This week I’m preparing the first batch of Harlequin 128K kits, and I’m using the issue 2D for that.
Now some people asked if the Harlequin issue 2D has the exact same timings as the original toastrack has.
Some say the Harlequin issue 2D has 48 timings and thus is different, but others say issue 2D should have toastrack timing
So I ran some demos and games with border effects, and IMHO I found out there are minor differences.
Let’s compare the video output.
Systems used
Upper: original toastrack, lower: Harlequin issue 2D.
The perticular toastrack board on the upper photo had a bad keyboard membrane, hence I’m using the external mechanical (which makes it sound cooler) DK’Tronics keyboard 😉
The Harlequin board on the lower photo is inside some kind of toastrack case.
First test: the timing test I got from Dan
Upper: original toastrack, lower: Harlequin issue 2D.
There is a huge difference. No idea which one would be the correct one as both have some strange details.
Second test: NMI 3 demo – part 6 ‘Splitty Overscan’
I selected 128 timings when I ran this.
Upper: original toastrack, lower: Harlequin issue 2D.
It’s quite clear that the upper photo (from the original toastrack) runs as it should.
Third test: Vectron
Upper: original toastrack, lower: Harlequin issue 2D.
No noticable difference here.
Conclusions
IMHO there are timing differences.
I received this info about timings:
“Harlequin issue 2D has the same timing as toastrack,100%.
Toastrack use 17.7345MHz crystal, divide by 2.5 to get pixel clock = 7.0938MHz
Issue 2D use 28.375MHz crystal, divide by 4 to get pixel clock (HCa) = 7.09375MHz.”
But how I read that is that the first line contradicts what is said in the last 2 lines: there is a minor timing diffenerence.
Update May 2019
After writing this article I found out that the slight differences were actually caused by the CPU used.
It seems that there are small timing differences between Z80 CPUs.
So the architecture of the Harlequin 128 rev 2D board makes it 100% toastrack and +2 timing identical, but the used Z80 CPU can cause a slight difference, which is often neglectable.