All the simulator options is same between MMSIM13 and MMSIM15, for gmin=1e-14, the output ac response is different when the gnd is not connected. why?![]()
↧
Simulation results is different for the same circuits when "No connections to node `0'" between MMSIM13 and MMSIM15
↧
eliminate multiple comments in the excluded signals in IMC
Hi everyone, I'm working with IMC and when i read multiple Vrefines files the excluded signals get multiple comments, i need to know if there is any way to eliminate the extra comments ? Thanks![]()
↧
↧
RE: Simulation results is different for the same circuits when "No connections to node `0'" between MMSIM13 and MMSIM15
Why are you trying to simulate without the ground connected? In both cases you're probably getting numerical noise (unless I've misunderstood your question, which isn't surprising given the sparsity of information in your post). Seeing the circuit and the simulator output log might help here... Andrew.![]()
↧
RE: Keyboard shortcut for running simulation
Hmm. This is very suspicious - your post is identical to this post (the link is to my reply). That's also identical to this even earlier post (link is also to my reply). The original poster from 8 months ago has a number of other posts so I'm sure is genuine but this and tutu10 seem rather strange to me. Regards, Andrew.![]()
↧
RE: do I have to write full path for my simulation model?
There are several approaches you could take here: Specify the model file as $MODELDIR/my_model.lib.scs Specify the model file as my_model.lib.scs and then add $MODELDIR in the Include Paths in Setup->Simulation Files - this has been supported since IC616 ISR9 (before that ADE didn't understand use of UNIX env vars in this field) Specify the model file as my_model.lib.scs and then also set the environment variable SPECTRE_DEFAULTS before starting Virtuoso: setenv SPECTRE_DEFAULTS -I$KITHOME/models/spectre (for example). This adds the specified argument (-I in this case - see "spectre -h" for more details) to the spectre command line whenever spectre is launched). It then gives the include path to be wherever you want. Regards, Andrew.![]()
↧
↧
Running AMS simulation on multiple Std. cell
HI Experts, Needs some suggestion to automate the below flow, ( or may be you can suggest some better way) Suppose we have multiple cell for inverters, and we have created respective function model of those cells. We are ruing AMS simulation to verify the models, ( creating config view and running simulation) I am looking a way to just create a one TB for inverter and it can run simulation for all inv flavor. I have the amsDMV flow based on simulation results to verify the model correctness. Thanks, Nasser![]()
↧
Tab Key issue in PCB Designer 17.2.050
Hi all, Since the update to the 17.2 version, I have been encountering an issue in PCB designer professional with the tab key which toggles the selection between the different selected elements of the filter. The software always reacts as there were more key stroke than one.I have found a workaround with the RMB menu and toggle select command, but it requires more clicks than the tab key stroke. I hoped this would have been solved in the previous hotfix (050), but apparently not. This issue is not even listed in the CCR. Do you also get the same behaviour? How can I submit a CCR by myself? Thanks for your help and kind regards, Nicolas![]()
↧
multiple Y-param results for varactor
Hello , i have built a varactor based on the theory shown bellow, on one hand i need to sweep control voltage on the other hand i need to have Y-param for every value of control, and convert it into capacitance,so we have here multiple SP simulations for each value of CONTROL voltage, is that the correct order of simulating for creating the Capacitance vs Control plot,as shown bellow? i know that we can use Goldengate to convert Y-param into capacitances (for certain frequency) Thanks![]()
↧
RE: multiple Y-param results for varactor
Dear yefj, I am not sure if I fully understand your question "so we have here multiple SP simulations for each value of CONTROL voltage, is that the correct order of simulating for creating the Capacitance vs Control plot,as shown bellow?". However, if you are asking how to generate capacitance data as a function of control voltage for various switch settings, this can be done with a series of small-signal AC simulations. It is not necessary to use S or Y parameters. I have outlined a methodology to accomplish this a few times. The most recent description, which includes a sample test bench and computation, is at URL: https://community.cadence.com/cadence_technology_forums/f/rf-design/38996/capacitance-vs-bias-voltage-curve-for-ferroelectric-varactor Did you see this forum posting and does it answer any of your questions? Shawn![]()
↧
↧
Simultaneous Load-pull and source-pull simulation in cadence
Hi there, Happy holidays and new year. I wonder if cadence has an option similar to ADS to simultaneously simulate source-pull and load-pull for simultaneous conjugate matching. Only simulating load-pull returns meaningful contours for the optimum load conditions. However, doing the same thing on the source side returns few points and does not really seem to optimize the input power. For these simulations, we use the adapter from the rfExamples library. Is there a different setup to simulate the source pull conditions? And also are there any chances to perform both source-pull and load-pull in one run of sims in one testbench? Many thanks in advance, Hamid![]()
↧
RE: Simulation results is different for the same circuits when "No connections to node `0'" between MMSIM13 and MMSIM15
Hi, Andrew. Thanks you for your reply. The reason why I simulated the circuits without the ground connected is that I find the simulation results is different when I simulated the comparator input noise in different spectre version. So I take a easy filter circuits without the ground connected to observe the noise level. Here are some details about the comparator input noise simulation. The first figure is the schematic of comparator, and the second is the tran analyses setting. The last is probability of comparator output voltage large than 0.5 in different input voltage. So I suspect whether the mechanism of the noise generation is different in different spectre version.![]()
↧
Copying p-cell from one library to another.
Hi all, I tried copying a p-cell from our tech library to a local library. But the p-cell is copied as a normal cell. Can anybody tell me how to copy a p-cell from one library to another library. Thanks, Ganesh Doddipatla.![]()
↧
RE: How to install old versions?
Bump... if anyone is willing to share Hotfix_SPB17.20.047_wint_1of1.exe with me, I'd much appreciate it.![]()
↧
↧
RE: do I have to write full path for my simulation model?
Hi Andrew, thanks for your suggestions. I followed your approach 3 by "setenv SPECTRE_DEFAULTS /net/artery_proj/bowhao/Process_Files/Model_Files/umc55nm_LP". and it works (log file displayed content of "SPECTRE_DEFAULTS"). i specified model file as "sim_model.lib.scs" (full path is: /net/artery_proj/bowhao/Process_Files/Model_Files/umc55nm_LP/sim_model.lib.scs ), and run simulation. while simulator failed to find it. did i missed anything, or did I set SPECTRE_DEFAULTS correctly?![]()
↧
RE: do I have to write full path for my simulation model?
You omitted the -I before the /net/... You should have done: setenv SPECTRE_DEFAULTS -I/net/artery_proj/bowhao/Process_Files/Model_Files/umc55nm_LP The -I is necessary to specify the include path. Regards, Andrew.![]()
↧
RE: Copying p-cell from one library to another.
Hi Ganesh, Assuming you're really talking about a PCell, this can't happen - it can't convert itself into a normal cellView by doing a copy, because it's an attribute within the cellView itself. Most technology libraries just have layout pcells (occasionally schematic pcells but less frequently) - did you copy just the layout cellView? What exactly did you copy, and what is the behaviour you're seeing after the copy? You've really not given enough information here. Regards, Andrew![]()
↧
RE: Simulation results is different for the same circuits when "No connections to node `0'" between MMSIM13 and MMSIM15
I don't understand how simulating without the ground connected is useful - circuit simulators need a ground node as a reference. There's not really enough information to go on here to explain the difference. It might be that the default noisefmin changed to 1Hz (I can't remember exactly when that happened). The best thing here would be to contact customer support as then you can hopefully provide the input.scs and simulator log files (I asked for the simulator logs earlier). Your picture above isn't really enough as it doesn't explain how it's connected up, and even then I doubt I'd be able to get to the bottom of it without seeing more information. Regards, Andrew.![]()
↧
↧
RE: Copying p-cell from one library to another.
Sorry Andrew for the above description. Actually, what happened is when I am instantiating that PCell in another cell view it is showing PCell Evaluation failed. So I thought that it is copied as normal cell. Could you tell me what could be the mistake I made. FYI The cell I copied is a capacitor PCell (vncap_b) from our tech lib to my local library doing a test case for migration. Thanks, Ganesh Doddipatla.![]()
↧
RE: Copying p-cell from one library to another.
Hi Ganesh, First step I'd take is to do Verify->Marker->Explain and click on the text where it says pcellEvalFailed. What does it report? Regards, Andrew.![]()
↧
RE: do I have to write full path for my simulation model?
Hi Andrew, it works after I added "-I". thanks a lot. (for other viewers: it's a uppercase i, not lowercase L). one more thing is: if i launch 2 virtuoso individually (based on 2 different process), setenv works for both virtuoso and then spectre will find model file under exactly the same path. is there anyway to set different path for different virtuoso? so i can identify process model by path (instead of model file name), and keep model file name unchanged.![]()
↧