I our PDK models we do have the option to select between different values for the CDF parameter "DFM_options": "minRule", "DFM", "DFM+Analog", "custom" .
By default it is set to "minRule" but we want to have "DFM" for all instances that have this CDF parameter.
How can the schematic Check&Save function be extended to check for the "DFM" to be set for all the instances which do have the "DFM_options" parameter ?
Can you provide a code template ?
Good morning,
I'm running MC mismatch and want to restrict the mismatch variations to a certain hierarchy defined by a config file.
This config file in itself is again part of another hierarchy defined by another config.
It seems not possible to select the instances for mismatch directly by clicking the config view:
WARNING (ADEXL-1667): The following instances will not be added as DUT because they are not valid subcircuit instances
/iloopfilter/imsde_wbadc_lf_fmdab_loopfilter/ilf_fmdab_st0_top (msde_wbadc_lf_tb_lib:tb_loopfilter_cmrr_cl:1)
The "specify instances for Monte Carlo" menu allows to select between "schematic, master, subcircuit, text " but not "config" (?).
How can I do this ?
Am I forced to first create a flat hierarchy?
groet
Hans
I'm trying to make viva look good using a white background and the default colors are kind of lousy (in my opinion). I read in the manual:
----------------------------------------------------
viva.graph colorBank string "default"
You can use this variable to customize the set of trace colors in a Virtuoso Visualization and Analysis XL session. The default value of this environment variable is the default color bank that is specified in the display.drf file.
Note: This variable does not change the colors defined in the display.drf file. Also, when you plot simulation results from ADE, the color bank from the display.drf file is honored.
The color names should be separated by a comma, a semicolon, or a space. The color name can be a list of colors defined in the SVG color keyword, such as steelblue, azzure, or the RGB hexadecimal value of the color in the #RRGGBB format.
viva.graph colorBank string "red; blue; green; yellow"
viva.graph colorBank string "#ff0000, #0000ff, #008000, #ffff00"
----------------------------------------------------
So I tried putting:
envSetVal("viva.graph" "colorBank" 'string "red blue green yellow")
into my .cdsinit but nothing about the waveform colors changed. Note that I have other viva related stuff in my .cdsinit that does produce changes in viva so I know that all works. I also note that the documentation says "when you plot simulation results from ADE, the color bank from the display.drf file is honored" which strikes me as a bit confusing i.e. even though I define the colorBank in .cdsinit does the display.drf file still take precedence? I did take a look at the display.drf file and couldn't figure out where/what the color bank was.
Regarding the waveform thicknesses, I find "fine" to be too thin and "medium" to be too thick (and obviously all the rest of the options just waay too thick). Is there a way to hack a file (maybe display.drf) where I can monkey with line thicknesses?
Hello,
I am trying to copy the layout from one cell view to other without pin and label. I have came across the command to delete the pin. Is their any skill command to delete labels from layout too?
Hi,
I did set up my .cdsplotinit to call a script, which allows me to print into PDFs. That is working fine for single schematic pages.
When doing a hierarchical plot from a schematic, it seems like for every plot a new print job is spooled into the print queue.
For this case I do have the workaround, to concatenate all jobs (postscript files) into a single postscript-file and generate a PDF (which is overwritten during each job process) - by the last job I do have a PDF containing all schematic pages from the hierarchy. The problem is, that the concatenated postscript file still exists and I have to delete it manually to get an empty file for the next print-job.
Question 1: Is it possible to have only one print job started for a hierarchical plot, containing all schematic pages ?
This would simplify the plotting and I would not need to concatenate print jobs.
Question 2: Is it possible to get a trigger/function after the last print job that allows me to delete the concatenated postscript-file ?
I can see the log in the CIW window that says "INFO (SCH-1272): Successfully plotted 3 cellviews" and I would make use of this function if possible.
I don't know how to inform Cadence about an error in their documentation, so I just thought I'd mention that there is a typo in the Virtuoso Visualization and Analysis XL SKILL Reference manual (Product Version IC6.1.7 Feburary 2017) regarding the awvPlotWaveform function on page 129.
It describes details for the argument "?showSymbol" but it should really be "?showSymbols"
Hi! I would like to change the color of the text used for the simulation annotations in the schematic views. For instance, the color of the voltages printed at each node when using "Results->Annotate->DC node voltages".
In the symbol view these labels are specified as "IILLabel" objects, with value "cdsTerm("pin_name")". Where can I change the color used to display these strings? Is it a layer in my display.drf? (if so, how can I determine which? my PDK has thousands of layers!). Or is it maybe a setting in ~/.Xdefaults?
Thanks and regards, Jorge.
Hi Team
Can I please confirm something about simulation and Calibre PEX?
I am currently running a transient simulation and a small part of my circuit is using calibre PEX for post layout simulation purpose.
During that simulation, which takes a long time, I created a layout2 view for that small part of the circuit and changed some drawings and PEXed again. Then I realized that the PEX generated Calibre view is updated, even though the simulator is using the previous PEX result.
I believe this update of Calibre pex view will not interfere with the current simulation in running, because the data from the previous PEX is used when I click "netlist and run" and the simulator stop using the PEX result once the simulation actually started,
Can you please confirm?
Thanks
Allen
Hi everyone,
I'm now running an analog-to-digital converter project, all the circuits are ideal (vcvs built op-amp, switch, quantizers...) to verify the function.
So if I run MC mismatch using ADE-XL for this ideal circuits, every time I got the same waveforms and final Quantizer output code, this is what I expected.
The problem occurs when I started to put some real circuits in the schematic, for example just an inverter in standard library (the inverter is not connected to anything, just put in the schematic), and now for MC
situation the results are different, some nodes' waveforms now start to have slightly difference, but I don't know where those mismatches come from.
Hello,
I am doing the monte-carlo simulation using ocean Script. Whenever, we run the monte-carlo simulation and when it gets completed results of that run got stored in a separate folder with name "adexl" in the current working directory of the same cell view folder.
Is their any command in ocean that allows to access the result stored, so that further analysis can be done on that without performing the monte carlo simulation again??
In VIVA, when you expand a bus it yields (top to bottom) bus<LSB> ... bus<MSB>. I need the order reversed i.e. MSB on top. Is this possible?
And so we come to the part where you chime in and say "no, you lose".
Hi All,
Is there any way to add or configure the column in Monte Carlo result page, e.g., Summary & Yield ? I would like to add the column which calculates the Std Dev/Mean .
BR
Hello,
Is there a way to delete multiple instances of a cell automatically in Layout L/XL? (like select * -> Delete)
I am manually deleting PC end cells from my layout from the right and left edges and have a huge design with multiple (>7) power domains. So, it is becoming very difficult.
Please provide me a quicker (smarter) technique for this.
Thanks,
Kashif
Hi
I have a circuit with differential output and would like to run differential load pull. How should I set up the simulation? In single ended mode, I use the PSS simulation and a PortAdapter block, finally plotting the constant power contours (selecting the input terminal to the PortAdapter itself). I am not sure how to properly set-up the simulation for the differential. Can you help, please?
Many thanks.
Luca
Hi all!
I have a question about termOreder when we are using ansCdlSubcktCall netlist procedure.
I have device ptap1.
There is auCdl section of this device:
cdfId->simInfo->auCdl = '( nil
netlistProcedure ansCdlSubcktCall
instParameters (r w l)
propMapping (nil r R)
namePrefix "X"
modelName "ptap1"
termOrder (TIE SUB)
componentName "resistor"
)
and I've created test schematic (see attached picture)
And when I export this schematic in cdl, we have following netlist:
.SUBCKT ptap1_a a1
*.PININFO a1:I
XR0 net02 a1 / ptap1 r=262.847 w=780.00n l=780.00n
MN0 net4 net6 net5 net02 sg13_lv_nmos m=1 w=150.00n l=130.00n ng=1
.ENDS
.SUBCKT ptap1 TIE SUB
.ENDS
The question is, that in auCdl section we have termOredet TIE, SUB (and auCdl view also) and I expected, that in my netlist I'll see XR0 a1 net02, but in real netlist I see: net02, a1.
If I will use ansCdlCompPrim netlist procedure - all works fine:
.SUBCKT ptap1_a a1
*.PININFO a1:I
RR0 a1 net02 $[ptap1]
MN0 net4 net6 net5 net02 sg13_lv_nmos m=1 w=150.00n l=130.00n ng=1
.ENDS
.SUBCKT ptap1 TIE SUB
.ENDS
Could you pls. tell me, how to resolve this issue?
Hi. all.
I'd like to plot frequency characteristic of capacitance in spectre.
In AC simulation, I get current of capacitor (Ic) and its imaginary part ( Im(Ic) ).
Now tring to calculate Cap = Im(Ic) / omega (omega=2*pi*freq)
I can't find how to devide Im(Ic) by freq or omega.
I find 'xval' can plot simulation frequency, but dividing doesn't work.
Could you help, please?
Best Regards
Kawai
The oceaneval capability in spectre seems to be very poorly documented. It is mentioned indirectly in a couple places in the manual, but that's all I can find.
I would like to know if it is possible to declare and make use of intermediate variables when constructing these expressions, like one can do in OCEAN. For example,
vclip = clip(vtime('tran "VOUT") 3n 4n)
vmax = ymax(vclip)
Is there a way to do something like this with oceaneval? Thanks in advance!
Hi,
All is in the title. I can not run Quantus QRC because it greyed in QRC menu and don't know why.
Someone can help?
Best regards,
wighou