Page 1 of 1

When the Schematic is misleading....

PostPosted: Mon Apr 02, 2018 11:47 am
by Mark0x01
I've finally decided it's time to get some of my collection to an operational state, and first up is an Osbourne 1.

After spending a few days treasure hunting - I'd forgotten where it was stored as it has been sitting in the same box for the last 20 years.
Once located, I unpacked it and checked it had survived the rodents - luckily they only ate large holes the cardboard box and paper packing around it.

Being a knowledgeable electronics bloke, I dismantled it and set it up to reform all the caps, especially the tantalum ones that usually explode on power up after a few years of storage.

So, with the schematic handy, I locate two suitable points to connect to - it has an optional DC supply input, so the ends of the isolating diodes were perfect for a couple of crocodile clip.

Applying voltage from my bench psu via a120 ohm 5 W resistor to both rails at once, I left it an about 0.2 volts for about an hour and slowly stepped it up to 0.5 volts.
I then ran the meter over the board to confirm things were as expected.

Problem - no voltage on some sections of the board. Check the schematic - R21 should be on the +12v rail, recheck - nothing!!
There are two connectors and leads from the power supply to a single plug on the board. Digging further I discover that these separate connections are indeed separate even on the main PCB.
Each of the two +5 and +12 pins feeds different board areas.
But this isn't what the schematic shows:
schematic.JPG (30.36 KiB) Viewed 10452 times

So maybe my older board is different, who knows, there is no indication this has ever changed in the manual. Never mind, lets move on.
So I simply plugged in the PSU and it's cables,which then connected them together as it would when running and started again.

The plan is to start with the main PCB and then do the PSU mains side separately, before wiring in the floppy drives, display unit and then re-doing the whole system from the start again when all together.
I'm looking to replace at least the PSU caps, since they are the most likely to give issues later on.

So once this is eventually up and running, it will be one down, and many more to go, with a DEC Rainbow next in the queue.

But I am not expecting it to be all as easy as step 1.

An interesting feature is the thermal protection - a thermal circuit breaker more commonly found in dryers and appliances, in series with the mains power.
This isn't mentioned in the technical manual other than as a spec, and poses a live mains hazard as the connectors aren't well insulated.


...Progressing steadily

PostPosted: Sun Apr 08, 2018 12:40 pm
by Mark0x01
Having experienced exploding paper foil, like Tezza with his Osbourne :wink: , and not wishing to let that "magic smoke" escape from mine, I have been proceeding with caution and will replace every electro and paper foil cap in the psu.

Using my bench supply, the main logic board and disk drives have been up to 5v, and I have had the display module running on it's own, with raster present, all without incident so far.

The PSU is still waiting on the next batch of caps to be ordered - RS Components is my usual source.
Matching the original electrolytic's is tricky, as they are all now obsolete. I did find some data to help pick a suitable replacement, but mostly guess work.

After much effort and search engine disappointment, I finally have schematics and parts lists for everything, including the PSU and monitor.

So progressing nicely so far.
One thing I can try out later is my recent purchase, an HxC floppy emulator, which will apparently does work on an Osbourne 1 with an adapter cable.

More Progress:

The size difference in the modern equivalent is quite considerable.
I haven't been able to get a spec sheet for the older one to compare the rated life.
cap_size.JPG (50.47 KiB) Viewed 10356 times

The X2 rated filter caps were definitely the worse for their age, and would have died in exciting ways.
The visible cracks are the danger sign, and have been replaced with new items the same type, and size.
X2_bad.JPG (39.19 KiB) Viewed 10356 times

So the Osbourne 1 now has been powered up fully without incident, however, there is now a logic board issue to resolve before I can progress further.
I have garbage on the display and the buzzer, so on to the next stage of troubleshooting the issues.

This one the later "Blue front" version, with a date label sticker in the power bay for 23-12-82.


Re: When the Schematic is misleading....

PostPosted: Sun Apr 08, 2018 3:12 pm
by tezza
Good stuff Mark.

I must say I'm a bit envious with the fact you have a DEC Rainbow (:

Re: When the Schematic is misleading....

PostPosted: Mon Apr 09, 2018 7:57 pm
by Mark0x01
Hi Tezza,

It did involve a considerable amount of actual "dumpster diving" when I was younger.
They had to be "seen to have been dumped" in the skip.
Luckily I was much younger then :)


Re: When the Schematic is misleading....

PostPosted: Tue May 22, 2018 10:22 pm
by Mark0x01
Progress has somewhat stalled on the repair.

The scope didn't reveal any stand out issue, so I'm going to make a no-op loop eprom to try to check the address bus.
The first hurdle was finding a suitable 2732 eprom. Unfortunately I could only find slow ones in my bits stock, so started searching.

Ali Express has just delivered five supposedly -200ns ones, and I'm dusting off my original ISA bus eprom programmer to test and program one.
I had to fire up an ancient Pentium II (slot version) pc with a real CRT monitor attached - no smoke signals :o - so I should be able to progress further soon.
I also dug up a 30+ year old UV lamp to erase them.

I have looked at upgrading my programmer, but the more modern ones won't do the 21/25 volts needed to program the older chips.
I'll look at an eeprom alternative at some stage, which looks to be the way to go.

I'm also considering creating a diagnostic eprom.
The earlier bios had better diagnostics, so I'm going to looking at that for inspiration. I just love the fact that full rom source listings used to be standard in the manuals, back when copyright only applied to the written word.


Old computers two, old fart nil

PostPosted: Mon Jan 28, 2019 11:46 am
by Mark0x01
Well things didn't pan out with progressing my troubleshooting.

Faulty test kit sends you in every direction.
The ribbon cable on my 40pin dip header was found to have multiple intermittent connections and wasted may hours when all the checking proved worthless.

So I now have a new cable ad will have to start again.

In the mean time I powered up the Rainbow, which needed a video and keyboard breakout as I couldn't find the monitor, which is where the keyboard plugs in.
Not sure if I actually had it, its been in storage for 20yrs.

The good news was no smoke, but it reports a z80 crc error, which is fatal so won't boot.
The manuals aren't kind enough to point a possible cause other that replace the system board, yeah right.

I really need a rom source list to see what it is actually testing, possibly the z80 rom, but that is only a guess.
Web searching hasn't shed any further light on this error so I'll just have to add it to the queue.