Using your Sega Dreamcast on a new TV
I had recently been playing my Dreamcast on my old TV, which had a VGA input. Using the DC VGA BOX, I had VGA and audio output from the Dreamcast and all worked perfectly.
In the last fortnight, I've picked up a brand new Sony KD49X8000D and it doesn't have VGA input! HDMI, Composite or Component... they're the options. It was time to browse eBay for options.
VGA to HDMI
Without doing any research, or reading the product description, I jumped straight in and purchased a cheap straight-through VGA to HDMI cable.
It was never going to work. It seems to only be for projectors where they transmit random signals via VGA plugs. Don't expect that an analogue VGA signal can just magically transform inside a cable to HDMI.
VGA to RGB
I then saw this cable at trash and treasure last weekend. It's got VGA at one end and RGB on the other. I thought that it might be component, so I paid the handsome sum of $2 to test it out.
I'd forgotten to test it, but just read online that it also wont work. The description of a similar cable here indicates that this is for specific projectors under specific circumstances.
MINI VGA2HDMI Adapter
I resorted back to eBay and looked for an 'active' adapter. I found this one, cheap, on eBay and thought I'd give it a go.
In the end, the cabling is a nightmare as there aren't nice small VGA cables available.
Super simple to connect. But it didn't work first off. Someone else has made this item work before, so I must have something wrong with my configuration. It could also be that my TV doesn't support the resolution over HDMI.
Uhmm... nope.... I then tried another HDMI cable and realised that this one seated in around 2mm further! And worked! Switching back to the other HDMI cable saw it working too: it just needed a bit more of a jab to make proper contact with the converter. Note! These converters are cheap and cheerful, so do be careful when hooking them up.
Check that out... you can see the TV LED at the bottom... that's the initial picture at 1:1 640x480. Very clear!
And then scaled... looking great. Sounding o..k.. a bit scratchy at times. Will check my headphone cable between the devices. Anyway. Success!
Final Note - Overheating...
This adapter is indeed cheap and cheerful! If yours stops working, then check my post here on how to apply a heatsink.
Burning M27C256B EPROMs with a Willem
This is another public service announcement. I've just wasted around 3 weeks trying to burn an EPROM which shouldn't've taken more than 1 minute. I purchased a Willem Parallel Port 5.0 PCB burner from eBay (cheap, I know) and although I could happily read chips, I couldn't write them.
The settings were correctly configured for the M27C256B chips. You need 6.2v to write and 12.75v external. This means you can't use the USB to power. This PCB 5.0 is dual power, so it should've selected the correct source.
Every time I tried to burn I got the following error: Error at 0x000000 chip = 0xFF buffer = 0x58. Note that all three of those values would change, a lot. 0x24 and 0x46 were frequently seen yelling at me.
Turns out that the 'latest and greatest' version of the software doesn't select the correct power source. Downgrading to 0.97ja (as per the screenshot above works fine.)
So if you're getting the errors as I was above, then use the 'alternate' version that should be on the driver CD that you were provided with.
NEC PC-8201A Portable Computer – Black Screen
Here's a new acquisition. It's an all-in-one portable machine from the early 80s. It was sold cheaply as non-working and, indeed, it non-works.
I hadn't realised how big these things were!? I was expecting something around half the thickness and maybe 60% the width/height. It's actually a really nice machine and I do like the form-factor. This item came quite scuffed and 'used'. All to be expected and nothing to worry about. I purchased it since they seem to have a great array of I/O ports and programmability. They run Microsoft BASIC off-the-shelf and have 32kb of storage.
Initial Inspection
First signs of tear-it-down-before-turning-it-on showed up as soon as I checked out the battery pack. The batteries weren't in there, but there was a LOT of corrosion. The circuit board in the battery pack even seemed to have components missing? Did they corrode and disintegrate away?
After a good scrub, all was much cleaner... I had already decided to pull the main machine apart first though. Corrosion has a magical property of travelling inside metal (just like electricity) and conquering whatever it can get its green rusty mits on.
Of course, it had spread. The terminal and battery pack isolation switch were the main victims. Fortunately, it didn't seem that the tracks or other components were affected?
Both Isopropyl alcohol and cleaning wipes were used to clean up the board. Light scrubbing was also required in the hard to get areas.
Powering it up
After all the cleaning and soldering above, it was time to bite the bullet. I found 4 random AA batteries and slotted them into the pack. My multimeter showed 6 volts, so I slapped it in an gave it a go.
Flicking the power switch on the side gave no response. I had read online somewhere that the first thing to check is the "Back Up Power" switch. I initially thought this meant "use the internal battery", but it doesn't. It actually means "use the batteries you've just inserted". With this switched over, the power switch on the side saw the unit come to life... well... there were at least a few electrons flowing.
Black Screen Of Death
For now, we're going to disregard the damage to the LCD. you can see that something pointy has taken out a sizable area of the screen... but 95% of it will still be usable prior to needing to replace that.
After the power switch was flicked, the screen produced the result you see above. Trouble-shooting this is pretty straight forward. As per the instructions here, press Enter, then type BEEP and press Enter again. If you get a beep, then your mainboard is fine and your screen is crap. My mainboard was not fine!
The next bit of googlin' found this result on diagnosing and repairing a black screen. Turns out one of the 245 chips was only half-functioning. I love that the author also found errors in the service manual. Either way, the repair-man found a defective IC and replaced it... as that I had no way of determining what was defective, I started replacing everything.
Component Shopping List
From the service manual, I found the following replaceable chips at my local electronics store.
Original Chip | Jaycar's Version |
---|---|
40H139 : U1, U53, | 74HC139 (ZC-4848) |
40H386 : U51 | 74HC86 (ZC4835) IS NOT PIN COMPATIBLE! Jaycar therefore doesn't have a replacement. |
40H245 : U13, U42 | 74HC245 (ZC4870) |
40H373 : U15 | 74HC373 (ZC4874) |
40H367 : U??, U?? | 74HC367 (ZC4872) |
40H138 : U2, U3, U4, U5, U6 | 74HC138, but Jaycar doesn't stock it! |
40H175 : U11, U12 | 74HC175, but Jaycar doesn't stock it! |
603 Transistors | BC547? |
115 Transistors | BC556/560? |
Note: The above list is a work-in-progress. I'll update this as I continue identifying and replacing parts.
Find and Replace
Removing components from PCBs is always a challenge. Even with the best solder-sucker, you can still have a very hard time making sure every pin is free. In the end, as long as you don't need the existing IC, snipping legs and then removing those can be a much better option.
I've used sockets for each IC I'm putting in as well... I've no idea if this whole board is pointless, so I'd prefer to waste a socket more than an IC. I started with the 245 that was mentioned above and got nowhere... still the blank screen... I then replaced a 367 and had no luck either. This was really a crap-shoot and a very slow process. From here I thought there had to be a better option... so I consulted the service manual.
Troubleshooting via the Service Manual
I really should've started here when diagnosing the issues in this machine but, when I had first read it, I saw the word 'Oscilloscope' and got scared off. I don't have one of these, so I didn't think to pursue this area of investigation. I had since come back to it after realising that replacing each of the ICs was going to take a REALLY long time.
On page 54 you actually get to test each of the pins on the CPU itself. It indicates the pin and the expected voltage. Note that the CLK pin is meant to flip/flop and this can't be seen from a multimeter. All you can really look for is a 'half' voltage, as the flicking effectively provides 50% 0v and 50% 5v totalling 2.5v. All of the pins on my CPU came up OK as per the values stated!?
From page 85 of the manual, you get instructions on what voltages are meant to be seen from the power supply board when the computer is powered on. All but one of these turned out to be OK. Pin 10 was meant to be greater than 7.5v, but mine was only just over 7.0v. The manual then tells you that U2, Q16, C1, or D7 of power board are bad. Ok... this is good! Let's replace everything we can.
Something weird happened next. I didn't have the required components to repair the power supply, so I tidied up the desk for the evening. As I put the machine back together (aka. placed it into a pile of loose parts), I accidentally toggled the power switch.
Wait... what... we've got movement! The screen displayed a bunch of random rubbish. Sometimes a cursor, other times a heap of LOAD statements and at one point I could even type in a command... I didn't get a BEEP though.
The only thing I can attribute this to is that the power supply board received attention. I had shifted the caps around and also wiggled transistors. Maybe some of these had faulty solder joints? I should've measured pin 10 again here, to see if it all happened because of a 7.5v reading. Unfortunately, I couldn't get it back into such a state. I went ahead and purchased parts to fix the power supply. Unfortunately, I couldn't find the correct transistors at Jaycar to replace Q16 as per the instructions above.
I replaced all I could but still couldn't get the machine to do what I wanted. Actually, as I first started tinkering, the display of crappy text was satisfying. Towards the end, I hardly had a jumble of pixels. Toying with the ROM did seem to help... I wonder if the chip or socket area was damaged? Maybe static had fried it? Out of desperation, I attempted to swap some transistors around on the power board. Maybe that Q16 was the final issue. I did note that the 7.5v was still very intermittent. It didn't quite appear when the machine was half-working and only appeared at other random times... maybe the board was pulling too much current due to other issues and therefore was never actually going to let the power supply do it's thing?
Playing switch-a-roo with transistors revealed some information on the power supply. Swapping the 603s between Q2, Q16 and Q3 made either the speaker click, the relay click or nothing at all. To me this meant that one/some of the transistors were faulty. It didn't matter which as I didn't have a replacement. Trying to find actual 603s was impossible but, using the cross-reference tool here, I have just found out that BC556s are a valid substitute.
Or are they? Let's rewind a bit... I assumed that the transistor was a 2SA603... but it actually can't be. Using this complimentary transistor table, it seems that my transistors are 2SC2603 and 2SA1115. It would make sense that one was NPN and the other PNP. I therefore need BC547s (although they seem to be rated 100mA less and the pins are reversed!) for the '603' and BC560s for the '115'. Unfortunately, Jaycar doesn't have 560s, so I'll choose 556s instead, which are just have a higher voltage tolerance and a lower amperage. I might also order the correct components from eBay.
So, transistors... turns out that on page 108 of the service manual, the components are listed. I was totally correct on the transistor model... well, after I corrected myself. I therefore bought 20 of each from Jaycar.
Of course, I jumped the gun and didn't correctly review each component. The pins weren't backwards, they were actually out of order!
So, with the legs crossed, just enough, the order of the pins could be adjusted...
Whilst replacing the transistors, I also replaced the capacitors. Of course, one had to have exploded. Happened to be the 33uf that assists with the -5v output. I assume it's a cap that stabilises the voltage. I simply used solder to bridge the damage.
The results started off erratic... but then seemed closer to rendering a real screen?
And then it happened!
Keyboard didn't work.. screen was frozen... but shiiiiit... that's the actual main menu! I actually nearly have a real computer! I then replaced all the signal diodes on the power supply.
With the diodes replaced, the keyboard worked! Note that at this point in time, the board relay was not actuating on power on. It had previously, intermittently... but doesn't seem to be essential? Either way, as above, the menu worked and BASIC loaded. Unfortunately, once loaded, BASIC was useless with a screen full of junk. Any keypress resulted in Memory Full and a system beep.
I opened up TEX, typed in the obligatory sentence, pressed enter and the machine froze up... the work continues!
One step forward...
And then back to the beginning. I did something stupid over the weekend. On Saturday night I placed the ICs I had purchased over the top of the ICs they were to replace, so as to be able to see what ICs I still needed to replace/purchase. With these piggy-backed loosely, I closed the machine and put it aside.
Forgetting that I'd done this, I attempted to boot the machine the next morning... It presented the infamous black screen and not much else. As I shifted the machine to the center of the workbench, I heard parts move and instantly remembered what I'd done the night before. What a bloody waste of time I am; forgetting that I'd left the ICs in there and hence possibly doing all kinds of damage.
For the life of me, I could not get the machine back to where I was on Saturday afternoon. Such a let down! The only real goal was to hope that I'd only damaged mounted items and not the new items I'd purchased... then again, I could just re-purchase as I'll socket all the ICs.
IC Replacement
Solder suckers are fun, but Desoldering Braid is the best thing since sliced bread! Once you've removed the IC+legs from the mainboard, you'll find it quite difficult to remove the solder from the holes. Desoldering braid does this for you and does it quickly and cleanly. All you have to do is unreel enough to cover the hole and then press your soldering iron directly into the center of the area. The chemicals in the braid them somehow manage to attract the solder!
If it doesn't work the first time, then re-apply solder first. Only use the braid on holes that are full. If there's a gap, then the solder wont jump, it'll only flow! Also, the longer you hold the soldering iron there, the better. Watch as the solder flows into the braid and lift when it stops flowing up.
After removing the TC40H386P, and thinking that the 74HC86 was a replacement, I worked out that it wasn't. A hack was employed to get around this.
I lost a lot of faith with hacks like the above. Regardless, I ended up replacing half the chips on the mainboard and didn't get anywhere... it was time to wait for the oscilloscope to arrive.
Power Mac 7220: Optical Drives
The Power Macintosh 7220 was built cheap and used off-the-shelf PC hardware, including IDE CD and harddrives. Due to this, upgrading the CD drive is relatively easy... or it seems it would've been... when IDE CD drives were popular! Nowadays everything is SATA and so finding a good IDE CD drive can be difficult. You'll find them on eBay, but they'll be marked as vintage and people therefore think they can raise the prices.
Fortunately, in one weekend I managed to acquire a grand total of 6. I bought 2 on eBay (a standard PC version and then another since it was an actual Apple model) as I was sick of not finding any, but then stumbled across another 2 at Trash and Treasure (Australia's version of Flea Markets or Swap meets) the very next morning. The following post is an effort to detail the pain and suffering of finding out how (in)compatible the drives were.
LG CED-8080B
I actually found this at a thrift store last weekend for AUD$5. Initially, installed into the Mac with cables connected, it stopped the entire machine from booting. It was about to be shown the bin... but I chose to test it again, once I'd worked through the rest of the drives, and the bloody thing decided to function.
It wouldn't boot the Mac OS 8.0 CD that I had burned on Windows (but this seems to be a common trend... note that the CD image is now deemed unbootable!) As another test, I installed Toast 4.1.3 and found a blank CD-R. Using Toast's "Disk Copy" method, I created a new temporary CD volume and then dragged the System Folder over from my main startup disk. Toast burnt all this to the CD (with the bootable flag) and the CD booted!
Pioneer DVR-106D
Very plain-looking Pioneer drive. Expecting good things from a quality manufacturer.
Again, it wouldn't boot the image burned via windows. I therefore tried something different: I imaged the burned Mac OS 8.0 CD to the desktop, mounted the image and then attempted to re-burn the image to a new blank CD with the bootable flag set. Unfortunately, I couldn't actually successfully burn a CD with this drive. No configuration or speed setting (buffer-underrun protection included) would work!
Another point: ejecting the tray saw it get sucked straight back in again... a fault of Mac OS or the drive itself? If you were quick you could snatch the disc out.
Pioneer DVR-108
Quite similar to the previous Pioneer, this is a very plain looking drive. Based on the previous drive, my hopes weren't so high!
This drive also failed to burn every CD I tried. Even on standard settings (16x), it threw buffer underruns. Switching on the buffer-underrun protection did nothing to help. Burning at lower speeds (all the way down to 1x) with underrun protection still failed miserably.
Don't use Pioneer drives with old Macintosh machines. Seems that it takes too long to spin up and the machine fails to have the data ready?
At this point I switched back to the LG CED-8080B that I knew worked... just to make sure my process of disk imaging was stable. It happily burnt the disk image! Of course, on reboot the Mac just showed a question mark. It had no intention on booting my 'bootable' copied-twice ISO.
Note that, once back in the OS booted off the HD, I could still happily use the installer on the CD. It just wasn't bootable.
LG GSA-4167B
This one is listed as a Super-Multi. I'd usually be happy with this level of functionality, but the above results indicate that the Power Macintosh itself may not be able to cope with a drive that can spin too fast.
Either way, it still is a really nice drive... smooth, quiet and fast. The best part is that it burned the CD with no issues!... but it proved that my CD burning idea was incorrect. You cannot convert a non-bootable ISO burned under Windows into a bootable ISO this way.
Note that this drive does not open with the case on. The tray face is too large to fit through the space provided in the front panel of a Power Mac 7220 case! This is sad.. it's the best drive of the bunch, and I didn't feel like hacking it to make it fit!
Apple GCC-4480B
Smooth drive. Happily imaged an 8.6 CDR ISO in a matter of minutes. Note that it does not have a drive activity LED on the front.
The image was then burned back to a CDR with the bootable flag and the image booted! The flashing disk with question mark showed up for a split-second and then it convinced itself to boot!
Feeling motivated, I used this drive to try and re-build the Mac OS 8.0 ISO into a bootable CD, but the drive also refused to boot it. Turns out that ISO itself is to blame?
This drive also doesn't fit correctly behind the face of the Power Mac 7220 case. Due to the missing front plate on the drive, the eject button doesn't mate and therefore the drive is rendered useless when the case is put back together.
Apple CR-583-B
This is the original drive that came with the machine. This has a little sticker over the eject button (but the button still works) and an activity light which is obscured by the front case.
This drive happily booted the OS 8.6 ISO that was re-bootable'd using the process above. This contradicts my initial impressions of "not being able to boot a burned CD", as ... well ... it does boot them. You just need to make sure the burned CD is 'correct'.
Of course, this drive is not a recorder... hence why Toast was telling me that no recorder was found. I even rebooted the machine to find the recorder... I can assure you that rebooting did not turn it into a recorder!
Which Drive?
After trying out everything above, I had to settle on a drive. I didn't want the original as it didn't have the 'tabs' to hold a CD in when the drive is mounted vertically. I also wanted to use an Apple drive, but the one's I had wouldn't really work with the case or weren't a burner. Instead I chose the black LG and modified it's tray face-plate to fit through the Power Mac's front fascia. This required a little filing on the ends of the tray.
In the end, it works like a charm!
B75 helps out Qube on the Maryvale Paper Train
This freight train operates in South-East Victoria. It runs from the terminal at Dynon and carries containerised paper products to and from the Australian Paper mill at Maryvale.
Qube operates the service and is often changing the motive power around. They have a good selection of locomotives, but recently have suffered a few failures in their more powerful units. Due to this, there's been a random selection of lesser-powered locomotives, meaning more units and more variety.
23th January 2017
The word had spread that B75 had been leant to Qube and would be leading the nightly service. It usually passes South Yarra just before 8pm, but, of course, was running late last night. Fortunately there was just enough light to snap it.
I was hoping it'd take the down Frankston line, but it instead stayed on it's usual down Pakenham. You could hear it coming.. the B sounded awesome with the S and VL behind it making just as much noise.
Nice and dark at 8:32pm, but I'm still happy with the result.
24th January 2017
Turns out this might be a regular event! The G classes must still be out of action. Nothing like a bit of sunlight to help out evening photography.
25th January 2017
This time I tackled the train head-on from platform 2 at South Yarra station.
Still looking just as clean and tidy as ever.
26th January 2017
This was to be the final day that B75 was helping out... It seems that the drivers wanted to do something special, and so instead of just driving the consist back, they ran B75 around at Morwell and had her lead back. It was also Australia Day, so there were flags hanging from the wing mirrors.
I really want to thank Qube for putting this show on!
SCT begins Melbourne to Brisbane freight services
SCT has always had an agreement with Aurizon to provide a small loading for north-bound freight services. This has recently expanded into a full-sized consist and SCT are now running the services themselves. The very first Melbourne-Brisbane service ran on Saturday the 21st of January 2017, departing the SCT Laverton terminal just after 8pm.
Thanks to daylight savings, there was a chance to actually get a photo around Sunshine Station. Aptly named in this circumstance... The area provides a view of the BG and SG tracks, with a set of grain silos as a backdrop. The usual V/Line and Metro services provided entertainment whilst waiting for the freights.
The Melbourne-bound XPT bolted through, only a little bit late.
Aurizon then came through with 7BM7, only to later have an extended stay in Wagga Wagga. I do love that they cleaned the grime off the front logo.
7MC2 started to depart Dynon, but the XPT also jumped. The XPT was therefore given priority.
7MC2 then spooled up. From Sunshine, you could hear it accelerating from stand-still over in Tottenham.
It sounded and looked fine as it passed... but it failed not too much further down the line, crawling into Seymour. It didn't continue its trip to Junee until the next day when two other locomotives joined it.
Up next, the SCT consist. SCT011 was attached at SCT Laverton to pull the whole train out of the yard and onto the main line. It then detached and SCT004 had the honours of pulling the train around the bend at Spotswood and over the triangle. From Sunshine, the train could be heard well before it was visible, as the wheel screeching over the triangle was tremendous.
It had an interesting consist, with the crew car 7 cars back from the locomotives.
And that was a wrap... the sun hardly held up for the late departure... I might try and get it earlier down the track next weekend. It's really great to see more colour on the eastern corridor.
The annual Christmas pilgrimage – December, 2016
Just like the christmas trip in 2015, this trip just past for 2016 Xmas wasn't overly different. Still a great location with a lot of sightings. An early start out of Melbourne meant an easy trip onto the Hume highway. I caught the southbound XPT just out of Kilmore and a southbound inter-modal in Seymour.
After a speeding ticket in Benalla, I b-line'd it straight to the Olympic Highway. It's a nice drive, off the beaten path, and away from influences to do stupid things on the road. First stop was at Harefield and I was greeted by an IRA 44 Class! Qube has taken ownership of a few of them and has been using one for the shuttle into Junee.
Not much was happening in the yard, so it was a bolt further north to intercept the southbound steel train. I beat it through Illabo, meeting it on a driveway half way between Illabo and Bethungra.
From there it was off to Cootamundra to tackle the XPT. I tried to beat it to Jindalee, but it won. Instead I then backtracked and got it at the station.
At this time of day, between the two XPTs, there's always a quiet time when the sun is at it's peak and the rails are buckling. Due to this, I took time to have lunch and check out the pool at the usual motel.
Back to it... the XPT was fast approaching from the south and I wanted to get it at Jindalee this time.
And then something weird happened... an off-the-radar rail-grinder rolled through! I love being in the-right-place-at-the-right-time.
And of course, being zoomed-in and distracted, I hadn't realised that the 81+82 on the grain train were steadily already rolling towards me.
The next train was a bunch of QBXs heading south to Junee. I had enough time, so I thought I'd try something different at Wallendbeen.
That angle worked really well! I then caught them again past Jindalee on a driveway half-way back to Cootamundra.
They didn't slow down through Coota, so I kept to the limit and then got in front of them before Bethungra.
I had a really hard time keeping up with them once they were on the other side of the spiral. It's all downhill for them and me, but they were doing somewhere around 95km/h and made it hard to catch up when the speed limit was 100km/h. We were both fast-approaching Junee and I managed around a 50 metre advantage; just enough to park road-side safely and snap a shot above some lovely green pastures.
From here, it was Junee yard.
From there, it was back to Wallendbeen to catch the next southbound intermodal.
I got in front of it and made it to another driveway past Jindalee, but before Coota.
The sun was already starting to lose light, but there was still enough time to get the 2 QBXs that were coming in via Stockinbingal.
Nothing was next on the radar, so I checked out the yard at Coota. Turns out there was a bit of shunting going on to get a grain train to fit in the yard.
That steel wagon must have been defective as it was sitting by itself, blocking one road, to start with. It ended up being shunted into another road onto so that the arriving train could stash half of its wagons there. Pretty quick work actually. Nice to hear the 81s powering up and down when shunting a whole train to shift one wagon.
After this, it was dinner time... an early start on the radar showed, what should have been, a great opportunity for a cross at Yass Junction. Due to this I packed up and got on the road early. Thanks to my miscalculation, the pass actually happened around Goondah and the light was terrible. Instead I forged ahead to Yass and got the northbound grain there.
I really do like that angle. From the grain silo, instead of the platform, you can get a nice long telephoto shot of the descent into the Junction. From here, it was off to Canberra and there's not much going on nowadays with the Museum closed and the scrap metal train done.
On the way back to Melbourne, I saw an opportunity for a shot at Jindalee of a northbound grain train. This all lined up well ... very well ... as the-right-place-at-the-right-time happened again.
The grain train then appeared.
A quick stop was had in Cootamundra to snap the Lachlan Valley Railway sheds where they have a nice collection of railmotors and 47 class locomotives.
And then finally, after a pie at Culcairn, I caught up to a southbound freighter at Henty.
The dim weather on the way back turned much dimmer over Seymour where visibility actually dropped to around 10 metres. Most people parked on the side of the road; but I didn't see that as 100% necessary and a few of us kept on driving through the car wash.
Another awesome pilgrimage!
Power Mac 7220
I've recently dug back into Macromedia Director disassembling with a goal to help finish an engine that can run movies under ScummVM. I've always wanted to make Gadget run on random hardware... and for that matter, on newer versions of windows without issues. To get this done, I've had to create permuatations of Director movies to be able to decode the bytes. For example, text labels on Director stages/frames can have fonts, styles (bold, underline, etc..), font sizes, margins, borders, box shadows and text shadows. Each of these customisations react against the others and therefore all permutations need to be given to be able to render stages correctly. To do this, I've been using BasiliskII, but it tends to dislike loading Director after closing it.
So... that's the brief... what's the answer? Since I've gotten rid of my Quadra 950 and Power Mac 7200, I scoured eBay to find a replacement. Turns out that a Power Mac 7220 (aka 4400) (aka not-really-a-mac) was available; two units, two keyboards, two mice and a monitor, actually. I only wanted one unit, but got a good offer to take the whole lot.
Tear-down
All items were quite dirty and so a complete tear-down was required.
That's not sepia tone.. that's a solid layer of 'protective' dust. I usually use that excuse with my car, but not with computers. The entire system was caked, so it received a solid once-over.
Nothing better than finding spider eggs... bleach fixed that.
Apart from the external case screws, the rest of the machine came apart with brute force. It is said that this Macintosh is as close to PC as possible... so it's very PC-like regarding IDE devices and case screws.. but then the Apple comes through with a perfectly dismantle-able case.
The motherboard was slightly grotty! It all came up good after a wipe and vacuum.
Of note: No fan on the CPU? AT-style power connectors. IDE connectors, but it seems that they're the same channel when you follow the traces (i.e. a header for master and a header for slave.) Crystal sound chip? Does apple ever use this? Built in ATI Graphics. "Feature Option"? First RAM slot is for "Single Bank" whereas the second two are "Dual Bank"? (I've read that this means you can have a 32mb in the single and 64mb in the doubles.)
Power Supply Test
The power supply looked to be in good shape externally and I didn't feel like opening it. I decided to take the punt and switched the machine on.
It chimed! The hard disk then started booting... All was going well until I heard the disk park the heads. It's a really scary event for a hard disk as it only happens when the disk itself knows it's in trouble. This is an old Quantum 1.2g and so the fact that it parked meant its basic error checking kicked in. The booting stopped as the drive was beeping. Not an overly audible beep, but it seemed to be fighting requests from the machine to fetch any more data.
Subsequent boots failed to produce happy disk-reading noises... it was parked for good. I'll need a screen attached to be able to see what's going on... and I'll need a video cable for that.
VGA Cable
I used the forum post here as a reference to wire up a VGA cable. I used an old VGA cable that I had in my junk box and purchased a male DB-15 connector and ribbon cable from Jaycar. After stripping and tinning both sides, I mapped out the VGA cable. The Macintosh side was easy enough as the IDC cable was, from the red wire, 1, 9, 2, 10, 3, 11, etc...
Heatshrink came in very handy to keep everything isolated. Make sure you choose a size that is wide enough to slip over everything you want it to. ALSO make sure that you don't try to slip it over whilst the solder is still hot! ALSO make sure you remember to slip the heatshrink on before you solder the wires :)
I initially tested on an LCD, but the Apple Multiple Scan 15 Display M2978 also worked first time! But gets very blurry after 30 minutes of usage.
I shut the machine down and gave the HDD a little 'love tap'. Usually when the heads are parked (or stuck?), a tap can dislodge them. Of course, it could be that the magnet that locks the heads in place was just sticky... maybe the drive was wearing itself out trying to un-stick the heads. Either way... the small jolt did the job!
PRAM Battery
There's no chance the battery still worked. It's also quite difficult to find an off-the-shelf replacement.
I ended up at Jaycar and found a 3AAA battery holder. Using a bit of extension wire, I snipped the plug off the dead battery and wired it all together.
The cable was run under the CD-ROM chassis and placed right next to the HDD. Taped up tight also... any effort to prevent future leakage (and terminal shorting) is appreciated.
CD-ROM Drives
The IDE drive that comes with this machine did not like booting from CD. It might just be that my CDs were copies, or that the drive itself was faulty. I tried two other IDE drives, but had no luck trying to boot from them. I then tried a SCSI CD drive, but that wouldn't boot either. In fact, the SCSI drive needed the Sunrise CD driver to even recognise CDs. So, any IDE drive will work with the base drivers, but if you want to boot off a burnt CD then you will possibly need a real Apple SCSI drive.
Of course, I've learnt all the above before with my previous Quadra 950. It's amazing how quickly you forget these things. The Power Mac 7220 was also locking the IDE non-Apple drives and they were getting quite confused. If you soft reboot when the drive is locked then you can quickly have an unusable drive until you power down the entire machine.
Again, just stick to real Apple CD drives and real Apple CDs.
Actually, with further googling, it turns out that there might be ways to correctly burn and make bootable a CD image. Supposedly the System Folder needs to be 'blessed'. I would've expected this to be already the case in the bytes inside the ISO image, but supposedly not. I might try Toast on the Mac and burn a disk image and test from there. Here's a good guide.
Extra IDE Disks?
There's two IDC headers for IDE connections on the motherboard. Looking closely at the board, you can see parallel traces running between the headers, so it's pretty safe to assume that they're on the same bus. From this you can then assume that the CD-ROM header is hard-wired as slave with the HDD header as master. Disregarding this assumption, I tried to plug two HDDs on the HDD port.
I had an 80-pin IDE cable on hand, but it turns out these are 'keyed' with one blocked pin. This doesn't fit the header on the motherboard. Instead I found another IDE cable. The second drive is 200gb and was seemingly busted... clicking badly! I found another HDD, this time 500gb.
The Apple booted up both times in this configuration, but never saw the second drive.
SCSI Disks
I've always been interested to know how 50-pin cables convert to 'Printer port' DB-25 pin plugs at the other end... turns out that this motherboard has a SCSI controller and a 50-pin IDC header that has an adapter to an external DB-25 port!
Ok, so ... every second pin is skipped? Half-duplex over full? Totally interesting! Fortunately, I had a cable to convert it back to 50-pin and tested out my dual-drive external bay. All worked flawlessly, even though the second disk was incompatible... will need to install PC Exchange or something else... I don't even know what's on the drive!
RAM
The 200mhz version supports up to 160mb. So that's two 64mb DIMMs and one 32mb. The recommended spec is Unbuffered 60ns Non-ECC 3.3 volt DIMMs. Even though it is mentioned here, here and here that ECC RAM can be used in non-ECC motherboards, this does not hold true for older Power Macs. G5 and higher were able to support ECC.
Network Card
Both machines came with the same network card in them. It's a Communications Slot II profile card, with very little information on it. Most information on the 4400/7220 indicates that it either has a 'built-in' card or just comes with a standard 'Apple CS II Ethernet' card and the aptly named extension from the Mac OS CD will work.
After re-installing OS 8.6, I could see the driver in the extensions folder, but there was no network link light. The device showed up in System Profiler, but with very little information.
I then tried OS 9.1. It is the last out-of-the-box compatible version with my machine. During both installs, I couldn't boot from the CD, so I performed a clean install over the top of 8.6. Neither made the ethernet work.
The card has "BD-064 REV a." written on the top. "GSEP-M01" and "94V-0" on the back. Also "805-1614-A" on the face-plate. Googling for everything but the first item came up with zero results.
Finally, I found this post (6 months old!) from someone in the same predicament. The post author indicates that there's more information near the ethernet port. It sure as heck isn't on the board...
Oh what? It's on the face-place.. under the removable face-plate? Nice work. Actually, now that I think of it... the removable 'face plate' is only used for this type of PC-style case. In a usual Performa (or other Power PC), this piece of metal, which obscures the relevant information required, would not be obscuring the relevant information required!
The author of the post above reported the same model number, so I used the drivers he specified. I tried to use a floppy to copy over Sonic Systems EtherLAN 7.8 Drivers. This didn't work on either the first, second or third attempt. I kept getting serious disk errors on the Macintosh side, so I think the floppy drive is gone. I therefore resorted to wasting a CD-R and copied the driver, plus some other bits and pieces, over.
Installation was a breeze... the link light lit up right at the end of the boot process just before the desktop and then I tried Internet Explorer... poor machine ground to a halt! 48mb of RAM seems to be insufficient.
Macromedia Director
3 is throwing "An error of type 2 occurred". I haven't tried 4 yet. One step at a time!
Turns out that Mac OS 8 is the last version that Director 3 will happily run on. I've created a partition specifically for this version. Director will open on 8.1 and 8.5 (getting past that error above), but then crash when trying to load movies. Mac OS 8.0 is the safest version to use.
Mac OS 9.2.2?
Nope. The 'Tanzania' motherboard that this Macintosh is based on does NOT support anything higher than OS 9.1. OS9 Helper which I used on my previous 7200 does NOT support this model. Don't even try!
BeOS 4.5
Turns out this is one of the few Power Macs that can run BeOS! I also have the CDs. Will test this out in short order.
Linux?
Sure, why not. Instructions are all over the internet.
Overclocking
The motherboard hardly has a passive heatsink on the CPU, so it can't be getting too hot during runtime. This gives an option to consider overclocking it, as we can just put a fan onto it!
There's a few links here at apple fool indicating overclocking the PowerMac 4400/160 and the Tanzania Motherboard. The first article is a dead link but web.archive comes to our rescue here. The basic idea:
How to update you PowerMac 4400/160 to a 4400/200:
1. Locate the SMD-Resitors R1 and R9 on the Logicboard. There on the left side of the CPU.
2. Remove R1 and R9 carefully and solder a 10K resistor to R2 and 10K to R8.
3. Attach a Fan on the Heatsink of the CPU.
4. Yes, now it runs at 200 MHz.
Ok, but we have the 200mhz version. Thanks again to web archive, here's the link on the Tanzania Motherboard. The blue box, up in the top-left corner near the CPU is where our settings are (or the bottom-right area below on my photo.)
R8 and R9 are below the CPU OPTION pads with R21-24 above. The SMDs are tiny, so shifting them around will be a nightmare. By the table, we'd only need to shift R9 to R8 to get 220mhz and then, if that was stable, R21 to R22 for another 20mhz to 240mhz.
As that the components are already on the board, it can't be too hard to try... I also have a spare machine! Supposedly you can also change the base bus clock frequency from a 40mhz xtal to 50mhz and get a quick boost!
I'll tinker once I've got the machine stable.
IBM NetVista M/T 6825-M23
This poor machine was about to be sent to the recycle center... so I salvaged it and downgraded it. Getting Windows 98 to run was a challenge!
Windows 98 doesn't like more than 512mb RAM
Found this out the hard way... the installer would boot the CD, start installing and then fail randomly throughout the install. Once it actually made it to the first reboot and then failed to reboot. It threw an error on the dos prompt of which I failed to record. The machine had 512+512+128 in there (which was an odd combo anyway) and I reduced it to just one 512 stick. After this is worked fine.
Sound Drivers
I downloaded and installed around 20 different SoundMax drivers and had absolutely zero luck. 1 month later (when I finally had a use for the machine) I came back to get the audio going; I needed a full multimedia experience.
I checked out the motherboard and found that it had a AD1887 chipset... turns out that this was made well after the Windows 98 era and therefore finding drivers wasn't easy.
Fortunately, Totempole posted the same issue in the vogons forums! Unfortunately, his links are dead. A little googling found the correct file (although it stated it was for Windows 7 or higher?) which worked a charm! Here it is for safe-keeping: SoundMax AD1887 Driver for IBM NetVista Desktop Workstation (d63z32us.exe).
Note that if you want DOS Sound to work, then you need to specifically enable it.
And then you can play awesome games!
NVidia AGP Riva TNT2
This machine came with this card in the AGP slot. There's no on-board graphics in this machine.
It seems like 81.84 was the final NVidia driver to support Windows 98 but I somehow managed up with 71.84_win9x_english.exe installed. Display Properties shows that this card is an NVidia Vanta? Never heard of it. Running at 1920x1080 in Windows 98 SE is pretty random.. but great for the next trick.
Now to start producing Director 5 movies for this project and ... playing A-Train.
The annual Christmas pilgrimage – December, 2015
This post is only around a year late. I've recently done this trip from Melbourne to Canberra to visit the family for Christmas and, whilst taking happy snaps this year (well, 2016), realised I had misplaced the photos I'd taken last year (well, 2015). Turns out they were sitting in a folder on the desktop of my (now disused) Vaio Duo.
Anyway, the annual pilgrimage includes a stopover in Cootamundra. I've now stayed at the Cootamundra Gardens Motel three times and have never been disappointed. Coota heats up during the day and the motel even has a pool! You're also in ear-shot of the railway and level crossing on Gundagai Road.
Cootamundra is a great spot to stop. It's the junction of the "west" line to Parkes and the "main south." You get grain trains frequently as the station precinct includes grain storage facilities. You also get the east-coast steel trains and intermodals. There's the XPT and also the Griffith Xplorer (when they're running.) As I mentioned, it also gets very hot, so sunlight usually isn't an issue... also, there's the occasional thunderstorm at this time of year, so the backgrounds can be quite picturesque.
Speaking of picturesque, there's a lot of great photo opportunities to be had in the area...
Jindalee
Heading North out of Cootamundra, you'll cross the railway and then hit the 100 km/h zone. From here, you run parallel with the railway until North Jindalee Road. Turn left into North Jindalee Road then then left again into West Jindalee Road. This is a dirt road that happens to cross the railway where it deviates to negotiate the climb to Wallendbeen. From this bridge, if you're willing to drag your car over the dirt track (which is actually in great condition), you get a great view of the curves on both sides.
Between Jindalee and Cootamundra
There are numerous level crossings along the stretch where the railway line parallels the highway between Jindalee and Cootamundra. Most of these are private property driveways, so be courteous and do not overstay your welcome.
Berthong Road Level Crossing
From Cootamundra, take Temora Road north. Once in the 100 km/h zone, take the second right turn. It's a cross-roads and, after turning right, you'll quickly come to a level crossing with ample space to park. Berthong Road provides a great location to get south-bound trains at any time of day. It is also a great place to get north-bound, but the lighting is better in the evening.
Bob Dear's Crossing
I don't know who Bob was, but just east out of Cootamundra is a fully equipped level crossing. You can get a good angle from below track-level on west-bound trains.
North Bethungra Level Crossing
When approaching Bethungra from Cootamundra, the highway crosses the rails just north of the spiral. This location provides a great view in both directions; just make sure you choose the right side before the lights start flashing.
Bethungra Spiral
This helix was created to enable heavy trains to climb the gradient what was too steep for a straight run. The south-bound track doesn't run the loop, only the north-bound track does. The basic idea is that the track is constructed in a loop that gradually inclines, allowing trains to keep speed and climb at the same time. Tunnels are used to allow tracks to cross.
There is an access road (look for the Bethungra Waterworks sign) that will take you up to some amazing vantage points. Trucks and other machinery often use this road, so please make sure you park your vehicle out of the way!
Junee to Harefield
Junee is another junction-town like Cootamundra. It also hosts a range of accomodation and constant activity. There's a roundhouse also, which has a miriad of rollingstock hanging around in various states of (dis)repair. Qube has recently been running shuttles out of Junee to Harefield to build their consists which then run to Melbourne and Sydney. The rail from Junee to Harefield is relatively straight, but has some great curves mid-way.
Fortunately, you can then cruise down the highway and meet any train again at Harefield itself. There's usually always activity here with a train being loaded or shunted.
So, this was 2015... I've got a wad of photos from 2016 to post now... but I'll let this lot settle first.