Combustion problems and detonation
(Applies to any engine, but the B series is particularly prone to it).
Combustion problems caused by erratic burning and/or detonation of fuel rather than controlled and even burning shorten engine life and also cause losses in economy and/or power. A number of terms are used to describe them, including pinking, pinging, detonation. There are a number of causes, with low octane fuel and distributors not performing to spec, especially wear in the centrifugal advance mechanism, being the leading culprits.
While there are several types of combustion problem, they fall into two general categories of phenomena which are loosely described as detonation which are:
- spontaneous combustion which occurs independently of the ignition spark (diesel action), and
- pre-ignition, brought about by the spark occurring too soon.
Technically they are quite different phenomena, but sometimes they are linked. Some of the causes are common to both.
For simplicity, the following doesn’t differentiate between the two, as to the untrained ear it can be hard to tell the difference.
Note: Run-on (engine continues to run – somewhat erratically – when the key is turned off) is a combustion problem which is covered in a separate page: Run on.
Combustion – how it should be:
The fuel/air mix in the combustion chamber is supposed to burn evenly rather than explode all at once, and should spread out from the spark plug at a steady rate across the chamber. A certain amount of swirl or turbulence in the fuel/air mix is needed to achieve this. If there is too much turbulence, instead of spreading evenly across the combustion chamber after the spark ignites the fuel, the last bit of fuel/air in the chamber can explode spontaneously … which we hear as pinking or detonation.
Peak useful combustive effort is achieved when the spreading flame front reaches its peak pressure at 12 degrees after top dead centre. The point at which the spark appears is timed to achieve this. That the spark is timed to appear anything up to 30 degrees before TDC imdicates that it takes quite a bit of time (relatively speaking!) for the flame front to take in most of the fuel compared with it all detonating at once.
For a given amount of turbulence, the speed of the flame front depends mainly on the pressure in the cylinder – the less pressure, the slower it burns. Hence at low pressure (high vacuum) the spark needs to be even earlier – and so you have a vacuum advance.
Because the speed of the front is independent on engine speed, it is necessary to make the spark appear sooner (ie advance its timing) as the engine speed increases. At higher speeds, the cylinder doesn’t fill completely so the reduced pressure tends to offset the increase in engine speed. Because of this counterbalancing effect, the amount of advance needed doesn’t keep increasing, but instead reaches a maximum figure after which engine speed increase is fully offset by reduced filling.
If the peak pressure is too late, then engery is wasted, mostly as heat going to the cooling system rather than as mechanical energy in the piston. If it is achieved too early, then most of the energy is pumped back into the moving piston, increasing its temperature and hammering it against the bore … once again, we hear this as pinking/detonation.
Causes of detonation are:
- Wrong octane petrol for the compression ratio. Higher octane petrol is more stable, so less inclined to explode spontaneously. 96 octane will work up to 10:1 with ease if everything else is right, even up to 12:1 has been possible with modified squish and very careful set-up.
- Wrong advance rate (too much advance too soon) – with higher compression or lower octane fuel the engine will need less advance around 1500 – 2500 RPM, with about the same total advance at higher revs.
- Too much squish for the compression ratio:
Squish (the squish of fuel/air trapped between the piston top and the bit of the head face which overlaps the bore and ejected with great speed into the combustion chamber as the piston gets to the top) is a design feature to inject the right amount of turbulence to get proper burning.
Higher compression also promotes turbulence, so the general rule is that as the compression goes up the squish should go down. If you have an MGA1500 head on an MGB for example, the chamber volume is smaller (=higher compression) giving at the same time more squish due to the overlapping head face area being larger. Squish is further increased due to the bigger bore size. So the result of combining all these will be most likely combustion irregularities
To limited extent, too much squish can be ameliorated by changes to the ignition timing, but the ideal solution is to reduce the squish whilst retaining the compression and then re-profiling the distributor.
- Inadequate cooling of the head, esp round the spark plugs. Check head gaskets for obscuring the coolant holes and that the hole pattern is the same in block and head. Check the casting in the water jacket around the spark plug holes for featuers which might reduce heat transmission to the water jacket.
An alloy head works better in this respect but it would be an expensive way of curing detonation, and may not solve it at that – it depends where the problem is.
- Wrong heat range spark plug. N9Y or equivalent is standard for 9:1 on 98 octane (what the engines were designed for). Cooler plugs may be needed for lower octane and/or higher compression. N8Y is good to use, above 10:1 it may be necessary to use N7Y
- There is one remaining peculiar circumstance in which pinking can occur. If the engine has the wrong vacuum unit there is a distinctive symptom of pinking on part throttle which disappears or diminishes on full throttle. I have yet to find any other cause of this symptom (though there is always a first time!). (Refer to the page on vacuum units for a full explanation and diagnosis.)
- Hot spots on the head from carbon or small machining dags (eg when the head is faced but the chamber edge is not radiused off where it meets the newly machined head face)
- Even lean mixture can be a cause.
The order for checking detonation causes is:
- Check spark plugs are right heat range – examine for signs of overheating
- Check carb needles are correct spec & not badly worn at the thick end and that the mixture is not wildly out at idle. If the needle is worn at the thick end, richen it up 2-3 flats on the adjuster and see if that makes any difference to the detonation. If it does, buy new needles.
- Check compression ratio approximately using a compression tester. Roughly, every 20psi is 1 ratio. (Not 15psi as you’d expect since the process is adiabatic not isothermal – the air heats as it is compressed – and so the simple form of Boyles Law does not apply.) Now you know what you are working with.
- Check the ignition advance rate (see notes below) and adjust if necessary.
I’d check where the advance is before doing anything. Re-profiling the advance curve very easy, and a lot easier and cheaper than removing the head, let alone replacing it. I’d always regard checking the advance curve as the first step. All that’s involved is measuring with a timing light what the timing is doing now (vacuum disconnected) and then tightening up the advance springs in the distributor a bit until it does what you want. In higher performance engines a spring change may be necessary, but not for what you are doing. Rough rule of thumb: 9:1 compression look for 25 deg total advance at 2000RPM and 30 deg after 2500 RPM, 10:1 compression no more than 20 deg total advance at 2000RPM reaching 30 deg after 3500 RPM.
This is covered in more detail in Distributor advance rate adjustments – compensating for high mileage wear and for modified engines: Setting up ignition timing for modified engines.
- If compression higher than standard, try one grade lower in spark plugs.
- That’s the end of the simple checks, after that it’s a case of remove the head, check the squish area/chamber size/head type/cooling holes/hot spots etc.
The later MGB heads had the same chamber capacity as the early MGB/MGA1622, but had a shallower chamber with less squish area; most had the same valve sizes, but some had bigger inlets so one of these would be the one to get if you could. Refer my web page B-series Cylinder head markings for notes on the various MGB heads.
Lowering the compression ratio is an option for addressing some of these problems. However, I’d be reluctant to give away compression ratio as a cure to detonation – compression increases both power and economy. The B series engine functions perfectly happily on 9:1 to 9.5:1. 10:1 is very viable with a little care and setting the advance curve correctly and even 11:1 is reasonably comfortable on 96 octane if everything else is optimised.