I've been getting very random 'explosions' in my air box for a while when starting in the mornings. The last one even had the neighbours curtains twitching it was quite loud. I've put this down to a leaky injector leaking gas back into the plenum. Anyway, the injectors are 5 years old now so decided to replace them. However, unlike the old ones, the new ones (OMVL reg fast) don't have a take off point for the Map pressure sensor, but come with a T piece to T into the 12mm gas feed. I've fitted this and all seems to be running well but there is a smell of gas at this T piece where the allen screw is. Can someone tell me the best way to seal this up please?

identifing what exactly is leaking would be the best start - soapy water will allow you to locate it, or alternatively a gas leak detector spray, if its the hose connection then if its as tight as it will go you might want to renew the 12mm hose rather than risking breaking the pressure sensor as the old hose may be too hard to shape properly round the sensor.

It may not be the injectors that were leaking - if your vapouriser has a vacuum line it may be worth checking that its not leaking gas through that way into the inlet, either put the open end of the vacuum line into some water when its running and check for bubbles (the other end should still be on the vapouriser) or check if you have any pressure coming out of the nipple on the vapouriser (while the engine is running in both cases)

Hi Brian, the current leak is from the new T piece I've just fitted. The 12mm hose is all new too. It's leaking from the 'take off' grub screw in the T piece. I shall try sealing it with some plumbing sealer I have.

As for the original problem, it was far too random to be leaking from the reducer. If it was, then surely it would happen all the time?

My guess is that just one injector was leaking, and if the engine happened to stop with the inlet valves of the cylinder in a particular position, it would allow the gas to enter the plenum and be ignited by a 'wasted spark' upon starting the engine. This would certainly explain the randomness of the occurrence.

I've now changed both injector rails and time will tell if it's cured the problem

Can't suggest much else on the leak - you know where it is at least. With the injectors it depends, not so much on the position of the engine, but if it is leaking it will fill the manifold/plenum up with gas, this might dissipate with time, also the amount of gas would depend on the pressure present when you stopped the engine etc.

Easy way to tell if its happening is to look at the pressure reading in the software if you have access to it - if when its not running on gas it stays reasonably stable then you should be ok leak wise, if you can see it gradually dropping its still leaking somewhere between the vapouriser solenoid and the injectors.

Fairly random would expect to be injectors though, and given they are the age they are and fairly cheap to purchase probably best to swap those and see what happens as you have done.

Inlet valve position would have nothing to do with it, since the injector outlet is on the plenum side of the valve anyway.

It is becoming more difficult to buy OMVL rails with a built in pressure take off, so T piece is often included. Some T pieces are not a problem but it seems some suppliers incorrectly assume a standard type temp sensor will seal properly in certain T's. Does the grub screw / temp sensor tighten up OK? If not, or if it continues to leak with the paste you'll do better asking them to replace it with matched parts that are designed to seal against each other.

Reducer leaks, or at least the effects of reducer leaks, can also be intermittent.

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

But surely the position of the valves would come in to play here, because there must be a spark to ignite the gas which has leaked into the plenum? If not, how is the leaked gas being ignited, if it's not the 'wasted spark' on the exhaust stroke?

I hadn't noticed before but Brian told you before me 'not particularly the position of the engine'.

Indeed a spark must get to the gas in the plenum to ignite it and cause a backfire, but this isn't really relevant to position of the engine valves when the engine stops spinning, and in fact you don't know which cylinder (if any in-particular) are igniting the gas in the plenum. At some point, when the engine is spinning again, a spark is introduced to the mixture in the plenum but you don't know the exact point at which it occurs and any/all cylinders could do that.

A leak from any LPG injector, regardless of engine stood position (inlet valve position) will fill the plenum with gas - with inlet valve closed the gas can only go into the plenum (not into the cylinder), or with inlet valve open gas will fill the cylinder and plenum at about the same rate.

The problem is the gas leaking into the manifold in the first place, no point looking any deeper into it. But if you were curious as to how backfire could occur (given that carb engines have explosive mixture in the plenum all the time) it is due to combination of valve timing and wasted spark system designed for an engine where plenum only contains air.

Take it this isn't on your other problem vehicle, the Bongo (which hasn't been fitted long enough for leaky injectors / reducer to be expected)?

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Don't forget that a weak mixture will cause a 'spit back' through the inlet irrespective of valve timing or wasted spark system. Anyone who has ever tried to balance multiple carbs will know that the one with the flames shooting out of it is the one that is running lean. If there is a fuel/air mixture at even remotely close to the optimum mix will ignite in the intake manifold or plenum.

_________________
'96 Saab 900XS, AEB Leo, sold
'93 Range Rover 4.2 LSE, Lovato LovEco, sold
'98 Ex-Police Range Rover 4.0, Singlepoint AEB Leo, my daily motor
'97 Range Rover 4.0SE, eGas multipoint, a project.....

Proud to be a member of the YCHJCYA2PDTHFH club.

That's another good point Gilbert.

Also implies that a plenum flooded with gas mixture that is too rich to ignite (so no backfires) to start with will get leaner as engine is cranked as gas is pumped out and air is pumped in, until it reaches the point where gas/air mixture in plenum is correct to ignite and then engine aspects (valve timing, burn speed in cylinders, wasted spark) can do their bit to cause backfire.

Worth saying that for a petrol injected engine that is being cranked on petrol, the petrol system should provide rich enough mixture (on it's own, regardless of extra fuel in terms of LPG) to prevent slow burn in the cylinder from causing flame presence at wrong side of inlet valve IF cam timing isn't also a contributing factor.

Slightly off topic but remembering when I had the fuel injected BMW 750 V12's, one oldish bloke told me 'Its alright having 12 cylinders but it is getting them all balanced that is the problem'. He was harking back to the days of multiple carbs on a V12, so I told him 'Not a problem with fuel injection'.

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Thank you for your posts, it does help me get my head around what's actually happening.
It does make me wonder why this scenario isn't happening more often, given that any leaky injector can fill the plenum with gas. Or maybe the design of my engine makes it more susceptible for some reason.
Anyway, it's only been 3 days but so far so good, no sign yet of any more unwanted explosions, I'm not sure how much more my air-box could take!

No Simon, no such problem with the Bongo yet, although I have used the same set up and injectors on this one too.



[Backfire: The term derives from parallel experiences with early unreliable firearms or ammunition, in which the explosive force was directed out at the breech instead of the muzzle. From this came the use of the word "backfire" as a verb to indicate something that produces an unintended, unexpected, and undesired result.]

Trouble is something leaking won't always leak - think how taps go, sometimes they will drip, other times they will be fine.

If the injectors (or for that matter the reducer) is gummed up with heavy ends then the chances of something sticking are much higher - if you have any damaged seals then bits that have been pulled off may sit down and seal sometimes, but not always hence sometimes its fine, others not. Other variables like the temperature of the injectors can play a part, hotter injectors less likely to be problematic in this respect, if they were the problem. similar if theres a small hole in the reducer it may not always leak for reasons you may not know.

3 goes does sound like you might have some luck fixing it though. Simon did a better job of explaining what I was trying to say originally!

I've now had the new injector rails on for a few weeks, and so far no backfires
The leaky T connector finally stopped leaking after some ptfe tape and paste was applied.
However, a new problem has come to light. I've done a couple of long trips recently and during these the LPG has switched itself off (leaving the level/changeover switch bleeping) and I've had to switch the ignition off then on again, after which it runs fine until the next time. Driving home from Gatwick yesterday (300 miles) it happened 4 times. It doesn't appear to happen on full acceleration, but while cruising at 70-80 mph. Filters were changed relatively recently, and the old ones appeared quite clean. Any ideas anyone?

Its logging a fault of some sort - you don't say what system is actually fitted (ecu) but do you have anything to connect to it with?

The common issue is if the coolant level drops low enough to stop flow through the vapouriser (ie you have a coolant leak that needs fixing) in which case it will drop back to petrol. So checking that as a first point wouldn't be a bad idea (if the vapouriser is high up in the engine bay it may be airlocked? Ideally you want it somewhere low down, but not always possible).

If you can get an idea why its doing it it may be worth reseating the connector to the problem component - I've had injector errors that disappeared once reseated the connector, but your somewhat stabbing in the dark without the software.

Are the injectors you fitted of the same design as the injectors you removed (both plastic bodied)? Did you fit the nozzles from your old injectors into the new injectors (or removed nozzles from new injectors if old didn't have nozzles fitted)?

If they're of the same design (except for the pressure take-off) and you fitted nozzles from the old injectors I wouldn't expect calibration to need to be much different to how it needed to be with old injectors - Of course the proviso here is that calibration was correct with old injectors. If you're fitting injectors of a different design or the injectors have no nozzles then there will definitely be need to calibrate.

I meant to write the above on another thread, didn't realise I was posting on the wrong thread but never mind because it's relevant here too!

The switching back while cruising could be due to a range of reasons such as actual lack flow of water through the reducer meaning the reducer gets too cold, or a faulty reducer temp sensor (common on OMVL reducers), dodgy tank/reducer solenoid coil, the ECU detecting another faulty component such as map sensor / gas temp sensor / injector coil. Sometimes when an ECU gets old it generates phantom errors and sometimes phantom errors can easily be got around by disabling the LPG ECU's error detection. Or if this is on a vehicle which disables cylinders when it senses a misfire, the LPG system will switch back to petrol when it sees no petrol injection pulse on a cylinder.

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Thanks Simon.
The new injectors were the same as the last ones, and fitted with the nozzles which were ready fitted (2.5mm I think)
I've been watching the reducer temp while driving and it seems to stay pretty hot so I think I can rule out a coolant problem, also no loss of coolant from the tank.
When plugged in, it showed 'injector errors' so I've cleared them.
I will wait to see what happens, but because the fault only seems to show on long drives I will have to wait a while for the next one

Cheers Brian, it's a King ECU, Magic 3 reducer, and OMVL dream XXI injectors.

Petrol injector errors or LPG injector errors?

Do you ever get a misfire, or engine warning light flashing on the dash?

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Also is it the same injector number mentioned if it comes up again? or several different ones?

I do get the occasional misfire, mainly when under load at low revs in 2nd or 3rd gear. This started happening after a change of camshafts (with slightly different inlet timings) so maybe I should've re-calibrated at the time but didn't wish to upset the apple cart as generally it ran pretty well. I've kept the original calibration that Tubbs helped me with several years ago. (anyone heard from him??)
However, the above is a separate issue, there is no apparent misfire related to the current problem.

I don't get any engine management light on the dash, and no faults registered on the cars diagnostics.

I think the errors showing were the lpg injectors, all of them listed as having had an error. Because it hasn't happened again since clearing them (not been on any long journey) then obviously I'll have to wait and see as regards to which errors pop up next

You'd probably do better posting up your current calibration and adjusting it to suit, rather than doing a auto calibration. You should be able to save your existing config anyway (would be a very sensible idea before trying to alter anything!).

I'd wonder if you have a bad connection somewhere thats been shaken if all the injectors are showing a fault - the battery feed to the lpg being the likely one i can think of that might effect all of them, though it could be a sign the injectors are getting on a bit and throwing random errors.

I don't know the King ecu but some will give you a time the error occurred - are these all logged at the same time or different? (its done in hours since on the ECU I've seen it on rather than an actual time, but should give you an idea if they are all happening at the same time)

I would assume with the injector change that you had the same injectors before though?