ZX Spectrum Harlequin rev.G
27.07.2015
A bunch of enthusiasts decided to develop a clone of ULA-free ZX Spectrum, but with precise timing.
It appears that they succeeded and after the fourth generation of the PCB, the final release is in the world.
The link is here Superfo. Unfortunately it's in Spanish.
I also decided to poke in it a little.
I have several reasons. I like ZX Spectrum, because I grew up on it and I feel a certain nostalgia to it.
I wanted to produce RGB output according to an article in Amateur Radio A09-1990, but the project failed to work.
Then I came across a Harlequin project with RGB output and format interchangeable with the original PCB of Spectrum and it was decided.
Task, to put on my desk functional ZX Spectrum with RGB output must be met at any price.
The original project is in OrCAD, but I use Altium.
So Export / Import followed.
With great effort I have succeeded, so now I can poke in it.
I finished 3D models of all components. Some I have downloaded, others I had to draw in CAD.
3D models are a fantastic thing. Although this is terrible job, but then mechanical collisions are visible immediately. In the company where I work, we do not do it otherwise anymore.
Definitely it will reduce the number of discarded prototypes.
It can be seen that the Superfo team didn't go this way. PCB would withstand one more revision as some components are in conflict, but you can live with that.
J2 is slightly in conflict with the U32 and U34.
J1 is slightly in collision with D9 and R47.
J5 and J6 connectors collide with PCBs. They should be shifted more toward to the edge of the board. Like this they are a bit lifted, but in practice it does not matter.
It is clear to me that most of you will say that I am a nit-picker, but I like it when things are perfect.
So what's next.
I think putting a resistor into the socket is a bit exaggerated (U47 8x470R in the housing DIL16), but well, aesthetically it looks like everything in the IC socket.
DC / DC power supply LM2596 is unnecessarily oversized (3A) When it was decided to use SMD, there could be some up to 0.5A and thus considerably smaller.
Bridge D12 does have bad pitch leads, but no problem to swig the leads a little .
It is also good to take care of the printing not to interfere outside the PCB. For connectors, they are not fixed. It´s possible that they could overlook it while preparing the model for production
and imprint secondary PCB in the wrong place.
In the project, I reduced the library. Some parts have the same housing with different names.
I canceled DIL16 housing for resistors and replaced it with a classic DIL16 (one fewer component in the library).
Otherwise, I physically did not change the PCB, so if someone would generate Gerber data, he should get the same result as the original project.
After all there are Gerber data in the original project, you can compare them.
The whole project in Altium can be found below.
If you ask why I did not correct these bugs, so for two reasons.
1) I kept the original state of the revision G.
2) I'm not going to produce a revision G.
That what I'm going to work on, is on the next page.
cztomeco@gmail.com