Lucas 157SA Ignition switch; remove (and repair maybe)

Remove and strip ignition switch.
I decided to investigate the switch. In the early cars the electrical section (containing the key-activated moving contacts and loom connections) detaches from the lock/key section which is itself bolted securely (with tamper resistant bolts) to the steering column. The contact and switch section is held inside the locking section by a single small screw in the case. Being a Lotus, access will not be easy. To see the screw you will be lying in the driver's foot-well looking up at the underside of the dashboard. From this position the ignition switch is hidden behind heater trunking which can be pulled out of the way. The screw is then located on the left hand side of the switch, carefully angled downwards so that screwdriver access is limited by the lack of space between dash and screw! Its an easy strip apart from access. A little more thought could have made it easier, but as it is there is very little room to use a screwdriver! I managed eventually using a screwdriver bit. Luckily it wasn't very tight.

I just managed to reach the screw with an Irwin screwdriver bit- using its magnetic screw holder as a thumbscrew- even this is really too long

The switch is covered by a plastic boot which pulls back to reveal the switch body and wire contacts at the rear. Unscrew (and do not loose!) the tiny screw that retains the contact section inside the lock section. 

This tiny screw holds the ign switch in place- hard to see- here seen using my endoscope!

The contact section then pulls out of the lock which stays on the steering column.

The contact section then pulls out of the lock which stays on the steering column.Boot top left

Front of contact section- note profiled hole that accepts the turning rod from the lock section.

Looking inside lock section remaining on column. Contact actuating rod visible inside.

I marked all the wires and terminals so I could refit before trying to sort out the switch. Obviously this is  not meant to be a serviceable part- its supposed to be replaced if faulty, but as I'm waiting on word of the availability of a new one, I may as well try and fix the switch I have... and see if indeed it could be the cause of my problem.
I started by numbering the contacts 1-7 clockwise from the large moulded ridge. These are arbitrary numbers simply to help reassembly.

 I then stuck masking tape labels onto the leads and removed them

This released the back section. Its a Lucas 157SA switch widely used and available from a variety of sources at around £20.

The switch has 7 blade terminals distributed between 4 busses- these are marked with the embossed numbers 1, 2, 3 and 5. There is no number "4". Wires connected to the same bus are permanently connected together regardless of the position of the switch. The busses are held on by rivets which penetrate through the base and form the contact pads on the inside of the switch. 

The contents are held in by depressed ridges around the periphery. These can be bent back gently with pliers or by tapping with a punch on a hard surface.

Retention indentations folded back

...and once these are released the inner section can be gently prised out.

Inner section coming out of switch body

The switch separates into two, an outer contact part which bears the wire blades on the exterior and contact spots on the inside (lower) and an inner rotating section which bears a rotating copper contact pips (upper). This last rests on top of, and engages with, a white plastic moulding. This is also moveable and covers the spring-loaded ball bearing that provides positive switch location in the "Off", "Aux" and "Run" positions as its turned. I didn't raise this plastic moulding as I didn't want to disturb this ball bearing.  All contacts/pips showed a lot of carbon presumably from arcing and all the contacts were very dirty. There was a lot of wear to the final contact in the set of three. Its very satisfying to at last find something definitely wrong and this could be the explanation for my troubles. I really don't want to risk refitting a potentially faulty switch unless there is a problem getting the replacement, but since the thing is apart it seems sensible to have a go at repair- even though I may not actually use it! The separated switch is shown below. I think the single copper pip (the one that shows all the arcing) is probably the one responsible for cranking although I can't find a description of how this switch works so I'm not sure what should be connected to what in each key position. I eventually worked this out and present the results of my deliberations below.

Lower- rivet bases forming contact pads inside the switch. Contacts are made by a rotating set of 3 copper pips (upper). Note damage to 1 O'clock contact which activates the solenoid for starting and dirt on the lower left contacts which energise the ignition system when cranking. Upper- copper contact strip with three contact pips. Single pip lower heavily carboned. It has also become flattened, as has the topleft which is I think responsible for powering the Ign circuit whilst cranking (see on)

The contacts cleaned up reasonably well,  but looking at the lower set in the picture below shows considerable wear on the the last two of the set of three contacts. The two wires blades serving these contacts are nos 5 and 6 in my nomenclature- these are I believe the ones used when the starter is cranked and blade connections here are white/red (solenoid) and yellow (aux).  Its hard to see but it appears that the end contact has worn down on one side creating a ridge which in turn might be preventing the key from rotating the copper contact pip fully onto this contact. It has worn quite groove and there is a telltale mark from where it has sat bridging the two pips. This would I think leave the aux circuit in play when cranking.

The pips also cleaned well but I did notice that the lower pip (the one that is responsible for activating the solenoid) had become flattened- I was able to pop it out again using a punch from the rear.

Note worn contacts (here lower left) and the tell tale mark that shows how the copper pip has been positioned across two contacts instead of moving onto the final point

The lower surface of this unit is covered with a loosely held fibre/paper profile. If this is removed the contacts can be cleaned more efficiently. I filed the worn contact down with the aim of turning the ridge into a ramp so that the copper pip could ride more fully onto the contact. This still reveals the low spot visible in the picture below but hopefully will allow the start to crank without shorting.

Contact filed to allow better pip movement. Low spot still evident

Switch operation
I wanted to understand whether these problems might explain the symptoms and this meant understanding how the switch works.

The switch has 7 blade connections which I numbered 1-7 so I didn't get them muddled when I refit. These numbers are shown around the outside of the switch in the diagram below but they are really only for my own convenience. The 7 blade contacts attach to only 4 busses and the embossed numbers Lucas provide for these are shown on the inside of the switch drawing. Since wires joined to the same bus are permanently connected, there are essentially only 4 types of connection:

1. Lucas terminal number 1 (my blade number 7) directs feed to the solenoid only.
2. Lucas no 2 terminal (my blade numbers 5 and 6) is connected to the battery and will supply power to be distributed through the moving contact as appropriate. This terminal also connects to other permanently live items such as the courtesy light switches.
3. Lucas terminal 3 (my blade numbers 3 and 4) supplies power to the ignition coil/distributor LT circuit, fuel pump.
4. Lucas terminal 5 (my blade numbers 1 and 2) directs power to things powered in aux position- e.g. radio.

The moving contact has three "pips" which I have represented as the points of a triangle in my diagram. For clarity the triangle just points at the positions concerned although the actual pips would lie over them! In this view the switch is seen from behind and so the "ON" direction is anticlockwise. All the components are shown superimposed so the contacts behind the busses are marked with solid circles, positions where there is no contact pad and therefore cannot make a connection are shown as empty circles.

Position 1: Contact is made with power bus 2 connecting it to the Aux bus 5 to power radio etc. Ignition circuits are dead.
Position 2- turning one space on connects power bus 2 to both ign and aux bus so that the motor can run as well as allowing all aux items to receive power.
Position 3: Power bus is connected to ignition circuits and solenoid for starting. In this position the aux circuit is dead.
The off position isn't shown but is shown if position 1 is rotated clockwise by 2 contact positions. The triangle would be vertical with top and bottom left points on the fibre cover between the busses, the bottom right pip would lie over the vacant space behind bus 2 and thus contact nothing. In this position power will flow only to any item connected to bus 2

If this interpretation is correct then starting the motor should clear bus 5 completely, allowing all power to flow to  ign and solenoid only. Since my switch wasn't reaching this position, the aux circuits were never fully disconnected and this would drain power away from the ignition circuit. Further the connection to the ignition circuits is made through a different contact pad on bus 3... looking at the switch contacts as found when it was dismantled shows that this contact pad was actually quite dirty and so could have been a bad connection anyway.This might provide an explanation for the absence of spark while cranking.

I reassembled the switch in the reverse order, but frankly I don't trust it and so I will wait for a new switch to be delivered before I reassemble the car. Hopefully the next instalment of this blog will show better news!

... it didn't! The new switch is delayed for a week so I bench tested the repaired one using a multimeter and got all the expected connections. I reassembled it into the car and tried again- however still no improvements- the behaviour was exactly as before. This time though I tested the voltage at the coil: in the ign position this was a healthy 12.7 V... but once the key was pushed round to "start" this fell immediately to 0 V (solenoid white/red wire disconnected to avoid voltage drop across the starter)! This I think at least confirms the problem lies in the switch so that's very good to know- its a little puzzling as to why this has happened though as there was no obvious breakage or disconnection inside that part.

I also took up one suggestion from the forum; there is a tag on the solenoid which becomes live when cranking- so its possible to run a lead from that directly to the positive side of the coil to supply power in the crank.  If the key is released then this will stop the feed but it might help with restoring the spark in the crank at least.

Spare positive source blade contact on solenoid and my temporary wire connection to positive coil- note I put a double blade terminal on the other end so that I could leave all the coil leads connected.

I tried that and have to report at least a partial success: The spark was greatly improved and although it was still not fat, perfect and regular (prunes?) it was much better! Hopefully the new switch will help but in the meantime I will check the wiring from switch to coil.

I was somewhat depressed though because I am still getting no trace of the motor wanting to fire up- despite fuel* and (now) a presumed spark I'm getting no detonations at all!
*However  bearing in mind the lack of coil voltage when the key is in the crank position; this must mean that power is also cut to the fuel pump and if so the patch wire wouldn't replace it. This means I have probably never had both spark and high pressure fuel and I guess the take home message is to be patient and await the new switch! Well, without encouragement I'm not enthusiastic about doing any of the other jobs on the car- and anyway its Mechanex at Sandown park tomorrow so at least there are a few things to occupy me!


Blast- Having waited over a week for the new switch to arrive from SJSportscars, it came today  and its the wrong one! There are at least two types of 157SA from Lucas; the one I needed with a brown fibreboard base and 7 spade contacts (and in fact as illustrated on SJS website) which is part no 39415 or 579085; and a second type with white ceramic base and only 5 pins; SQB136 or PRC8230. The SQB type could probably be made to work if the wiring is modified slightly, but I would rather have the original- also the SQB is usually only about 2/3 the price of the 39415 so if this is the new equivalent replacement its poor value.