), or putting the braided hoses (fitted by previous owner) in the bin.

Steve 

[PodOne]

Hi Steve

Just a thought or two.

Given it seems the MU is the last and most possible culprit why no resemble it all and see if the symptoms remain the same (which is most likely). Then fit another good MU and see if the issue is resolved? 

Is there anywhere the fuel can still drain out of the hoses once the pressure is lost after 24h? Is it worth disconnecting the MU return and see if there is any further return of fuel as you could approximate the volume within each pipe and compare. 

Andy

[EGCD]

8 minutes ago, PodOne said:

Hi Steve

Just a thought or two.

Given it seems the MU is the last and most possible culprit why no resemble it all and see if the symptoms remain the same (which is most likely). Then fit another good MU and see if the issue is resolved? 

Is there anywhere the fuel can still drain out of the hoses once the pressure is lost after 24h? Is it worth disconnecting the MU return and see if there is any further return of fuel as you could approximate the volume within each pipe and compare. 

Andy

Thanks Andy I'll try disconnecting the MU return after it's all back together and been run for a while. I would try another MU if I had a good one......

Steve 

[john.r.davies]

The return line from the M/ u is not like that on moderns.   They send full flow to the injectors and the EMU tells them how much to use, the rest returning.

Under Lucas Pi, the M/u determines the flow, and it divides up that to each injector.  The return is merely for the inevitable leakage of fuel from around the bobbin and rotor.   That line has no valve in it, so it could in time allow back flow and a head of negative pressure at the injectors, causing them to aspirate air.

John

[PodOne]

24 minutes ago, john.r.davies said:

The return line from the M/ u is not like that on moderns.   They send full flow to the injectors and the EMU tells them how much to use, the rest returning.

Under Lucas Pi, the M/u determines the flow, and it divides up that to each injector.  The return is merely for the inevitable leakage of fuel from around the bobbin and rotor.   That line has no valve in it, so it could in time allow back flow and a head of negative pressure at the injectors, causing them to aspirate air.

John

Hi John 

Get the return on a modern EFI set up with a fuel rail and PRV which on my MX5 set up would keep the fuel system pressurised for weeks.

If the Lucas PI return deals with inherent leakage within a normal system from the bobbin and dragging in air into the lines at what rate? Are you saying it could be the cause of the poor starting as the fuel gradually drains from the MU unchecked due to the lack of a non return valve.

Andy

[john.r.davies]

Yes that's what I said.

[astontr6]

On 11/6/2022 at 7:49 PM, Tom B said:

I have a 1970 TR6 and it behaved just like the OP – difficulty starting. I did some work on the injectors that others may find useful/informative.

The thread on the injector to the fuel lines is ¼” BSPP. The flat is also ¼” Whitworth (13.34 mm) which is why it is difficult to find a suitable spanner in your toolchest (I had one at the bottom..)

The car has a mix of injectors 73125 A,D and E plus one Injector, 73045A, which is different to the others in that the flat on the injector is not 1/4" Whitworth but probably 3/8" Whitworth (18mm).

If you run through air from a compressor it will be damp. The bodies seem to rust quite easily so it is best that they are re-installed and used promptly after refurbishment

The O-rings are available from Polymax (the go to place for O rings)

4.47mm ID x 1.78mm CS FKM (Viton) 75  BS008V75

17.12mm ID x 2.62mm CS FKM (Viton) 75  BS115V75

(Suitable for Gasoline (including high alcohol content) temp range -10C to +150C)

I replaced all the O rings as I was uncertain of the heritage. The larger ring is for the insulator blocks and the smaller for the injector. I use a bit of silicon grease to aid installation. (I appreciate that will wash off with the petrol but it makes life easier during re-assembly)

Some had filters in the body some didn’t. When I have the strength I may look around to see if there are any suitable replacements.

The ring at the bottom of the injector near the tip that flies off when you remove (advice from a previous poster on performing this operation in a plastic bag is extremely good) is known in engineering circles as a “snap ring”. I have not tried to buy any but a search on that term brings up appropriate results.

The injector pressure was set at 51 +/-1psi. I tried various ways to set/measure the pressure but the one that worked the best was also the simplest. I have a cheap “H***s” compressor (actually it works quite well) that has a digital display. I made up an adapter from a ¼” BSPP cone fitting and an old Schrader valve with the guts removed (shown in the photograph) and powered it from a battery. It quickly pumps up to pressure and when connected to the injector gives a good display of the pressure. Because the compressor runs off a piston, and the pressure is modulated slightly, the injector happily puffs away about 3 times/second. We therefore have a dynamic measurement (as recommended by another poster).

I did the injectors in turn and didn’t allow the fuel lines to drain (they were supported upright). After each injector replacement the engine was started and run until all cylinders were firing. If this is followed there should be no need to “bleed” the air out. I did notice that as each one was re-set the starting became easier. If the poppet valve pressure is too low (as some of mine were) then the sealing against fuel draining away is poorer. The “before” injector opening pressures were 26, 53.5, 53, 39.5, 37.5, 40 – quite a variation. However, when running, I have to say the engine seemed rather good, but the starting was poor. It seems better now I have re-set the injectors but time will tell.

I use my Son's high pressure cycle pump which can go up to 150psi I have made up an adaptor so that I can screw on an injector and look at the pressure gauge to see the blow off pressure. I have been doing this for many years. The thing that I have learnt is that if the injector loses pressure very quickly do not use it as the spray cone will not be very good. I have even rejected injectors that have been newly re-conned by the specialists??? I have never had a query from them. The returned one always seems to be spot on?

Bruce.

[Tom B]

3 hours ago, astontr6 said:

I use my Son's high pressure cycle pump which can go up to 150psi I have made up an adaptor so that I can screw on an injector and look at the pressure gauge to see the blow off pressure. I have been doing this for many years. The thing that I have learnt is that if the injector loses pressure very quickly do not use it as the spray cone will not be very good. I have even rejected injectors that have been newly re-conned by the specialists??? I have never had a query from them. The returned one always seems to be spot on?

Bruce.

Bruce, I found that it was possible to set the injectors quite accurately - I did try a cycle pump, but for various reasons I found the tyre compressor easier to use. One reason was that sometimes the injector needed working before it settled to a constant value. As I mentioned, it worked the valve making it pop a few times/second. After dismantling I guess that the components need to find the best working position. However, if I had the patience the cycle pump would have worked as you note. Thanks for the point about the injector losing pressure - that point was the subject of the original post! I'll set up to check that in future. 

All the best

Tom

[astontr6]

3 hours ago, Tom B said:

Bruce, I found that it was possible to set the injectors quite accurately - I did try a cycle pump, but for various reasons I found the tyre compressor easier to use. One reason was that sometimes the injector needed working before it settled to a constant value. As I mentioned, it worked the valve making it pop a few times/second. After dismantling I guess that the components need to find the best working position. However, if I had the patience the cycle pump would have worked as you note. Thanks for the point about the injector losing pressure - that point was the subject of the original post! I'll set up to check that in future. 

All the best

Tom

hi Tom,

I think the type of cycle pump is a key issue. When I first tried this method there were not cycle pumps on the market that could pump up to 100 psi today that has all changed. 100 +psi is the norm. I had to use a car foot pump with an adaptor fitted to it. Although a cycle pump has a pressure gauge fitted to it which is not calibrated, it is a good guide and shows how good or bad the air loss is.

Bruce.  

[EGCD]

10 hours ago, astontr6 said:

The thing that I have learnt is that if the injector loses pressure very quickly do not use it as the spray cone will not be very good.

Bruce.

Bruce, do you know what the acceptable pressure decay rate is? How long to drop from 50psi? What does good look like?

Thanks, Steve

[astontr6]

33 minutes ago, EGCD said:

Bruce, do you know what the acceptable pressure decay rate is? How long to drop from 50psi? What does good look like?

Thanks, Steve

I have always looked at the pressure gauge needle on a good injector the needle will move slowly down to zero. Usually around 20 t0 25 seconds on a bad injector in under 5 seconds. You have to remember that using air for a pressure test is very server test in comparison to a liquid.

Bruce.

[Mike C]

It seems to me that whatever causes this issue  occurs over a time frame of several weeks, not seconds or minutes.

[EGCD]

6 hours ago, astontr6 said:

I have always looked at the pressure gauge needle on a good injector the needle will move slowly down to zero. Usually around 20 t0 25 seconds on a bad injector in under 5 seconds. You have to remember that using air for a pressure test is very server test in comparison to a liquid.

Bruce.

Thanks Bruce, my injectors are all very good then, taking many hours to get to zero.

Steve

[astontr6]

4 hours ago, EGCD said:

Thanks Bruce, my injectors are all very good then, taking many hours to get to zero.

Steve

A point to note is that the original O rings had a swelling agent in them which reacted with petrol to give a 10% increase in size, that's why the original ones used to often come out looking square section in size. I am not certain that any of the PI restorers use O rings that do this? One of my jobs as a Bought Out Buyer was to buy millions of O rings for OEM and J Payen gasket sets.

Bruce.

[barkerwilliams]

I have a PI. and have just come late  to this thread and have had a quick read through and have noticed that I could find no reference to the fuel tank,

In a TR the fuel tank sits very high, remove the feed to the MU and fuel flows from the tank however much fuel is in it. At no stage can the MU leak back to the tank. Similarly the injectors are only a couple of inches below what would be the fuel level in a full tank  and could generate little back pressure to leak back to the MU and the tank. 

Does you starting issue relate to the quantity of fuel in the tank?

If you were feeling really keen a full tank and lifting the rear of the car six inches  would ensure the injectors and lines were below the level of fuel in the tank and could not leak back and any poor starting could not be attributed to the injectors draining back.

There are two types of fuel pump a turbine pump or a gear pump. A turbine pump uses its rpm to generate the fuel pressure and usually operates at the 14 volts whilst running and when starting may have 10 volts are less and consequently spins slower. A gear pump  pushes the  fuel through meshed gears and is less prone to reduced pressure on low voltages. Once again if you were really keen you could disconnect the wiring  on the fuel cutout switch and temporarily connect a second battery to determine if low voltage whilst cranking is affecting your pump.

In one of the posts above I saw connecting an airline to the injectors to check the  pressure at which they open. If you are going to try this then point the injector into a bucket with a  couple of inches of water in and with the injector 6"  above the surface observe the spray pattern on the water surface. I have had injectors the do not fully open and only produce a semicircle of spray, good enough for hot running but reduced atomisation when cold starting

Because the TR has a low pressure injection system it does not atomise the fuel very well with a cold manifold particularly if the injecting - valve opening timing is not good and the fuel is not instantly drawn into the turbulent air entering the cylinder when a good fuel=air mix is not possible. Once the car is running the heat in the manifold masks the issue. It is normally combatted by the choke over richening the mixture inorder that than a larger volume of fuel can vaporise to a critical explosive mix.

As in a previous post start the car and once running for a few seconds turn off and leave for a few days to eliminate any issues of heat soak into the injector lines and creating a vapour lock.

Do you have any talcum powder? Liberal dusting around the metering unit fuel pump, pressure regulator and pipe  unions is the one of the best trackers for small fuel leaks that typically evaporate before they leave drip marks.

 

[EGCD]

On 12/2/2022 at 11:25 AM, barkerwilliams said:

I have a PI. and have just come late  to this thread and have had a quick read through and have noticed that I could find no reference to the fuel tank,

In a TR the fuel tank sits very high, remove the feed to the MU and fuel flows from the tank however much fuel is in it. At no stage can the MU leak back to the tank. Similarly the injectors are only a couple of inches below what would be the fuel level in a full tank  and could generate little back pressure to leak back to the MU and the tank. 

Does you starting issue relate to the quantity of fuel in the tank?

If you were feeling really keen a full tank and lifting the rear of the car six inches  would ensure the injectors and lines were below the level of fuel in the tank and could not leak back and any poor starting could not be attributed to the injectors draining back.

There are two types of fuel pump a turbine pump or a gear pump. A turbine pump uses its rpm to generate the fuel pressure and usually operates at the 14 volts whilst running and when starting may have 10 volts are less and consequently spins slower. A gear pump  pushes the  fuel through meshed gears and is less prone to reduced pressure on low voltages. Once again if you were really keen you could disconnect the wiring  on the fuel cutout switch and temporarily connect a second battery to determine if low voltage whilst cranking is affecting your pump.

In one of the posts above I saw connecting an airline to the injectors to check the  pressure at which they open. If you are going to try this then point the injector into a bucket with a  couple of inches of water in and with the injector 6"  above the surface observe the spray pattern on the water surface. I have had injectors the do not fully open and only produce a semicircle of spray, good enough for hot running but reduced atomisation when cold starting

Because the TR has a low pressure injection system it does not atomise the fuel very well with a cold manifold particularly if the injecting - valve opening timing is not good and the fuel is not instantly drawn into the turbulent air entering the cylinder when a good fuel=air mix is not possible. Once the car is running the heat in the manifold masks the issue. It is normally combatted by the choke over richening the mixture inorder that than a larger volume of fuel can vaporise to a critical explosive mix.

As in a previous post start the car and once running for a few seconds turn off and leave for a few days to eliminate any issues of heat soak into the injector lines and creating a vapour lock.

Do you have any talcum powder? Liberal dusting around the metering unit fuel pump, pressure regulator and pipe  unions is the one of the best trackers for small fuel leaks that typically evaporate before they leave drip marks.

 

Thanks @barkerwilliams some good and different ideas there. The starting problem is the same regardless of whether the tank was full or near empty. I have checked the injector nozzle spray pattern and they are all good, and all relieve between 50 and 55 psig. I've also previously eliminated any potential fuel vaporisation induced issues due to heat soak with the braided fuel hoses by doing the cold shutdown with the bonnet open, leaving for a couple of weeks and then trying a start, but still the same.

I've since done a few other changes and currently just leaving the car for a couple of weeks before starting again to see if they have any effect. I've changed the injector O rings to Viton ones, although the injectors were supposedly refurbished just a couple of years ago. The Metering Unit has been completely rebuilt with all new Viton seals, O rings and diaphragm and the outlet NRVs refurbished, and calibration checked. The NRV refurbishment could have a big effect as they may not have been sealing too well, see before and after pictures, the sealing faces were cleaned up with 4000 and 7000 grade wet and dry to make them seal better.

Initial indications are that the car starts really well, but as I say I am just leaving it for a couple of weeks now.

Steve

 

[michaelfinnis]

On 12/2/2022 at 11:25 AM, barkerwilliams said:

I have a PI. and have just come late  to this thread and have had a quick read through and have noticed that I could find no reference to the fuel tank,

In a TR the fuel tank sits very high, remove the feed to the MU and fuel flows from the tank however much fuel is in it. At no stage can the MU leak back to the tank. Similarly the injectors are only a couple of inches below what would be the fuel level in a full tank  and could generate little back pressure to leak back to the MU and the tank. 

Does you starting issue relate to the quantity of fuel in the tank?

If you were feeling really keen a full tank and lifting the rear of the car six inches  would ensure the injectors and lines were below the level of fuel in the tank and could not leak back and any poor starting could not be attributed to the injectors draining back.

There are two types of fuel pump a turbine pump or a gear pump. A turbine pump uses its rpm to generate the fuel pressure and usually operates at the 14 volts whilst running and when starting may have 10 volts are less and consequently spins slower. A gear pump  pushes the  fuel through meshed gears and is less prone to reduced pressure on low voltages. Once again if you were really keen you could disconnect the wiring  on the fuel cutout switch and temporarily connect a second battery to determine if low voltage whilst cranking is affecting your pump.

In one of the posts above I saw connecting an airline to the injectors to check the  pressure at which they open. If you are going to try this then point the injector into a bucket with a  couple of inches of water in and with the injector 6"  above the surface observe the spray pattern on the water surface. I have had injectors the do not fully open and only produce a semicircle of spray, good enough for hot running but reduced atomisation when cold starting

Because the TR has a low pressure injection system it does not atomise the fuel very well with a cold manifold particularly if the injecting - valve opening timing is not good and the fuel is not instantly drawn into the turbulent air entering the cylinder when a good fuel=air mix is not possible. Once the car is running the heat in the manifold masks the issue. It is normally combatted by the choke over richening the mixture inorder that than a larger volume of fuel can vaporise to a critical explosive mix.

As in a previous post start the car and once running for a few seconds turn off and leave for a few days to eliminate any issues of heat soak into the injector lines and creating a vapour lock.

Do you have any talcum powder? Liberal dusting around the metering unit fuel pump, pressure regulator and pipe  unions is the one of the best trackers for small fuel leaks that typically evaporate before they leave drip marks.

 

 I don't know what the pressure drop across the metering unit is, but the Lucas pump delivers over 100 psi, and the injectors open at about 50 psi, so the pressure at the injectors must be in excess of 50 psi. That is at least as much as a modern indirect fuel injection system, so hardly a low pressure system.

[6TR]

On 11/1/2022 at 5:12 PM, EGCD said:

I’ve trawled the forum for ideas to help with my starting problem, and tried most things, but still have difficulty starting after I’ve left the car for anything more than a week, I am hoping someone here can help. I’ve had the car for several years but it definitely seems to be getting worse. I am not expecting it to start instantly, but other fuel injected 6’s I know seem much better.

Symptoms are that the engine spins up fine, but wont catch until at least 3 cranks of about 15 seconds, and then it only fires on a couple of cylinders and requires chasing with the starter spinning before the others kick in.

It definitely seems like a fuel problem and apart from this the engine performs well and is in great condition, rebuilt by TRBitz just a few years ago. Warm and hot starts are no problem.

I’ve tried starting with the ‘choke’ fully in and all positions up to fully out, with all combinations of accelerator pedal position.

So far I have done the following:

·       Fuel enrichment lever on MU checked and set up correctly

·       Fast idle cam checked and set correctly, even tried with it disconnected.

·       Refurbished injectors fitted, new injector body seals

·       Injector spray pattern checked and a good cone produced

·       New Sytec OTP019 fuel pump fitted this year replacing Bosch pump

·       Fuel pump delivery pressure set to 105psi with calibrated gauge, and also checked during cranking

·       Increased gauge direct load and earth cables to fuel pump via separate relay

·       Pre and post fuel pump filters changed.

·       Large bore pipe feed to fuel pump

·       Fuel is super unleaded

·       New plugs, red rotor arm, leads (Distributer cap has not been changed), Luminition magnatronic electronic ignition and new coil

·       Timing set using retard from pinking under load method

·       High torque starter motor last year

·       Battery is good

·       Metering Unit Nos 2 and 5 Banjos change, from Moss.

·       Metering unit replaced by previous owner in 2006

·       Metering unit vacuum tongue test ok

·       Vacuum hoses and clamps changed

·       Compression about 165psi, all cylinders

·       Kent TH5 Fast Road Camshaft RR1562FR and Unleaded stage 2 cylinder head E67/L/2 fitted by previous owner

·       Braided fuel hoses fitted by previous owner, but I don’t think these are the cause as starting is difficult after a shutdown with a hot engine after normal driving, or after a shutdown with a cold engine once all cylinders are firing, ie not a heat soak fuel vaporisation in hoses issue.

I also let the fuel pump run for 20sec or so before cranking.

I’ve read about non-return valves in each of the Metering Unit fuel outlets, which if they were leaking could let the fuel leak out of the injector hoses, which would cause the problem.

However, I can't find any information on them, and can't see where these are in the Lucas PI handbook publications, Moss or Rimmer catalogues. Has anyone any information on these?

Sorry about the long post, but you can see how much I have done to get to the bottom of it!

Any other ideas before an expensive MU rebuild at Prestige Injection?

Steve

I notice a significant improving in starting after fitting a new battery. The petrol pump sounds like it has more 'zing' when the ignition is switched on and far fewer cranks are required before the engine catches as compared with the old battery. Hope this helps.

[JohnC]

On 11/18/2022 at 6:49 AM, john.r.davies said:

The return line from the M/ u is not like that on moderns.   They send full flow to the injectors and the EMU tells them how much to use, the rest returning.

Under Lucas Pi, the M/u determines the flow, and it divides up that to each injector.  The return is merely for the inevitable leakage of fuel from around the bobbin and rotor.   That line has no valve in it, so it could in time allow back flow and a head of negative pressure at the injectors, causing them to aspirate air.

 

Hi John,

If I follow your logic (which make sense to me at least), that would point to the culprit being one of the two large O rings in the MU, or the diaphragm seal. That would let fuel leak from the MU, into the return line in the case of the diaphragm and one of the O rings. I know the entire CU/MU has been reconditioned, but it's not unknown for an O ring to get damaged in reassembly. Worth a look?

JC

[JohnC]

On 12/11/2022 at 10:37 PM, EGCD said:

Initial indications are that the car starts really well, but as I say I am just leaving it for a couple of weeks now.

So can we expect a Christmas Eve present (update)?