_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

I am good with the scanlines! That retro look is gold.

They are inevitable due to the fact that I am not interlacing, so every second line is missing.
Although not an official "mode", it is really 240p instead of the typical 480p.

What I am referring to is the retrace lines.
If you look at the last photo, which is not on the previous page, you can see seven diagonal retrace lines.

These happen when the grid is too hot. The beam is still pushing during the vertical retrace.
I have to dig into the adjustment. It is usually near the flyback, but on this PCB I don't see it yet.
I will eventually find it and lower the "grid" or "screen" voltage to hide those lines.

I also did consider interlaced output for a full 640x448 display, but this complicates the video circuit in several ways.
First, I am using only 32K per buffer. 640/8=80 * 224 works out to 7 bits (X) and 8 bits (Y). Nice and clean.
448 vertical lines would now require 9 bits for (Y) and 64K per buffer. That means 4 * 32K and inverters.
Also, the drawing of odd and even frames makes the graphics more complex, and the 6502 in the OS is doing a lot already!

My display will do 80 x 28 characters, and that is plenty for what I need to display.
This is also a fully bitmapped system as you can see by the test image, so it will be able to do a lot.

Brad

Here is the frame buffer segment of the video circuit.
Nice and simple, but also very fast.



The 6502 writes to the one video memory while the other is shifted to the CRT display.
On the VBlank, the 6502 flips the buffers. This lets the OS continue to work as the SyncGen hammers out the screen.
The SyncGen is made of several counters and 688 comparators.

Dual buffers are probably overkill for this project, but it's not like I am afraid of adding more ICs!

Brad

Not a lot of point fine tuning the CRT until it's in the frame; all that iron around it is going to affect its magnetics, at least a little. We used to make frames for walls of colour broadcast monitors, and preferred aluminium for the frame for exactly that reason. (Though to be fair, we we _very_ picky about colour accuracy, geometry, and convergence...)

Neil

p.s. Will you be using a small steam engine to provide a generator with motive force? I feel there is scope for one

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

And a crew of three: driver, fireman, and player.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

Actually... if it were steam powered, it would have to be Calliope, no?

My goal was to have no single part (main unit or stand) weight more than 100 pounds and I may actually hit the goal.
The unit will be easily moved up and down a standard stairwell by two people without steroids.
The stand / floppy drawer is removable in two sections and not part of the original design, so that helps.

The main frame is the heaviest component and I would guess it at around 50 pounds so far.
I will toss it on a scale if I have time this weekend.
Wood panels will be second heaviest bit and then finally the CRT and frame.

Now compared to the equally sized upright piano, I am coming in at 10 times less the weight.
On that note, I could load all 16 channels with individually sampled piano hits (one per note) and have essentially piano quality emulation!

Using all HC logic, I expect the power requirement to be low.
I was running one of those huge boards on a 500ma 5v wall wart.

The 1541 floppy guts and the CRT eat up about 2 amps at 12 volts, but they have a dedicated power supply.
I will probably run the works off a small ATX 500 watt computer power supply that puts out 5 and 12 volts.

Oh, and the amp is 200 watts, so it also has it's own PSU.
It's driving dual 250 watt 6.5 inch speakers and all of it maybe weighs in at 20 pounds.

Box dimensions of the unit are 70 inches width, 24 inches depth, and 60 inches height.

Brad

I had a meeting with "Admiral Computers" this afternoon and they helped solve one of the problems I was having with the current design.
I was using a single 170K floppy from their V20 home computer, but this required a lot of disc swaps between OS loading and the samples.
Since each of the 16 channels can hold 512K worth of sample data, this became really tedious.

CEO Jacques Leimart pounded his fist on the desk and proclaimed... "you shall have dual half meg disk drives"!

And just like that, two of these beauties came off the 8050 disc drive production line....



Not only that, but they are supplying the complete controller board that allows a much faster IEEE parallel interface.
Now I can load an entire channel from a single disc in under a minute. Just amazing!
With dual drives, users can leave the core OS disc in Drive-0 and load songs or samples from Drive-1.

I have also received the keyboard assembly from their Japanese factory, and it is great.
Also the same one used in the V20 home computer.

I updated the model to reflect the new drives and keyboard.
The keyboard cutout will be made this week using the high precision angle grinder.



Still searching for a wood panel source, but no rush.
Will keep working on the frame as long as winter is a thing of the past out here.

Radical Brad

The keyboard assembly arrived today, fresh from the Admiral V20 production line.
What a great mechanical keyboard, and a simple Row + Column matrix interface.
I made a paper template so I can transfer the dimensions to the keyboard front steel panel.



I will hack out the opening with a grinder, keeping inside the lines and then tune it with a hand file.
The function keys will probably be the most used keys with apha keys being mainly for song naming when saving.

Brad

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

It's 1980, I have not yet heard of such a drive!

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!