exxos's Atari, Amiga & retro forum

Stephen, I have a fair bit of 'PC GEM' software in zip archives.

PM me if you want me to send it over. :)

OK, will do!

Wonders..... "Can I replace the PC-GEM desktop fonts with the Atari versions?..."

stephen_usher wrote: 21 Feb 2023 20:13 Wonders..... "Can I replace the PC-GEM desktop fonts with the Atari versions?..."

good question,
I read somewhere that Atari GEM uses FNT in Intel format

Cyprian wrote: 22 Feb 2023 09:16

stephen_usher wrote: 21 Feb 2023 20:13 Wonders..... "Can I replace the PC-GEM desktop fonts with the Atari versions?..."

good question,
I read somewhere that Atari GEM uses FNT in Intel format

It does indeed, as that's the GEM font format.

I've used the STonX emulator's "tosfonts" utility to extract the fonts from the ROM image.

However, I may need to rebuild the GEM display driver with the new fonts as it holds the "standard" fonts internally.

I had a bit of energy this evening and so thought about going back to something I've had in the background for years.

Probably about 10 years ago, whilst renovating a BBC Micro which had the "DataGem" database ROM installed I accidentally installed the DataGem PCB into the ROM socket backwards as there was no orientation marked upon it. When I turned on the Beeb, well, you get the picture.

At the time I left it for dead and put it on a shelf. During lockdown I saw the PCB and thought I'd have a look at it. I first removed the EPROMs and tested that I could read them (there are two plus four other chips), they worked.

So, now to the other logic chips. The software company had sanded the markings off all four of them but I was able to make out enough on three of the chips that they were 74LS133, 74LS02 and a 74LS00. The fourth chip was completely blank. I desoldered them all and tested all the chips I could identify and they all checked out. This meant that the one chip I couldn't identify was obviously the one which was killed! Typical!

Anyway, tonight I thought I'd take another look and try to work out what this other chip could be using the traces on the single-sided PCB. The more I looked the more it looked like triplets of inputs/outputs, very similar to the 74LS00. Of course I couldn't guarantee that it was a quad NAND gate.

So, I stuck it into the 266 EPROM programmer and tested the chip against every 14 pin 72LS chip in the database. The closest match, with one error, was the 74LS00. It looked like channel 1 was constantly high.

Searching my chip stock (how have I got so many 244s and 245s?!) I found some LS00s, just a few... Phew! :-)

I soldered sockets in and put the three known good original chips in and a new 74LS00.


I got my original, trusty Beeb down from its place on the back of the sofa-bed, tested that it still worked, fitted the PCB and switched on... Boo-Bip! Yay!

Not only that but...


And it does indeed work!


Anyway, I'm not sure what I'm going to do with it, but I do know that it's operational now.

On another note, my outbuilding is coming along now and I can finally get an accurate measurement of the internal dimensions as the builders have lain the first of the internal wall blocks.


The size inside (excuse the Imperial measures) will be 23' 3" x 8' 3". This has given me an opportunity to properly size the railway layout and workshop spaces. With the planned 16' layout, it's going to make the workshop area a bit tight...


I may have to shrink it a bit.

After last night's BBC shenanigans I decided to check on the BBC B+ 128K and BBC Master 128 as they've not been down from the loft in a couple of years and the B+ seems to like to have RAM chips die whilst in storage.

The B+ was happy and no issues. The Master, however...

Initially the power supply was bouncing, i.e. on for a short time, off for about half a second and back on again, etc. Unplugged the PSU from the motherboard and it just about limping into life with the correct voltages. Plugged it back in and after a few more chugs it finally started up. Everything else looked in order, except the I key and space. The I didn't work at all and the space was very intermittent. So, I squirted good old Electrolube EML down the step, waited a minute and the I key sprang to life without even having to press it lots of times. A similar squirt down the space bar switch and all good again. Electrolube EML works miracles!

Anyway, I turned my gaze to the power supply. This machine is an eBay purchase and the previous owner had "recapped" it in 2017 apparently. Well, looking closer at the PSU and the wiring rang alarm bells. He'd broken one of the ground wires somehow and poorly soldered it and just wrapped it in cheap electrical tape which was unsticking itself. He'd managed to strip both the screw and socket of one of the location positions and the gromet protecting the wires coming out of the PSU from chaffing was missing. uh-oh!

Anyway, I stripped the PSU down and found that the "recapping" was just the RIFAs. Also, the start-up cap connections to the circuit board looked rather iffy. This was confirmed when I tried to desolder it when it just pushed through, dislodging dried flux. It seems that it was held in with flux and no solder! Anyway, I replaced it and checked the other caps. They looked good quality and quite late in date for a BBC Master as some had 1988 date codes. Given the issue seemed to be a start-up one and there was an obvious fault I didn't bother looking at replacing them. I didn't have a spare 2200uF 25V cap anyway.

Reassembly was the opposite of removal, as Haynes manuals like to lie and tell you. On the way out I put electrical tape around the inside of the hole where the wires come out to prevent chaffing and put back the cable ties. I also retapped the hole on the securing bracket and replaced the M3 screw.

I also re-did the ground wire soldering and fitted heat shrink on it. Everything now safe.

Well, that was an unexpected repair.

Ahh yes, good old Electrolube, it's got me out of few switch related problems over the years :)

I was wondering if you had an iffy start up cap, a problem I've seen a few times now, but it sounds like a lot more was needed, but good that you found it and got it repaired and gave the PSU a good sort out.

:girldance:

I've not had the energy to do much stuff recently with the inflammatory bowel disease flare up. So no more progress on getting the disk images completed on the Apricot F Series.

I have converted a couple more model locomotives to DCC control. The only one which gave difficulty was the "Mainline" Class 56. The original wiring all hung off the motor and was a dog's breakfast from the factory. Goodness knows how it didn't short out. So, I had to rationalise it all. Still, all done and the directional lights work.