| ID: | 4820 |
| From: | Bob Paddock <gr...@gmail.com> |
| Date: | Fri, 16 Apr 2021 08:57:15 -0400 |
| Subject: | Re: [pcb-rnd] feature removal RFC: the pscalib action |
| in-reply-to: | 4819 from rn...@igor2.repo.hu |
| replies: | 4821 from Bdale Garbee <bd...@gag.com> |
On Fri, Apr 16, 2021 at 7:37 AM <rnd2@igor2.repo.hu> wrote: > I did a lot more research and resolved this. Thanks to Bob for explaining > the history of this feature! Your welcome. >I think this is a fair tradeoff as you > need to calibrate your printer/process only once, or at least rarely, Ambient temperature and humidity play a part in this process, so absolute precision, when that level is required, will still need to calibrate at the time of use. The other complication is the heat from the printer itself would distort the process with transparencies. Today there are far better alternatives to the UV Sensitized boards to require this level of precision. > while you will probably print more often. So the rare side became a bit > more work and the heavily used part got easier. Exactly. > Another alternative is full manual calibration, using a static ps file. > I'm attaching a dump of the original calibration page generated by Bob's > code. I like it, there's only one thing I'd do differently: I'd do both > the horizontal and the vertical ruler in the center of the page instead of > left/bottom, because if they get too close to the margin they may be lost. Today I'd do both. They were in that location to detect where the printer really printed, vs where it said it printed. It also comes back to how things get distorted. The distortion is not uniform from the heat process of printing itself with toner based printers. This is the manal 'paper' equivalent of per-panel, per-board and per-part fiducial targets in the actual layout. At each level the distortion is different.
Reply subtree:
4820 Re: [pcb-rnd] feature removal RFC: the pscalib action from Bob Paddock <gr...@gmail.com>
4821 Re: [pcb-rnd] feature removal RFC: the pscalib action from Bdale Garbee <bd...@gag.com>