180nm Calibre PEX error，especially Calview setup error

[VictorWu]

Hello everyone,
I'm using Charter_018BCD PDK. I made a DAC and have performed DRC, LVS successfully. But when I performed PEX whose output is CalView, I ran into some problems. The error show as follows:
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v nmoscap_1p8 ivpcell). It will remain unconnected."
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v pmoscap_1p8 ivpcell). It will remain unconnected."
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v nmoscap_6p0 ivpcell). It will remain unconnected."
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v nmoscap_6p0_dw ivpcell). It will remain unconnected."
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v pmoscap_6p0 ivpcell). It will remain unconnected."
"error: Could not find pin mapping for terminal "SUB" of cell (chart018ull_hv30v pmoscap_6p0_dw ivpcell). It will remain unconnected."

I don't understand what these errors mean, and what should I do to eliminate these errors.In my circut, there is no pin "SUB". I will attach some screenshots of errors and warnings.

Dose anyone kown how to solve these problems?
Thanks!

 

[timof]

There may be different root causes.

The most likely one, is the cell/device terminal pin names mismatch between Calibre and Virtuoso.

For example, one may have the pins named PLUS and MINUS, and the other - POS and NEG (or ANODE and CATHODE).

Such mapping is established by a file called calview.cellmap.

Read this:

**broken link removed**

--- Updated Feb 2, 2021 ---

I suspect that calview.cellmap methodology was developed way after 180nm technology was developed, so some old PDKs may not have this file available, you have to create it by yourself, "manually".

 

[VictorWu]

There may be different root causes.

The most likely one, is the cell/device terminal pin names mismatch between Calibre and Virtuoso.

For example, one may have the pins named PLUS and MINUS, and the other - POS and NEG (or ANODE and CATHODE).

Such mapping is established by a file called calview.cellmap.

Read this:

**broken link removed**

--- Updated Feb 2, 2021 ---

I suspect that calview.cellmap methodology was developed way after 180nm technology was developed, so some old PDKs may not have this file available, you have to create it by yourself, "manually".

There may be different root causes.

The most likely one, is the cell/device terminal pin names mismatch between Calibre and Virtuoso.

For example, one may have the pins named PLUS and MINUS, and the other - POS and NEG (or ANODE and CATHODE).

Such mapping is established by a file called calview.cellmap.

Read this:

**broken link removed**

--- Updated Feb 2, 2021 ---

I suspect that calview.cellmap methodology was developed way after 180nm technology was developed, so some old PDKs may not have this file available, you have to create it by yourself, "manually".

Thanks for your replay. I don't have calview.cellmap but I have a file named calview.xxxx.map, I think they are the same. I will attach a screenshot of calview.xxxx.map behlow. As you can see, for nmoscap_1p8, (nil SUB "dummyGlobalNet!"), but for nmoscap_1p8_dnwell, (sub SUB), I think this is the reasom why nmoscap_1p8 have errors but nmoscap_1p8_dnwell don't. What should I do to fix this, hope you can give me some advice.

 

[timof]

Yes, this is the format of calview.cellmap file (the file name may be different, of course).

I would suggest the following steps, to do in parallel:

1. Ask the foundry to provide an up to date file (and explaining the error you see).
2. Ask your Cadence support engineer for help.
3. Ask your Mentor support engineer for help.
4. While you are waiting for a response from actions 1-3 above (you may never hear back from them - depending on the status of your company with these vendors), do some experiments. This may be risky, as you may get the error go away, but internally, there may be a problem (with accuracy, or with correctness of the device connectivity, for example).
Please read the link i provided earlier - it offers a script for automated file generation, and points to the specific section in specific document, in Mentor documentation, for a detailed format description and explanation.
I am not an expert in this syntax, so can't help you right away.

--- Updated Feb 2, 2021 ---

Another option - is to generate a DSPF file, and use it for SPICE simulation.
There will be no such errors, as DSPF file will be using Calibre-generated names.