The header tank cap doesn't need to be on particularly tight - I keep mine just tight enough to ensure it won't unscrew itself from engine vibration, and I don't lose any coolant.

If you are losing coolant, there are two things prone to failure on the P38 diesel:

1) Radiator - the plastic tank at the top of the radiator has a tendency to crack. My experience shows that their life expectancy is about 1.5-2 years (at about 10-20K miles/year).

2) Water pump bearings - they fail every 50K miles or so, possibly less if you still have a viscous fan on the front of it.

Both are very common problems on P38s. I wouldn't be suspecting a cracked head until you have thoroughly checked both of those, and have done a chemical leak test to detect traces of combustion gases in the coolant.
 
Well,I tried to bleed the radiator today and two things happened with which I'm not confortable with:

- The top hose (the one that comes from the right side of the cylinder head) seems very soft like it never has any coolant in it. The bottom hose (the one on the left side) also felt soft but after pushing it several times it eventually became filled with coolant.

- After letting the engine run for a good time, with the tank cap off, there was no more water bubbles coming into the tank, from the bleeder tube. I took the bleeder tube off and there was a nice constant jet of water coming from the rad. But, after a few more minutes the coolant started to boil. I was under the impression that this should not happen, since there's no load in the engine. I know that the cap raises pressure and rises the boiling point but I've never seen it boiling with the engine running at idle speed. Is this normal or could this indicate something that is not working correctly in the cooling system (water pump or viscous fan) ?
 
When you say boiling the temp gauge in red / header tank boiling / or top hose out of head bubbling , the more info the better!
 
I meant that the coolant in the header tank was boiling. Temperature gauge was still a little bit to the left from top center.
 
Mate let it cool right down fill header tank to top level mark put the cap back on, remove the left top hose/ driver side ensure you can get as much water into the rad with a jug funnel and hose when running out of the top of rad put hose back on. Start car leave running for a few minutes after a few minutes remove bleeder hose on rad side not header tank side make sure water is comming out not air if water is comming out then you should have no air make sure its a proper flow ! Then put back on bleed pipe recheck header tank level hey presto you have filled up bled system ....
 
Last edited:
When you look at the radiator from the front, there is a small bleeding pipe in the top left. Take that off (carefully, the bit where it attaches on the radiator is a bit fragile). With it disconnected, pour coolant into the tank until you get a steady stream of coolant (and no bubbles) coming out from that bleeding pipe on top of the radiator.
 
Well,I tried to bleed the radiator today and two things happened with which I'm not confortable with:

- The top hose (the one that comes from the right side of the cylinder head) seems very soft like it never has any coolant in it. The bottom hose (the one on the left side) also felt soft but after pushing it several times it eventually became filled with coolant.

- After letting the engine run for a good time, with the tank cap off, there was no more water bubbles coming into the tank, from the bleeder tube. I took the bleeder tube off and there was a nice constant jet of water coming from the rad. But, after a few more minutes the coolant started to boil. I was under the impression that this should not happen, since there's no load in the engine. I know that the cap raises pressure and rises the boiling point but I've never seen it boiling with the engine running at idle speed. Is this normal or could this indicate something that is not working correctly in the cooling system (water pump or viscous fan) ?

For what it's worth, I have just replaced the anti freeze in my diesel, I had no trouble bleeding it, it did not boil at idle with the cap off as I kept it topped up and squeezed the air bubbles out. It helps if the nose of the car is higher than the rear.
Does your RAD get hot all over when it boils?
 
If you put the heater on when the coolant is hot and it works OK it will prove the water pump is working.
 
Last edited:
There are two types of radiator on the diesel - early and late. The early ones have the bottom connection on the bottom. On later ones the bottom connection is on the top, if that sounds Irish.
My 2000 is a doddle to fill up. Just fill, run engine on idle, top up, take car round block, final top up. Job done. It seems all the hassle is with the earlier rads.
If you put the heater on when the coolant is hot and it works OK it will prove the water pump is working.

How early p38's as far as I know have both hoses going to the top of the rad, as stupid as it seems should have one on the bottom for intake but some dimwit is stupid and put them both on top the early's I've seen are 95 ????
 
I have edited my earlier post.
I think I was confusing diesel and petrol rads.
 
Last edited:
How early p38's as far as I know have both hoses going to the top of the rad, as stupid as it seems should have one on the bottom for intake but some dimwit is stupid and put them both on top the early's I've seen are 95 ????

All P38's have both RAD hoses going in the top and yes it's absolutely fecking stupid, guaranteed to leave any **** in the system at the bottom of the RAD to block it.
 
If you put the heater on when the coolant is hot and it works OK it will prove the water pump is working.


Hot water flows through the heater matrix whatever the heater settings are.
Just feel the pipes under the bonnet to check if there is flow.:)
 
I bleeded the rad, again, checking that there was a constant flow out of the bleed hose.
I connected the A/C fans to work on maximum speed and took for a drive. Waited for the engine to be up to temp (85ºc) and gave it a good thrashing. The engine started to heat up and reached 120ºC with the gauge on the red, again. One thing I noticed is that the blower, that was on Max. heat, stopped blowing hot air.
I'm thinking that if I have a HG/head failure the combustion gases will fill up in the heater matrix, since that is the heighest point of the system ? Or does this just mean I have an airlock ? It was blowing hot air before starting to overheat.
 
Would of thought the header tank is the highest point as that's where you fill it up, so has the header tank dropped water since you filled, its best to make sure .best not to go driving off if you think there is a air lock why don't you just leave it ticking over until your sure if the heater is cold then would suggest wouldn't it unless your bend motors are shagged. So did you let it go cold and restart and check the rad bleed hose again it only takes a second or to. To see if that's dropped water, I'm sure it says in rave to rev @ 2000 revs to circulate the water quickly ie fill up before as it was up do you do this are you doing it to procedure ...
 
Last edited:
Are you getting plumes of steam out of the exhaust? Did you tighten up the header tank cap hard or just leave it only barely tight enough to ensure it doesn't unscrew itself from vibrations? Was the header tank pressurised when it overheated?

If heater was blowing hot and then went cold, that means you had coolant that subsequently went missing.

Have you double-checked all your coolant hoses? Particularly the less than obvious ones at the back of the engine that go to the heater matrix - there is a section that goes right between the engine and the bulkhead. It's worth getting the car up to temperature with the header tank cap tightened up, then switch the engine off and listen for hissing noises and look for steam.

If you are suspecting you might have a blown head gasket or a cracked head, do a chemical leakdown test that detects exhaust gasses in the coolant.
 
I bleeded the rad, again, checking that there was a constant flow out of the bleed hose.
I connected the A/C fans to work on maximum speed and took for a drive. Waited for the engine to be up to temp (85ºc) and gave it a good thrashing. The engine started to heat up and reached 120ºC with the gauge on the red, again. One thing I noticed is that the blower, that was on Max. heat, stopped blowing hot air.
I'm thinking that if I have a HG/head failure the combustion gases will fill up in the heater matrix, since that is the heighest point of the system ? Or does this just mean I have an airlock ? It was blowing hot air before starting to overheat.

The heater matrix is not the highest point but the pipe entry through the scuttle is and if the water boils the steam will migrate there and possibly stop the flow. As I said, I had no problem filling and bleeding mine so it sounds like you have HGF or a cracked head assuming that the rad gets hot all over:(
Expansion cap needs to be screwed dpown tight to seal, on mine it goes tight 2 full turns before it seals.
 
Last edited:
The radiator is heating evenly now.

I have no experience in sniff tests, can someone tell me how they work ? Do you have to test the vehicle with the engine running or do you just need to test coolant that may have been exposed to HG failure ? If you need to have the engine running I'm thinking that the problem may not reveal itself since it seems to overheat just under heavy load.
 
The radiator is heating evenly now.

I have no experience in sniff tests, can someone tell me how they work ? Do you have to test the vehicle with the engine running or do you just need to test coolant that may have been exposed to HG failure ? If you need to have the engine running I'm thinking that the problem may not reveal itself since it seems to overheat just under heavy load.

the sniff test will still show HC in the water whether idling or having been thrashed then brought back home.
 
I think a pressure test could prove better as what if its not a hgf through cylinder.A pressure test would find a leak any where as the gas test will only give a result of the one option.
 
I think a pressure test could prove better as what if its not a hgf through cylinder.A pressure test would find a leak any where as the gas test will only give a result of the one option.
Unfortunately, experience shows that pressure testing does not always find a cracked head. It needs to be done with the head off under high temperature and high pressure to be sure.
 

Similar threads