I have the same issue as Carl here: <https://skywa...
# xycet
Ted
03/31/2022, 7:11 PMI have the same issue as Carl here: https://skywater-pdk.slack.com/archives/C01TLV579C5/p1643427056467499
Xyce as a problem with the original hspice:
Xpch_inv enable_bar i_enable_hv vdd_hv vdd_hv sky130_fd_pr__pfet_g5v0d10v5 w = 50 l = 1 nf = 4 ad = ' int((nf + 1)/2) * W/nf * 0.29 ' as = ' int((nf + 2)/2) * W/nf * 0.29 ' pd = ' 2 * int((nf + 1)/2) * (W/nf + 0.29) ' ps = ' 2 * int((nf + 2)/2) * (W/nf + 0.29) ' nrd = ' 0.29 / W ' nrs = ' 0.29 / W ' sa = 0 sb = 0 sd = 0
And the version that came out of Xdm:
Xpch_inv enable_bar i_enable_hv vdd_hv vdd_hv sky130_fd_pr__pfet_g5v0d10v5
+ PARAMS: W=50 L=1 NF=4 AD=' int((nf + 1)/2) * W/nf * 0.29 ' AS=' int((nf + 2)/2) * W/nf * 0.29 '
+ PD=' 2 * int((nf + 1)/2) * (W/nf + 0.29) ' PS=' 2 * int((nf + 2)/2) * (W/nf + 0.29) '
+ NRD=' 0.29 / W ' NRS=' 0.29 / W ' SA=0 SB=0 SD=0
'-hspice-ext all' did not help. Is xyce not able to use device parameters in expressions the way hspice can?
e
Eric Keiter
04/01/2022, 3:33 PM@User Thanks for the information. I would have predicted that Xyce could handle this, but possibly there is a bug I need to track down. I’m a bit swamped at the moment, but I’ll look at this as soon as I can.
Eric Keiter
04/04/2022, 11:20 PM@User I wanted to follow up on this. You are correct, Xyce doesn’t currently handle handling inline parameters in the subcircuit line. I have entered an issue in our bug tracker about this. In the meantime, there is a workaround.
t
Ted
04/04/2022, 11:22 PMI actually have some spice parsing and re-writing tools, and I've been able to modify the netlist before using Xyce.
e
Eric Keiter
04/04/2022, 11:24 PMGreat! I am guessing you are doing the thing I was about to recommend. Basically, if you add a .param statement inside the subcircuit it can handle this.
Eric Keiter
04/04/2022, 11:24 PMSo if you have: Xpch_inv enable_bar i_enable_hv vdd_hv vdd_hv sky130_fd_pr__pfet_g5v0d10v5
+ PARAMS: W=50 L=1 NF=4 AD=' int((nf + 1)/2) * W/nf * 0.29 ' AS=' int((nf + 2)/2) * W/nf * 0.29 '
Eric Keiter
04/04/2022, 11:24 PMchange it to:
Eric Keiter
04/04/2022, 11:24 PMXpch_inv enable_bar i_enable_hv vdd_hv vdd_hv sky130_fd_pr__pfet_g5v0d10v5
+ PARAMS: W=50 L=1 NF=4 AD=' int((nf + 1)/2) * W/nf * 0.29 ' AS=' int((nf + 2)/2) * W/nf * 0.29 '
.param AD=' int((nf + 1)/2) * W/nf * 0.29 '
.param AS=' int((nf + 2)/2) * W/nf * 0.29 '
Eric Keiter
04/04/2022, 11:25 PMThe .param statements inside the subcircuit will take care of it.
t
Ted
04/04/2022, 11:25 PMYeah, that makes sense
Ted
04/04/2022, 11:25 PMIs there a better place for reporting bugs and requesting features? Now that I've been using Xyce I've run into a few things that are annoying
e
Eric Keiter
04/04/2022, 11:26 PMBut, this shouldn’t be necessary, so I’ve entered an issue about it, like I mentioned. It should actually be easy to fix. All the expressions inside the subcircuit are able to find everything on the PARAMS: line. I just have to make it so that expressions on the PARAMS: line can refer to each other.
Eric Keiter
04/04/2022, 11:27 PMThere are a few places. You could enter an issue at our public github page: https://github.com/Xyce/Xyce/issues
t
Ted
04/04/2022, 11:27 PMok, I'll add them there
👍 1
e
Eric Keiter
04/04/2022, 11:27 PMOr you could ask about it in our google group: https://groups.google.com/g/xyce-users
Eric Keiter
04/04/2022, 11:28 PMThe github issue tracker would be fine. Thanks!
t
Tim Edwards
04/21/2022, 1:54 PM@User: I thought to look today, and I didn't see any issue posted on the github issue tracker, so I went ahead and created one. On a side note: would you like me to formally log an issue about the lack of level 3 diode model support?
e
Eric Keiter
04/26/2022, 5:18 PM@User thanks; I have entered the level=3 diode issue into our internal issue tracker but I don’t mind (unlike one of my colleagues, lol ) having a redundant one in the external github tracker.