Blowing exhaust O rings and manifold gaskets.

When I got this car it was making a lot of noise.  The choke snuffle caused by the cold start system  was combined with an agricultural blow from the exhaust.  Clearly I needed a new one which I got from MGOC.  I had this fitted by my local garage...  but they couldn't eliminate the blow-by at the head end.  Luckily. .. As they had done the work they didn't fail the car in the mot on this point,  making it an advisory only.   However, since then the blowing has been getting worse and now it's about as bad as the old one. Its time to investigate, and given the problems I'm going to remove manifolds and check/clean all the joints before fitting some nice new gaskets. I've already covered removing the carbs but as everything is running at the moment I don't want to disturb them and will remove carbs and linkages as a unit.  This is a general pic of their setup before removal.

Disconnect the cables (leaving the throttle adjuster set nut undisturbed). All the nuts used in manifold and exhaust mounts are 1/2" AF. They should be brass as steel will corrode irreversibly in place, usually fracturing the stud  if you attempt to unscrew them. I was warned at the dealers that this can result in a cumbersome and costly job drilling the manifold to remove any stud remnants and fitting a helicoil in the hole! I soaked all in WD40 overnight before I attempted to undo any. I was very lucky in that all the nuts were brass and unscrewed relatively easily- only one was welded onto its stud, and unscrewed from the head as a unit.  It was easily dealt with (below).

I started by loosening the 4 1/2" nuts holding the carbs on and easing the carbs forward to clear the studs.  I held the carb/linkage group together and fitted it onto a spare manifold so that the linkages remained in contact.

Carbs and linkages stored on spare inlet manifold

Removing the carbs revealed the tatty heat shield and heat isolators

Heat shield with isolators and upper gaskets

I removed the isolators and the paper gaskets on top of them

Which exposed the lower gaskets between carb isolator and heat shield.  These seemed thicker than paper and there were traces of black gooey sealant around the port. 

I removed the manifold heat shield steady bolt to lift the heat shield off the manifold.

spanner to remove heatshield steady

This exposes the gaskets below the heat shield, which fragmented and spread itself between the heat shield and inlet manifold, so parts were stuck to both surfaces. Remnants of more black gooey stuff on both sides.

Intake manifold with gasket remnants

Gasket remnants on rear of heat shield,  again this looks like a more substantial gasket than the paper type so I need to check paper form will be OK here

Rear of heat shield- more gasket remains

Before removing the exhaust manifold I supported the pipes on a stand before...

I loosened the manifold to pipe nuts from underneath using a 1/2 socket on a long extension.  1/2" drive is fine for all but the two nuts trapped between block and pipes.  I reached these using a 1/4" drive socket on a long wobble bar. 

Removing exhaust clamp rings from underneath the car.

I could then loosen the manifold bolts,  there are 6 of these, the 4 central studs have large thick washers overlapping both intake and exhaust manifolds.  The two outer studs have small washers. These bolts were much tighter and one came out as a stud/bolt combi

Manifold bolts.  Note left hand end bolt is unscrewing as a stud/nut combo and large washer on the next bolt spanning both manifolds.  In this pic the intake manifold has yet to be removed.

I removed the intake manifold first (disconnect vacuum advance pipe and servo hose). Then pull exhaust manifold forward and up out of the pipe clamps. The exhaust manifold showed clear evidence of blowing from the pipe manifold junction.

The head side of the manifold wasn't too bad- it didn't seem to have been blowing but the finish on it did seem rather rough and lumpy with some raised edges around thecstud holes which I filed flat. 

Rough surface on mating flanges.

Gasket fragments left on block. Its not too clear in the pic but there was a lot of oil down the side of the block and tracing this back it appeared to be coming from the right hand side cover. This is secured using a bolt through a cup washer and flat 'O' ring- in my case the "O" ring was clearly not sealing and oil was spreading down from under the bolt and across the cover and onto the block. Of course this doesn't mean that the cover itself isn't leaking as well so I will change its gasket as well as reinstating the "O" ring.

Block after manifolds removed- debris around ports but oil spreading out in a fan shape from the right hand cover retaining nut- the 'O' ring beneath this has failed. Possible leaking from left side breather cover. 

Oily side to the motor- is the cover leaking too?

The exhaust sealing rings are visible in the ends of the pipes in the pic above and these are clearly the source of the problem. The sealing rings have a metal cased side and a carbon side- the metal cases were uppermost when I removed the manifold and the right-hand side has actually separated from the ring which just lifted out!

Metal side of sealing ring detached from carbon ring.- It just lifted right off.

The lower part of the sealing rings were only loosely held in the ends of the pipes and lifted out easily with a screwdriver. Here are the two old rings, the right hand side (lower in this pic) has separated from the metal casing and appears burnt and porous. These are clearly the reason for the blowing.




New rings below- shown each way up to illustrate metal and carbon sides.

The fact that these rings have two different sides begs the question which way up should they be fitted? Checking the web shows arguments for both ways: Some saying the metal side is there to contain the flexible carbon side and they should be fitted into the pipe so that the flexible carbon can be compressed upwards around the pipe and into the intake manifold whilst its base is supported in the metal cup resting in the exhaust pipe collar ring. In contrast others say that the metal is there as a heat protection and should be fitted against the hottest surface which is the manifold. This means that there are clearly cars out there being driven about with these seals in either orientation and it may actually not matter which way round you fit them. I have to say that I favour the first method even though in my car they were fitted the opposite way round. 

Looking at the manifolds above its clear that the tapered openings in the manifold (which accept the sealing rings) are poorly cast. The left-hand side in particular seems less distinct than the right. Secondly both openings are heavily encrusted with carbon- either from exhaust blow-by or more likely debris from previous sealing rings. My take on this then is that my car has had the sealing rings fitted carbon side up in the past. This allows the flexible carbon section to mould into the manifold tapers and compensate for any poor/rough moulding in the opening. However, this has left remnants in the tapers of the manifold openings and these deposits (combined with a poor casting on one side) have prevented the metal surfaces from seating and sealing properly into the cone when the rings were used the other way up. I guess one other conclusion to take away from this is that whichever way you do fit these rings, if its working then  its probably not a good idea to change it unless you are prepared (as here) to remove the manifold and thoroughly clean the ring- mating tapers in both exhaust and manifold.

Exhaust manifold. .  Experience shows its a great idea to chase the threads on the manifold studs whilst it's off! 

I cleaned both tapered sections thoroughly using a brass wheel and Dremmel, polishing with Emery cloth. Even so the difference in casting quality between the two openings can be seen above- the rhs looking rather smoother than the left.

I also had to address the issue of the block studs- the recalcitrant nut was removed from the end manifold stud with a nut splitter, and all threads were cleaned with a brass wire wheel before refitting. I removed all studs from the inlet manifold so I could clean around the stud bases and remove any lipping around the holes where the alloy might have been distorted. One stud was partially unscrewed and it's thread in the manifold was damaged.  I retapped that, cleaned all the studs and  refitted them all at the correct depth.

I cleaned up the block mating surface and refitted any studs I had removed (coarse thread into block). I had cleaned up most of the oil on the outside.

Block cleaned up and studs refitted

Given the amount of oil on this side of the block I was worried that the side panels might be leaking so I removed both side covers to renew their gaskets.

Remove the breather hose

Loosen the retaining nuts

The front was fitted with a cork gasket,   the rear with a rubber one.

There were indications that both could have been leaking.  I cleaned up both sides of the covers  as the paint was peeling.

Breather panel

The left hand side panel also incorporates the engine breather outlet which is fed via an internal wire mesh acting to trap any solids and condense oil mist  so that it can return to the sump. However over time this can clog with carbon and there is no way to strip it out for cleaning. It is possible to remove it and then re-braze the case back together- but as mine is a relatively low mileage car I doubted that such drastic steps would be necessary. I filled the breather compartment with Diesel EGR cleaner and let it soak for a few hours with occasional agitation. I could then pour the used solvent and any loosened muck out of the housing before repeating the process and finally flushing through with carburettor cleaner.
I also cleaned the mating surfaces in the block (I blocked the holes with tissue whilst I cleaned the surround to prevent rust and grit entering the motor).

Tissue blocking the access panels whilst the muck is scraped off the block surround.

I used Wellseal to stick the new gaskets to the covers and then gave a light coating to the block facing side and refitted them  using new sealing rings below the cup washers.  The front cover with its longer bolt was simple and fitted easily.  I torqued it to 6 ft-lb.  The rear cover was far more problematic,:  Firstly the rubber gasket is too small as supplied and slips out of place as you try to fit it.  It has to be stretched a few times before it will fit and stay on the cover whilst the bolt is tightened.  Secondly,  and the most irritating problem, was the cup washer.  Although this accepted both the bolt and sealing washer individually,  it wouldn't take them together.  This is a cheap part and I think q/c isn't that good.  The hole can be off-centre and this conspires with any deformation in the sides of the washer to mean that the rubber can't fit down around the bolt and inside the cup.  This it MUST do because the uncompressed rubber gasket plus the thickness of any protruding sealing washer conspire to  make the bolt too short to engage its threads! A ridiculous problem to have and one which took ages to sort, filing the bolt hole and creating more space inside the cup by grinding the walls.

Cup washer and sealing ring

It's definitely worth buying a few of these washers and checking them for fit with both bolt and washer together before doing this job. I managed it in the end tightening them to 2ftlb. Finally I repainted both covers in brushable black Car Plan engine lacquer. 

The heat shield did look a little damaged around the port openings where the surface was scuffed.  This may well be why it was fitted with the black sealant around the gaskets. I cleaned and sprayed the heat shield in VHT aluminium paint. This might not help reflect any more heat but at least it looks better. 

I also had trouble with the exhaust manifold-to-pipe collar studs.  The threads of these should be 5/16 NC18 and I'd cleaned them out with a wire brush wheel and so expected the new nuts to fit easily by hand.  However the new brass nuts wouldn't screw on easily with hand pressure and soon jammed- behaving like they were cross threading. The new nuts threaded easily onto inlet manifold studs so I was sure that they were correct. I really didn't want to strip the studs on the manifold and so spent a while trying to match the thread here to other formats in case they were not the expected UNF/UNC sizes. They didn't match any other likely candidates so I concluded that they were just poorly cut or damaged NC18. I just tightened the nuts anyway which proceeded without problem once you know you have to force them. There is no set torque for these nuts so I tightened them evenly and will check that they seal the pipes once the motor can be started.

I remounted the manifolds with a new metallized manifold gasket-the first I'd bought was a cardboard structure which didn't seem adequate. I refitted the inlet manifold and then added a carburettor gasket using just a little blue Hylomar around the port on the side facing the heat-shield where the surface was scuffed. I did similarly on the other side of the heat shield below the carburettor heat isolators. I used no hylomar on the sides facing the heat isolators or between isolators and carbs. Finally, I positioned the carb/linkage assembly in place and bolted it down to compress the gaskets.  The nuts are very hard to start as access is very limited behind the linkages, especially the lower nut on the front carburettor.  The best way of getting at that is to tilt the carb/linkage pair upwards to give more room below but it's still awkward. Once the nuts are on its just a case if reconnecting the breather hoses,  fuel hose and cables ( not forgetting the two carburettor linkage-to-heat shield springs.

The motor started easily,  there was quite  a lot of smoke but this was WD40 burning off the manifold and pipes.  After this cleared there was no sign of blowing.  I need to take the car on a run and check/ tighten the pipe to manifold bolts as required once everything h as warmed up and settled. 

Upgrading Starter: Fitting Powerlite hi-torque starter motor.

A recent battery failure during the cold weather reminded me of just how much juice the original Lucas starter required. I fitted a new battery, but to preserve its life, and hopefully improve starting I decided also to fit a Powerlite Hi-torque starter motor. This is one of the cheaper Hi torque motors, I don't think its the geared type which cost rather more. The distributor has to come out for access to the starter motor so if you are contemplating fitting electronic ignition then this is probably a good time to do it. I am (or rather I was), but it turned out I have the wrong distributor so this will have to wait for a later post. 

Raise the front of the car on ramps to assist access from below and disconnect the battery.

Since I am removing the distributor its important to set the motor at a known position so I set it at tdc on compression in no 1 cylinder. The rotor arm is then pointing at no 1 spark plug lead

Motor at tdc...

 Distributor-rotor position

rotor arm points at no 1 plug lead- in my case up at 45 deg and towards the front of the car

To preserve timing on the distributor- (in case of early abandonment of the swap), I removed the distributor by releasing the 2 distributor clamp-to-block bolts NOT the distributor timing clamp. I'm coming at this car from a Lotus- remember all those gags about access in those... the necessity of adopting the "lotus position"???- Well I can now officially confirm that access to the distributor and starter on the MGB surprisingly is even worse! Given the amount of space in the engine compartment, it is truly astounding how all the components have been arranged to block access to each other! Best I can recommend is a short and thin 7/16 ring spanner.

Distributor- adjustment screw top but do not release this one

Removing both bolts lets the distributor pull out giving improved access to the starter beneath

Distributor removed- dipstick also out

 The starter is held by 2 bolts- the upper also retains a cable clamp

Starter upper bolt and attached cable clip

 The lower starter bolt is virtually inaccessible- I can't even photograph it from above or below! I found that it can just be be reached using a small 9/16 socket on a long wobble bar extension. Luckily it wasn't tight.

Check the starter connections- these are all made to the solenoid part of the starter and are positioned underneath the motor. They appear at first sight chaotic. There are 4 ring terminals to the main starter terminal post. One of these is the battery cable- I'm not sure what the others are. There are also two spade connections; 1 of these (the larger in the black plastic cover) is the solenoid energising trigger, the second (green/white wire) supplies full voltage feed to the coil for the "cold start" system for starting. The Powerlite unit doesn't have this second  12V take-off and the makers' suggest simply leaving this second wire disconnected and do without the cold start boost. I'm not really happy with this so I will return to consider it later... probably earlier if I have starting problems.

Starter connectins- 4 ring terminals on main post and 2 spades

second view

9

Third view- spades clearer here?

 Disconnecting these wires was straightforward.  solenoid trigger

Disconnecting both spade temrinals

I tied the 4 ring terminals together with a cable zip tie so I remembered to reconnect them all.

 I removed all wires before loosening the lower starter motor bolt and removing the starter itself.

This is the new starter- It has a single spade triggering connection in addition to the main copper terminal. Note that although it is smaller and lighter than the original , It is a different shape and this will complicate fitting.

Side view copper contact on top, spade contact on side

End view

The starter has a plastic drive gear- which isn't great but this was by far the cheapest starter I could find and I guess this price saving has to be reflected somewhere. Note that the fitting holes are different- one is a simple hole for the bolt whilst the other is open making it a slot.

Front view plastic drive gear; hole-and-slot mounting points

I assumed that the slot mounting is to be fitted downwards so that the weight of the starter will hold the motor on. It has the added advantage that the lower (and inaccessible) bolt doesn't necessarily need to be taken out completely in order to remove/refit the starter. However there are two problems with this- firstly the design of the new motor means that it has several bulges on the side, when positioned this way up against the block, the new motor blocks access to the bottom mounting bolt more effectively even than the original! Access is now virtually impossible and you need to get the bolt in to close to the right depth before slipping the motor onto the stud to minimise the number of turns needed to tighten it after installation.  You can just reach it approaching horizontally underneath the motor using a 9/16  1/4 drive socket on a  long extension wobble bar and turn it, but you can't really exert much force. If anyone knows a better way please let me know.

The second problem is that this moves the terminal to the top of the motor and although the solenoid trigger and three of the ring terminals can cope, the battery lead is annoyingly just a cm or so too short. Normally this would not be a problem since I could slacken the mounting clamps underneath the body and ease the cable through for a cm or so. However, in my case I have just spent some £500 having the under-body professionally cleaned and sealed (Rustbuster of Spalding). This has coated the cables and clamps in sticky goo. As I doubt they removed the cables and pipes to treat behind them, they have effectively sealed the starter cable, fuel pipe and brake pipes to the under-body.

Starter cable in underbody clamp. Note the black goo now sealing it to the body

If I were to move the cable I would break this seal, potentially allowing rust-promoting water to sneak in behind the cable and rot the body along the cable line. I don't therefore want to move the cable and this means I will need to crimp on a short extension at the solenoid end.

NB although both bolt access and cable length problems are solved by installing the starter the other way up, you cannot do that because in this position it blocks the distributor so its essential to do it the harder way!

A word here about battery/starter cables because I found it confusing. There are 2 types of cable commonly used in this application: battery and welding cable. Both are bundles of copper strands inside an insulating coat. Welding cable has a finer strand which means firstly, its more flexible and secondly, more of the interior cross-sectional area is actually occupied by copper. Consequently, a welding cable can take a higher current than a battery cable of the same nominal internal copper cross section. Welding cable also has a higher quality of insulation and so is sometimes considered preferable for locations subject to weathering. However the MGB uses battery cable. For most cars battery cable of 16mm2 cross sectional area (110 amp) is fine- but for cars with the battery at the rear, the intrinsic resistance of a  long cable run could reduce voltage at the starter. So to keep cable resistance down cars like the MGB use 25mm2 battery cable (170amp).

Joining lengths of cable, or adding new terminals can be done using solder-on or crimp-on terminals. I use crimp-on as I find it nearly impossible to get enough heat into a thick copper cable to melt solder without also melting the insulation! I use a cheap hydraulic crimper from eBay as it can be used in position on cables fitted to the car. I think the secret is to crimp first using a bigger dye. I use the 35 size, and then do a second crimp using the dye of the correct size (25). I should also add that buying terminals can also be problematic in terms of both size and amperage capacity. My advice is always buy cables and terminals matched by cross-sectional area only. Some  suppliers only describe their terminals by cable diameter. This is often of no use since its not clear if they are talking about the internal cross sectional diameter of the conductor bundle or the whole cable including its insulation. Other suppliers use an AWG size which doesn't help at all, or give only an amperage figure for the cable and this is often confused with welding cable of the same diameter. This approach does reduce choice of suppliers, but at least you know that everything should fit. If there is a choice of amperage rating for the terminals I always buy the highest amp rating at the size I need. I bought a length of 25mm2 battery cable and joined it to the original using butt-join crimp connectors and added a new eyelet terminal. The joins are then covered in two layers of heat-shrink tubing. I needed to extend the cable by 25 cm to reach the starter post terminal. Be careful not to over tighten the connections to the solenoid post- it is copper and easily stripped!

I obtained some advice on reinstating the cold start system from the MGB owners club- I am indebted to Paul Hunt for the link to
http://www.mgb-stuff.org.uk/coilboost.htm where I obtained the relevant diagram reproduced below.



This method adds a new starter relay to work in tandem with the existing relay. The existing starter relay is one of two cylindrical relays that operate the starter and ignition circuits. Both are located on the inner wing on the right hand side and there is room to mount the new relay beside them.  NOTE I found that the original source (link above) was wrong, at least as it applied to my MGB, because it described the starter relay as connected to the brown/white wires. This is incorrect, that is the ignition relay! The starter relay is connected to the red/white wires. You must connect the new relay to the starter relay as you want to supply the 12V coil feed only during cranking. If you use the ignition relay you would be supplying 12V to the coil at all times that the ignition is on with presumably drastic effects on the 6V coil.

Given the confusion in wire colour its tempting to try and identify these relays whilst the battery is connected by assessing how they are supplied with 12V and under what circumstances they output that voltage. All I can say is DON'T! Be VERY careful with these relays! They feature live (brown) connections in ridiculously close proximity to earth (black) connections and they are not fused. Do not even think of touching anything here unless the battery is disconnected. I omitted this and a momentary flash contact meant I'd melted a good 6 inches of earth wire and had to unpick the loom and replace it! Luckily there was no more melting beyond the first 6 inches and no other wires were damaged. I strongly suggest that you rely on the wiring diagram.

I bought a new 30A  switching relay with a plastic base which accepts the spade terminals and holds them in position- minimising the chances of an accidental short.

New relay in position with wired base. I used a piggy-back connector to pick up a 12V output from the starter relay (energised when the ignition switch is turned to "start") and feed this via a red wire to both the relay activation (terminal 85) and switched power in (terminal 30). Relay activation earth (terminal 86) was connected to the chassis using the relay mounting screw to avoid over stressing any of the existing earth wires. Switched power output was connected via a green wire from terminal 87 to the existing coil boost wire formerly connected to the old starter motor.
Note the unpicked loom! I have ordered more cloth loom tape to rebind the loom in this region.

I tested the operation of all these wires off the car to make sure everything behaved as expected, and tested again on the car to check that the green wire was only live during cranking. Finally  the green wire was connected to the coil feed via the existing spade connector and I made a last check of the insulation round all the new spades I had fitted.

In view of my mix up with the distributor I opted at this stage not to fit the Powerspark unit but simply cleaned, lubricated and reinstated the points distributor fitting new points, condenser, plugs and plug leads.

Reassemble and re-time and everything so far OK. Cranking performance is much improved.