timestep too small(mixing circuit)

[vick123y]

Hi,all!
I have a mixing circuit(using AD734).When I tried to simulation,it shows that Error:

|   Output from instrument analysis  
|   |   info: tmax is automatically adjusted from 1e-005 to 9.76257e-009
|   |   TRAN:  Timestep too small; time = 2.02631e-007, timestep = 1.22032e-018: cause unrecorded. 
|   |   doAnalyses: timestep too small 
|   |  
|   |  
|   |   tran simulation(s) aborted 

When I do this :Select Simulation >> Interactive Simulation Settings >> Under initial conditions select set to zero,there is no error then. However,there is still no graph showed in the oscilloscope.I don't if there is anything wrong with my correction.
I have uesd multisim8.
I am a freshman so I don't know what the problem is.Maybe there is somthing wrong with my circuit.I hope someone can help me!
Thank you very much!
Here is the circuit.

帖子被vick123y在10-10-2007 11:48 PM时编辑过了

[lacy]

Hi there and welcome to the forum,

I have found one thing so far. Double click on the IC and check under the fault tab. When I loaded your circut into Version 10, the fault tab had the "OPEN" circle checked and all the pins had a check mark beside them. I don't know if this is happeneing in your version or not, but you should check about that.

As far a circuit operation, I can't seem to get the desired output. I have checked the datasheet and done some comparision with what it says should be going on with it, but I can't get them to match up no matter how I set up the simulator Correct me if I am wrong, but you want to Multiply X1 and Y1 and then divide that by your denominator which according to the datasheet is 2. If this is correct then I should have 20.48 Mhz at the output and I just can't get that.

[lacy]

Alright,

After numerous go arounds I think I may have something for you to try. Granted I can't be exactly sure it will work, but you can give it a shot.

1) Set Initial Conditions under the Simulaor setting to USER DEFINED. Set TMAX to 1E-005

2) Set Integration Method to Gear

3) Just as a precaution Find IT1 and set it to 1000

4 ) Also as a precaution, Find ITL4 and set it to 10,000

I have Version 10 and this is what the convergence assistant came up with (except for the last 2 which was found on a previous problem). I can only hope this works for you. There may be other options that may be better than this, but this is all can come up with at the moment.

These options seem to give me a correct signal at the output with no simulation errors.

Message Edited by lacy on 10-11-2007 09:46 AM

[vick123y]

Thank u for reply lacy!

But I can't do it well,there is still some problems.

Can u put a your corrected circuit on the internet?

[lacy]

Unfoertunely, that isn't going to possible. I am using Version 10 and you are using Version 8, therefroe even if I did post it you would not be able to open the file. I don't have access to Version 8. Maybe Nestor or one of the other Team Members have acces to it and can assist you further. Sorry, I can't be of more assistance.

[lacy]

I am just curious about one thing and maybe the NI Team can answer this and maybe it might help you also.

Is there a frequency range limit with Multisim? If so, what is the maximum frequency that Multisim can comfortably handle?

Your frequency is so high is this circuit and I am wondering if it may be hitting some sort of ceiling limit that Multisim can handle. I noticed that sometime I would get a good signal from it and other times i wouldn't. This make me wonder about the maximum frequency that Multisim can handle in Transient Analysis mode.

[lacy]

I went back and look at your circuit again. There were a few other things I changed that I failed to memtion that could be of some help. I will outline them below:

1) I placed grounds on your input sources (don't know if this was problematic, but Multisim likes its sources to have grounds)

2) Removed the Digital/Virtual ground from the IC and replaced it with a normal ground. (this chip is analog and not digital, don't know if was causing problems or not). This would be on Pin 10.

These are the only circut changes I made and the rest was through the settings. One thing I noticed is that you have 1uF caps on the input sources. This would greatly attenuate your signal a 6 Mhz. Not sure if this was done intentionally or whether it was an error. In order fro your signal to pass without attentuation you would have to lower these capacitors to a value in pF range.

In my previous post I was concerned about a maximum frequency limit of Multisim, but you might want to look at the datasheet for this IC and check it out. The way I read it (and this may not be correct) it is ouput limited to 10Mhz. Therefore if you multiply your 6 Mhz inputs and divide by 2 you end up with 20 Mhz which is out of range for this IC. Look at this and let me know if I am reading this correctly.

This is all I know. I wish I could post this for you and let you try it. But our file formats are incompatible.

Message Edited by lacy on 10-14-2007 12:56 PM

[lacy]

I apologize for an error in my last post. Your input capacitors are O.K. I was thinking backward. Capacitive reactance varies inversely with the frequency and capacitance. I don't know where my mind was at that moment, but it wasn't with me. Even I get confused on occasion but it is rare ( Laughing)

[vick123y]

Thank u!I am in China,so I can't reply immediately.The terrible time differences!

I've got the graph now.You know,I am new in circuit design.

Can u tell me the turth about the part of filter.Can it work normally？

[lacy]

I am no expert at circuit design either. I would still consider myself a Novice approaching Intermediate as far as my knowledge of design is concerned. I seem to learn something new everyday.

As far as your last statement, you'll have to forgive me for not understanding what you are asking there. Could you eleaborate on exactly what you are wanting to know?