why won't a flip flop made out of gates work?

[Russtsa]

Hi everyone!

I'm evaluating Multisim 10 for purchase,  and the first thing I tried was building a J-K flip flop out of gates (similar to a 74107). It doesn't work anything like it should (attached). The 74107 that came with Multisim works like a charm. Could someone please tell me what I'm doing wrong? I'm using NI Multisim 10 version 10.0.144. My OS version is Microsoft Windows XP Professional Version 2002 Service Pack 2.

When I hit the simulation switch, it appears like it is going to work for about 1 cycle and then it just dies. I can't justify the cost of this application if I can't even model a simple flip flop out of primative logic components. I hope someone can set me straight.

This site won't allow me to attach an *.ms10 file so I've attached a screen shot of the circuit instead.

Thanks in advance!

Russ

[Russtsa]

I forgot to mention that I'm running the "Education Edition" in evaluation mode.

[lacy]

I am usually pretty good at finding problems with circuits, but this has me stumped. I tried your circuit with my Multisim 2001 Power Pro and it did not work the way it should. It just sat there oscillating at a very high frequency reguardless of the clock frequency. I tried rearranging it to no avail. I tried using CMOS parts intead of the individual gates and had the same result. I even went one step beyond this and found another circuit off the web and it did exactly the same thing with both individual gates and CMOS parts.

There is just something about digital oscillators created this way (which is what this would be if both inputs are tied high) that multisim does not like. The only reason I can come up with as to why the JK flip itself would work and this circuit would not is in the way the Component versions are modeled. They may be modeled other than with digital gates (i.e transitors, diodes, current switches)

I hope maybe someone has a solution for this. I would like to know what, if anything, someone else may find because now I am curious about this too.

[Russtsa]

Thanks for trying, Lacy!

I ran into another related issue. I've been using Electronic Workbench
5.1 for years. It's getting a bit old, so that is why I decided to
evaluate Multisim. I modeled up a 2 bit counter in Electronic
workbench and got it working. I then modeled up the same 2 bit counter
in Multisim and guess what? I spent almost all day trying to get it to
work... Nothing like beating a dead horse!

Attached are two screen shots. One shot is of the circuit running in
Electronic Workbench, the other shot is the same circuit TRYING to run
in Multisim. As you can see from the O-scope in Multisim, it's got a
good clock input, but the A and B outputs are pulled high and totally
inactive.

I hope the dev's are monitoring this forum and respond with a
solution. I really like Multisim, but I need it to give me reliable
results. It's pretty sad when a predecessor application works better
than the upgrade. I keep hoping I'm just doing something stupid.

Does anyone know if there an evaluation version of Multisim Base
available for download? The only one I could find is Multisim
Educational.

[lacy]

Well I am one for beating a dead horse until it is just a greasy spot on the ground. I will continue looking (from a circuit perspective) for a solution or workaround. If it is in the Software itself or in the internal settings, then I will leave that to someone with a little more experience on that subject.

You are right about one thing, this should work. This is basic logic circuit analysis that is taught in school. You start by learning about individual gates and how they tie together to do other more sophisticated funtions. If this doesn't work correctly, then how can you use Multisim to learn or teach this. This definitely need to be reviewed by their research/engineering dept. Maybe they can come up with an easy solution for everyone.

[digitalguy]

Your circuit won't work because you have both the SET input and the RESET input of the D flip-flop tied to ground in the Multisim circuit. The little bubbles on the flip-flop schematic diagram indicate the inputs are active low.

[_user32]

Here is an equivalent JK FF.  The reference for this is found in 'Digital Design' by Waverly.

I also put the circuit on ftp://ftp.ni.com/outgoing/JKFF_Equivalent.ms10.

[lacy]

I was experimenting with the circuit Multisim_2bitCntr.jpg and I came up with at least one solution. I understand that a previous poster replied that it wouldn't work with the Set and Reset tied to GND. This is true, but it still would not work for me in Multisim 2001. To get it to work I had to add pull-down resistors (1Kohm) to both the input and outputs of the D-Flip Flop. Also, I had to tie the input of the inverter/not gate to ground. Just a note, I left the inputs to to Set and Reset of the D-FF open and not tied to anything.

After doing all these things, My waveforms matched the EWB 5 version waveforms perfectly. I am going to experiment with the original circuit in a simular manner and see if that is what is needed for it. I will post my results if I am successful.

[digitalguy]

Here is my version of Russtsa's 2 bit counter as simulated on Multisim 10. I also left the SET and CLEAR inputs open (not good practice in real life) and tied the input if the inverter to ground. Works fine.

[Russtsa]

Thanks Lacy and digitalguy for the help with this!