Clocking writeM

[Pillages]

When I update writeM using combinational logic, everything in CPU works fine. However, in order to plug CPU into Computer, my understanding is that each pin needs to be clocked. The problem is that when I try and clock writeM, it updates one clock cycle too late. Anyone have any insight, helpful hints on this problem?

[cadet1620]

Pillages wrote

When I update writeM using combinational logic, everything in CPU works fine. However, in order to plug CPU into Computer, my understanding is that each pin needs to be clocked. The problem is that when I try and clock writeM, it updates one clock cycle too late. Anyone have any insight, helpful hints on this problem?

You don't need to clock any signals in the CPU. The clocking is all handled by the DFFs deep down in the Bits within the RAM.

--Mark

[sophia]

I'm confused about the same thing. In the CPU, I update writeM using combinational logic, but if I don't clock the signal before passing it in to the Computer chip's Memory, then I get the error "This chip has a circle in its parts connections".

I was passing the output of writeM through a DFF to clock the signal into Memory -- but then it's one cycle behind and nothing ever gets written to memory... :(

My CPU and Memory chips both pass tests, and my Computer output is all correct except for the memory (using the clocked writeM).

Any suggestions would be very much appreciated. Thanks!
Sophia

[cadet1620]

It sounds like there is an unclocked path from writeM to inM. This is probably through Memory.hdl.  Is writeM used in any way to control the output from the RAMs, Screen or Keyboard? It should not be involved with the output side of Memory.

If you want, feel free to email me your CPU, Memory and Computer HDL files and I'll let you know what I see.

--Mark

[sophia]

Hi Mark,

That's exactly what was going on -- I was trying to use writeM in the Memory chip to toggle output.

Thank you SO much!
Sophia