Nov 162013

One of the first modifications I’d decided to make was the change to an adjustable reaction plate for the torsion bars. In part the decision was due to the enormous trouble I’d had removing the torsion bars and reaction plate.

Also, even though the front suspension should only need to be set up once, if there was some settling of the suspension after the rebuild, subsequent fettling would be far easier. So I purchased an adjustable reaction plate from Rob Beere and followed Bob Skelly’s excellent installation guide.
– PDF Version

The bolt tubes on standard reaction plate are flush with the outer edges .... unlike the adjustable plateI’d planned to install the front suspension and torsion bars on two previous occasions. However, both times, progress had been thwarted due to some other fitting ‘difficulties’ that had been encountered. The first when installing the IRS and subsequently the engine.

So it shouldn’t have come as a surprise that fitting the reaction plate would be equally challenging! The first problem was the adjustable reaction plate was approximately 3-4mm wider than the original. The tubes for the bolts securing the plate to the underfloor channels protruded much further beyond the outer edges.

Rob Beere suggested using a pry bar and the need for a tight fit, which may well need hammering to ‘persuade’ it into position. If this didn’t work, the ends of the tubes could be ground down slightly to fit. No matter what I tried I couldn’t get it to fit and so had to resort to the latter.

The large Allen bolts are fitted first. Some paint repairs are now needed due to the tight fit and the need to tap the reaction plate into positionEven so, it still required hitting home with the nylon hammer. The various attempts to get the reaction plate to fit resulted in some damage to the paint work, which will need to be repaired.

Fortunately there are a number of other adjacent areas that still need to be touched up, where the chassis was attached to the support frame during painting. So these can all be tackled at the same time before the exhaust is fitted.

It was surprising to see that the new clutch slave cylinder had started to show some surface rust, even in the short time since the transmission was installed. I’ll have to treat it with some Dinitrol hard wax asap.

Bob’s instructions suggested tightening the large Allen key bolts once the upper bolts had been inserted. However I had slight alignment issues with all the mounting bolts and the torsion bar ‘ear’ brackets. Once the Allen and upper bolts were tightened, it was impossible to fit the remaining bolts and brackets.

The torsion bar 'ear' bracket and the upper & lower mounting bolts were all fitted before everything was tightened upI found it was necessary to have everything initially finger tight, which enabled a screwdriver to be inserted in bolt holes to pry the other mounting holes in the frame into alignment with those in the reaction plate.

The fitting order that worked for me was the large Allen bolts followed by the ‘ear’ brackets, the upper bolts and finally the lower pre-cut bolts.

Only once all these were in place could everything, except the bolt through the ‘ear’, be fully tightened. It is worth reiterating that:
i) the Allen bolts need to be tightened before the adjusting cam is fitted, as the nut securing the cam obstructs access to the head of the Allen bolt
ii) the ear brackets needs to be at the top of their permitted travel before tightening the lower pre-cut bolts.

Labels were added to mark the steps in the adjusting camI also followed the advice of labelling the cam steps and then painting the outer face with some clear lacquer. However I didn’t bother highlighting the edges of the steps as I thought this was a bit of overkill.

With hindsight, I think not adding the highlights was a slight mistake. It would have provided a better visual guide to ensure the step of the cam is parallel with the edge of the torsion bar ‘ear’ bracket.

It’s not a major problem, provided there’s sufficient light when setting the cams. If I were to do it again, I’d use two bright, contrasting colours to paint alternate step edges.

I’d not been looking forward to fitting the torsion bars. I hadn’t been able to dismantle them in the conventional manner, described in the various service manuals. There wasn’t even a slight hint of movement in the torsion bars despite some very hefty blows wielding a club hammer. In the end, as an act of self-preservation, I conceded defeat and removed each side of the suspension as single units.

Time for some (dubious) Maths – the torsion bar setting link
The shock aborber is replaced by a fixed length link to provide a datum point when setting the torsion bars. This should then give the correct ride height, although the adjustable reaction plate would then come into its own if it needed subsequent tweaking. The setting link for the early cars was 17 13/16″, however this had increased to 17 31/32″ for the S2 cars.

I’d obtained some replacement torsion bars at Stoneleigh but hadn’t realised at the time that almost all new torsion bars are ‘uprated’. The standard bars are 0.77″ in diameter while the replacements were 0.85″. As a result, the bars will be stiffer, so using the recommended setting link length would result in the ride height being too high …. but by how much?

A plot of Classic Jaguar's recommended setting link lengths against Torsion Bar diametersAfter some research I found that Classic Jaguar in America had produced a chart with recommended setting link lengths for various torsion bar diameters.

Unfortunately they don’t have a figure for 0.85″ bars so I thought I’d plot their recommendations in order to determine the link length required. The graph wasn’t what I was expecting, with a linear relationship between the setting link length and the torsion bar diameter.

Hmmmm! Perhaps I’m missing something as I thought the torsional stiffness or angular deflection of a solid bar was inversely proportional to the diameter to the power of 4. Still, without anything better to work from, using a linear calculation the setting link length needed was 43.1cm.

Fitting of the torsion bars
Replacing the shock absorber with the setting link provides a datum point for setting the ride heightThe calculated length of the setting link should give me roughly the correct ride height (fingers crossed etc). So I chose to set the reaction plate cam to the mid-setting ‘4’ and will be able to raise or lower the ride height if it’s not exactly right. With the setting link in place and the ‘ear’ bracket locked at setting ‘4’, the rotational positions of the front and rear splines in the suspension are fixed.

The torsion bar has a different number of splines at each end – 25 at the rear and 24 at the front. This provides a high resolution vernier adjustment, allowing the torsion bars to be set very accurately and therefore the ride height. The fitting of the torsion bar is now a matter of trial and error, rotating the bar by one rear spline at a time until the front splines are perfectly aligned with those in the wishbone.

A rotation of one rear spline is equal to 14.4 degrees while it needs 15 degrees of rotation to move on by one front spline. Another way of looking at it is when the bar is turned by one rear spline, the relative position of the front splines is altered by 0.6 degrees, in the opposite direction to the direction of rotation. The front splines will align perfectly for one of the 25 possible orientations!!

The torsion bars need to be passed rearward all the way through the 'ear' bracket. The torsion bars were protected to avoid the splines damaging the paint of the barsI had passed both splined ends of the torsion bars through their corresponding mating pieces a dozen or so times until I was satisifed it would only need three or four solid blows to hammer them home.

The torsion bar need to be passed rearward through the rear ‘ear’ mounting and then forward again until the front meets the splined hole in the lower wishbone. However the splines were still too tight a fit. It was necessary to carefully file the spline faces on the torsion bar until it only took one firm tap to fully engage the splines.

This enabled the torsion bars to be pushed forward by hand until the front was 1mm or so from the rear face of the wishbone. A tap with the hammer would then bring the bar up to the wishbone, at which point it was possible to determine if the splines were correctly aligned. I used a 12″ pointed concrete chisel for a drift, so the point could sit in the indentation at either end of the bars.

The mistakes I made were:

  • Smothering Copperslip over the front splines on both the bar and within the wishbone
  • Blindly accepting the view that it’s a matter of trial and error to find the best fit

The Copperslip did a splendid job of masking whether the splines were properly aligned and so it was all wiped off. The best time to apply it was once the correct orientation had been determined and the front splines had just engaged.

I followed the advice of adopting a methodical approach of rotating one spline at a time until an exact fit was achieved. After completing one full rotation I wasn’t convinced I was any the wiser. The correct orientation had probably been missed under the cover of Cooperslip!

It was only at this point did I sit down and work out the Maths of the relative 0.6 degree movement of the front splines for a rotation of one rear spline. A couple of minutes of thought up front would have saved several hours of grief and frustration with a club hammer! Armed with that knowledge, it was then quite easy to quickly home in on a small area of splines spanning the best fit.

As an example:

Front spline need clockwise rotation Result of rotating anti-clockwise by one rear spline Eventually an exact alignment is reached

In the left photo, gaps can clearly be seen between the splines. The front splines need to be rotated clockwise to close these gaps. The middle photo was taken after the torsion bar had be rotated anti-clockwise by one spline. The gaps have clearly been reduced.

Eventually an exact or best match is achieved. Although I found when viewed from the lower inboard (7-8 0’clock) the front spline alignment would look spot on. However when viewed from the top outboard position (1-2 O’clock), gaps would be visible.

I think this is because the angle between torsion bar and the wishbone isn’t exactly at 90 degrees. So the lower inbound splines start engaging before the top outbound splines. Hence why gaps are still visible from one view and not the other!

Finally the torsion bars were both in and I’ve now less fear of tackling them again in future.

Nov 122013

The horns suffer a harsher environment that a lot of the other component as they're located low down at the frontA pair of Lucas windtone 9H horns was fitted to the Series 2 E-Type, one emitting a high tone and the other the low tone. The excitation of the air column is achieved by vibrating an internal metal diaphragm, with the frequency of vibration and the shape of the horn snail or trumpet determining the note produced.

The switching frequencies are carefully chosen to produce a major third musical interval (spanning 4 semi-tones). Together they set up beat frequencies producing a tremolo affect and a perceptibly louder sound. In the case of the 9H, the low tone switches at 392Hz and the high tone around 494Hz, producing a G and B respectively.

Great in theory, however both my horns were stamped with an ‘H’ on inside of the trumpet indicating they both produce the high tone. Well, they would, if they both worked! One of them only produced a sound for a split second before falling silent. The only recommended external adjustment that can be made is the contact breaker gap via a small screw.

The horns were shot blasted (after blocking up the inner trumpet!) but only the working horn could be painted at this stageRather optimistically I thought it would be just a matter of readjusting the gap to get it working again. Alas, there was something more seriously wrong inside so only the good one was repainted at this stage.

One of the problems with the horns is the two halves are press riveted together. I’ve not been able to find anyone who supplies these rivets so, even if a repair is possible, it won’t be an ‘invisible’ repair.

I’m thinking of using something like Chicago screws but first I need to get inside to find out how it works and if it’s possible to change the frequency. It’s a voyage of discovery from here as I’ve not found any information on the horn innards.

The rivets were drilled and then punched out – the rivet inside the trumpet is slightly shorter than the others so I’ll have to remember that when ordering fixings to hold it back together. The two halves can then be carefully separated as the diaphragm was sandwiched between two thin, wax impregnated gaskets which are quite fragile.

Drilling out the rivets Horn carefully split in two Metal diaphragm removed

The rivets had to be drilled and then punched out to split the horn

The horn split in two - the right hand side has no moving parts and is just the horn snail or trumpet

The diaphragm removed, showing its ferrous attachment and disc operating to operate the contact points

The diaphragm was then removed to reveal the inner workings. Attached to the centre of the diaphragm is a ferrous cylinder so that its movement can be controlled by the rapid switching on and off of an electromagnet.

When current is applied, the ferrous cylinder and therefore the diaphragm is drawn towards the electromagnet. As the diaphragm nears the end of its travel, a disc around the ferrous cylinder hits the base plate of the contact breaker, opening the points. The electromagnetic field then collapses and the diaphragm returns to its natural position and the process is repeated.

Operating the base plate of the contact breaker to open the point. The points were cleaned with some 400 grit wet and dry paperThe resistance of the contact points was around 7 ohms so a light rubbing with 400 grit wet and dry soon got this down to 0.8. Although I wouldn’t have thought this would stop the horn operating. I think the problem is with an external screw fitting which the service manual suggests should not be touched.

I’m fairly sure it has been adjusted at some stage as it’s screwed tight against the ferrous attachment. Therefore stopping any possible movement in the diaphragm.

Of the two external adjustments, the small screw adjusts the contact points gaps. The service manual states that this does not adjust the tone and is only to take up wear in the points. The central screw adjustment, with locking nut, only limits the length of travel permitted by the diaphragm so if it did effect the tone it would only be marginal (ie for fine tuning). I doubt it would give anywhere near the variation to recalibrate it to the low-tone.

Hmmmm …. stumped. Going back to first principles, due to the lack of tone adjustment in the electric components. The tone must be controlled mechanically but the spring rate of the diaphragm is fixed. Therefore the only two things that I can see that would effect the output tone are the mass of the ferrous diaphragm attachment, which would naturally impact the switching frequency, and the shape of the trumpet.

Neither of these two can be changed (easily) with the parts I have in front of me! I’ve found a restorer of old horns, Taff The Horns, who might have a non-working low-tone horn to provide a donor for a transplant. Otherwise plan B is to purchase a repro horn from Holdens for about £40!

A whole post on horns without a reference to a sketch by the late Peter Cook!!

 Posted by at 8:02 pm  Tagged with:
Nov 082013

The ‘refresh’ of the alternator turned into a full rebuild because the rotor’s slip ring needed to be replaced, as it was heavily grooved, and the main rotor bearing had seized. In fact, the slip ring was found to be cracked on the underside once it had been removed. Fortunately all the components needed for a rebuild are still readily available.

The wiring diagram for the Rootes Parts Lucas 11AC alternator conversionAs previously mentioned, the Rootes Parts website covers the conversion of the internal electrics to a more modern, self-energising design.

The revised electronics energises the rotor field winding from the alternator’s output, rather than directly from the battery. This produces what Rootes term a ‘soft start’. When the car is started, the output from the alternator is zero, so a much reduced current flows through the field winding as it is wired in series with the more resistive Ignition Warning Light bulb.

The initial current in the field coil is approximately 10% of the normal operating current. However as the alternator output increases up to a typical 14.3v to charge the battery, the voltage either side of the warning light becomes equal, effectively removing it from the circuit energising the field winding. As there’s no voltage drop across the warning light, it goes out and the field winding is now powered directly from the 14.3v alternator output.

Being self energising also has the added benefit that if the belt fails, the alternator output would fall to zero and the field winding current reduces back to 10% of the normal operating current. Thus avoiding the field winding being burnt out.

The slip ring plate needed to be replaced but first needed the field wire connections un-solderedThree screws fix the split ring plate to the rotor. Once removed, the plate can be pulled away from the rotor to reveal the ends of the rotor field winding, which pass through the rear of the slip ring plate and are then soldered at the front to the brass slip rings.

It took a high output soldering iron to melt the solder joints, probably due to the age of the solder, enabling the slip ring plate to be removed.

The rotor’s field winding wire is insulated by an enamel coating and then covered by a fibreglass tape so these were protected by masking tape before the rotor was bead blasted, in preparation for painting. The rotor was painted in numerous coats of Burnt Copper and then Clear VHT Engine Enamel paint. VHT paint was chosen due to the high operating temperatures reached and the proximity of the alternator to the exhaust manifolds.

Painted in Burnt copper VHT Curing in the oven Winding re-lacquered

The rotor was painted with Burnt Copper VHT Engine Enamel paint and then a VHT clear coat

The VHT paint needed to be cured in the oven to obtain its full protective properties

The field winding was re-enamelled and the ends re-sleeved to ensure the winding does not short with the rotor body

Once the paint had dried, it was cured in the oven at around 100 degrees Celsius for about 45 minutes ….. and, rather sheepishly, the oven then left for a day or two until the aroma had gone! The fibreglass insulation sheaths covering the ends of the field winding had deteriorated and eventually broken free with all the disruption getting the slip ring plate off.

I found an electronics supplier, Brocott UK, who supply the 2mm fibreglass sheathing as well as bottles of motor winding enamel and new fibreglass tape. After the old tape was removed and the winding cleaned up, it was re-coated with a generous coating of enamel. The new sheathing had to be held in place by a small amount of Araldite were it enters the rotor as I needed to ensure the field windings couldn’t short against the rotor body.

The field winding was re-taped which was a fairly fiddly job, as the 8 prongs of the rotor get in the way. I suspect the winding was originally wound and taped before the two halves of the rotor were fitted together. Before progressing any further, two important tests were made; i) the winding resistance and ii) confirming the field winding wasn’t shorting with the rotor. The winding resistance was 4.2 ohms which is slightly above the recommended upper limit of 4 ohms but I’m happy to leave as is.

Checking the winding wasn’t shorting New slip ring soldered and fitted

A multimeter confirmed the field winding has not shorted with the rotor body. The field winding resistance was also measured at 4.2 ohms

The field winding was soldered to the new slip ring plate and the field winding covered with fresh glassfibre tape

Attention now turned to the fitting of the main rotor bearing. The original had seized and so a replacement bearing kit was ordered. The new bearings are sealed for life units which have a slightly lower maximum RPM. However I was assured they would be fine unless driving at full RPM for prolonged periods.

The outer cover plate and then the rubber ‘O’ ring are placed in the front alternator housing before the bearing is tapped into place. Finally the rear cover plate is inserted and the whole ensemble retained in place by a circlip. I couldn’t compress the parts sufficiently to fit the circlip, due to the thickness of the new rubber ‘O’ ring, and had to resort to clamping them between two suitably sized washers using a bolt. With the bearing installed, the rotor was tapped into the bearing with a nylon hammer.

Parts of new bearing kit Compressing the ‘O’ ring Rotor press fitted into bearing

New bearing kits are available. A rubber 'O' ring sits between the front cover and the drive end housing. The rear cover is secured by a circlip

It was necessary to compress the 'O' ring in order to fit the circlip at the rear

Once fitted the rotor was gently tapped into place with a nylon mallet

It was only at this point that I found a spacer that is clamped between the inner bearing race and the fan & pulley did not fit. The inner diameter of the new bearing cover was too small to enable the spacer to be fitted. So the whole bearing had to be removed and re-fitted using the original outer cover!

Alternator’s internal electronics
The stator's three field coils attached the stator to the rectifying diode housingThe rectifying diode housing and attached stator were removed from the rear housing by undoing the external nuts on the three terminal posts. Once again I had great trouble de-soldering the joints, this time connecting the stator windings to the diode housing.

The problem is that typical soldering irons, used in electronics, don’t have sufficient output to melt the solder. This is mainly because the relatively large gauged wire used dissipates more heat than the soldering iron can deliver and the diode housing is designed to act as a heatsink.

So I’m now the proud owner of 4 soldering irons, with increasingly higher output ratings! The highest being an 80W iron normally used for soldering stained glass window frames – it was ideal!

The Lucas 11AC alternators use a three-phase bridge rectifier, containing two sets of three diodes, to convert the AC output from each of the three stator windings into DC. Essentially the output of this type of rectifier is close to a DC supply as it is the sum of the positive components of the three AC voltages, which are 120 degrees out of phase from each other.

The conversion involves adding a further set of three diodes to provide an additional DC output to power the rotor field winding. The self-energising alternator no longer requires a 3AW relay so the AL terminal can be re-used for the secondary DC output.

Each terminal post has a variety of insulations and metal fittings – their fitting order was carefully noted as each part was removed to ensure they were correctly refitted during the rebuild, as shown in the photo below.

Notes taken during disassembly Parts assembled for the conversion

Details notes were taken during the dismantling to avoid problems during the rebuild

The numerous parts needed to rebuild the rectifying diode housing, including the additional diodes for the feed to power the field winding

The main points to note are the B+ and AL terminal posts must not short with each other or the alternator casing. With this in mind, the order and purpose of insulators becomes more obvious. The diode housing consists of two halves into each is pressed a set of three diodes. The two halves also act as part of the circuitry as well as a heatsink.

The left hand side is connected directly to the alternator housing via the Earth terminal and provides the common anode connection for the three (negative) diodes. Similarly the right hand side provides the common cathode connection for the three (positive) diodes, which is the B+ output.

The original AL terminal is connected to the wire connecting a pair of the rectifying diodes. This will be replaced during the conversion so the AL terminal is then connected to the output of the three new diodes.

Although the six button diodes look identical, the outer casing acts as the anode for the three inserted into the left hand heatsink (their cathodes being wired to the stator coils) while it acts as the cathode for the other three. Therefore it is essential to ensure they are installed into the correct side of the diode housing. Fortunately most multimeters have a diode test function so it’s easy to check which is which.

The same 2mm glassfibre sheathing for the rotor winding was used to insulate the 16swg copper wire connecting pairs of diodes on each half of the diode housing. Again, the 80w iron made the soldering a doddle. The next step was to install the additional three 6 amp diodes to provide the output for the field winding.

Their anodes are soldered to the anode posts of the positive diodes, delivering the B+ output, and their cathodes are all connected to the AL post. I crimped and then soldered the cathode wires into a 3/16″ eyelet terminal which fits neatly onto the AL post.

Wiring of the standard rectifying diodes Additional diodes to power field winding

There’s not a lot of room between the rectifying diode housing and the rear alternator body so insulation was added to the leads of the additional diodes. Several trial fittings were made to ensure the diode housing did not foul within the alternator body, before finally soldering the diodes in place.

In the meantime the resistance of the stator coils had been checked to ensure all was well before the stator was cleaned up – wire brushing to remove all the rust and then repainting. The stator was then reattached to the diode housing by soldering the coil field wires.

The bearing kit came with both front and rear bearing so the latter was pressed into place and the bearing completed by fitting a felt washer followed by a metal washer and finally a square, spring washer. The corners of the spring washer lock against the alternator housing and hold the other washers in place.

Rear bearing components Housing alignment markings

Two things I wasn’t sure about was whether to soak the felt washers in oil and what grease to use for the bearing. In the end I soaked the washer and opted for a polyurea grease. Hopefully it should be ok! The stator and diode housing could then be refitted to the rear alternator body, taking care to add the various insulating washers.

New slip ring brushes were fitted before reuniting the two halves of the alternator. The two alloy halves have a small dimple marking which both need to be aligned with the narrower and deeper groove in the stator.

The rebuild was completed by fitting the spacer on the rotor shaft followed by the Woodruff key, the cooling fan and finally the pulley.

Electrical terminals at rear Front & rear halves reunited

The only decision left is whether to get the alternator professionally tested off the car or wait until the engine is running …..