LPG Discussion Forum

Smeel of LPG I think
Page 1 of 1

Author:  jookul [ Mon Apr 07, 2014 8:33 am ]
Post subject:  Smeel of LPG I think

Hi Guys

Looking for some guidance.... of recent I have noticed what I think is a smell of gas coming from my engine bay. It does seems to be coming form the side where the evaporator is fitted and I noticed when the car is warm and I have finished my journey.

But here is the clincher I can't tell for certain is it a smell of gas, or a smell of something mixed with a gas smell, so I'm saying it kind of smells of gas but I can't be certain.

So here is the question... I have read that when you run a car on LPG you can get a wiff of gas from the engine bay as time goes by , just the same way a petrol engine will have some petrol type smelsl coming from it after time.

Is this true?.
Also I have recently read that if a car is tuned too rich you can get similar smells.

I have taken the car back to my LPG guy who said it was nothing but a smell of heat, he also said if there was a gas leak the system would shut down.

I have also sprayed leak detection spray in that area and it didn't show up anything mind you I just remembered the car was not running, though I could get the smell at that time

I am running a Mercedes CL420 1998 on a Prins VSi System

Author:  LPGC [ Mon Apr 07, 2014 10:18 am ]
Post subject:  Re: Smeel of LPG I think

Smeel eh? I almost mis-read that so was nearly concerned you could smell LPG in the engine bay there for a minute :lol:

It is possible to smell gas at the exhaust if an engine is running rich on gas and the cat isn't up to temp, but this is very unlikely and besides, you're smelling gas in the engine bay not at the exhaust? It is possible to smell gas at the air filter when an engine has just been turned off, but again very unlikely with a sequential LPG system.

Heat doesn't smell except it you burn the hairs up your nose! Things can start to smell when they get hot, for example oil etc gunged on the outside of the engine can start to smell, crankcase breather pipes can smell. Incidentally on Merc V8's there is a breather pipe from the inside of the middle of each cam cover that runs towards the back of the engine - If one of those has come off you might smell crankcase fumes?

You've been around the reducer with leak detector and haven't found any leaks. The VSI system probably won't have a vac line from the reducer so that would rule out gas leaking from the reducer's vac connection into the inlet manifold which might be notable by a smell when the engine was first switched off. Unlikely the injectors are leaking and usually hard to detect a drop on pressure with Prins after engine switch off (without interface and software) because Prins vapour pipes are usually hard rubber so the simple pipe pinching method isn't do-able.

Can only really advise going around it all with leak spray, not just the reducer pipe connections but all connections and all parts of the system, e.g. the reducer could leak from somewhere about it's body not just the pipe connections.


Author:  jookul [ Tue Apr 08, 2014 4:53 am ]
Post subject:  Re: Smeel of LPG I think

Thanks for the advice. I had it at my lpg guys place yesterday and while I got a wiff of what I think is gas yesterday morning when I got to work. I could, nt really get anything abvious while the car was running ag the lpg garage..... at leat anything I could put my finger on. But I will got aroubd again with spray. I was thinking of using soapy warter this time. Incidently my mpg was quite low on this tank but I has increac my RC setting and also my off set was not set correctly and I was doing a lot of short journeys so for now ill putbit down to that.

Do u c any problems using soapy water for leak detection

Author:  rossko [ Tue Apr 08, 2014 10:00 am ]
Post subject:  Re: Smeel of LPG I think

Have a good waggle at the hoses with your soapy water over the joints, seeps can come and go suprisingly easily.

What makes you think RC and offset factors needed changing?

Author:  LPGC [ Tue Apr 08, 2014 11:21 am ]
Post subject:  Re: Smeel of LPG I think

You didn't mention any drive-ability issues before so I'm wondering if the mixture really was set incorrectly before the adjustments? Was changeover between petrol and gas smooth before? Is it now? How did the guys who made the adjustment ascertain that a mixture adjustment was necessary - Did they take it for a drive and switch between petrol / LPG and back etc both whilst driving and at idle and take their time in doing so? It is possible to adjust both the mutiplier (RC_Inj) and offset (Off_Inj) in Prins to give effectively the same mixture at idle but such an adjustment could affect the mid range and top end fueling, so the adjustments they made to multiplier and offset should have necessitated a drive in the car to at least confirm the adjustments made were correct.

A 1998 Merc CL420 may not have OBD2 and may have 0.8 to 1.6 Lambda probes. The former may mean the only way to check calibration is by using the switching fuels method whilst monitoring Pinj, time consuming and requiring a decent length drive in the car, the latter meaning the OpenLoop Prins settings should probably be disabled. I hope the guys didn't pay lip service to 'calibration issues' but this might be the case if they didn't take the car out for a good drive with laptop plugged in...

Another reason why I am pessimistic is because I have noted on many Mercs, especially those from around 1998 and especially those with 0.8 to 1.6 Lambdas, that they can be slow to steer mixture according to lambda readings particularly at idle and believe this is due to Merc knowing that a fluctuating mixture (as with usual quickly flipping lambda readings) doesn't provide as smooth an idle as a more constant mixture. If idle mixture goes rich the idle speed will increase until the idle valve decreases air flow and vice/versa, so I believe it is Mercs thinking (as a luxury marque) that they will beat the competition in idle smoothness with slower idle mixture adjustment.. The problem here regarding LPG tuning is that if the fuel trims have previously adapted over a long time to running with incorrect LPG calibration, any adjustment made to mixture is likely to give lambda lean or lambda rich readings until the fuel trims re-adapt (which due to the above can be longer than on most vehicles). This can give the impression to the installer that the current LPG calibration is correct and maybe even that there is some problem with the petrol system, when what they really need to be doing is running the vehicle on petrol for a fair time until the fuel trims are re-learned on petrol before attempting to adjust the LPG calibration.

Other things that can often go un-noticed on these engines but are a bit more noticeable running on LPG are cam sensor issues, old and slow to respond Lamdas (this being besides the slow to respond ECU's as mentioned above), broken Lambda heaters, long hoses between LPG injectors and manifold nozzles (particularly on these engines as the idle speed is only around 600rpm and the manifold design doesn't lend itself well to getting those large Keihin injector rails close to the manifold).

Many years ago I converted an E420 which had a slight issue of a very occasional single misfire at idle that occurred only slightly more often when running on LPG. I diagnosed a slow to respond Lambda and failing cam sensor and informed the owner. The owner wanted a second opinion so took the car to a couple of guys he knew and trusted, these were ex Merc main dealer mechanics who had set up by themselves with all the proper Merc diagnostic gear. These guys did their diagnostic but incorrectly blamed the LPG system (even though the problem also occurred on petrol) so the owner came back to me saying it was my fault.. I checked it over again and found the same Lambda and cam sensor issues, so he went back to the Merc mechanics, whom this time told him they couldn't do any further tests until the LPG system was completely disconnected as the LPG install interfered with their checks (a load of rubbish since when running on petrol the LPG system doesn't affect the petrol system at all). So against my better judgement but to keep the customer happy I disconnected the LPG system and back he went to the Merc mechanics who still couldn't find anything wrong. So the customer decided to go with my advice and had them change the Lambda probe and cam sensor anyway, and this fixed the slight issue completely, but only after all the messing about! The owner still has the car 10 years later and no longer goes to the Merc mechanics but I generally see him once a year (when he remembers) for his LPG service.


Author:  jookul [ Sun Apr 13, 2014 8:47 am ]
Post subject:  Re: Smell of LPG I think

Hi guys

Thanks for all the great replies. Of recent I havent smelt the so called gax smell but I will have another poke around with some sopy water.

In relation to the tuning, and this is a slighty dicferent question. I have benn fiddeling around with the rc setting on my car for the lzst few weeks usinv my installers software. The reason for this is because I was and still am not too confidenr that my car has been set up correctly....when my installer set up the car the RC was set to 93. And I didnt feel it was running very well. On my instance anx wifh some trial an error its now 107, but these changes were made without drivng the car. I dont why but I would like to check if the car is running to lean, but check corectly. With this in mind I was thinking of buying the Bexprins software and doing this myself. No when I mentioned this go my installer he said the the VSI Princs system can tell if an aftermarket softwatre package has been installec and if I have a pro lem they will null and void my warrenty. First of all is this ture.

Secondly I have done a bit of reacing on it, but what is the correct way to tune this car bearing in micnd it does not have an OBD connection so it wil, have to be done by driving, just az a reminder the car is fitted with a Prins VSI and its a
4.2 L mercedes CL420 1998

...... and hopefully the leak is a thing of the past

Author:  LPGC [ Sun Apr 13, 2014 1:03 pm ]
Post subject:  Re: Smeel of LPG I think

I would say that you really want an RC of at least 100, most Prins I see are set with an RC between 105 and 115. Don't forget you can lower the working pressure if you otherwise find you need a lower RC...

You need to find a combination of RC, pressure and offset that gives unchanging Pinj when you switch fuels at all loads when running closed loop (up to maybe 3/4 load not full load) - And which also gives a rich mixture when under high load (if it goes open loop).

Bare in mind that offset effects low load more than high load, whilst RC and pressure effect the overall mixture... So it can be better to set up for medium loads (aim for RC of 110 and adjust pressure until fueling is correct) with no offset and then go on to check high load fueling, then offset, then re-confirm medium and high load fueling.


Author:  Gilbertd [ Sun Apr 13, 2014 5:28 pm ]
Post subject:  Re: Smell of LPG I think

jookul wrote:
I was thinking of buying the Bexprins software and doing this myself.
I thought Bexprins was a free download and the only time it has been mentioned on here is when people have used it and it has rewritten their settings screwed everything completely. If you have access to the real thing, continue to use it.

Author:  rossko [ Sun Apr 13, 2014 6:49 pm ]
Post subject:  Re: Smeel of LPG I think

Naw, Bexprins is a commercial product. With no idea that anything needs tickling anyway, it strikes me it'd be better to spend the money on Mercedes diagnostic software.

Author:  jookul [ Wed Apr 16, 2014 5:31 am ]
Post subject:  Re: Smeel of LPG I think

Hi guys

thanks for the replies.....my rv
c is currently set to 107 and rc at 0. The car is running quite well though my mpg hss gone up a vrry small bit from when the rc was 105 and offsrt of 7.

I suppose I need to check the timings as mentioned by Simon, also one of things I really want to do is confirm if the car is running rich or not. My installer revkons he can see this by watch the hidh low change of the lambda sensor and seeing ifbit reads a 1. It
Is that the only way to confirm. I know the set up is crucial on the rc, but how can I confirm if the car is lean or not, and I suppose im intredted in confirming im not doing any long term damage

Author:  LPGC [ Fri Apr 18, 2014 1:13 pm ]
Post subject:  Re: Smeel of LPG I think

rossko wrote:
Naw, Bexprins is a commercial product. With no idea that anything needs tickling anyway, it strikes me it'd be better to spend the money on Mercedes diagnostic software.

I agree that spending the money on the Merc software would be a better bet - Already said there might be nothing wrong with the calibration / mixture... but I am still a bit surprised the installer set RC below 100. Also wonder if this Merc has 0-1 or 0.8 - 1.6 Lambdas. If the installer said he's looking for near 1v at WOT this would imply it has 0-1 Lambdas, unlikely he got that wrong but if he did it might be running lean near WOT. These old Mercs are a bit like P38's in the respect that Lambda issues can go un-noticed for a long time...

Jookul, You should be able to figure out how to calibrate it from what I've said in previous posts, if not then in all honesty it's not looking promising that you'll manage to calibrate it any better than the installer.. I said you can compare Pinj when running on petrol to Pinj when running on gas, but you also need to confirm lambda flick (not just 'lucky one-off flicks) each time you switch fuels or the Pinj figures won't mean much. Confirm top end fueling is correct by checking lambda voltage at high load - but for all this you will first need to know if the Merc has 0-1 or 0.8-1.6 probes, and you'll also need to bare in mind the slow mixture adaption I mentioned on Mercs (wait until the Lambda does flick properly.

Not sure why you doubt it was setup much wrong given that your original issue was 'just' a potential leak? Perceived slight differences in mpg can be very wrong - the only way to confirm actual mpg is comparing mileage obtained from a full tank over many tank fulls of very similar driving. The differences you've made to calibration probably haven't made much difference to fueling... if the system was calibrated anything like correct before it is likely you've just made a few % difference to fuel trims. How long are your pipes between injector rails and manifold? Did you adjust the pressure setting when adjusting the multiplier and offset? These Mercs idle at around 600rpm, can be more of a problem than most vehicles when combined with the longish pipe lengths we usually see on Mercs but this also seems unlikely to be a problem since you didn't mention engine hesitation as an issue...


Page 1 of 1 All times are UTC
Powered by phpBB® Forum Software © phpBB Group