Time to build a 540B-1 replica.

[CommodoreZ]

I've begun assembling my 540B-1 board! Various parts shipments have been coming in for a few weeks now, most should be in within the next week or so. The remaining components to arrive are mostly passives, that slightly uncommon transistor, and the 8T26's. I have yet to make the jumper modifications nor cut any traces.

I've been going based on Jeff's build log on how to populate it. Thank you for providing such detail in your reports, it's been very helpful.

540b partial.jpg (2.65 MiB) Viewed 14942 times

With any luck, I'll be ready soon enough to install it in my C4P. Question is, when the time comes to test it, will I have to change out any of my ROMs to accommodate the shift from a monochrome board to a color one?

20200809_054450.jpg (1.89 MiB) Viewed 14942 times

Also, I've unfortunately discovered that my current EPROM programmer, a TL866II+, does not have the voltage requirements to burn a 2716 (it's about 7V shy of the required VPP). I'd drag out my old Willem, but I don't trust that thing to burn anything other than my house down at this point. For the sake of testing, if I had to pull the character ROM from my monochrome 540B board, what are the chances it would work?

[dave]

Interesting. I didn't realize that the TL866II+ couldn't program a 2716. You can still use a 2816.

Or, you can try making an adapter that breaks out the programming ping (VPP, pin 21) with the rest of the pins straight through (e.g., a ZIF socket with pin 21 bent up). You could then separately apply a 25V source between VPP and ground, and otherwise program normally.

You could get fancier, and include a circuit to fool the pin detect circuit on the programmer, and/or use the VPP signal from the programmer to gate 25V to the chip.

The build looks awesome so far. Jeff's build log is a gem.

Dave

[gekaufman]

Happy to burn 2716's, and have a working 540B that I can test them in. Hopefully you're in the USA as postage to elsewhere is crazy.

Same roms for color or monochrome.

The TL866 and later version TL866II+ both won't handle 2716's properly.

- Gary

[Mark]

I think you could burn the contents to half of a 2732 which programs at 21V, depending on what pin 21 is wired to on the 540 . Same read pinout/number of pins. Perhaps use the other half for the OSI "hires" character generator ROM with a switch?
But then the TL866II+ only gets to +18v VPP... so that wouldn't help.

-Mark

[gekaufman]

Make sure to use a 2732a as the 2732 programs with 25v.
Just duplicate the 2716 image in both halves.

[CommodoreZ]

While I do have 2732A's, those have the same limitation as the 2716's in a programmer that can't provide more than 18V on VPP. I lack any 2816's to try that approach. Smallest EEPROM I have is a 2864.

I've attempted adapting my TL866 with an external voltage source and disabling the pin detection, but I've been unsuccessful at burning a 2716. I'm going to see if I can find where I put the Willem and give it a shot.

The 8T26's arrived today, as did the 2N5225's. Connectors and remaining passives arrive tomorrow.

[Mark]

Looks like someone figured out a solution to the low programmer voltage with a bit of hardware...

https://www.eevblog.com/forum/blog/eevb ... eview/650/

TL866 25V mod

TL866.gif (20.29 KiB) Viewed 14890 times

[CommodoreZ]

I found that schematic, but significantly lower resolution. Might be worth a shot. I did find my Willem 5E and test 2716 and 2732A. No dice. Couldn't get it to burn anything, which wasn't a surprise.

I mentioned the trouble I was having to a few friends, and one offered to burn a 2716 for me and lend me his older series programmer for awhile. I picked it up last night, meaning I've got a character ROM. I also found my right-angle molex connectors for the bus! All that remains now is the last batch of passives which are in the mail.

I'm curious how common it was for 540B-1's to be built up in the color interface configuration. I've only seen photos of one or two in action. I would have imagined that it was an inexpensive and desirable upgrade for users to pursue, but I get the impression monochrome was more common. How off-base am I?

[Mark]

All 3 of the 540B's I've seen in the wild have been fully populated for color output, however I haven't seen a lot of software that took advantage of it.

Display quality was an issue and monochrome monitors were easier on the eyes when trying to read a 64 character display on composite video. Is that a 'B' or an '8'? The 540A sold with the 500 CPU in the C2 was monochrome only. The color on the 540B was always a bit of a kludge and didn't look that great. The big phase adjust pot on the back of the system to change the modulation interference didn't really help. The cheapest color displays were still 3x the price of monochrome and were generally modified color TVs.
I think Plot BASIC supported the color display but I can't recall any other OS version that would even scroll the color RAM with the character display!

[CommodoreZ]

Behold, a populated 540B board!

540B populated.jpg (3.45 MiB) Viewed 14861 times

Spent most of the day on it.

20200816_012946.jpg (1.81 MiB) Viewed 14861 times

I thought this particular bodge wire was fun.

When I was adding R74 to my 502 board, I noticed this particularly egregious mess as I was adding a socket for the cassette I/O

20200815_191256.jpg (1.91 MiB) Viewed 14861 times

I know my repair job there isn't perfect, but it will do.

20200815_194121.jpg (1.95 MiB) Viewed 14861 times

After I was done with that, I tested the monochrome 540 board to ensure that my changes didn't disrupt functionality before I started testing the new one.

I think I've got a few intermittent connections on my trimpots which has me concerned just based on a quick video test not showing much of anything. I'm going to hook it up to the scope soon enough and go through the various waveforms and test points in the Sam's document to work my way up to viable video. This will be interesting, and either fun or frustrating depending on how much is wrong.