Page 4 of 5

Re: D612 PNT One of these days...

PostPosted: Tue Apr 25, 2017 8:08 pm
by ACourtney
It seems that no one is going to work out the cause of the smoke with the information I have given so far, so I will give a couple more clues.

When Fozzza mentioned water residue in the fuel tank, I was thinking along those lines when I discovered the real cause. Not from water left over in the tank from the cleaning, as Chris had thoroughly dried it and treated it with an epoxy fuel tank sealer system. However, I did wonder if some water might have been introduced by new fuel, or from condensation forming in the tank. I decided to test this idea by disconnecting the pipe from the tank and feeding the fuel pump straight from a can of fresh petrol. It turned out that the smoke wasn't down to water in the fuel tank, but it was during the course of that investigation that I discovered the real cause.

And the final clue I will give is that having read Johno's recent post, I don't think he will see the same issue.

Re: D612 PNT One of these days...

PostPosted: Sun Apr 30, 2017 11:51 am
by All That Glisters
So are we looking at the fuel tank filler hose/breather pipe allowing water back into the tank?
I'm not convinced that this would result in your problem because petrol and water don't mix and the fuel pickup pipe pulls water from the bottom of the tank and fills the carb float chamber until the engine splutters and stops. You have not had that problem but your cryptic clue seems to be pointing us this way (unless you mean a different 'Johno's recent post').
I'm not losing sleep over this but it is becoming a 'two pipe problem' for me!
You're just a little tease really aren't you.
Regards, Mark

Re: D612 PNT One of these days...

PostPosted: Thu May 04, 2017 6:05 pm
by ACourtney
I thought that I should clear up the mystery before I disappear on holiday next week. I am afraid that no-one has got near the answer, but maybe I can come up with another (easier to solve) mystery and give the mugs away at my open day in July.

It seems that everybody followed the same line of thinking as I originally did, that the white smoke must be steam due to water getting into the combustion chambers. When I first introduced the mystery back at the end of March I said that 'John Roe had visited my workshop whilst I was puzzling over the problem and made some comments that were extremely pertinent' the most pertinent of his comments was that "white smoke can also be caused by over-fuelling'", "when the mixture is too rich you get more water (steam) and carbon monoxide, as well as unburnt fuel (moisture) in the exhaust gases". I initially dismissed that as the CO level was spot on at tickover - but then again I wasn't getting the smoke at tickover!

The other clue that I had missed was the dampness of the pistons when I took the head off (see my posting of April 11). The pistons were all wet with fuel. I had totally ignored that clue as I was looking for where water could be getting in :roll:

So following the idea that water must be getting in somewhere I decided to try some fresh clean fuel (see my posting of April 25). I decided to disconnect the fuel line from the tank just before the fuel filter and connect a pipe from a spare can of fresh fuel. First I warmed up the engine to see if the smoke was still there (it was). To my surprise, when I disconnected the fuel line, a spurt of fuel shot up from the pipe a good six inches above the end of the pipe. Now the end of the pipe must have been at least a foot above the level of the fuel in the tank, so the only explanation was that the tank has become pressurized. I quickly stuck the pipe back on (I don't like jets of fuel splashing around the engine bay!) and mopped up the fuel. I then went around to the back of the car and removed the fuel cap. We started the car again and this time no smoke. I put the fuel cap back on and within a couple of minutes it was smoking again.

So the cause of the smoke was the pressure in the tank. As the car warmed up the exhaust will have warmed up the fuel tank increasing the vapour pressure and evidently the tank was not adequately vented (hence my comment about Johno's post). In fact that came down to my installation of the MGF fuel cap, so I must take the blame for causing the problem as well as taking so long to find it.

The SU fuel pump is closely matched to the SU float and needle valve so that it supplies fuel at the correct pressure - if you have ever fitted something like a Facet competition electric pump then you will know that you need a fuel pressure regulator before the carb(s). The diaphragm of the SU pump will let any pressure from the tank through as it just supplies a pressure difference when it pumps the fuel. This extra pressure, must have been enough to just pressurize the float chamber without causing it to overflow. The vent pipe comes vertically out of the top of the float chamber, so it must be possible to raise the pressure by a couple of inches of head before it overflows. A couple of inches of head may not be much, but it was clearly enough to create a big difference between the mixture at tickover, when the needle is sat well down into the emulsion tube, and the mixture as soon as you opened the throttle. With the fuel cap off, I needed to richen up the tick over mixture a fair bit, so even then the pressure must have been pushing more fuel through.

The real cause of the problem was down to my fitting of the MGF fuel cap. It totally sealed the tank. The arrangement with the MGF fuel filler neck was that the pipe that provides the vent for the Mk2 Metro fuel tank becomes a fuel return pipe for any back splashes caused during filling. Owners of Mk4s, with Rover Metro tanks will know that they have a fuel return pipe and a vent pipe. The tank breathing was easily rectified, by teeing in a second pipe and fitting a little non-return valve at its top.

ImageDSC_0025 by Alistair Courtney, on Flickr

ImageDSC_0026 by Alistair Courtney, on Flickr

With those fitted I would like to be able to say that the car sailed through the MOT, but D612 PNT still had a trick or two to play.

Re: D612 PNT One of these days...

PostPosted: Thu May 04, 2017 9:18 pm
by benofbrum
ACourtney wrote:The real cause of the problem was down to my fitting of the MGF fuel cap. It totally sealed the tank. The arrangement with the MGF fuel filler neck was that the pipe that provides the vent for the Mk2 Metro fuel tank becomes a fuel return pipe for any back splashes caused during filling. Owners of Mk4s, with Rover Metro tanks will know that they have a fuel return pipe and a vent pipe. The tank breathing was easily rectified, by teeing in a second pipe and fitting a little non-return valve at its top.

Presumably the non return valve lets excess pressure out, but stops air being pulled in. Does that not mean that the tank will be under negative pressure when the fuel and tank cool down? This could mean that the petrol pump might not be able to suck fuel through against this vacuum.
Thanks for giving us the final answer, though. Another potentially useful diagnostic tool.

Re: D612 PNT One of these days...

PostPosted: Thu May 04, 2017 10:08 pm
by Hans Efde
I am a lucky guy to have a Metro turbo tank under my Midas. It has a return line and an opening for venting. On this vent tube I have placed a Mocal airvent, which is situated in my boot (convertible). It will let out and in air, but a floating ball prevents fuel coming out. When the car is upside down in an accident, it closes completely.
Image

To prevent fuel smell in the boot, the vent line goes outside again:
Image

Re: D612 PNT One of these days...

PostPosted: Fri May 05, 2017 7:59 am
by ACourtney
benofbrum wrote:Presumably the non return valve lets excess pressure out, but stops air being pulled in. Does that not mean that the tank will be under negative pressure when the fuel and tank cool down? This could mean that the petrol pump might not be able to suck fuel through against this vacuum.
Thanks for giving us the final answer, though. Another potentially useful diagnostic tool.


The non return valve works in the same way as Hans' describes. It has a floating ball that allows air to pass slowly, both in and out, but it is closed by a fast movement, such as a gush of fuel from fast cornering (Metro Mk1 issue!), or the car rolling over. The one I found is sold for quad bikes and is a lot cheaper than the ones they sell for cars. I guess that quad bikes are more likely to get rolled than cars!

The last couple of times that I had fitted MGF fuel caps to Golds they must have had Metro Turbo tanks as they both had separate vent and fuel return pipes. It didn't occur to me that Chris's tank, a standard Mk2 tank with just the one pipe, would need an extra vent. Still I have learned the lesson and now have a small stock of Y-pieces and one-way valves.

Re: D612 PNT One of these days...

PostPosted: Sun May 07, 2017 2:04 pm
by fozzza
Even though I was excluded from the competition because Alistair had been working on Desmonds fuel tank at the time I visited the factory, not in a month of Sundays would I have linked white smoke with over fueling, nor a mechanical fuel pump being bye passed by a pressurized tank :roll:. I will have to go to the corner of the room with my dunces hat on.
At least its a fault I can prevent happening when I get my MK3 on the road.

Re: D612 PNT One of these days...

PostPosted: Sun May 07, 2017 6:00 pm
by Stuart
fozzza wrote:Even though I was excluded from the competition because Alistair had been working on Desmonds fuel tank at the time I visited the factory, not in a month of Sundays would I have linked white smoke with over fueling, nor a mechanical fuel pump being bye passed by a pressurized tank :roll:. I will have to go to the corner of the room with my dunces hat on.
At least its a fault I can prevent happening when I get my MK3 on the road.


I'd better join you in the corner too :oops:

Re: D612 PNT One of these days...

PostPosted: Fri May 26, 2017 6:01 pm
by ACourtney
Having cured the white smoke problem we thought it would be plain sailing for the MOT, but D612 PNT still had a couple of tricks up its sleeve.

Although the white smoke had gone the engine still wasn't ticking over nicely, or rather I could get it to tick over nicely by adjusting the mixture screw, but as soon as you blipped the throttle or drove the car the tick-over went lumpy, or faded away to a stall. One disadvantage of having a D plate is that the it needs to pass the emissions the emissions test (post august 1986 for a D plate), whereas a Q plate car only needs to pass the visual smoke test. The lumpy tick-over was going to be an issue.

Another job had appeared as well. Since we had been driving the car up and down the estate road the front ride height had dropped. There were a couple of speed humps on the estate road then and even more now - they seem to be breeding! When I came to pump the front hydragas units up I had a problem with the OS unit. The caps had been stuck on both units and took lots of applications of penetrating fluid and gentle persuasion with pliers before they would come undone. The NS unit then pumped up okay, but I couldn't get the pump nozzle to seal on the Schrader valve. In the end I took the unit out and then saw that the Schrader valve was so rusty that the the lip of it had a lump missing. I tried filing it smooth, but realised that I would need to remove so much metal that the pump fitting would open the valve before I had it fully screwed on. So even if I did get a seal I wouldn't be able to remove the pump without losing pressure.

Fortunately, I have spare Schrader valves and some brass adaptor fittings (as I've been experimenting with re-gassing some units). So in the end I cut off the old valve, soldered on an adaptor bush and then screwed in a totally new Schrader valve.

ImageDSC_0108 by Alistair Courtney, on Flickr

That just left the matter of the emissions test. I had a chat with my friendly local MOT tester (Ernigrip motors in Unit 2a) and we agreed to try a couple of things before the test. I had a spare MG Metro SU carb with the standard needle fitted. So by swapping over the whole piston, needle and suction cup we were able to do a quick change (However, this quick fix meant that I didn't spot the real cause of the lumpy tick-over problem). That got the emissions into the pass zone and gave a steady tick-over, but of course it would be too lean above tickover for the 1380 engine. Never mind it was enough to get a pass, which was all we needed as the next step was to take the car to a rolling road and sort out the running properly.

Re: D612 PNT One of these days...

PostPosted: Mon Jul 03, 2017 9:01 pm
by ACourtney
So D612PNT had gained its ticket, but getting the fuelling right was still a problem. I could set the mixture to the correct CO range and get it to tick-over smoothly, but as soon as the throttle was blipped open the mixture went out of the window and the tickover would be lost. Sometimes the tickover would go up and other times it would drop and the engine would stall.

I had suggested to Chris that we should take the car to a rolling road to get it set up properly and that seemed the next logical step. So I booked a slot at The Pitstop near Brize Norton. The Pitstop is run by Ralph Saunders, who used to run the rolling road at Oselli back in the days when they were the place to get your A-series tuned. I don't know how many years experience Ralph has with the A-series engine, but he was already Oselli's recognised A-series guru when I needed my Midget's engine reboring back in 1986.

As D612PNT was not running properly we decided that I should trailer it to The Pitstop. Once the car was unloaded and strapped down to the rollers, it was warmed up and Ralph did some basic checks. The timing was pretty well where it should be, but the mixture was still all over the place. I was beginning to whether I had done something wrong when I rebuilt it, such as pinching an o-ring. However, as soon as Ralph had the SU's suction cup and piston apart he spotted the cause of the problems. The needle had some fine pitting on one side, just near the top i.e. where the needle is metering the fuel at tickover. On the HIF SUs fitted to the Metro the needle is spring loaded and can rotate. So whilst I could set the mixture with the piston more or less steady for tickover, as soon as the throttle was blipped the needle would turn around presenting a different flow area for the fuel to come through.

ImageDSC_0384 by Alistair Courtney, on Flickr

So it wasn't something that I did causing the problem, rather something that I didn't do. In my defence, please take a look at the photos of the carb in my first post - the carb needed a lot of cleaning and the needle and psiton were some of the cleanest bits. So I gave them a wipe over with a petrol soaked rag and didn't look too closely at them. I suspect that when the engine had been standing there had been a puddle of old fuel sat in the carb - there is always a puddle of fuel left in the bottom of an SU carb when you switch off, hence the tendency for them to run on a bit. That puddle had probably turned into a brown goo around the back of the needle. The fuel probably etched some zinc out of the brass and that would have left the affected area weak and crumbly. With modern pump petrol containing increased levels of ethanol this sort of thing could be an increasing problem. Ethanol also etches into aluminium, which nearly every carburettor under the sun has been made from.