help in simple logic problem

[zhaniko93]

https://www.electro-tech-online.com/attachments/untitled-jpg.50082/
https://www.electro-tech-online.com/attachments/untitled-jpg.50083/

I have a counter connected to to latch throught DeMUX
the problem is:
look at the first picture, when I click on A (input button which is zero), the first latch will go high and THEN second will go low and while this THEN, 1 is "sent" (transformed to 0 with NOT gate) from the first one and 1 from the second one (transformed to 1 with NOT gate), so in this tiny interval, 01 is being sent to Demux and it sets the latch (01 means the first pin, so it sets first pin to 1 and the latch is being set). After this long discussion I hope you get the point. How to solve it? Simply speaking, the to latches form a "counter", which, after increasing to 2, on next step it quickly goes to 1 and then to 3. how to remove this 1????? So on the second pic there should be 1 on D (look at the first one)
I have attached LOGISIM file for those who have the program

 

[MikeMl]

I notice that you have your clock edge wired to both the T input and Clock input of the T flip-flop. I know nothing about your simulator, but in any real logic family, this is a real NO-NO! Look up the Set-Up and Hold requirements for a T-flop.

 

[zhaniko93]

Yes but it didn't solve the problem (see attachment), thanks anyway for correction!

 

[MikeMl]

That seems a silly way to build what appears to just be a 3 bit ripple counter? What is using the 2 bit demux supposed to accomplish?

The glitch comes from the fact that the demux is decoding a two-bit ripple counter, which if you consider some reasonable propagation delay through the T-flops causes the decoding of transient states. A cardinal rule of logic-design is "NEVER decode the bits of a Ripple Counter" If you must decode the states of a counter, then use a synchronous counter that counts in a Gray code, and that will guarantee glitch-less decoding.

 

[zhaniko93]

That is for bigger project, All i want this to do is to count in decimals. just put high the first pin of demux, then second, then third and then forth. a very simple problem which am I STUCK in. after 2 it just jumps to 1 for maybe picoseconds or less, but is enough for any logic "step", and the latch changes. I want this inaccuracy to be removed. heeelppppP!!!!!

 

[zhaniko93]

ok sorry i didn't know the rule and can you suggest any book or something for such "rules" and whole digital electronics? I want to learn digital electronics!

 

[zhaniko93]

and plz can you give the schematics of the "Grey code counter"? the counter which will fit my circuit

 

[MikeMl]

Read this thread. It'll get you started.

 

[MikeMl]

Also, a CMOS 4017 decade counter internally is a glitchlessly decoded 5bit Johnson counter which counts by tens. Download the data sheet and study its internal wiring...