Zel's Motoring Adventures...Peugeot, Renault, Rover, Trabant, Invacar & A Sinclair C5 - 19/04 - HVAC Preemptive Investigation...

[Zelandeth]

Having discovered such plentiful quantities of mould in the boot of the Jag I decided to go in there armed with a bleach solution before going any further.  Hopefully I've now eliminated as much of it as possible.  There's never really been any sign of damp in the boot since I got the car anyway so I'm hoping it's just a sign of a historic leak.

The carpeting which covers the fuel tank was utterly disgusting you can see in the above photos, so this was pulled out and put through the washing machine.  I'd like to do the same for all the carpeting in the boot, but it's all quite firmly glued down so that's not really practical.  The fuel tank trim though didn't clean up half bad - a lot better than I was hoping for really.  That piece is just glued in at two spots so is relatively easy to remove.

I know 90% of it is hidden behind the spare wheel, but I feel better in the knowledge that the carpet doesn't contain its own ecosystem any more.  I do note that even after a trip through the wash that this carpet still smells quite strongly of petrol, so I am leaning more towards the thinking that the smell in the boot may indeed simply be residue from a leak some point in the distant past.  In which case getting rid of it entirely may well be challenging without stripping out and replacing all the soft surfaces.

Will get that put back in later today hopefully.

While I was in there I figured it was about time that I actually put the battery clamp back on.  It has been rolling around in the boot since I got the car.  I suspect on account of the fact that you need to assemble it *before* you attach the battery leads.  A fact which I'm sure has resulted in many XJ-S owners spending a not insignificant amount of time with untethered batteries when that realisation has dawned right after the owner has reset the clock and reprogrammed all their radio stations.

I also discovered that the threads on one of the retaining hooks were damaged preventing it from being tightened up properly.  This was resolved by grabbing a few random nuts out of the drawer of random fasteners to use as spacers allowing me to use an undamaged area of the threads.  Not pretty but it's hidden away under a cover normally and does the job of ensuring that things are secure.

I also made an adaptor out of a bit of brake pipe offcut to hook up the battery vent tube properly. 

Probably the first time that's all been properly bolted in place in a while.

Not a huge battery for the size of engine really is it...at least the leads are satisfyingly substantial though.

The strap over the top would originally have held in place the vent assembly which used to sit over the top of the cell access caps on the original battery which are obviously not accessible on this modern maintenance free type.

Then the covers were all put back in place and given a wipe down.  Looking a bit less disgusting now I think.

The fuel pump cover really wants a lick of paint too, but that's about item number 39409385398 on the priority list.  Likewise there should be two press-fit clips to hold the cover over the battery on.  They are quite a pain to remove though so rather than replace the missing clips like-for-like I'll probably put a couple of thumbscrews there which won't require wrenching on a relatively brittle bit of plastic to remove.

 

[Broadsword]

Seeing those photos of the XJS boot gives me memories of my own XJS. There is no two ways about it the design is a absolute mess.

[Zelandeth]

The pipework and wiring in the vicinity of the battery is indeed a bloody mess.

The rear carpeting has now been reinstated.

[Zelandeth]

Not much been going on car wise the last couple of days as we've been going flat out to try to tame some of the greenery around our place.  By mid afternoon today we were nearly dead, but have probably got through slightly less than an eighth of the required work.  Need to get the skip swapped out before we get much more done.

Most of that came out of this corner.

Number 6A is seeing daylight for the first time in probably the best part of a decade.  I would usually go to slightly greater lengths to compact what's in the skip, but about 80% of that is sodding hawthorn so I'm not going near it.  I was wearing welding gauntlets and a leather apron while loading it and I still got cut to ribbons.  Most of the next loads will be far less pointy nonsense which I'll be far more willing to spend time flattening.  There is a bit of random stuff I'll stuff in a few of the gaps tomorrow as well.

This evening I wanted to get something on my own to do list ticked off so decided to get the oil and filter on the van changed.  I don't actually usually use the sump plugs on many of my vehicles much these days, I usually just use the Pela to drain it out the top.  The van is the exception to that rule just because it's so easy to get to the drain plug.  I do have to remember not to use my usual drain pan though as this thing takes 7.5 litres of oil...so draining it into a 5 litre drain pan will end badly.

Plenty of room for bigger containers though.

Oil filter is the most fiddly bit on this to be honest.  You have to remember to have a drain pan on hand when you crack the lid off the filter housing as about two pints drain out of the filter housing, oil cooler and the lines. 

It's really awkward to get everything out without making a mess as the filter is attached to this thing...

...and as soon as you tip it anything resembling sideways (which you have to to extract it around the suspension) it then dumps another half pint or so of oil out.  This is the fourth time I've done this, I was being careful and still managed to make a mess.

There's also obviously no way to prime this thing with oil before restarting the engine.

I did give the housing a good wipe down to ensure any gunge was removed though before putting the new filter cartridge onto the base plate, replacing the outer O-ring (new one supplied with the filter along with a new copper washer for the sump plug) and putting it back together. 

Fourth time and I actually put the cover on the right way up!  Doesn't actually make any difference but the text being upside down bugged me...

Sump plug back in, seven litres of oil went in (I know from prior experience it drops markedly after the first start due to the oil filter and cooler needing to prime), and once the engine had been run to circulate things the final 3/4 or so of a litre was added.

I know this oil only has a couple of thousand miles and six months on, but you wouldn't know it!

The ability of it to serve as black paint is demonstrated after the container has been decanted into my waste oil bottle!

I do seem to still have an oil leak on the offside front of the engine that's running down over the alternator frame.  I had thought it was the cam cover when I got the van but I changed the seal for that a year or so ago.  Next obvious candidate would be the oil filler neck.  I'll have a closer look shortly to see if I can more accurately pin down where it's coming from.  It doesn't ever drip but it would be nice to have the engine looking nice and dry when she goes in for the MOT.

I note that I've been doing an awful job of remembering to note the mileage down in the maintenance spreadsheet for the van so made a point of taking a photo of the odometer this time so I couldn't forget.

Have the dates I've done pretty much everything down, but mileage is missing from quite a few which is annoying.   Yes changing it every six months is probably overkill given how few miles I do, but it's been done for the last thirty years so would be a shame to stop now.

At least I didn't manage to pour oil up my sleeve this time!

 

 

[Zelandeth]

The Jaguar XJ-S is officially absolutely the most aggressively awkward car to work on I have ever owned.  Yes, even compared to the Activa.

Today I decided to tackle the rather ragged looking belts.  I was fully expecting this to be exceedingly awkward simply because of how deeply buried down the front of the engine they are. 

I underestimated quite how awkward this job was going to be.

While it *looks* awkward, this doesn't give quite the full picture...there are two things which make the job torture.  Firstly is that the belt tensioners on this car are the threaded type rather than the automatic spring based ones you tend to see on modern cars.  So you can't just pull a spring back...you need to wind the tension off manually.  Doesn't sound like much when you say that you needed to essentially unscrew a nut a couple of inches...until you realise that the most accessible adjuster you can get about 1/8 a turn a time on...the second one you can get about 1/16th of a turn, requiring the spanner to be rotated through 180 degrees each between each tiny movement, spending 10-30 seconds of flailing around to try to get the blasted thing back on the nut.  Oh, and periodically drop the spanner into the bowels of the engine bay.  I did find someone else's 9/16" socket on top of the front subframe though!  The tensioner on the alternator has to be done in a similar way, but *entirely* blind.  My arms are a good foot too short to reach from above, and unless the car is on ramps or jacked up, the air dam is precisely where my head wants to be and you can't get anything close to a look at what you're doing.  The fact that this everything in the vicinity of the front of the engine is drowned in oil makes the job disgusting as well as awkward.

First order of business was to scrape my knuckles to ribbons on the radiator find while wrangling the belt off around the fan.  Then discovered that the replacement belt I had for that was a totally different size.  On the plus side while it's pretty old, the fan belt is entirely serviceable.  So I've noted down "replace when possible" in the maintenance log.  The alternator belt actually looked absolutely fine (has a 2015 date code)...however hadn't been tensioned...like at all.  I could rotate the alternator pulley with a single finger.  So I just tensioned it correctly and left it alone.  This had absolutely nothing to do with me turning the the tensioner when I went to take the belt off and having utterly run out of patience by that point, having spent the last twenty minutes trying to wind the tension off!

While the alternator belt looked to be absolutely fine, the fan belt serviceable but old, the same couldn't be said of the what I'd say is the most important belt - the one which drives the water pump and power steering pump.

Brace yourselves...

Yeeeeeaaahhh...that was alarmingly close to a very expensive failure.  If that belt had let go, bye bye coolant circulation...one cooked engine.

That one at least did match the replacement!  Getting the sucker on however was a gigantic pain in the tail.  I must have spent nearly an hour wrestling with it before I got the thing on all three pulleys.  Then it was a simple matter of spending another hour winding all of the damned tensioners out again.

I was curious to see if this would have made the charging system behave itself properly...

Yes!  That's with everything electrical turned on only seconds after starting the engine.  Previously this would have required me to bring the revs up to get it to move out of the lower red section on the gauge, and it would tend to drop back again at idle with any heavy load on requiring the same to be done again.  It now seems to be doing exactly what I'd expect.  As soon as you get anywhere near 1500rpm we have full charging voltage shown even with everything on.  With all the loads turned off, it jumps straight up to 14.4V as soon as the engine starts now.

What an utter pain of a job though!

I had planned to have a look at the distributor as well today...however after this messing around I was pretty much out of afternoon.

I'm still slightly surprised that the garage that put the radiator in didn't also raise the condition of the belts as an issue.

[Saabnut]

I had a 1987 pre-HE XJS V12 back in 1997/1998. Loved the car but hated working on it, even with the fully equiped workshop (it was the first year I had the workshop). When the gearbox failed I found a superb Jag specialist up by Banff I think it was (he ran a Mk2 Pickup as a garage hack). He sorted the car for very reasonable and decided the Jag was a keeper. Sadly the specialist passed away soon after, and after several minor jobs on the XJS I decided that the pain of working on it outweighed the pleasure when it was working and it was swapped for a 205 GTi 1.9

Loved the car, hated working on it. We did maintenance on a 3.6 XJS for a customer for many years, and that was not bad at all for working on, but a disappointment to drive after the V12, nice enough but not a V12. Would I have another one? I was really tempted when the one you have came on the market, but decided one visit to Hell was enough. A car to admire and, yes, lust after but never to own again for me.

[Zelandeth]

46 minutes ago, Saabnut said:

...but decided one visit to Hell was enough. A car to admire and, yes, lust after but never to own again for me.

I think there are several ways in which the HE is even worse as the cooling system is even more complicated and of course the degree to which it makes getting the plugs out difficult is legendary.

It really does feel like a car made by committee...there's absolutely no reason 90% of things need to be so difficult if the teams designing various systems had ever bothered to speak to each other.

[PhilA]

10 hours ago, Zelandeth said:

It really does feel like a car made by committee...there's absolutely no reason 90% of things need to be so difficult if the teams designing various systems had ever bothered to speak to each other.

How can they talk when the other team is on strike?

[Zelandeth]

Have just been out to get groceries.  Getting a brief chirp from the power steering/water pump belt just as I turn out of the drive, but I'm going to give the belt a few trips out to bed in before fiddling with it, the tension seems in the right ball park at least.

The temperature gauge seems to sit fractionally lower now, not sure if that's my imagination, though the belt that was on there before was so loose that it was riding right up the groove in the pulleys, so would have been turning things slightly slower.  If enough to actually make a difference I've no idea. 

Charging system now seems to be behaving properly I'm glad to report.  The car is also noticeably quieter at idle.  I really didn't realise how much belt whine/hiss there was before until it was gone.

[LightBulbFun]

given how knackered the old belt was I wonder if it was also slipping somewhat, which would mean the pump would not be driven as fast?

 

BTW I saw you posted about this on your version of this thread on Retro rides, but yeah I was thinking the same I love how you have 2 motoring extremes  

a Model 70 and a V12 Jag, (and you actually use both of them when you can!)  

2 completely different ends of the automotive spectrum, amusingly both made by famous sports car manufactures

I still want to see a Picture of the Jag and TPA together so you can say "yeah got the Jag and the AC out for a photoshoot" 

and I can bet you people would not be thinking of a Model 70 when ya say that, and then you can show them the photograph and see the look on their face

 

[Zelandeth]

Nice little package just turned up for the van. 

This won't be getting fitted immediately, not least because I'll need to get someone in to remove the windscreen first - I'm not even entertaining the idea of taking out or fitting a screen that size myself, needs modifying to match the RHD wiper layout too.  However it's going to need to be done at some point so it made sense to get the panel in stock before they become hard to find.  Will probably get the bonnet ordered next month as that's the other panel I obviously need (the skin on my one is stuffed on both front corners).

[Zelandeth]

I've had something of a critical patience failure with the Invacar sitting immobile in the garage.  It's bloody well getting sorted.  Not in six months time, now.  Realistically I'm not much closer to a replacement hub than I was in September.  A few leads but nothing has yet materialised. 

It's been a while, so here's the current situation. 

[] On 2nd September 2019 I went to get a slow puncture on the nearside wheel sorted. 

Just needed the rim cleaned up a bit, took five minutes to sort.  Right up until this happened.

Closer inspection revealed the rest of the studs on this wheel to also be knackered.  You can actually see the upper right one is bent in the photo.

[] Three of the studs unscrewed from the hub without too much protest.  Except for the one which had already snapped.  Which snapped twice further before ending up snapping totally flush with the hub. 

[] At this point I made a blundering mistake.  Rather than taking the hub off and sending it to a machinist to be sorted, I figured I'd have a shot at drilling out the remains of it in situ.

This was foolish for a couple of reasons.  The first was that I had utterly underestimated how hard to source a replacement hub would be if I could things up.  Secondly was that I was doing this folder into a cramped, poorly lit corner of my garage.  So there was absolutely zero chance of me drilling it out straight as I couldn't see what I was doing and by the time I had the drill set up had already lost feeling in my left leg.  I'd also underestimated quite how awkward this job was determined to be.

By the time the stud was removed the hole was distinctly oval.  That is the point at which I started seriously looking for another hub - and that's where we are still.

It took a bit of experimentation, but we managed to figure out that threads in the hub are 3/8" BSF, and there are currently three high tensile bolts holding the wheel on.  I'm not 100% happy with the way I've had to use the original wheel nuts (they're bog standard Mini ones) to centre the wheel, I really need to get some proper conical washers. 

Having thought about it a bit I don't see any problem using bolts rather than studs and nuts.  I can see that they'd use the studs when the cars were in service s it's going to be far quicker to swap out a stud if the threads get mangled than change a hub.  However I'm hardly going to have the wheel off every weekend, and with me being careful rather than monkeys in Kwik-Fit it should be absolutely fine.

The remaining one is getting drilled out to the first size necessary to get the hole properly round and is then getting tapped and an appropriately sized bolt will then be fitted.  It's not a perfect solution but in the real world it should be absolutely fine.  Plenty of cars far heavier and more powerful do just fine with three wheel nuts or bolts just fine...and I really can't see a 400kg 20bhp three wheeler having the slightest issue because the fourth bolt is a fractionally different size.  So long as the wheel is properly clamped against the hub it will be just fine.  The only issue I can possibly forsee is the balance being thrown off a little if the one bolt is slightly heavier, but realistically I'm not going to be doing hot laps around the 'Ring...it'll be fine.  There's enough vibration from the drive and the engine as it is I'm probably not likely to even notice if there's a slight imbalance from a wheel.

This isn't going to be a long term solution.  Once the world returns to some semblance of normality I'll be pulling the hub and sending it to a machinist to be sorted out, probably adapted to take standard press-through Mini studs.  However as all of the places I'd be looking to send it are currently closed, I'm forced to improvise.  Everything will be very carefully monitored but I'm really not expecting issues once the wheel is bolted on.

Plus if the usual way of the universe continues, the moment I get a workaround sorted another hub will materialize as if by magic in tomorrow's post.

Drilling things went okay today, the hole I'd already got wasn't far off round.  However then this happened.

Ah.  It's going to be one of *those* days is it...

Apparently my tap set is made of cheese.  I'll get a hopefully better one ordered and then we'll come back to this.

Meh, plenty of other things I can do today.

First up came under the heading of "unexciting but worthwhile" in that I made a run round both the van with some grease.  The passenger door on the Jag has creaked like a prop from a Haunted House since it arrived on fleet.  The bonnet has been sticking a bit on the van lately too.  So I went round and greased all the door, boot and bonnet hinges, the pivot points for the gas struts and the latches.  I didn't realise how badly the bonnet on the Jag was binding before, it now opens and closes silky smooth, no creaking or groaning.  The doors on the van now snick closed with the gentlest push as they should.

I do need to adjust the bonnet closure on the Jag a bit, the driver's side rear corner sits a bit too high and bounces around a bit when driving.  Slight echoes of the Lada there...though in that case you need to leave a little slack or the cable will snap the first time you pull the release handle...and if that happens you're in for a world of pain getting in.  Been there, done that, have the scars to show for it.  You really don't want to have to deal with a Lada Riva with a snapped bonnet cable unless you have a spare bonnet on hand and can just take a grinder to it.

Next up I wanted to check the tension on the new belt I'd fitted, it's had a couple of trips out to bed in, seems generally fine but has been giving a brief chirp of protest from the power steering when first backing out of the drive.

Did seem a little loose, so time to nip it up a bit. 

The lock nut is moderately awkward.

The adjuster that I needed to wind out to tension things up on the other hand is bloody fiddly.

You have a working arc of about 10cm between the exhaust downpipe which limits your space upwards and the top of the power steering reservoir itself which gets in the way downwards.  Between each movement you need to rotate the spanner 180 degrees and have a 50% chance of dropping it into the bowels of the engine bay.  It's fiddly.  No access from above unless I attack the spanner with the grinder as there's a metal coolant line in the way.

Simple enough though, just fiddly.  Seems to have sorted the squeak, I'll check it again in a week or two, but it should be fine now I think.  Need to go back in to swap the fan belt once the new one shows up anyway.

It was starting to spit with rain at that point, but I wanted to get one more thing ticked off.  The rear fog lights weren't fitted properly and were pointing inwards towards each other.  Made the back of the car look cross-eyed.  It was bugging me.

There are two large washers and a shakeproof washer on the fitting bolt, these had been fitted wrong.

Both of the penny washers were below the mounting bracket which was pulling the light down too far.

Instead there should be one on each side of the bracket, like so.

Problem solved.

Looks immeasurably better.  Just need to sort the exhaust tips and the back of the car shouldn't look off.

The heavens then opened so that was that for the afternoon.

Will see what the weekend holds when we get there.

 

[LightBulbFun]

Yay very happy to see some work/progress done on TPA its been far too long  

im happy to see at least 1 invacar owner here is able to do so something with theirs  

while most people were locked down with their cars, so have at least been able to make use of the opportunity to get work done there, of course for me the lockdown happened while mine was/is about 100 miles away! so there's sweet FA I can do regarding fettling it!

although I shouldn't complain really I know others are having a much tougher time with it all

(being stuck at home being able to do sweet FA about anything has been the status quo for the past 8 years or so for me so this whole lockdown is not really any change for me!)

of course also happy to see the other members of the fleet being looked after!  

[spartacus]

I would say you've got nothing to worry about with your plan for the hub repair, I don't think the larger bolt will make a noticeable difference to balance at that distance from the centre of the axle.

I reckon you knew that though.;-)

You're flying with these projects, an inspirational thread.

[Zelandeth]

I don't know about flying... it's taken me since September to really make any headway with the snapped stud!  Most of the stuff on the Jag has been relatively small bits of fettling.  The more in depth stuff will be starting soon when I start dismantling large amounts of the engine bay.

Got a really good MPG readout today!

Yeeeaaah...the poor computer is a bit confused.  Some idiot (yes, that would be me) managed to leave the interior lights on yesterday and flattened the battery.  The resulting voltage dip when I went to try to start it this afternoon wiped half of the memory...it has zeroed the fuel used but not miles covered...hence the ridiculous average MPG listed.  Battery has been left on the charger overnight.  Hopefully it will have survived the experience as I could do without having two cars with stuffed batteries right now.

Hooking the battery charger up was the extent of car related things today as we wanted to make good use of the relatively cool weather to continue attacking the garden.  I think we have now successfully nuked all the Hawthorn at least.

This evening I did have a little project planned though of the computing variety.  Those of you who follow that side of my ramblings will probably recall that a couple of weeks ago I upgraded the CPUs in my main workstation PC to a pair of Xeon 5365s, the most powerful chip supported by the machine.  While this went smoothly enough I wasn't too happy with the thermal performance I was seeing afterwards.

Given that the X5365 chucks out 150W of heat when all four cores are flat out (and I have two of them...this thing is an effective space heater!) I was expecting it to run warmer than before, but the temperature readouts at full load were worrying.  I was seeing around 40C at idle (was usually mid 30s with the previous CPUs), but it was climbing well into the 70s under full load.  Noticeably though the CPU fans weren't really doing much, which tells a story as in this system the fan speed is regulated based on the readback from a thermistor on the CPU hestsinks, rather than the CPU temperature itself.  This told me that the bottleneck was more the ability of the heatsink to pull heat away from the CPU package than the ability of it to actually shed the heat.  I wasn't sure if these reported temperatures were accurate, but a couple of system lockups during heavy CPU load did seem to suggest things weren't happy.  Not surprising if the temperature readouts were accurate given the maximum allowed opersting temperature for the X5365 is 63C...

Initially I tried replacing the thermal compound which came with the new CPUs with some of a better quality (Arctic Silver MX4), which did knock a couple of degrees off but didn't really make a huge difference.

It was brought to my attention by @Lightbulbfun that the CPU coolers used in the Mac Pro 3.1 are some dimensionally identical to those use in the version 1.1/2.1 machine, use the same fittings etc, but contain three heat pipes rather than one.  Ten minutes rummaging on eBay sourced a set of Mac Pro 3.1 heatsinks for £10.

This morning a surprisingly large box arrived which contained these.

Yes, I did straighten those fins before I fitted them.

They aren't a totally drop in replacement.  You need to remove the air guide from the one (the one to the left above), and fit a couple of foam pads to help prevent the airflow from bypassing the heatsinks.  I was able to just transfer all but one over from the old ones in this case.

The only other thing I had to do was bend the pin header for the lower (B) CPU heatsink thermistor plug on the motherboard over by 90 degrees as the leads on the new heatsinks are about an inch shorter than the originals.  I found that by having this plugged in parallel to the board it was possible to have it hooked up without any tension on the leads or needing to make an extension cable. 

I was expecting this to improve cooling performance, but I wasn't expecting quite such a big difference!

Idle temperatures are now in the high 20s to low 30s.  Under full load it's now levelling out at far more sensible numbers.

The temperatures are shown at the top of the screen.  They're monitoring one core temperature on each physical CPU.  High 40s to low 50s, far more sane...and well clear of the maximum allowed figures according to the data sheets.

Another difference which is really obvious is that you now hear the fans ramping up almost immediately when you drop a heavy load on the CPUs now...that's good as it shows the heat is being transferred efficiently to the heatsink. 

This is a really good demonstration of how much more there is to heatsink design than just a block of metal and some fins...the hestsinks here in the "before and after" situation were to the untrained eye identical.  Same number of fins, same sized copper mounting block, same physical size...the differences are all in the internal architecture...and the difference in performance is staggering!

I left it running flat out for about an hour this evening and the temperature never budged from the figures shown above - numbers I'm perfectly happy with.

Overall a good result, especially for a tenner and half an hour of time.

So yeah...if you want to upgrade a Mac Pro 1.1/2.1 to run Xeon X5365s, it's definitely worth spending a few quid extra upgrading the CPU hestsinks to the later versions while you're at it.  Means literally about ten minutes additional work if you do it at the same time as the CPU upgrade itself.

[LightBulbFun]

1 hour ago, Zelandeth said:

I don't know about flying... it's taken me since September to really make any headway with the snapped stud!  Most of the stuff on the Jag has been relatively small bits of fettling.  The more in depth stuff will be starting soon when I start dismantling large amounts of the engine bay.

Got a really good MPG readout today!

Yeeeaaah...the poor computer is a bit confused.  Some idiot (yes, that would be me) managed to leave the interior lights on yesterday and flattened the battery.  The resulting voltage dip when I went to try to start it this afternoon wiped half of the memory...it has zeroed the fuel used but not miles covered...hence the ridiculous average MPG listed.  Battery has been left on the charger overnight.  Hopefully it will have survived the experience as I could do without having two cars with stuffed batteries right now.

Hooking the battery charger up was the extent of car related things today as we wanted to make good use of the relatively cool weather to continue attacking the garden.  I think we have now successfully nuked all the Hawthorn at least.

This evening I did have a little project planned though of the computing variety.  Those of you who follow that side of my ramblings will probably recall that a couple of weeks ago I upgraded the CPUs in my main workstation PC to a pair of Xeon 5365s, the most powerful chip supported by the machine.  While this went smoothly enough I wasn't too happy with the thermal performance I was seeing afterwards.

Given that the X5365 chucks out 150W of heat when all four cores are flat out (and I have two of them...this thing is an effective space heater!) I was expecting it to run warmer than before, but the temperature readouts at full load were worrying.  I was seeing around 40C at idle (was usually mid 30s with the previous CPUs), but it was climbing well into the 70s under full load.  Noticeably though the CPU fans weren't really doing much, which tells a story as in this system the fan speed is regulated based on the readback from a thermistor on the CPU hestsinks, rather than the CPU temperature itself.  This told me that the bottleneck was more the ability of the heatsink to pull heat away from the CPU package than the ability of it to actually shed the heat.  I wasn't sure if these reported temperatures were accurate, but a couple of system lockups during heavy CPU load did seem to suggest things weren't happy.  Not surprising if the temperature readouts were accurate given the maximum allowed opersting temperature for the X5365 is 63C...

Initially I tried replacing the thermal compound which came with the new CPUs with some of a better quality (Arctic Silver MX4), which did knock a couple of degrees off but didn't really make a huge difference.

It was brought to my attention by @Lightbulbfun that the CPU coolers used in the Mac Pro 3.1 are some dimensionally identical to those use in the version 1.1/2.1 machine, use the same fittings etc, but contain three heat pipes rather than one.  Ten minutes rummaging on eBay sourced a set of Mac Pro 3.1 heatsinks for £10.

This morning a surprisingly large box arrived which contained these.

Yes, I did straighten those fins before I fitted them.

They aren't a totally drop in replacement.  You need to remove the air guide from the one (the one to the left above), and fit a couple of foam pads to help prevent the airflow from bypassing the heatsinks.  I was able to just transfer all but one over from the old ones in this case.

The only other thing I had to do was bend the pin header for the lower (B) CPU heatsink thermistor plug on the motherboard over by 90 degrees as the leads on the new heatsinks are about an inch shorter than the originals.  I found that by having this plugged in parallel to the board it was possible to have it hooked up without any tension on the leads or needing to make an extension cable. 

I was expecting this to improve cooling performance, but I wasn't expecting quite such a big difference!

Idle temperatures are now in the high 20s to low 30s.  Under full load it's now levelling out at far more sensible numbers.

The temperatures are shown at the top of the screen.  They're monitoring one core temperature on each physical CPU.  High 40s to low 50s, far more sane...and well clear of the maximum allowed figures according to the data sheets.

Another difference which is really obvious is that you now hear the fans ramping up almost immediately when you drop a heavy load on the CPUs now...that's good as it shows the heat is being transferred efficiently to the heatsink. 

This is a really good demonstration of how much more there is to heatsink design than just a block of metal and some fins...the hestsinks here in the "before and after" situation were to the untrained eye identical.  Same number of fins, same sized copper mounting block, same physical size...the differences are all in the internal architecture...and the difference in performance is staggering!

I left it running flat out for about an hour this evening and the temperature never budged from the figures shown above - numbers I'm perfectly happy with.

Overall a good result, especially for a tenner and half an hour of time.

So yeah...if you want to upgrade a Mac Pro 1.1/2.1 to run Xeon X5365s, it's definitely worth spending a few quid extra upgrading the CPU hestsinks to the later versions while you're at it.  Means literally about ten minutes additional work if you do it at the same time as the CPU upgrade itself.

 

 

ohh very nice work  

its worth noting that the software you use to measure CPU temperature is measuring the temperature of the CPU die from the dies internal thermal diode,

where as the 63c rating you get from intel is for the tCase ie how hot the heat spreader on the CPU can get

the tJunction (thermal junction) max of most CPUs including your X5365's is 100-105c so your still in the clear at 70c  

although of course bringing the temps down is always a good thing! and I have to say that is quite an impressive drop in temperature 

not that I think your measuring this but as general a thing thing to watch out for on these systems is if your measuring CPU temperature is a specific reading called "relative to prochot" this is a temp reading that works in reverse as a count down, the closer you get to 0c the closer your CPU is to overheating until it hits prochot at which point a little Overtemp LED will light on the logic board and the CPUs will start to throttle themselfs, (so this prochot reading will go DOWN as the cpu's get warmer, hence why I know your not reading that one!)

the logic board has a number of handy diagnostic LEDs dotted around the place, including 1 by each memory slot that will light up red if theres a problem with its respective memory slot

as a side note the MP2,1 which was the X5365 version of the MP1,1, also used the 3 heat pipe heatsinks, but the "2,1 specific" ones are quite rare, as the MP2,1 was a BTO version of the 1,1 and was very expensive at the time! so not many get parted out etc and are not any different to the 3,1 ones performance wise

plus the MacPro2,1 (and the 3.2Ghz configuration of the MacPro3,1 for that matter) also used liquid metal thermal interface material, which is always "fun" to try and clean off 14 years later! 

(it does however tickle me however when I see the PC Gamer people talk about how liquid metal thermal "paste" is the best new stuff when Apple used it 14 years ago!)

 

if you want a fun party trick, use some fan control software and manually override the fans to maximum speed, it shifts enough airflow that it will provide enough pressure to stick paperwork to the front of the machine

 

one thing I'll do tomorrow is grab a fun screen shot of all the system sensors a/my Mac Pro has, its quite impressive just how many they are, apple really kitted them out, and they are all monitored by a separate chip on the motherboard the System management controller (SMC) to regulate stuff like fan speed etc

(dont worry back when I owned your system as well upgrading the main system firmware from MP1,1 to MP2,1 firmware for the quad core CPUs I also upgraded the SMC firmware to MP2,1 SMC firmware to better handle the higher heat output, most people do the MP1,1-2,1 main system firmware (BIOS in PC speak) upgrade, when fitting Quad core CPUs but the SMC firmware upgrade is often neglected/not done because it involves getting freaky with an EFI shell/command line LOL)

[Zelandeth]

I've seen how many sensors there are... I'd love a cheat sheet to figure out what's what - though I don't imagine that the IDs that are listed by libsensors equate to anything in the manuals.

I can't see how the thermistor cables could be rerouted...it literally just comes directly out the base of the heatsink.  The only reason it's an issue at all is that it's a fair bit shorter than the original ones.

I've seen the fans at full output - it locked them all into full speed when we had the lockup issue - enough airflow to thoroughly de-dust the case... meaning everything else in the room acquired a fine film of PC case lint!

[LightBulbFun]

12 hours ago, Zelandeth said:

I've seen how many sensors there are... I'd love a cheat sheet to figure out what's what - though I don't imagine that the IDs that are listed by libsensors equate to anything in the manuals.

now that im on the desktop I can post that screen shot!

 

once I get some more RAM for the MP1,1 I have in my collection, ill fire that up with the same software and hopefully you can figure out what's what  

(as the above screenshot is from my daily machine my MacPro4,1/5,1)

[Zelandeth]

Yeah, the problem I have is that I don't get "human readable" names for most of them, just the hardware ID as reported by the sensor monitoring library.

Some are pretty obvious, others less so!

I can of course assign names to them, but that's only useful once I know what's worth monitoring!

[Zelandeth]

So the gardening has continued, pretty much all day today so nothing else was really on the cards.

This was empty yesterday morning.

Before we started (forgot to actually take one...so courtesy of Google Street view as of a couple of years ago):

Where we left off last week:

Wider angle...

(This is the skip load from last week).

Now:

Pulling the van into the space here has always involved folding the mirror in and barging through the hedge to some extent.  Opening the driver's door was never on the cards.  Backing out the garage with the Invacar always involved picking greenery out of the offside window runners afterwards too.

Bit different now!

We will be extending the full height fence all the way round to the house once we've finished clearing the greenery.  I'm planning to have a little gated store where that pile of junk is as that's where I tend to keep stuff that's waiting to go to the tip.  Would be nice to have somewhere that's out of the way for stuff that really doesn't need to be in the garage too like wash buckets, and somewhere the garden waste bin can't blow over would be nice.  So the fence will come out to be flush with the end of the house, will nicely disguise the rear gate too.

The overgrown Leylandii and whatever that bush is will also be going as they're out of control and beyond taming.  There's a gaping hole in the back of it too where a tree has died back historically too. 

We will be planting some things in here, but they'll most likely be fruit trees rather than supposedly decorative hedges.

Looks a bit of a disaster area just now, but once we're done it will be so much better and easier to actually maintain.

Seven days ago we didn't even know that lattice fence existed!  About 80% of the greenery from the driveway to the far fence was either hawthorn or had hawthorn intertwined with it.  We are very bloody glad to see the back of that.

Before the obvious question is raised, yes I've climbed up there and had a good dig around to check for any bird nests before we started pulling things down.  The only one found hasn't been occupied for at least the last two years.

My back is absolutely killing me!

[Zelandeth]

Having been attacked by the reel of vacuum hose I picked up for the Jag today for the umpteenth time when I opened the garage door I figured it was time that I fitted some of it.

Most of the lines on there look like this.

These were the ones I was successful in swapping out today.

A couple were larger diameter ones which I didn't have in stock - there was enough slack in them though I was able to just cut off the perished/stretched ends and refit the hose.  I'll get the right hose in stock and then replace them in due course.

I discovered that the little elbow connectors that attach to the throttle body and fuel pressure regulators are a very loose fit.  I was planning to stick a couple of hose clips on there to tighten them up but sadly can't find any hose clips of an appropriate size.  Will need to get some ordered in.

I also only managed to change about 30% of the hoses.  The rest are utterly buried and are pretty much utterly inaccessible.  I can't even begin to tell where half of them go beyond "vanishes under the inlet manifold." I have to assume a lot of the lines were attached during engine assembly before the inlet manifolds were fitted.  Another job to add to the list while they're off then.

I've also noticed that I have a very slight weep from the radiator bleed screw. 

I'm guessing this should have a copper washer or an O ring under it (currently absent), can someone confirm this?

I noted that the nearside tyre looks slightly low again.  This reminded me that I'd been needing to investigate why towards the end of last year my compressor took out the RCD when I plugged it in one day.

I was hoping it might be something simple like a fault in the pressure switch or a chafed wire...no such luck...it appears that something pretty major has gone awry as there's a more or less dead short between the motor windings and the case.

Can't really justify replacing it just now...while it's nice to have air in the garage it's hardly mission critical.  A thought sprang to mind though...a while ago I stumbled across an ancient tiny little compressor hooked up to a probably even more ancient motor.   It's only a baby one, but I had established that it would at least move air. 

Cue some scrabbling around the garage for things I could scavenge fittings from to actually attach any form of outlet hardware to it, finding a plug (as of course I'd nicked that for something else) and cobbling things together.

At which point it became apparent that I'd left the tank drain open.  This had then seized.  After a little gentle persuasion it started moving...then sheared off.  Balls.

Eventually finding a bolt that was the correct size plugged that, and we could see if the teeny tiny compressor could fill the receiver in a vaguely sensible time.  If it's going to take an hour to get it up to 50psi that's not really going to be useful.

The belt tension needs setting up properly as I need to nudge it with my foot to start it just now, but once it's kicked over it sits there going "dugga-dugga-dugga..." quite happily.  Surprisingly quietly too.  The vast majority of the noise is actually from the intake, so a muffler would solve a lot of that.

Less than a minute in...

Looks promising.  About five minutes in we were up to about 70psi, so that's entirely usable .  I'll pull everything out of the garage at some point so I can actually get at it and then have a think about how to actually properly pipe things up.  Long term I'll need to look at getting a new compressor as this isn't going to have the legs to keep up with any real demand, but it should at least let me do things like checking my tyre pressures reasonably quickly.

Will probably look to install a little shelf above the receiver and use this as an excuse to get some polished copper and brass in the garage.  I'll look to add a fan of some sort to blow over the compressor body too, even though it only got vaguely warm during today's test.  The hose was quite warm though which was the main reason I called time when I did as it's obviously not really suitable for use in that position.

I was backfilling the tank through the outlet through a random extension hose because I simply wanted to see whether this would even work, the correct inlet and outlet will be used once I set stuff up.  No reason I can't use the existing pressure switch and regulator etc as there's nothing wrong with it as far as I can tell. 

...Glad I saved this one from the skip at the side of the road now!

[LightBulbFun]

1 hour ago, Zelandeth said:

This reminded me that I'd been needing to investigate why towards the end of last year my compressor took out the RCD when I plugged it in one day.
I was hoping it might be something simple like a fault in the pressure switch or a chafed wire...no such luck...it appears that something pretty major has gone awry as there's a more or less dead short between the motor windings and the case.

its a good thing your a little bit more sensible than me

as id of probably just disconnected the earth wire while making damned sure not to touch the thing LOL

(obviously I wouldn't leave it/use it regularly like that but if there something that needed blowing up urgently I could see myself doing the above LOL)

[Zelandeth]

2 minutes ago, LightBulbFun said:

its a good thing your a little bit more sensible than me

as id of probably just disconnected the earth wire while making damned sure not to touch the thing LOL

(obviously I wouldn't leave it/use it regularly like that but if there something that needed blowing up urgently I could see myself doing the above LOL)

I've lifted a ground on more than a few occasions to test a theory and on a couple of occasions to get a job finished.  However there are limits to what I'll consider...and a large entirely metal cased appliance that I have to physically squeeze past to get in or out of the garage (and that I have to lean over to turn the plug off)...just nope.  There are plenty of far less stupid ways to kill myself.

[Angrydicky]

Garden looks good. Plenty of space to build a nice double garage now

That tap is appalling. They should be hardened steel, that clearly isn't! Selling crap like that is taking the piss to be honest.

[Zelandeth]

1 hour ago, Angrydicky said:

Garden looks good. Plenty of space to build a nice double garage now

...With our planning department?  Hahahahahahaha!

Yeah...even a carport is utterly out of the question. 

What we will be doing is levelling the area beneath where that stinking great pine tree currently is (that will be going later in the year as it's just massively too big and is undermining both the driveway and the footway, and heaven forbid it came down in high winds as it would flatten either our or our neighbours houses).  Once that's done we'll be getting grids laid down on the whole grass area so it's also available as parking without getting torn up.

Yes... that's me basically admitting to myself that I'm very unlikely to get the fleet down to anything less than four.  Not because I want to own four cars...but because I really can't bring myself to part with any of them!

[] Invacar: Well, it's such a bonkers little thing isn't it.  While I've had faster, smoother, better handling, more economical, more comfortable cars... I'm not sure I've had another one which is quite so fun to bomb around in.  Wouldn't necessarily want it as my only car, but as a second car that reduces driving to its most basic components it's hard to beat.  Free tax, MOT exception and virtually free insurance are bonuses.  Oh, and the fact that it actually fits in my ridiculously narrow garage helps.

Not saying it won't ever sell it, but it would have to be something pretty special to take its place, Morgan 3-Wheeler is pretty much the only real contender in my mind just now.  That's a £30K car though, so yeah very much wish list stuff...If I could afford that though I'd probably also be able to afford to rent a bit of storage to keep TPA though!

[] Van: Really can't see it going anywhere.  Let's face it, neither me nor the family really batted an eyelid at the totally unexpected near £600 bill at the last MOT.  I think that tells the story really.

Just feels so right for me driving it too.  Plus watching people in the supermarket car park panic when I thread it through gaps they've just struggled to get through in their Kia Picanto never gets old!   

Only major change I can see maybe happening *if* the right stuff just happened to turn up would be a power unit upgrade to OM.605 power.  Not actively looking though, purely something I'd look at if the parts turned up.

[] Jag: I'd really only planned for it to be a bit of fun and to probably look to move it on relatively quickly.  That logic lasted about half an hour into the drive home (which was the point at which it decided to run on all twelve cylinders).  It just feels such a special thing to drive...granted I've not really got stuck into any of the particularly horrible jobs yet, but so far it's been worth the price of admission...and the price of fuel.

[] Xantia: It's from that period where Citroen had figured out how to make a truly staggeringly competent car.  It's quick, it's comfy, it's roomy, it's quiet, it's effortless to drive irrespective of if pottering around town, floating up the motorway for hundreds of miles or blasting across a twisty B-road.  It's both interesting and easy enough to live with to be a daily driver.

This is one which is probably the most likely far to change into something else - though I can't immediately think of many cars which would be viable to take the position on the fleet.  Volvo 240 (would have to be a very early one though as I far prefer the first generation interior), Volvo 850, Citroen GSA or maybe a late 70s to 80s Mercedes.  Like Volvos they're just a car I've never had the chance to try.  Oh, or a Saab 9000.  Being honest a 9000 is probably about the closest car in the world to the Activa in character now I think about it.  Have had four 900s, but only even driven one 9000.

[Zelandeth]

A few people have suggested that it might be possible to replace the motor on my currently dead compressor.  I'd originally discounted this myself as being a direct drive unit I expected things to be pretty model-specific.  It's only one of those around £100 including a bunch of basic tools jobs anyway so not worth throwing a huge amount of money at.  However it's worth looking into depending on what state the compressor side of things is in.  I particularly wanted to examine the bore on account of having discovered that the cheap piece of recycled biscuit tin that the air intake was made from was sufficiently poorly made that there was a good 1/8" gap between the air cleaner element and the back of the housing.  So none of the air was actually being required to go through the air cleaner.  I discovered this *after* doing a bunch of soda blasting and painting a car in the garage!  As such I was...curious...to see how much of this gunk it had ingested.

Had to admit to being a bit curious anyway...I've had a few refrigeration compressors apart before so I was interested to see what differences there were in the construction.

Off with its head!

The valve plate definitely shows that it's been breathing more than just clean air...This is the side which faces down into the cylinder, there's a reed valve which sits over the central hole allowing air into the cylinder through there, but forcing it out through the three smaller holes above it.  Two of these had a good 1/16" of solidified gunge on them, and the inlet reed valve had about the same thickness of paint on the outward facing surface. 

Um...Yeah...there has been some contamination then.

That's a large blob of what appears to be a mixture of soda blasting media and compressor oil.  I wasn't sure if I wanted to push the cylinder down to examine the bore at this point!

Actually a lot better than I honestly expected.  I can *just about* feel the most obvious score you can see in the above photo, but that was the worst of it.  With a good clean I reckon it could live to fight another day with a replacement motor - if it's cheap enough.

I knew the older one I was monkeying around with yesterday wasn't performing its best down to how much the belt was slipping, and wanted to see how it compared actually working better.

There was no way to get the tension set properly on the belt with the slots in the base it's currently mounted on being where they are, however I just happened to have a slightly shorter belt in stock, which meant I was able to line things up better.  Also meant I could actually bolt both things down with four bolts rather than the original two.

Which allowed the tension in the new belt to be set correctly.

Doesn't that look better?

It's happy to start reliably now, even with 100psi of head pressure on it - which there would never be on startup anyway (think the pressure switch cuts in at around 60psi), and there's an unloader built into the feed piping to the receiver if I were to set this up right.  With the belt no longer slipping this performs a lot better, honestly not seeming that much slower than the compressor it's currently serving in place of.  It's massively quieter too, which I'd gladly accept it taking a bit longer to get up to pressure as payment for.

Further experimentation and investigation of what my options are will undoubtedly follow.

Only other thing I really got done today was to give the leather in the Jag another going over with the leather feed.  I think this must be the sixth or seventh treatment, and it's actually just about starting to not immediately get absorbed.  The rear seats in particular felt very dry when I got the car - not that anyone is ever really likely to be in the back given how hard to is to get in and out of!

Looks better for it, though I do at some point need to give them a really good scrub down with the cleaner as they're still quite grubby in a few places.

There are a couple of small areas that could do with some attention...so a question to those of you who have dealt with interior trim repairs before (it's a new field for me).  Are these things that a specialist would be able to repair, and if so is it likely to be at a reasonable cost?  I've never really had a car where it was worth asking questions like this before!

Rear seat has three spots where the stitching is failing.

That last one is probably the most important as it's really clearly visible from outside the car.

The driver's seat also has a spot where the backrest adjuster lever has rubbed a hole in the side of the seat base.

I imagine that's a very common issue on the XJ-S as it's always going to rub in that spot.

Also there's a bit of wear to the backrest bolster on the same seat.

I'm not massively worried about that...it's a 78K mile car and there are going to be signs of wear from that...but if it's something that can be made less obvious for reasonable money I'm not going to discount doing anything about it.  Just seems that little things like this can potentially both add value long term, and do make the car a nicer place to be.

Looking at those photos really has highlighted how grubby the seats still are...I really do need to sort that out...

[MorrisItalSLX]

The thread Jaguar used to sew the seats together is quite rubbish as TADTS. The front seat bases on my XJ have been completely re-sewn at some point, but the seat backs and rear seat now need doing.

An automotive trimmer would be able to re-stitch the seats, condition/colour the worn areas and replaced the holed piece. But as you say, is it worth doing as I imagine it wouldn’t be cheap and they are only little issues.

[Jikovron]

With that compressor, motors are possible to buy but they are surprisingly expensive, I looked at an equivalent single phase motor for my band saw and given the £75 for a Tec motor I ended up taking one from a cheap pillar drill, suffice to say it ran the wrong way but I just turned the housing back to front rather than try messing about with the run cap. Thinking about space maybe the compressor vessel could mount on the ceiling and pipework run down to a take off PCL and the old unit somehow mount to the vessel or feed it from a shelf, I've considered similar but my weedy 6L compressor is infinitesimal! 

[Zelandeth]

59 minutes ago, Jikovron said:

With that compressor, motors are possible to buy but they are surprisingly expensive, I looked at an equivalent single phase motor for my band saw and given the £75...

...Thinking about space maybe the compressor vessel could mount on the ceiling and pipework run down to a take off PCL and the old unit somehow mount to the vessel or feed it from a shelf...

That's pretty much what I found yesterday evening.  Cheapest I found was around £50 from some random Chinese supplier, which really doesn't seem worth it when Toolstation have a brand new equivalent of the whole thing available for £99.98.

Moving the receiver to somewhere above ground level is a good idea actually.  There's a bunch of dead space around the garage door framework where it could easy enough be slotted in, while still keeping the tank drain accessible (in fact easier than it usually is as it's a pig to get to with the thing on the ground). 

Here is where I'm thinking.

Can't see piping things up being difficult.  I was always planning to move it away from the door and pipe air to somewhere nearby as space is tight enough there as it is.  Just need to figure out what fittings I need.

[Zelandeth]

Done a bit more digging into the compressor today as much for my curiosity as anything.  I'd expected some sort of coupler between the motor and compressor...however it quickly became apparent that it's all one chunk.

Also became quickly apparent that the crankshaft seal has been leaking, the whole rotor is covered in sticky horrible oily gunge.

Getting the fan off the end of the shaft was an absolute pig of a job as it was clearly designed with a keyed fan in mind...however the manufacturer had literally just hammered one on there that was small enough that it was jammed in place.

The receiver isn't all that heavy in itself, so I think the idea of mounting it remote to the compressor is a decent one.  Pressure switch, reg and water separator (not currently fitted) can be kept at eye level, and a drain down valve can also be remotely piped in.  Reckon a pipe off the bottom of the tank and a ball valve will be more reliable long term than the fiddly arrangement it originally had anyway.

It does take up a good old chunk of floor space so I like the idea of it being shoved somewhere out of the way.

At this point I was intending to tidy up from that before moving onto another job - however successfully sidetracked myself en route and still haven't remembered what I actually went into the garage to do in the first place.  I stumbled across an ancient bottle of Auto Glym Bumper and Trim treatment which I think I've had since I worked at the garage (which I moved on from in 2005).  This reminded me that I noted way back when I first picked it up that the plastics and bumpers on the Jag were looking really dull and lifeless.  I didn't realise how bad they were until I started treating them.

I honestly thought the strip under the radiator grill was grey not black.

The pillar air vents speak for themselves I think.

This thing was a pig to do.  In hindsight I should have pulled it off the car, would have been quicker I think.

I always assumed that grill would be plastic... actually appears to be die cast aluminium.  Every day's a school day.

I always forget how much just treating all the plastics and bumpers can make a car look so much tidier.

The one bit of plastic I didn't touch is this one.

Reason being is that I want to go in with a tiny brush and touch in the lettering on the panel between the reversing lights.  I'm not sure if this being silver or black is a HE/Pre-HE difference, but I prefer the look with the text standing out a bit.

The other bit of silver detailing I've seen on a couple of photos of early cars is on the surrounds around the instruments.

Like this (image grabbed from a classified listing somewhere - trade-classics judging from the image watermark - a while ago, and used purely for reference).

I know it's technically not correct for a HE, but I think that the silver detailing really looks nice and breaks up what is otherwise quite a big slab of black.  I need to pull the panel again because the warning light for the rear fog lights has packed in (again, dodgy lamp holder), so might add a bit of early car detailing there.  It fits with the overall black with silver piping theme outside too.