Re: Zel's Fleet Blog...Jag, Citroen, Mercedes, Sinclair & AC Model 70
Posted: Sat May 02, 2020 1:14 am
A few things rolled into one again as I've done a poor job of keeping things up to date lately.
I needed to head out a couple of days ago to pick up some prescriptions (it would be really nice if they would just let me order things a couple of days earlier so I didn't need to make four separate trips for four separate medications!), of course I took a small, sensible car didn't I?
Well...kind of.
To be honest I don't really want to take the Jag out in the wet until I've sorted the rust on the nearside wing on the Jag and refitted the wheel arch liner, so it was the logical choice.
Now a brief Segway into tech as happens now and then in this thread.
About a year ago I was generously gifted an Acorn RiscPC 600 by a fellow fan of these machines. A RiscPC had been on my wish list since I first encountered one in the late 90s. I have several Archimedes machines as I've always been a huge fan of both Acorn and Amiga computers, having grown up with them.
Don't ask me to pick a favourite... they're very different machines, though from a purely technical perspective Acorn probably gets my vote. They were so, so far ahead of their time architecturally. While I enjoy using Workbench on the Amiga, 3.0 is what I really made my first *real* steps into computing with, I can't deny it's quite clunky in a lot of ways. RiscOS on the other hand still feels slick now...while it might lack modern visual sparkle in some ways, it still works well. It was very much its own thing in the time and to this day nothing has behaved in quite the same way. I think it's the massively efficient use of screen space in RiscOS which makes me want to bash my head against a brick wall when using Windows 10.
While I've several Archimedes machines (A3000, A3020, A4000, A5000) and relatively recently found an A7000, I've never managed to track down a RiscPC at a price I thought was sensible. So when the opportunity to get this one came up I didn't hesitate for a millisecond. Hopefully one day a BBC Micro or Master will also materialise.
Scruffy though it may be, I was very glad to have added this to my collection of interesting (to me anyway!) computing hardware.
Unsurprisingly this machine needs a bit of TLC.
While I didn't have time to give it too much attention when it initially arrived, I did give it a quick once over. This immediately revealed that the RTC/CMOS battery had been leaking, a common issue on Acorn gear, which can do quite a lot of damage if left unchecked.
Thankfully despite having made a mess there was no actual damage visible.
So the battery was snipped off the board, the alkaline gunk neutralised and then the motherboard cleaned before it was put into storage to wait its turn in the queue.
After patiently waiting a full year, it's now getting some attention.
We appear to have the 2Mb VIDC Memory upgrade, a StrongARM CPU upgrade board (not sure exactly which one yet), 32Mb of RAM and the 80486 expansion board (which makes me happy as the idea of being able to run DOS/Windows software in a window on a RiscOS desktop appeals to my sense of ridiculous).
Also enjoyed seeing a TI branded 80486 rather than an Intel one, don't think I've come across that before.
Starting status: Dead. Well not entirely. On powering on there's a click from the speaker, the hard drive spins up (with that unmistakable "Nnnn-Wrrrriiimmmm...tick-tick-tick-clack" anyone into computers from this era can ID as a Connor hard disc), the power and hard disc LEDs light as expected (disc activity light goes out once the hard disc has spun up), the network adaptor status LED lights, and the keyboard LEDs blink showing me it's getting power, but that's it.
No soft synthesized "Boooop" sound effect, no seek of the floppy drive, no screen output, nothing. These machines are able to output quite a few POST error codes via flashing the activity light on the floppy drive, however nothing showing here.
If the CMOS data has become corrupted or left in a state which could prevent the system booting, it's possible to do a factory reset by holding down delete when powering it up. Didn't make any difference.
One of the best things about these machines is that you can get down to the motherboard (and even get that out) without needing any tools. The "slice" based case really is a clever idea.
At this point I stripped the machine down to it's bare bones and tried known good memory to no avail.
Moving on from there I made sure that all the correct voltages were all present, which they were. Nothing for it then but to pull the motherboard out for a closer examination to see if I could find any tracks broken by corrosion from the battery leakage, especially with the proximity of it to the RAM sockets. While I was there it made sense to pull every socketed chip, clean them up and reseat them.
It didn't take long for a likely candidate to turn up... though not for the reasons I was expecting. This is the chip which handles most of the system I/O, probably being vaguely analogous to the northbridge chipset on a modern conventional PC.
While I'm not entirely sure how this happened (best guess someone dropped one of the CPU cards on it), we have four bent and one broken pin. I could see that one was already broken, but I decided to see if I could get the remainder to bend back into shape enough to ensure they weren't actually touching each other.
To give a better idea of the sort of nanosurgery I'd need to do to repair the pin, here's a fingertip for scale.
I didn't expect anything, but I figured I'd see if we had any change in behaviour with the short removed at least. I was staggered when the monitor woke up, and the floppy drive chunked its way into life for the startup seek. Still didn't boot though, was still non responsive. We did have some sign of life on the display though which was huge progress.
I know that as the infamous battery is right next to the video memory expansion slot and that contact issues there are common. Sure enough wobbling the memory stick in the VIDC memory expansion slot made the pattern on screen change.
A quick test to see if the problem is with that memory is simply to remove it. The machine has onboard video memory, you just don't have access to the highest resolution and colour depth modes without the expansion fitted.
Powering up then resulted in a buzzing noise from the speaker...and the unmistakably obtrusive chatter of a Connor hard drive booting up an OS echoing around the room. I was staggered to see this then appear.
Now something is still very unhappy given the video issues that are visible...but given that there's physical damage to one of the main ICs on the motherboard, getting it to boot at all seems need miraculous to me. Seems to be 50/50 as to whether I get the above display or this...
This artefact is actually a bit more complicated than it might at first look, as it's not just the screen display being folded over - it does actually seem to be getting drawn that way as the application bar scrolls left or right as you mouse over it, so it seems that we are in fact in some strange, probably invalid screen mode. It's also not just the monitor getting confused, the signal being fed in is a normal VGA signal.
Realistically I need to track down either a replacement IOMD21 chip or a parts donor motherboard before I can go forward with this. There's just no way on earth that I can repair what I've got with the equipment and experience I have. I'm just not set up for nanosurgery like this.
While I had the workbench clear I set about wiring up the additional instrument pod for the Invacar, just need to bolt it in place and add the senders for that. My intention is for it to sit roughly here.
The exact angle will be chosen to minimise reflections in the face of the gauges. They're just cheap random ones I've accumulated over the years in there at the moment, but I'll try to replace them with more period correct Smith's or similar ones as time goes on and I find appropriate gauges at autojumbles etc. I'm not willing to pay what seems to be the going rate for them on eBay these days. This will be getting attached to a metal bar that runs under the dash rather than to the plastic itself, so won't put any additional strain on the already brittle plastic or anything like that. It will also be removable in the future without leaving any visible evidence if I decide to get rid of it.
I've no real interest in the absolute values that these gauges will show me, I'm far more interested in observing what the baseline for "normal" is for my engine, then any sudden deviation from that norm will be cause for investigation. I may feel that in the more distant future that it's unnecessary, but I'd like to be able to keep tabs on things until I've got at least a couple of thousand miles covered trouble free. Given I do have plans for a long trip in mind which will probably involve a lot of sustained high speed driving, I feel that it's a good idea to have the pod there at least until that's happened. If I was purely trundling around locally I'd be less bothered.
-- -- --
I thought that I'd resolved the issue with water finding its way in through the driver's door on the Jag...
...Apparently not. Looks like I'll be having to pull the door apart again then. We will have to see if I get bored enough to tackle that next week.
This afternoon I decided it would be fun to stand on my head under the dash of the Invacar for half an hour.
Before anyone says anything, I'm waiting for the package with the ring terminals small enough for these terminals to arrive from eBay, I got fed up of waiting...when they arrive I'll terminate the wiring properly.
I've simplified the wiring into the car a bit by just thing the illumination and power supply to the gauge pod together, so it will just be lit whenever the ignition is on. So all I'll need is an ignition switched live, a ground (which will be attached to the bracket once I've got ring terminals), and the two sender lines. "Keep it simple, stupid!" Is an approach I like.
Hopefully the wiring will be done this weekend. Need a couple of adaptors for the senders (also in the post) but no reason I can't get the wiring done.
Doesn't look too out of place.
Will obviously look better once I get some period gauges in there rather than the cheapie Chinese nonsense currently in there. These have been in the garage for goodness only knows how long though so seemed the sensible choice.
Will serve the purpose for now though even if they're a little cheap and nasty.
I needed to head out a couple of days ago to pick up some prescriptions (it would be really nice if they would just let me order things a couple of days earlier so I didn't need to make four separate trips for four separate medications!), of course I took a small, sensible car didn't I?
Well...kind of.
To be honest I don't really want to take the Jag out in the wet until I've sorted the rust on the nearside wing on the Jag and refitted the wheel arch liner, so it was the logical choice.
Now a brief Segway into tech as happens now and then in this thread.
About a year ago I was generously gifted an Acorn RiscPC 600 by a fellow fan of these machines. A RiscPC had been on my wish list since I first encountered one in the late 90s. I have several Archimedes machines as I've always been a huge fan of both Acorn and Amiga computers, having grown up with them.
Don't ask me to pick a favourite... they're very different machines, though from a purely technical perspective Acorn probably gets my vote. They were so, so far ahead of their time architecturally. While I enjoy using Workbench on the Amiga, 3.0 is what I really made my first *real* steps into computing with, I can't deny it's quite clunky in a lot of ways. RiscOS on the other hand still feels slick now...while it might lack modern visual sparkle in some ways, it still works well. It was very much its own thing in the time and to this day nothing has behaved in quite the same way. I think it's the massively efficient use of screen space in RiscOS which makes me want to bash my head against a brick wall when using Windows 10.
While I've several Archimedes machines (A3000, A3020, A4000, A5000) and relatively recently found an A7000, I've never managed to track down a RiscPC at a price I thought was sensible. So when the opportunity to get this one came up I didn't hesitate for a millisecond. Hopefully one day a BBC Micro or Master will also materialise.
Scruffy though it may be, I was very glad to have added this to my collection of interesting (to me anyway!) computing hardware.
Unsurprisingly this machine needs a bit of TLC.
While I didn't have time to give it too much attention when it initially arrived, I did give it a quick once over. This immediately revealed that the RTC/CMOS battery had been leaking, a common issue on Acorn gear, which can do quite a lot of damage if left unchecked.
Thankfully despite having made a mess there was no actual damage visible.
So the battery was snipped off the board, the alkaline gunk neutralised and then the motherboard cleaned before it was put into storage to wait its turn in the queue.
After patiently waiting a full year, it's now getting some attention.
We appear to have the 2Mb VIDC Memory upgrade, a StrongARM CPU upgrade board (not sure exactly which one yet), 32Mb of RAM and the 80486 expansion board (which makes me happy as the idea of being able to run DOS/Windows software in a window on a RiscOS desktop appeals to my sense of ridiculous).
Also enjoyed seeing a TI branded 80486 rather than an Intel one, don't think I've come across that before.
Starting status: Dead. Well not entirely. On powering on there's a click from the speaker, the hard drive spins up (with that unmistakable "Nnnn-Wrrrriiimmmm...tick-tick-tick-clack" anyone into computers from this era can ID as a Connor hard disc), the power and hard disc LEDs light as expected (disc activity light goes out once the hard disc has spun up), the network adaptor status LED lights, and the keyboard LEDs blink showing me it's getting power, but that's it.
No soft synthesized "Boooop" sound effect, no seek of the floppy drive, no screen output, nothing. These machines are able to output quite a few POST error codes via flashing the activity light on the floppy drive, however nothing showing here.
If the CMOS data has become corrupted or left in a state which could prevent the system booting, it's possible to do a factory reset by holding down delete when powering it up. Didn't make any difference.
One of the best things about these machines is that you can get down to the motherboard (and even get that out) without needing any tools. The "slice" based case really is a clever idea.
At this point I stripped the machine down to it's bare bones and tried known good memory to no avail.
Moving on from there I made sure that all the correct voltages were all present, which they were. Nothing for it then but to pull the motherboard out for a closer examination to see if I could find any tracks broken by corrosion from the battery leakage, especially with the proximity of it to the RAM sockets. While I was there it made sense to pull every socketed chip, clean them up and reseat them.
It didn't take long for a likely candidate to turn up... though not for the reasons I was expecting. This is the chip which handles most of the system I/O, probably being vaguely analogous to the northbridge chipset on a modern conventional PC.
While I'm not entirely sure how this happened (best guess someone dropped one of the CPU cards on it), we have four bent and one broken pin. I could see that one was already broken, but I decided to see if I could get the remainder to bend back into shape enough to ensure they weren't actually touching each other.
To give a better idea of the sort of nanosurgery I'd need to do to repair the pin, here's a fingertip for scale.
I didn't expect anything, but I figured I'd see if we had any change in behaviour with the short removed at least. I was staggered when the monitor woke up, and the floppy drive chunked its way into life for the startup seek. Still didn't boot though, was still non responsive. We did have some sign of life on the display though which was huge progress.
I know that as the infamous battery is right next to the video memory expansion slot and that contact issues there are common. Sure enough wobbling the memory stick in the VIDC memory expansion slot made the pattern on screen change.
A quick test to see if the problem is with that memory is simply to remove it. The machine has onboard video memory, you just don't have access to the highest resolution and colour depth modes without the expansion fitted.
Powering up then resulted in a buzzing noise from the speaker...and the unmistakably obtrusive chatter of a Connor hard drive booting up an OS echoing around the room. I was staggered to see this then appear.
Now something is still very unhappy given the video issues that are visible...but given that there's physical damage to one of the main ICs on the motherboard, getting it to boot at all seems need miraculous to me. Seems to be 50/50 as to whether I get the above display or this...
This artefact is actually a bit more complicated than it might at first look, as it's not just the screen display being folded over - it does actually seem to be getting drawn that way as the application bar scrolls left or right as you mouse over it, so it seems that we are in fact in some strange, probably invalid screen mode. It's also not just the monitor getting confused, the signal being fed in is a normal VGA signal.
Realistically I need to track down either a replacement IOMD21 chip or a parts donor motherboard before I can go forward with this. There's just no way on earth that I can repair what I've got with the equipment and experience I have. I'm just not set up for nanosurgery like this.
While I had the workbench clear I set about wiring up the additional instrument pod for the Invacar, just need to bolt it in place and add the senders for that. My intention is for it to sit roughly here.
The exact angle will be chosen to minimise reflections in the face of the gauges. They're just cheap random ones I've accumulated over the years in there at the moment, but I'll try to replace them with more period correct Smith's or similar ones as time goes on and I find appropriate gauges at autojumbles etc. I'm not willing to pay what seems to be the going rate for them on eBay these days. This will be getting attached to a metal bar that runs under the dash rather than to the plastic itself, so won't put any additional strain on the already brittle plastic or anything like that. It will also be removable in the future without leaving any visible evidence if I decide to get rid of it.
I've no real interest in the absolute values that these gauges will show me, I'm far more interested in observing what the baseline for "normal" is for my engine, then any sudden deviation from that norm will be cause for investigation. I may feel that in the more distant future that it's unnecessary, but I'd like to be able to keep tabs on things until I've got at least a couple of thousand miles covered trouble free. Given I do have plans for a long trip in mind which will probably involve a lot of sustained high speed driving, I feel that it's a good idea to have the pod there at least until that's happened. If I was purely trundling around locally I'd be less bothered.
-- -- --
I thought that I'd resolved the issue with water finding its way in through the driver's door on the Jag...
...Apparently not. Looks like I'll be having to pull the door apart again then. We will have to see if I get bored enough to tackle that next week.
This afternoon I decided it would be fun to stand on my head under the dash of the Invacar for half an hour.
Before anyone says anything, I'm waiting for the package with the ring terminals small enough for these terminals to arrive from eBay, I got fed up of waiting...when they arrive I'll terminate the wiring properly.
I've simplified the wiring into the car a bit by just thing the illumination and power supply to the gauge pod together, so it will just be lit whenever the ignition is on. So all I'll need is an ignition switched live, a ground (which will be attached to the bracket once I've got ring terminals), and the two sender lines. "Keep it simple, stupid!" Is an approach I like.
Hopefully the wiring will be done this weekend. Need a couple of adaptors for the senders (also in the post) but no reason I can't get the wiring done.
Doesn't look too out of place.
Will obviously look better once I get some period gauges in there rather than the cheapie Chinese nonsense currently in there. These have been in the garage for goodness only knows how long though so seemed the sensible choice.
Will serve the purpose for now though even if they're a little cheap and nasty.