Norfolk Garage
Baffling gearbox issues and electrical gremlins
I’LL start this month with a confession. In last month’s column I described in some detail the problems I encountered with a Series III gearbox which was refusing to select first and second gear after having fitted various new components. I thought I had tracked the problem down to badly-machined baulk rings and a short road test seemed to confirm this: first and second still seemed rather stiff and notchy, but I put this down to all the relevant parts being brand-new and needing to bed in. So I rang the owner, arranged for him to collect the vehicle a few days hence, and got on with finishing my column.
What is the worst noise in the world? It’s the sound of a vehicle being driven back into the workshop, ten minutes after you have sent it out. The owner informed me that the vehicle had travelled only a couple of miles when first and second became almost impossible to select. The 2.5 diesel is a flexible old thing which will pull happily enough from almost no revs, and he had managed to drive back using only third and fourth. I took the vehicle for a short test drive which was enough to confirm that we still had a major problem, made grovelling apologies to the owner and parked the vehicle at the back of the workshop while I tried to work out how to fit another unscheduled gearbox removal and strip-down into my hectic schedule.
I still clung to the forlorn hope that the components might bed in with use and spent nearly an hour driving up and down the yard, changing from first to second and back. At the end of this exercise I had taken most of the skin off my left palm and the gearbox was no better than when I started. Sometimes first and second would engage with no trouble, other times they refused to go in at all. So out the gearbox came, and I soon had a workbench covered in oily bits. At this point I had absolutely no idea what might be causing the problem: all I knew was that it affected first and second gear equally badly, and all the other gears were fine.
One possibility was a problem with the interlocking plungers between the selector rails. The Series gearbox is of the ‘three-rail’ type, with three separate selector rails (first/second, third/fourth and reverse), each with a selector fork attached to it to slide the relevant gears in and out of engagement. The problem with this design is that it is theoretically possible to select two gears at once, which will not do the gearbox internals any good at all. To prevent this Land Rover fitted sliding plungers between the rails, engaging in notches in the rails and with a sliding pin across the centre of the middle rail. Engaging any one gear pushes the plungers across and locks the other two rails solid. It is a simple system and seldom gives trouble but I checked it anyway, in case a sticking or damaged plunger was physically preventing the first/second gear rail from moving. I could see nothing wrong.
Next up for inspection was the first/ second selector hub. This has a sliding outer member which engages with the dog teeth on the gears themselves, locking them to the mainshaft. When checking one of these hubs you have to be very careful as there are three detent plates between the inner and outer portions, each containing a strong spring and ball bearing. Move the outer member too far and – ping! – balls and springs all
over the place. The trick is to squeeze the detent plates between the baulk rings, which have square bosses on the inside face to stop them rotating too far (more than half a dog tooth either way) within the hub. These square bosses have a second function which is to keep the detent plates in place. The sliding action of the hub was absolutely faultless.
The hub and gears had come from different sources. Had one of them been machined to the wrong dimensions? Each gear slotted happily into the hub with plenty of clearance all-round and no evidence of anything catching. But as I was fiddling about with the gears and hubs I noticed something odd. As I mentioned last month, the dog teeth on the gears (and the corresponding teeth on the baulk rings) have pointed faces to help them slide into engagement. The baulk rings will guide the dog teeth into the hub within half a tooth of correct alignment, and the gear and hub will then twist very slightly relative to each other so that the two can fully engage with minimal effort.
At least that is the theory. On these components I could feel that if the alignment was slightly out in one direction the gear and hub would readily align themselves so that they could be pushed together. Slightly out in the other direction and the two would lock solid. I fetched my most powerful magnifying glasses from the workshop, carefully studied the components and finally found the answer. The ends of the teeth on the hub outer member had been machined off-square, leaving a small protruding lip on each tooth. Very hard to see, but just enough to snag the dog teeth on the gears and prevent them from engaging.
To be sure of my diagnosis I compared the hub with some old ones I had lying around and also with another brand-new one which I had managed to borrow. All of these had perfectly symmetrical teeth, and the gears engaged smoothly no matter how hard I tried to provoke them into baulking. The final step was to lash up a temporary spacer to hold the components onto the back half of the mainshaft, so that I could test the entire first/second assembly on the bench. With the borrowed hub fitted the gears engaged smoothly, but when I substituted the suspect hub, engagement became near-impossible. A replacement hub took a bit of tracking down, as most of those on sale were of the same manufacture as the defective one, but I found one eventually. Now all I have to do is put the gearbox back together again, which I will do as soon as I have finished writing this. Third time lucky? I do hope so.
fitted with various UK market-only electrical trickery to comply with the regulations.
Land Rover used two different systems. Readers may remember that I had a Defender 300Tdi in a few months ago which had lost its dipped beam headlights for no obvious reason. That vehicle was fitted with the later of the two dim-dip systems, which uses a special relay to activate the dim-dip: when the relay fails it cuts the feed to the headlights. I was able to bypass the relay by cutting two wires and joining them together. The vehicle I had in my workshop now was fitted with Land Rover’s earlier attempt at dim-dip, which uses system voltage and the oil pressure warning light to determine whether or not the engine is running, and feeds the headlights via a voltage regulator to achieve the necessary dimming.
The whole lot is contained in a large finned aluminium box behind the instrument panel. On this vehicle the electronics had gone haywire. The control unit was no longer picking up the oil pressure input, and the trigger voltage for the dim-dip was somewhere very close to the system voltage with ignition on and engine not running. Switching on the sidelights still ‘armed’ the system as it was supposed to. When the indicators were switched on, the slight spike in system voltage every time the bulbs went out was enough to trigger the dim-dip system and illuminate the headlights, causing a voltage drop which immediately prompted the control box to switch the headlights off again. I should add that the headlights came on at full brightness, not dimmed, so I would guess the voltage regulator inside the control box had shorted out as well.
All this explained the reluctance of the headlights to extinguish when I switched to sidelights: the switch itself was not faulty after all. The good news was that for this earlier type of dim-dip system the cure was even simpler than for the later type. I simply pulled the multi-pin plug out of its connector on the side of the control box, and all lighting functions returned to normal. A functioning dim-dip system is not, and never has been, part of the MOT test requirements, and as far as I am aware its use has never appeared in any edition of the Highway Code. It was truly a solution in search of a problem, and its absence will not be missed, as I’m sure a lot of people would agree.