Just to be clear, you are saying that you didn't bother to read how to use the tool and it didn't work for you?
If that isn't correct, please distill down which step did not work and use gitlab. Please don't abuse this list. We realize that gitlab requires authentication. This is on purpose. Please report one problem per issue report on gitlab. Not an essay. While well intentioned, it is not useful. If the first statement is correct, please read the documentation for for the tool is supposed to work at docs.kicad.org. Seth On Sun, Feb 9, 2025, 2:44 AM Rafał Pietrak <[email protected]> wrote: > Hi everybody, > > This is my MMA (Monkey Mode Attestation - was "verification", but MMA > sounds better :) of channels tool in KiCAD v9. My apologize for abusing > the list with this report, but gitlab refused my registration. > > So: > 1. MMA means, that user did not read the manual but choose to > press/click "at random" to get expected results. Pls take into account, > that although I tried to write down all my "klicks", occasionally I > might have failed to remember to do so. Particularly in cases leading to > nothing. > 2. First I've pressed tools->MultiChannel->RepearLayout ... I was > prompted to select the RuleArea to "copy from". > 2.1 I clicked to "something" ... nothing happened. May be some sort of > alerts panel popping up would be in order here? > 2.2 Surely the Place menu has "Place->Rule Area", so I added two rule > areas on PCB: one covering my "original sheet to copy from" selecting > one subsheet in "placement", the other "somewhere else" selecting the > other sheet as it's "placement". > 2.3 now RepeatLayout resulted in components rearrangement ... but > components of "slave" sheet-nr2 weren't moved into the other "rule > area". They stayed more or less where most of them already were - they > only moved slightly to reflect "master" arrangement. This is strange for > an MMA tester - why bother to "draw" the other rule-rectangle ... just > listing the "slave" parts of the schematic when the initial "rule area" > is drawn should be sufficient. Then again, why draw anything at all - > the "master" channel is already correctly laid drawn and the "binding to > slaves" is based on linked "schematic sub-sheets or block", drawing > anything seem superficial. > 3. On one occasion I did clicked the design into a state, where the > process of copying worked "the other way around", meaning that my > intended source was used by KiCAD as target for placement and vice > verse. I couldn't find any way to revert that, nor any means of > examination (E) of current state of "repeat layout" tool. > 3.1. Whenever I try to examine (E) a selected "rules area", there is no > mention which one is "the source" and which one is "the target". If > there is a way to check it, for MMA tester it's not so easy to guess how. > 4. after drawing "rule area" on schematic diagram and then doing "update > PCB", the report of "update process" does not contain anything (as if > nothing was exported). So, how does user know if those areas can be used > by PCB or not? > 5. examining (E command) a selected rule area on schematic does not let > putting a name to it. It should be under user control which area's are > the same and which are different and their name is a natural way to > allow for that. > > My project is in the attachment together with excerpts from "Design of > CDC6600" book as found at > > https://cseweb.ucsd.edu/classes/sp10/cse240c/Papers/DesignOfAComputer_CDC6600.pdf. > > My KiCAD files attached are based on register design found on page.26 of > that book. > > The other excerpt (pages 108-109) I'm including to suggest an > alternative implementation of "sub-sheets" in KiCAD, which may result in > some optimization (simplification) of multichannel implementation. The > "trick" is to allow a symbol of "array of blocks" on the schematic > diagram and have an assumed scheme for identification of a particular > block in such array. The problem here will arise from the fact, that > every such block may have a BUS of signals entering/exiting it. A bus > is a list (one dimensional array). An array of busses is a two > dimensional array. I'm not sure if KiCAD wire identification schemes can > cope with that now. > > I'm bringing the CC6600 drawing as an example because "sub-sheet" on > pages 108/109 of that book are presented as "arrays of identical blocks" > (graphically). Identical to the extend, that anybody doing engineer work > on the design would have much easier job when (say) transistor Q17 in > EVERY block/channel had the same Q17 designation. At the level of > schematic drawings this translated to the fact, that there will be just > ONE sheet (or block) to represent the content of block-nr-0. All other > blocks are the same and will NOT have designated schematics. > > Currently, KiCAD annotates all components on identical subsheets with > different labels (see my files: R1 on sheel1, R7 on sheet2). This has > it's merits, but does not work for the above "array of blocks" system. > One would need to have a "block repetition" explicit by some way of > components getting a "block identification" attached. One such way could > turn Q17 transistor label on "first block" into Q17a (or Q17.0, or > something), and into Q17b on second block, etc. > > With such "channel representation" as shown on CDC6600 design schematics > it should be easier for Repeat Layout tool to identify which components > belong to the same channel and for user to know what exactly "KiCAD was > thinking". User would just need to indicate a single component that the > process of copying the PCB layout should be repeated for every > channel/block. The "repeat engine" would then copy whatever is indicated > based on current location of that particular component in every channel > and would not get confused since the component would have exactly the > same annotation (like Q17). No selection of "rule area" on PCB would > then be required. > > I hope this tests and observations help. > > Still, if you already knew about MMA confusion with current > RepeatLayouts tool or had availability/plans-for "arrays of blocks" in > KiCAD, I sincerely apologize for the noise. > > with best regards > > -R > > -- > You received this message because you are subscribed to the Google Groups > "KiCad Developers" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion visit > https://groups.google.com/a/kicad.org/d/msgid/devlist/0b01257c-7c08-4c52-9c61-50774899e0c4%40electric-sheep.eu > . > -- You received this message because you are subscribed to the Google Groups "KiCad Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/a/kicad.org/d/msgid/devlist/CAFdeG-oaZd0VwGqXeJtrsdGYWZ%2B0C25-EDEsC82NeqT7v_SgEw%40mail.gmail.com.
