Look what I made - expansion tank cap tester

[Charless]

No sensible long term F is without a low level alarm; so what to do with the old expansion bottle?
Well you could butcher any old bicycle inner tube for the schrader valve and cut it off close to the tyre. Tap the internal metal reinforcing tube out of the small aperture on the tank and saw it off flush, then screw the inner end of the valve into the soft plastic, leaving no less than 10mm protruding, and epoxy it once in the final position.
Buy a cheap 2.5 bar pressure gauge off the internet with either a 1/4 BSP or 1/8 BSP bottom entry thread. If the former, tap out the internal metal tube reinforcing the large pipe aperture and simply screw the thread into plastic. If the latter you can screw it into the metal reinforcement tube with a little PTFE tape to get a good seal. I then supported the outside, lest it splits, with a jubilee clip
If I did it again I would go for 1/8 BSP bottom entry ( eg http://www.tcdirect.com 717-010. As it was I did 1/4 BSP 717-116).
You can then test tank caps till the cows come home and find there is nothing wrong with most of them - blowing off at 1 bar and holding nearly that pressure once the air pump has been turned off for a couple of minutes. Simples.

IMG_5117 (Medium).JPG (103.78 KiB) Viewed 4598 times

IMG_5119 (Medium).JPG (112.75 KiB) Viewed 4598 times

I have never believed that caps cause as much mayhem as they are blamed for, nor that a known good one needs replacing with an unknown one on a regular basis.

[ErikB]

Nice one !!

(btw: I won't go for the cheapest guage to get the impression I bought a reliable one)

[mgtfnut]

Well done , I made one some years ago. However, testing of many caps, new and old, over the years reveals that 2/3 at least are faulty.
I check mine once a year approx and it's been fine fortunately.

[Charless]

I am testing them by:
1. Pressurise the tank with air from a tyre pump until the cap vents at 1 bar +/- 0.1 bar.
2. Stop the pump and it should hold the pressure for a minute.
3. Remove from the tank and suck the cap to prove air can flow in easily as the sytem cools.

I have tested 6 caps so far, 4 on Fs and my two spares. I have had one fail by venting at 0.66 bar but holding 0.66 bar thereafter, and one which vented fine but then lost pressure down to 0.5 bar in about a minute with an audible hiss. Both 'failures' were on daily drive cars.
All six have been the 'good' style black ones with brass internals. Of the four clear passes at least 2 are at least 5 years old and one is 10 (along with it's cambelt and waterpump when I bought it!).

[Tipper]

Don't forget that once a 'safety valve' has lifted it can fail to reseat properly which can lead to loss of pressure for ever more. This is due to dirt/crud/gunk/anything else loitering around the top of the expansion tank getting on to the seat.

New ones should be OK but anything that has been in service especially if there has been HGF may never seat properly again.

[Charless]

I can do it, and do it again until I am happy about the test; which of course only proves the test! I suppose I could fill the cap under test with distilled water and try to wash the valve seat, but it seems to work repeatedly, exactly as it is supposed to, with most of them. Which reinforces my pre supposition that the caps, though cheap and easy to blame, probably don't make a huge difference to the bigger picture.

[Chris Tideswell]

A local garage tests all caps that they sell and at one point a few years ago the off the shelf fail rate was 7 out of 10!

[Charless]

Chris, I have heard of similar fail rates reported elsewhere so was quite pleased with my elderly 4/6 'success'!
Have you any thoughts about the tests I mentioned, can you think of any other aspect I should investigate. I did consider using a bucket of near boiling water to submerge the cap in so as to mimic the increased temperature of normal operation - becoming a bit involved!
It does seem strange that the two failures I have 'identified' are both fitted on well used daily drives without any cooling issues......

[mgtfnut]

Chris, I have heard of similar fail rates reported elsewhere so was quite pleased with my elderly 4/6 'success'!
Have you any thoughts about the tests I mentioned, can you think of any other aspect I should investigate. I did consider using a bucket of near boiling water to submerge the cap in so as to mimic the increased temperature of normal operation - becoming a bit involved!
It does seem strange that the two failures I have 'identified' are both fitted on well used daily drives without any cooling issues......

I up end the cap under pressure into a bowl of water to more easily spot any bubbles.

I don't think it strange that "older" pressure caps are better. I think Dieter once remarked that the tooling aged and the accuracy of the pressings diminished. Subsequent production and after-market caps must have been of poorer quality. I bought a new Halfords cap, tested it and it was useless. Took it back and after a long "discussion", got another new one. This behaved exactly the same, at which point I gave up New OEM type ones are much the same.

On the other hand, last year my low level alarm sounded and I had to remove the presure cap and continue my journey. I had the heater valve fully open, drove sympathetically with no more coolant loss to my destination (yes it was mild HGF). As long as the coolant is under 100*C, all will be well.

My other Jap 4x4 car has the original radiator cap on - D reg!

[Rob Bell]

I like this test rig Charles - very neat!

Will probably make my own from your instructions - seems like a useful thing to have in the garage given the unfortunately well-deserved poor reputation of the pressure caps.

In my years of MGF motoring, I have had one explosion of steam that was attributable to the cap being faulty. I think I probably got off lightly - but if the coolant loss is sufficient, then a faulty cap can ultimately lead to HGF...

[Charless]

There is no substitute for a low level alarm! (or two ;o))

[brandon mitchel]

What the primary purpose of that tester? I don't have one so I'm not familiar with it. And I don't know how to use it.

[Chris Tideswell]

What the primary purpose of that tester? I don't have one so I'm not familiar with it. And I don't know how to use it.

Welcome

The coolant cap has a pressure relief valve in the top of it and is prone to either fail or not work properly in the first place, if it does fail open then the coolant is liable to start to boil and evaporate. I shouldn't worry about not being able to test the cap just keep an eye on the coolant level an keep a spare cap in the boot just in case.

[Mr Teddy Bear]

Mine started to whistle shortly before I took her off the road at the end of June [ just as the good weather ended, so I got something right] I swapped it for a Quintin Hazel from Halfrauds. They used to be a reputable after market brand years ago. This whistle wasn't accompanied by any coolant loss, anyway the replacement cured the problem. Where/who is a reputable source for a spare?

cheers,
TB

[tonyduthie]

After fitting a new header tank cap gotten from Halfraulds some time ago for my wife's Rover 25 the fluid level started to go down abit which made me suspicious of a poor relief setting and so made a similar test apparatus so that I could do regular checks on both my F and 25. My experience has shown that the cheaper after market caps are not so reliable as the OEM one and have found none that have attained 1bar, all blowing at 0,5 to 0,75 bar and then not seating until the pressure has decayed to 0,4 bar. Recencently I got two new OEM ones for both cars and these on test did not blow until nearly 1.5bar and then reseated at 1.2bar. So for about £1 more I now always buy the OEM unit from Rimmers, it is such a critical item and experience has now told me that it is silly to use a cheap aftermarket item.

Tony

[Chris Tideswell]

I a agree that the OEM cap is the one to go for now that the quality control problems seem to have been addressed, good idea to buy it from somewhere like Rimmer's or B&G who will have a high turn over of stock to avoid buying an old cap that maybe inherently faulty.

[Charless]

One of the failed caps from a friend's F has now been replaced, this was the one that blew at only 0.6 bar. I took a hacksaw to it and found out why, and like 'Tipper' advised earlier in the thread - if it has boiled over replace the cap especially if radweld has been involved!
HISTORY: This X reg F has been owned by a not-particularly-MG-enthusiastic friend for about 8 years and annually serviced by non specialist garages in other parts of the country. It had never given any trouble even though at some stage one of them diagnosed 'overheating' and changed the brown temp sensor for a nice new black one - Kerching, thank you sir.
If they had bothered to test the rad fan afterwards they would have found it still didn't work (maybe they did). Perhaps they filled it up with Radweld as a precaution because its an F. It survived several years of rural non rush hour journeys without a rad fan or overheating until the day before his step daughter's wedding. When things were quite fraught enough without car failure, his wife was stuck in traffic doing frantic collections/deliveries - how typical of an MGF, explosions of steam from the expansion tank - tow it in. Some garage hero then found that the fan didn't work - I wonder why? So they ripped the old fan out and put a new one in for a couple of hundred quid - Kerching, cost you a lot more elsewhere sir. Did they test their work? Either way, more radweld should do the trick - they didn't even change the cap!

The first garage replaced the ECU temp sender with a black guage sender - so couldn't have tested whether it switched on the fan - it did not and never could have done so! The fan that was replaced was probably fine!

[Geoff.F]

My rig is simply the expansion bottle attached to a compressor with variable outlet pressure and gauge. Fill the bottle with boiling water to simulate engine conditions and each yearly OEM replacement and the on board spare have been satisfactory.
Geoff F.

[Dieter K]

Hi,
thanks to Charless for the idea.
Did a copy of the device. It works as it should. I've got a battery compressor so will take *the thing* to meetings for quick tests.
How it works:
https://www.facebook.com/groups/mgfregi ... 262113465/

btw. another guy here in D found a Renault part with the same thread and nearly the same pressure..
I'll test it and get back with results asap.

[robbie1003]

good idea, but issue I will have concern with is that the cap tested is in a cold condition so could alter your findings at operating temperature regarding blowoff pressure and the pressure it reseals. simple answer is change bottle to polo one, lasts a lifetime.