My friend is a bit of a gym junkie, and I am very much not. There is a category of product that these people often use, called “protein powder”. There are a bunch of flavours, and since my friend has a side gig importing these sorts of products, he had samples of every flavour in stock and offered to let me try them all and do a writeup. So, here’s my review of every flavour of Optimum Nutrition protein powder.

These are ranked in order, from (personal) best to worse.

Chocolate peanut butter: stirs in relatively easily. Very smooth, strong “not quite peanut” flavour, much like a Hershey’s Peanut Butter cup. I normally don’t like that, but it works really well with milk. A few small floaties, but nothing major. Leaves quite a lot of “debris” on the sides of the glass. Flavour is consistent all the way down, with no weird bits at the bottom.

Double rich chocolate: relatively hard to stir in. Really likes to sit in the bottom edges of the cup. Leaves a lot of globs on the sides of the cup, and floaties. Tastes great though – milk chocolate. Consistent, no weird aftertaste but some sweet residue. Still liked it even after tasting most of the flavours and getting a bit tired of all the tastings.

Extreme milk chocolate: easy to stir in. Lots of small floaties but no big globs. Came out of the bottom edges easily. Basically the same as double rich chocolate, but a more mild flavour and no sweet residue. Very smooth texture and consistent flavour.

Banana cream: somewhat difficult to stir in – likes to sit in the bottom corner. Basically impossible to get rid of the floating globs. Very strong “synthetic banana” taste (like a milkshake), which I happen to like. But it is strong.

Mocha cappuccino: somewhat difficult to stir in – sits in the bottom corner and leaves plenty of clumps. Gentle flavour, only coffee-flavoured as much as an “iced coffee” flavoured milk. Very smooth texture, and consistent taste all the way through.

Chocolate mint: difficult to stir in. Floats aggressively, sticks to corners, and forms a thick layer on top. Taste is quite good though – the powder smells like straight mint, and I don’t normally like mint, but you can barely taste the mint once stirred. If you’re a big mint fan, this is probably disappointing. Flavour is consistent, but the texture was one of the stickiest.

Chocolate hazelnut: very easy to stir in, and doesn’t leave many clumps or floaties. Taste is good, but overwhelmingly hazelnut rather than chocolate. Very smooth texture, and no “sweet” bit at the bottom – consistent flavour.

Rocky road: smells surprisingly authentic – chocolate with marshmallow. Easy to stir in, and didn’t stick to the corners or form clumps. Taste is not very distinctive, I couldn’t tell it apart from chocolate hazelnut if it weren’t side by side. Taste is smooth and consistent though, with no “sweet bit” at the bottom.

White chocolate: Stirs in very easily – no floaters or stuff hanging on to the side of the mug. Smooth, consistent taste all the way down. Quite malty with a fairly mild white chocolate taste – it reminded me strongly of the vanilla flavours. White chocolate is a very personal thing; I like actual white chocolate and also liked this flavour.

French Vanilla Creme: very easy to stir in, no clumps and nothing getting stuck to the corners or sides. Very strong vanilla flavour, and quite authentic too. Very smooth texture, which was consistent all the way down.

S’Mores: relatively difficult to stir in. Congeals and forms floating clumps that are hard to break up, but at least didn’t stick to the side of the cup. Does stick to the spoon slightly. Flavour is good, but odd – tastes like some sort of malty Arnott’s biscuit. Chocolate undertones. Flavour is consistent all the way down.

Strawberries and cream: easy to stir in, with none sticking to the sides, corners, or spoon. Smell is like strawberry Chupa-Chups turned up a notch, but taste is a lot smoother – good balance of creaminess and (artificial) strawberry, with a little bit of the “sharpness” associated with that flavouring. Small amount of floaties. Flavour very consistent throughout.

Strawberry banana: moderately difficult to stir in. Doesn’t leave large floaties or stick to the side of the glass, but does sit in the bottom corner. Texture is reasonably smooth but has a small amount of floaties. Taste is smooth and consistent, more strawberry than banana though.

Vanilla ice cream: easier to stir in. Very smooth texture, no globs on the side of the cup, or floaties. Flavour not great – strong vanilla flavour, but a slightly tangy undertone? And a slightly greasy mouth feel. Weird sweet patches while drinking. Was not happy about having to finish this sample.

Cinnamon roll: moderately easy to stir in. Didn’t sit in the edges of the cup but was hard to “thin it out” and it retained quite a thick (but smooth) texture. Does not taste like cinnamon, more of a weird off-brand vanilla, kind of malty? Left a lot of gloop on the sides of the cup. Taste changed while getting to the end, some of the cinnamon flavour seemed to settle.

Coffee: easy to stir in. A few small floaties and doesn’t stick to the cup. Flavour is more like iced coffee, which is good because I don’t like coffee. Mouth feel is very smooth – it really is like drinking iced coffee flavoured milk. Flavour was consistent all the way through.

Salted caramel: smells aggressively of caramel, and reminds me of the cookies and cream. Does not taste anything like it smells – it’s a bit like French Vanilla Creme, but there is definitely an artificial form of salted caramel on top. Almost like malt? Stirred in quite easily, only formed a few clumps at the top and did not stick to the sides at all. Texture is very smooth all the way down. Mild “sweet spot” at the bottom.

Cookies and cream: very easy to stir in. Did not stick to the corners or sides of the glass, and left no floaties. The flavour, however, seems practically identical to vanilla ice cream, with the addition of small granules of black flavouring. The granules did taste like cookies, but they sank immediately, leaving the flavour not uniform at all.

Delicious strawberry: moderately easy to stir in. Sat in the bottom of the cup a little bit, but not too bad. Left a small number of clumps that were impossible to break up, but otherwise smooth. Powder had an interesting texture – more like sand than flour (which most of the other flavours were). Flavour, consistency and smoothness basically identical to strawberry Nesquik. Unfortunately, there were occasional very bitter pieces even after mixing thoroughly.

Chocolate malt: very difficult to stir in. Very buoyant, so ended up with a thick layer of uncombined goop at the top. Also really likes to stick into the corners of the cup. The mouth feel of this floating goop is quite sticky and hard to dislodge – really needs work to get it all out of your mouth. Sticks to the spoon and leaves unmixed gobs down the side of the cup too. Flavour is alright, it really is what it says on the tin – chocolate with malt. Besides clumping issue, flavour was consistent all the way down.

Milk tea: easy to stir in, with only a small number of floating clumps and doesn’t stick to the corners of the cup. Flavour is awful – a very poor facsimile of tea, with a weird strong overtone that I can only describe as malt but sharper. The flavour is, at least, consistent throughout the drink.

1. Uninstall emacs, make sure tree-sitter goes with it.
2. Download tree-sitter-0.24.7-1-x86_64.pkg.tar.zst from the Arch Linux Archive
3. Install it manually: sudo pacman -U tree-sitter-0.24.7-1-x86_64.pkg.tar.zst
4. Add tree-sitter to IgnorePkg in /etc/pacman.conf.
5. Install emacs from source (go to the GNU website; download a tarball; extract; ./autogen.sh; ./configure; make sure it says Does Emacs use -ltree-sitter? yes; make -j16; sudo make install).

6. Use tree-sitter grammars at a suitably old version. This diverges based on your setup. I use casouri’s helper, so after cloning that repo, I need to apply the following patch:

       diff --git a/build.sh b/build.sh
       index b0920fb..ab70e13 100755
       --- a/build.sh
       +++ b/build.sh
       @@ -151,6 +151,9 @@ case "${lang}" in
       "proto")
       org="mitchellh"
       ;;
       +    "rust")
       +        version="v0.21.2"
       +        ;;
       "scss")
       org="tree-sitter-grammars"
       ;;
   

7. Then, run JOBS=16 ./batch.sh and cp dist/* ../tree-sitter.
8. Open a random Rust file to check that it’s all working properly. If you have syntax highlighting, it works.

Olive Oil Blend

Too oily.

Brine

The container was too big and hard to open.

Chilli

I think this was yummy, my favourite out of the chilli ones for sure. I do think this flavour is permanently ruined for me as a result of Clarisse’s tuna oatmeal “incident”.

Thai Red Curry

This one smelt exactly like the real deal. I, along with others, were all so flabbergasted by the accuracy of the tuna. However, the taste was much more bland than anticipated. I felt let down, I felt disappointed, I felt ashamed.

Fiery Tuna and Tangy Jalapeno

This one was okay. I did not enjoy the texture of the jalapenos. It felt like really gelatinous and the jalapenos reminded me of aloe vera.

Lemon & Cracked Pepper

This one was quite yummy. I love lemon, so this was good for me. I felt very sophisticated eating this one. I think this would be super yum on some crackers, although, I think I would sick of it so quick.

Onion and Tomato Savoury Sauce

It tastes like a Kmart version of Onion and Tomato, but I cannot pinpoint why.

Mango Chilli

Absolutely feral. The texture of the mango was so gross, it was like loose skin that had been soaked in acid. The mango tasted like a mango scented candle. I was very excited for this one too. Very upset.

Mild Indian Curry

It’s ok, worse than the Thai Red Curry.

Naturally Smoked

I feel so neutral about this one. It’s probably yummy in a sandwich or something.

Oven Dried Tomato and Basil

This one was so nasty. It tasted like basil with the essence of tuna. It tastes like if you went to a nice italian restaurant and your waiter was a feral cat who breathes in your face while you eat.

Sweet Chilli

I usually like sweet chilli flavoured things, but I think it was a bit too sweet and not enough chilli. It was ok.

Sweetcorn and Mayonnaise

MY FAV FOREVERRRRR so delicious and scrumptious in my tummy. I could eat this for ever and ever and it even has corn in it?!!! So fun when you poo out the kernels. The mayo also super yummy. No one else liked it though.

Zesty Vinaigrette

I love vinegar but this one nasty af dawg.

Springwater

A staple. You cannot hate on this one. I always mix it without condiments though, so I guess I don’t love it on it’s own.

Onion and Tomato

YUMMY. This is a childhood favourite. It was very nostalgic and brought me back to when I would bring it to school for morning tea and then struggle to open it and then spill it on my uniform and then my mum would get mad at me and then I would have to wash it but for some reason she kept packing it for me.

Thousand Island Dressing

I think they might be recalling this one, but this TECHNICALLY is my favourite flavour of tuna. I will say, it is not bang for your buck. It’s like 10% tuna and 90% sauce. So maybe I don’t like it as much as before. Also the recall warning is too scary, I will not be purchasing again.

Will’s Notes:

• Mango chilli: has a whole chilli, and chunks of mango. I try to be objective but this was a very strange clash. I appreciate the bravery. Mango has no texture and disintegrated.
• Springwater: much firmer texture than other chunk style flavours for some reason. Less salty, slightly less smooth and more fishy.
• Naturally smoked: VERY smoky, which I liked. Smooth texture, but not very strong tuna taste which might be a downside.
• Thai red curry: has whole curry leaves. Smells coconutty? Really does taste like jar sauce curry. Leaves don’t have a strong flavour.
• Onion and tomato savoury sauce: less harsh than the tomato and onion, and smoother.
• Zesty vinaigrette: visible herbs, very lemony but more chemical.
• Tangy jalapeno: chunks of jalapeno, and the right amount of jalapeno flavour. Not a lot of spice overall but the chunks do have a little spice.
• Oven-dried tomato basil: not sure about this one. A LOT of basil, not much tomato.
• Sweet chilli: very different to the chilli. It tastes like the subway sweet chilli sauce. Not spicy at all.
• Tomato and onion: more like a relish. Surprisingly harsh.
• Mild Indian curry: smells like butter chicken. Tastes like it too, but very mild.
• Chilli: contains a whole chilli, which has some kick. The tuna is also hot, but it dissipates quickly enough.
• Olive oil blend: very strong olive taste, and quite salty. Not a tonne of tuna flavour. Good texture.
• Lemon and cracked pepper: only really tasted the lemon, but it was good overall.
• Sweet corn and mayonnaise: needs mixing. Mayonnaise was weirdly tangy, corn didn’t seem to add to the flavour.
• Brine: bland and salty. A lot like the springwater, but a little better.
• Thousand island dressing: needs mixing. Very little tuna, pasty texture. Flavour is reasonably close to what’s on the tin, but texture is more like sauce.
• Spread: red mystery. A little tomato, a little onion, a little ranch? Mild flavour.

And for the two custom additions.

• With Nutella: the Nutella did not mix into the tuna at all, I suspect because the tuna is oily and the Nutella is water-based. As a result, there were two totally distinct flavours that didn’t mix in the mouth at all. So it wasn’t unpleasant - more like eating some tuna, then eating some Nutella.
• With peanut butter: this mixed in a little more, and was about as unpleasant a clash as you’d expect.

Before eating any tuna, we ranked how good we thought the various flavours would be, so we could compare them to our ranking afterwards. Here are the full results:

The bolts are part 12 in the diagram. The bevel pinion housing is 11, and the locker replaces parts 2, 5, 6, 7, 8, 9 and 10.

The threads situation had to come first, so I ordered some new differential bolts. Once those arrived, I tried to wind the old bolts out (into the axle case) but in the process I also noticed the securing wire around the perimeter was broken on both sides (there are two wires, one on each side of the cutout for the ring gear).

This is a picture off the internet showing the axle housing - you can see the two cutouts for the ring gear, and the bolts in situ around the perimeter.

Not only that, but I also noticed while breaking the bolts loose that two of them were bent, and badly enough that they would’ve damaged the threads in the axle case, which would’ve been very difficult to repair. I decided to just cut the old bolts short, so I’d only have to wind out the 10mm of thread already in the axle casing. This worked fine (about the only thing in the rear ATB installation that did), and I installed the new bolts.

I decided not to replace the locking wire, and use Loctite instead. They can’t back out once installed, since the nuts on the outside of the bevel pinion housing (part 37 on the diagram) hold them. The only time that’s risky is during installation, where it would be possible to wind the bolts further into the axle casing.

Once the new bolts were installed, I had another shot at installing the diff, and it still wouldn’t go in. It wasn’t even going as far as the studs, so it seemed like the bevel pinion housing was incompatible with the axle case. With this in mind, I rebuilt the Torsen unit into the old rear diff, figuring that the diff that came out would fit back in.

It didn’t.

In fact, it was exactly the same problem - something was getting caught up, and it felt like the ring gear was hitting the axle case. I did some research into the bevel pinion housing part numbers to check if the locker should’ve fit my housing. It turns out all three of the differentials I have (the two originals, and the one I bought that’s now in the front axle) have part number 219636, which is actually from a Rover P4. This is considered an upgrade, since it’s off a more powerful car and is stronger. The part that the manual says I should have is 528257, which is available but only second-hand from the UK. There’s also part number 219635, which is a late Series I diff. I believe either of these would likely be compatible with the Truetrac, but they’re not what I had.

I emailed Ashcroft to see if this was a known issue, and while they weren’t sure about my part number, they mentioned it was common that they had to grind down the bevel pinion housing to make things fit. That got me thinking, so I wrapped a layer of masking tape around the likely contact points on the differential to see if it was getting hung up on the housing itself.

Doing a test install and looking at where the tape got damaged, the problem was the corners of the head of the bolt protruding into the opening of the axle case! You can see a little bit of this in the axle casing picture above. So I ground the bolts down a little, and the diff could now get a lot further in.

The next problem was the bolts again. Whatever bent the bolts originally had also tweaked the axle case, so now two of the bolts stuck out from the casing at an angle at wouldn’t line up with the bevel pinion housing. I ran a bunch of nuts down the bolt and abused a socket and breaker bar to bend the casing back in line, and the diff slotted right in.

After that, it was the usual process of sealing things up, tightening it down, reinstalling the halfshafts, filling it with oil, and test-driving it. I’m happy with it, and now I have front and rear Torsens.

Apologies for the lack of photographs in this blog post, I simply forgot to take them as I was going.

Starting this post with a diversion: I’ve been using the Series II at least once a week, and it’s been working great. For example, here’s a picture from a car show I attended. I didn’t mean to enter it, I just wanted to go look at other cars. But the people directing traffic sent me to the exhibition area, which had the nice side-effect that I didn’t have to pay for entry. It doesn’t fit in that well with the polished-up sedans and sports cars, but oh well.

Diff rebuild

The differential rebuild was sitting for a while after I got it apart, since I didn’t really know whether I wanted to rebuild it back to factory spec or put some sort of air locker in.

I ended up going for neither option, and went with a Detroit Truetrac differential, which is a Torsen unit. I went this way for a few reasons. One is that I don’t have to worry about wiring up a compressor and switch, routing air lines, and running the air line into the differential casing (which would require drilling a hole in it). I don’t mind making reversible changes to the vehicle, but a hole in the differential housing is not in that category.

There’s a tonne of information elsewhere on the internet about Torsens vs locking differentials. My two cents is: a Torsen is better in some situations where you don’t want a locker, like a side slope, or snow. The main downside is that when a wheel is in the air, the wheel on the ground gets no torque. But in that situation, you can gently apply the brakes and get some torque to the wheel on the ground.

The second advantage is that it’s passive, so I don’t have to remember to switch it on. This is nice off-road – one less thing to think about – but it’s also nice on-road, where I often spin up the inside rear tyre in the wet. The skinny 6.00x16 bar-tread tyres aren’t much good in these conditions, and a Torsen diff would help with traction.

Once I’d made that decision I got on with the rebuild. I had the pinion depth set (see my previous blog post on this topic), so the next step was to get the pinion bearing preload sorted out. This took quite a while, since the combination of shims I’d ordered and shims I removed during disassembly didn’t get me to the right preload.

I ended up making a small jig with a mirror and a wooden block to polish down one of the shims (using the mirror as a very flat surface) to get it to the right thickness. You can see the nylon string I was using to measure the turning resistance. The workshop manual says there should be between seven and twelve pounds of resistance when measured on a spring balance, but then mis-converts it to “between 3.2 and 4.5kg” – I believe the latter figure should be 5.4kg, assuming the imperial value is canonical.

Once I’d got the pinion depth and preload correct, it was time to install the Torsen unit. This replaces the spider gears and associated casing. Unfortunately, I was unable to source the “correct” part for my vehicle, which is Detroit Truetrac part number 910A460 for the rear and 910A461 for the front. However, part number 910A400 is compatible with the 10-spline axles.

That part (910A400) has a few differences. The first is that it’s meant for the 3.54:1 differential instead of my 4.7:1. The main difference between the two is the pinion offset, which is fixable using a spacer ring sold by Ashcroft Transmissions.

Adding this spacer ring makes the ring gear “thicker” than it originally was, so I used some 1-3/8” bolts to secure the ring gear to the carrier (instead of the 1-1/4” bolts that were originally used). I don’t think this was strictly necessary in hindsight - the original bolts sat basically flush when installed (instead of having a few threads proud) so there was adequate contact even if it wasn’t up to a machinist’s standards. And the locking tabs make it impossible for them to walk back out. The torque spec isn’t excessively high, either (48Nm).

The other issue is that the 3.54 Truetrac uses metric bearings rather than the imperial-sized bearings used on the 4.7 differentials. Ashcroft has something for this, too – a pair of adaptors that can be used to fit metric bearings into the 4.7 differential bearing caps.

So you press the bearing race into these adaptors, then continue building the differential as normal.

Finally, there was another problem that possibly didn’t need to be solved. The “correct” Truetracs came in a front and rear version, whereas the 3.54 Truetrac is only available as a rear. My understanding is that the rear units can be used in the front, but they’ll be running “backwards” and the thrust forces from the worm gears will be towards the centre of the diff rather than the end caps. There’s also some debate about whether the amount of preload in a front Torsen is different, but that’s not something you can change so I ignored it.

Happily, there are guides out there on reversing Truetrac differential directions, including this one. I followed this, and, with the combination of the spacer ring and bearing adaptors, had a differential ready to go.

The only thing left now was to measure the crown wheel run-out, get the backlash between the ring and pinion gears correct, engage the locking tabs, and do up the bearing cap locking wires.

The locking wires in this picture are a little messy - they were my first attempt, I cut them back off and re-did them (as you’ll read below).

After I was happy with the differential on the bench, it was installation time. I wanted to re-seal the swivel housings anyway, both of which had developed a slow leak out of the bottom swivel pins (surprisingly, the large circular seal that runs on the inner swivel ball was fine). So I removed the halfshafts from the existing differential by opening the swivel housings and removing the front halfshafts. Normally, this is the long way to do it.

Anyway, with the Torsen differential installed, I had a problem. The manual that comes with the Torsen unit says that with the input shaft locked, it should be possible to spin one front wheel, and have the other wheel turn the opposite direction. However, the entire unit was locked solid, almost like something was seized up internally.

After sleeping on it (the installation process had been quite a lot of work) I decided that there was a fair chance I’d messed up the reversal process, and that I should remove the differential to see what was happening. So I took the diff centre back out, and disassembled it on the bench to get at the Torsen unit again.

It did seem seized - turning one of the axle splines didn’t make the other turn the opposite direction. So I started disassembling the Torsen unit itself (again), when I noticed that with the end caps off, there was no preload and I got the desired behaviour. So that’s a lesson for new players - the Eaton manual for the Truetracs does not consider that the preload in the Torsen unit might prevent the test from working.

So, an hour after pulling the rebuilt diff out, it went back in again. There wasn’t any change in behaviour but I decided to just test drive it and see what happens. Lo and behold, it worked great! It was much quieter than the old front differential, although I think the preload has a small effect on steering - I find myself having to make small adjustments around centre a bit more often. Perhaps I’m imagining it, and perhaps it’ll go away as the differential wears in.

I’m a happy camper with the Truetrac, and I’ve ordered a unit for the rear. I’ve disassembled the old front differential and figured out why it was noisy - the backlash is 1mm instead of the factory-spec 0.2mm, and the pinion preload is nowhere near high enough. But the gears are in good shape, and it’ll serve fine after being rebuilt.

One side-effect of the front-end work was that I got a chance to re-seal the swivel balls, differential housing, and differential drain plug. I’m happy to report that these are no longer leaking. The RTV I’ve found to work best is JB Weld Ultimate Black, which I put on the shims for the bottom swivel pins, the gasket between the differential and the axle casing, and I’m also using it as a thread sealant for the differential drain plug and the bolts for bottom swivel pins. So far so good.

I took it back to the same instrument repair place as last time, and when I got it back, the thing was completely dead. This is more of a problem than it sounds - the immobiliser trips if the instrument panel is missing or unconfigured, so the vehicle won’t start. It also won’t shift out of park, so it can’t even be towed when in this state. (There is a way to get it into neutral, but you have to get underneath it and physically move the transmission shift cable.)

Luckily, Land Rover had new replacement clusters in stock, so I simply ordered a new one and installed it. But things got a little more interesting after installing it on the vehicle.

The typical replacement procedure is to download the CCF (car configuration file) from the old cluster, then swap in the replacement cluster with the vehicle still powered on. Then program the backed up CCF into the new cluster, do an odometer sync and a VIN learn, and off you go.

This wasn’t going to work for me since the old cluster was dead. Thankfully the GAP IID tool does have a way around this: first, flash the instrument panel cluster to factory settings. Then you should be able to do an odometer sync and VIN learn like in the previous procedure.

This probably only works for new clusters - there is a separate process for used clusters, because the vehicle’s computers are (understandably) reticent about winding the odometer backwards (in the case where the used cluster has more kilometres than the car). As an anti-tampering measure, the odometer reading is backed up in several modules, even the lighting control module!

The other downside is that it loses a lot of settings - time, saved radio stations, and probably other things. That’s all fixable, but still worth knowing.

Anyway, that wraps up a quick update on the instrument cluster, and hopefully provides some useful information for other people that might be struggling.

The good news is, it’s all working now - although the sound it uses for the indicator tone is slightly different so that’ll take some time to get used to.

So I needed this tool. And there aren’t many left.

There was one for sale on eBay, but it’s unusably rusty:

That gauge block fastened to the top of the cylinder needs to be accurate to +/- 0.02mm, so clearly that’s not going to work.

After quite a lot of time looking, I sourced another, and figured I’d post the dimensions on this blog so that anyone else looking for this tool can save a bunch of time and money and simply make one (or get one made). It’s a simple thing - just a cylinder and a block. The original one has some fancy end caps and a long threaded rod to keep both of those bits together, but it’s not necessary.

Anyway, here’s some pictures of what I sourced. (That’s not rust on the cylinder, just some protective wax from storage.)

As you can see, this thing is in great shape, so I feel pretty good about posting the dimensions here.

Cylinder:

• Length 187mm +/- 1mm
• Outermost sections: OD 73.32mm +/- 0.02mm, width 15mm +/- 1mm
• “Shoulder” sections: OD 76.28mm +/- 0.02mm, width 24mm +/- 1mm
• Centre section: OD 76.28mm +/- 0.02mm, width 53mm +/- 1mm

The un-machined gaps between centre and middle sections are about 25mm.

The gauge block is 63.54mm x 12.26mm x 37.00mm, all +/- 0.02mm.

With this, I’ve been able to progress the differential rebuild - I have the pinion height set perfectly, albeit with a lot of faffing about since I don’t have the special tool for removing the inner pinion bearing that the shims go under.

I’m currently waiting for parts again - this time some shims for the flange end of the pinion, where I need a bit more distance to get the preload correct.

I didn’t think much of it until a recent trip, where after getting the car very wet, I parked it and restarted it shortly after only to find the entire instrument cluster dead - no gauges, no backlighting, nothing. Thankfully the car still ran fine.

The cluster was back to normal the next morning, but obviously this needed to be sorted out. I started with a wiring diagram (from the Haynes manual) which joined a few dots together - there’s one big connector at the back of the cluster, and one of the wires leads directly to the coolant level sensor (a simple on/off switch as far as I can tell). There’s also CAN wiring, which would talk to the air suspension. These disparate issues were all pointing to the cluster as the common failure point.

The bulk connectors on these clusters have some interesting failure modes as they age, including cracked solder joints and green crust from corrosion. So I proceeded to take the cluster out, which wasn’t too difficult.

After getting the cluster out, I opened up the white casing to see what I could find. I did spot a few tiny cracks in the solder joints where the bulk connector (green rectangle) is surface mounted to the cluster PCB, unfortunately too small for me to get a decent photo. I’m not set up for such a small solder pitch so I dropped it off at a local instrument cluster repairer.

I got it back a few days later, and haven’t had any issues since. The repair cost $150, so I think that justified sending it out instead of buying new soldering equipment and making this my first experiment in soldering anything this small.

I took the car for its annual safety inspection, which it passed with no real issues. I took it to a brake shop that also does these inspections, since I had a pulsating brake pedal from an out-of-round drum that needed some work. It also pulled to the left a bit.

After machinining one fo the drums and readjusting, these two known issues were fixed. It then passed the inspection on the first attempt. However, the shop noticed that there was a small oil leak from the hub seal, which was starting to contaminate the brake shoes.

Luckily it was early on, and the shoes didn’t need to be replaced, but I did need to get the seals replaced quickly to avoid a safety issue (and having to replace the shoes).

Hub seal replacement

This was a pretty straightforeard job, and I decided to do both front wheels while I was in there. The distance piece for the bearing sleeve, that the hub rides on, was something I didn’t replace during the initial restoration because I didn’t realise how important they were for the hub inner seal to work correctly. This is the better side - as you can see, they were both very worn.

The process was basically drama-free, except for the left-hand brake backing plate not wanting to separate from the stub axle, and me getting my thumb with a hammer in the process of separating them. One trick I found was to hang the backing plate off the steering arms, to avoid having to remove them (or hang them off the brake hose, which I don’t like).

Some more random photos from the process:

I took it for a run today, and it all seems to work well. The brakes need bleeding for some reason - possibly something the shop did, but that’s a very minor deal and I’ve got a friend visiting next week to help with that.

Solex carburettor reinstalled

I ended up purchasing a rebuilt Solex carburettor from the US, since I wasn’t sure my old one would be in a state where it could be rebuilt. I also got a rebuilt Zenith just as a spare, but I prefer the Solex since it’s easier to tune, apparently works better off-road, and is period correct for the car.

Instead of cutting/welding the relay lever to refit the Solex, I bought a second relay lever, so now I have parts on hand to use either a Solex or a Zenith. I also took the opportunity to replace the crappy Britpart plastic ball joint sockets with some higher quality metal items. I sourced these from Dingocroft, but searching for “2BA ball joint” led to plenty of options. That’s something I’ve been learning - if a part is generic, like a ball joint, search for the generic part rather than using Land Rover part numbers.

I’ve been doing longer drives with the new Solex since it’s been running so well, but on a very hot day I noticed that the oil pressure light was flickering at idle. I wasn’t too worried since it developed good oil pressure at higher RPM, but I still wanted to fix it.

The issue turned out to be the engine idle speed dropping too low at higher temperatures. From stone cold I measured it at about 650 RPM, and then after getting it up to about 60°C on a drive, measured it again and got 320 RPM. I adjusted it back up to 500 RPM per the workshop manual, but I’ll have to wait for another hot day to try again (I started having problems at 90°C).

Springs levelled

This is another job mentioned in the previous blog post that I’ve since completed, with the modified springs installed the vehicle now sits much more level.

Transfer case oil leak

There has been a slow oil leak from the bottom of the transfer case since rebuilding it, but it recently got worse and I finally decided to do something about it. I initially thought the leak was from the sump gasket, which is a notorious leak spot. I tried various permutations of Permatex sealants, but nothing was working right until I tried JB Weld Ultimate Black.

This helped a lot, but didn’t reduce the leak to zero. I did a little more diagnosis and found that there was a second slow leak around the area of the seal between the speedometer drive housing and the transmission brake backing plate. So I disassembled this and found that the seal for the output flange had worn a slight groove in the flange (some time before the rebuild). At the time of assembly I didn’t think this was enough to cause a leak, but I was wrong.

I found a Speedi-Sleeve of the appropriate size (part number 99162). After installing that and renewing the seals in that area for good measure, I no longer have a transfer case oil leak.

Oil pressure and water temperature gauges added

For my own peace of mind on longer trips, I’ve installed aftermarket gauges for oil pressure and coolant temperature. They’re installed in a way that can easily be reversed, although it did necessitate converting the vehicle to negative earth.

Besides the battery (obviously), there are only three polarity-sensitive parts in the vehicle:

• The ammeter (fixed by swapping the connections)
• The ignition coil (again, swapping the connections fixed this)
• The generator (fixed by “re-flashing” with battery voltage across the field coils)

I made up a custom bracket for the cabin installation, although I also needed a T-piece with weird threads in order for the original oil pressure warning light to work at the same time as the gauge.

Small radiator leak

I noticed a small, pinhole leak after a longer drive. I considered trying to solder it up from the outside with the radiator in situ, but I decided to remove the radiator and have it gone over professionally.

I’m glad I did that - turns out the last place did a rough job in a couple of spots, so it’s all been silver soldered and pressure tested again. Interestingly the drain tap failed the pressure test, so I’m just using a bolt. This has been working great since, even on a hot day in traffic after a highway run.

Coil moved back to bulkhead

This wasn’t a functional thing, but I never really liked how the engine bay looked with the ignition coil mounted on the valve cover. So I got a longer HT lead from the coil to the distributor, and now the coil is back where it should be.

I also installed an inline fuel filter - it’s not necessary since there are already gauze filters in the fuel pickup, pump, and carburettor inlet elbow. But it’s nice to see what’s going on.