Multisim and Ultiboard

cancel
Showing results for 
Search instead for 
Did you mean: 

singular matrix: check nodes...

I have a strange problem. I'm doing an electronic tic-tac-toe game, and I have the circuit working for just one position. I've put all that into a subcircuit and it's still working. But then I add the same circuit for a different position I get the whole "timestep too small" error and a whole lot of "singular matrix: check nodes adacin1:xu3.y#branch_1_0 and...". I've tried troubleshooting for a timestep error, but none of those adjustments help. It just seems strange to me that it works with one block, but add another one that's wired the same and it doesn't. I've attached the circuit, which includes both the subcircuits. Any help would be greatly appreciated.

James
0 Kudos
Message 1 of 10
(21,828 Views)
I would try to help, but for some reason I can't get your file to open. I have the freware version of Multisim 9 and it should recongnize MS8 files, but it isn't doing so in this case. Maybe you could try positing it a different way (don't ask me how, cause I don't know).
Kittmaster's Component Database
http://ni.kittmaster.com

Have a Nice Day
0 Kudos
Message 2 of 10
(21,810 Views)
The freeware version has a 25 component limitation, so my circuit must have more than that. Thanks, though

James
0 Kudos
Message 3 of 10
(21,807 Views)
I failed to mention that I have Multisim 2001 Power Pro. I can plug your circuit into it if I had a picture of it. If you are still looking for some help just let me know via the forum.
Kittmaster's Component Database
http://ni.kittmaster.com

Have a Nice Day
0 Kudos
Message 4 of 10
(21,796 Views)
Here's a screenshot of the main circuit and the subcircuit. When you try simulating it, it will only work if you delete one of the subcircuits. It's supposed to mark X or O depending on whose turn it was when the switch was pressed. Thanks for your help

James
0 Kudos
Message 5 of 10
(21,790 Views)
Oops, here's the screenshots
Download All
0 Kudos
Message 6 of 10
(21,790 Views)

Here's my findings and as simple a solution as I could come up with. Like you, everytime I hooked the CLK pins of both subcircuits together the "sigular matrix" error reared its ugly head. I deduced that by unhooking the CLK to the second Subcircuit the error disappeared. So, I tried hooking up a 0ohm resistor between the first and second subcircuits on the CLK line and the error disappeared and the circuit seemed to work O.K. This is not saying that it worked the way you wanted it to work only that the error message no longer appears. 

You can give this a try and hopefully you will have the same success. I hope I solved your problem and if you need anything else I'll be here.

Kittmaster's Component Database
http://ni.kittmaster.com

Have a Nice Day
Message 7 of 10
(21,775 Views)
Adding a resistor to the CLK line allows the simulation to run, but when a switch for that position is pressed the turn doesn't change, so it sets the wrong LED when the other switch is pressed. I think I can work with it, though. Thank you so much for your help.

-James
0 Kudos
Message 8 of 10
(21,752 Views)
I have been further analyzing this circuit and have come up with a better option that makes the circuit work the way you intended. Instead of my resitor idea, try running the CLK  lines of the Subckts to the inputs of a OR gate and the ouput of the OR gate to the CLK input of the main JK-FF. The drawback is that there will be an extra logic gate in your circuit, otherwise everything seems to work fine.
Kittmaster's Component Database
http://ni.kittmaster.com

Have a Nice Day
Message 9 of 10
(21,743 Views)
Yes! That is so coincidental that you suggested that solution. Last night as I was thinking about a workaround I came up with the same idea, and I was just about to try it but I thought I would check here first. How wierd. Thanks so much Lacy.

-James
0 Kudos
Message 10 of 10
(21,735 Views)