Electronics-Related.com
Forums

Altium Designer Pricing

Started by Tim Wescott April 2, 2014
On 4/3/2014 7:30 PM, Gerhard Hoffmann wrote:
> > I can easily move the VM from my workstation to the laptop and even > _running_ simulations can continue there. At the moment I'm in a > hotel room in Munich and could do productive work, if it wasn't for > s.e.d :-)
Yeah, that durn s.e.d!
> Altium said they would not support me running it on a virtual machine > (had some performance problems with design files on network-connected > partitions) but me running it on a VM was OK. I asked b4 I bought it. > Methinks they put a lot of trust on me, I could abuse it easily.
I guess I'm new to this. What is the difference between a "virtual machine" and a "VM"? I thought a VM *was* a virtual machine??? -- Rick
On 4/4/2014 12:33 AM, Tim Williams wrote:
> "Chris Jones" <lugnut808@spam.yahoo.com> wrote in message > news:533e17a9$0$16068$862e30e2@ngroups.net... >> There was some incompatibility between one of their licence server >> things and Windows 7 iirc, and I had stopped using XP by then. > > I've used both on Win7 without much trouble. Does seem to be written for > XP; the only 'gotcha' I found was, a fresh install of Win7 has desktop > scale set to 125%, which is simply ignored by the layouts of many > un-resizable dialogs, making buttons and such inaccessible. > >> Yes I eventually got it sorted out, but it seemed very rude of Altium to >> put me through that (using up several hours of otherwise billable time), >> when I know that some people use the pirated DVD and experience none of >> that crap. If I were Altium then I would always want people who pay for >> my product to get a better service than those who use it illegally. > > Free software may yet rule the day, but it has a long way to go. Vote > Piraten Partei? :^)
So what is wrong with KiCAD? -- Rick
On 4/3/2014 1:14 PM, sms wrote:
> On 4/2/2014 4:51 PM, Tim Wescott wrote: >> So, I've been looking over a few shoulders lately and seeing Altium on >> screens, and thinking maybe I should move away from EagleCAD. >> >> I looked on the Altium web site for pricing, and they're one of those >> "hand us one of your testicles and we'll tell you how much you need to >> pay to buy our product" sort of companies. >> >> I just want a budgetary price for buying one seat in the US. >> >> Anyone use it? Have a recent version? Know how much it costs? >> >> I think I may stick with Eagle. > > It's very difficult to get a straight answer out of Altium. > > Stick with Eagle. We are transitioning from KiCad to Eagle. KiCad is > okay but there are some PCB things that are difficult or impossible to > do with KiCad that are easy to do with Eagle. Try doing a PCB inductor > with a spiral coil on KiCad!
Why is that so hard? What else is difficult with KiCAD? -- Rick
rickman <gnuarm@gmail.com> writes:

> On 4/3/2014 1:14 PM, sms wrote: >> On 4/2/2014 4:51 PM, Tim Wescott wrote: >>> So, I've been looking over a few shoulders lately and seeing Altium on >>> screens, and thinking maybe I should move away from EagleCAD. >>> >>> I looked on the Altium web site for pricing, and they're one of those >>> "hand us one of your testicles and we'll tell you how much you need to >>> pay to buy our product" sort of companies. >>> >>> I just want a budgetary price for buying one seat in the US. >>> >>> Anyone use it? Have a recent version? Know how much it costs? >>> >>> I think I may stick with Eagle. >> >> It's very difficult to get a straight answer out of Altium. >> >> Stick with Eagle. We are transitioning from KiCad to Eagle. KiCad is >> okay but there are some PCB things that are difficult or impossible to >> do with KiCad that are easy to do with Eagle. Try doing a PCB inductor >> with a spiral coil on KiCad! > > Why is that so hard? What else is difficult with KiCAD?
I found a few things, some niggles but some more serious (but it may be just that I did not find the best way). Some examples: * It was too hard to create a footprint that was exactly what I wanted. You could get close quickly enough, but getting things exactly right seemed difficult. I was spending hours getting awkard irregular parts (like connectors) exactly right. * At the time the libraries were very fragmented and I did not like the way most of the footprints were done. For example silk screen outlines going over the pads. * I believe the design rule checking is incomplete, there are some constructs that are not checked properly. * vias are not "first class objects". It is hard to just sprinkle a thermal vias around a pour or a thermal tab say. They have to be routed as part of a net first and I think they can get removed or cause DRC errors. There is some workaround where they are created as individual "components". * 3D support is limited, there was no easy path to add a step file of the component to its definition and no easy way to export a step file of the finished PCB. (But now there is a nice script for export using other free tools). BUT some of the features are extremely powerful, checkout the new routing features. <http://www.youtube.com/watch?v=irqlrVUbjuQ>, <http://www.youtube.com/watch?v=CCG4daPvuVI>. It seems to be on a great path for the future. I have been following the developers list and all the changes made and proposed and the general direction look good. Altium development seems to be more focussed on cosmetics and subscription lock-in and extracting yet more money for paid add-ons. Kicad is cross-platform, 64-bit and written in screaming fast C++. AFAICT Altium is 32-bit only and I gather it consists of 15 million lines of object pascal dating back to the the 80's. Good luck rewriting that... There are important bugs going back years and for me it crashes several times a day. Sometimes I think they just add more and more stuff around a core that they no longer understand. Kicad Development has been boosted recently with major and ongoing contributions from CERN. <http://home.web.cern.ch/about/updates/2015/02/kicad-software-gets-cern-treatment> There is a major new release due real soon now (until recently you had to build from source to get anything not ancient). -- John Devereux
On 9/23/2015 12:26 PM, John Devereux wrote:
> rickman <gnuarm@gmail.com> writes: > >> On 4/3/2014 1:14 PM, sms wrote: >>> On 4/2/2014 4:51 PM, Tim Wescott wrote: >>>> So, I've been looking over a few shoulders lately and seeing Altium on >>>> screens, and thinking maybe I should move away from EagleCAD. >>>> >>>> I looked on the Altium web site for pricing, and they're one of those >>>> "hand us one of your testicles and we'll tell you how much you need to >>>> pay to buy our product" sort of companies. >>>> >>>> I just want a budgetary price for buying one seat in the US. >>>> >>>> Anyone use it? Have a recent version? Know how much it costs? >>>> >>>> I think I may stick with Eagle. >>> >>> It's very difficult to get a straight answer out of Altium. >>> >>> Stick with Eagle. We are transitioning from KiCad to Eagle. KiCad is >>> okay but there are some PCB things that are difficult or impossible to >>> do with KiCad that are easy to do with Eagle. Try doing a PCB inductor >>> with a spiral coil on KiCad! >> >> Why is that so hard? What else is difficult with KiCAD? > > I found a few things, some niggles but some more serious (but it may be > just that I did not find the best way). Some examples: > > * It was too hard to create a footprint that was exactly what I > wanted. You could get close quickly enough, but getting things > exactly right seemed difficult. I was spending hours getting awkard > irregular parts (like connectors) exactly right. > > * At the time the libraries were very fragmented and I did not like > the way most of the footprints were done. For example silk screen > outlines going over the pads. > > * I believe the design rule checking is incomplete, there are some > constructs that are not checked properly. > > * vias are not "first class objects". It is hard to just sprinkle a > thermal vias around a pour or a thermal tab say. They have to be > routed as part of a net first and I think they can get removed or > cause DRC errors. There is some workaround where they are created as > individual "components". > > * 3D support is limited, there was no easy path to add a step file of > the component to its definition and no easy way to export a step > file of the finished PCB. (But now there is a nice script for export > using other free tools). > > BUT some of the features are extremely powerful, checkout the new > routing features. <http://www.youtube.com/watch?v=irqlrVUbjuQ>, > <http://www.youtube.com/watch?v=CCG4daPvuVI>. > > It seems to be on a great path for the future. I have been following the > developers list and all the changes made and proposed and the general > direction look good. Altium development seems to be more focussed on > cosmetics and subscription lock-in and extracting yet more money for > paid add-ons. > > Kicad is cross-platform, 64-bit and written in screaming fast > C++. AFAICT Altium is 32-bit only and I gather it consists of 15 million > lines of object pascal dating back to the the 80's. Good luck rewriting > that... There are important bugs going back years and for me it crashes > several times a day. Sometimes I think they just add more and more stuff > around a core that they no longer understand. > > Kicad Development has been boosted recently with major and ongoing > contributions from CERN. > > <http://home.web.cern.ch/about/updates/2015/02/kicad-software-gets-cern-treatment> > > There is a major new release due real soon now (until recently you had > to build from source to get anything not ancient).
I'm on the KiCAD user mailing list as I am expecting to convert to it for my next project. They are all talking about the new release which seems to be out, but I'm not sure it is the actual "stable" release or a trial version before the official release. I read about a *lot* of issues people have with the product, but they all seem to be little things with easy fixes/work arounds. Some of the posts here seem to say KiCAD is not ready for prime time, but no one has given any real details that can be evaluated, or the problems are rather old and seem to have been fixed. As you say, CERN is pushing the software and making major contributions to the development. So today's KiCAD is likely not much like the KiCAD of a few years ago. -- Rick
rickman <gnuarm@gmail.com> writes:

> On 9/23/2015 12:26 PM, John Devereux wrote: >> rickman <gnuarm@gmail.com> writes: >> >>> On 4/3/2014 1:14 PM, sms wrote: >>>> On 4/2/2014 4:51 PM, Tim Wescott wrote: >>>>> So, I've been looking over a few shoulders lately and seeing Altium on >>>>> screens, and thinking maybe I should move away from EagleCAD. >>>>> >>>>> I looked on the Altium web site for pricing, and they're one of those >>>>> "hand us one of your testicles and we'll tell you how much you need to >>>>> pay to buy our product" sort of companies. >>>>> >>>>> I just want a budgetary price for buying one seat in the US. >>>>> >>>>> Anyone use it? Have a recent version? Know how much it costs? >>>>> >>>>> I think I may stick with Eagle. >>>> >>>> It's very difficult to get a straight answer out of Altium. >>>> >>>> Stick with Eagle. We are transitioning from KiCad to Eagle. KiCad is >>>> okay but there are some PCB things that are difficult or impossible to >>>> do with KiCad that are easy to do with Eagle. Try doing a PCB inductor >>>> with a spiral coil on KiCad! >>> >>> Why is that so hard? What else is difficult with KiCAD? >> >> I found a few things, some niggles but some more serious (but it may be >> just that I did not find the best way). Some examples: >> >> * It was too hard to create a footprint that was exactly what I >> wanted. You could get close quickly enough, but getting things >> exactly right seemed difficult. I was spending hours getting awkard >> irregular parts (like connectors) exactly right. >> >> * At the time the libraries were very fragmented and I did not like >> the way most of the footprints were done. For example silk screen >> outlines going over the pads. >> >> * I believe the design rule checking is incomplete, there are some >> constructs that are not checked properly. >> >> * vias are not "first class objects". It is hard to just sprinkle a >> thermal vias around a pour or a thermal tab say. They have to be >> routed as part of a net first and I think they can get removed or >> cause DRC errors. There is some workaround where they are created as >> individual "components". >> >> * 3D support is limited, there was no easy path to add a step file of >> the component to its definition and no easy way to export a step >> file of the finished PCB. (But now there is a nice script for export >> using other free tools). >> >> BUT some of the features are extremely powerful, checkout the new >> routing features. <http://www.youtube.com/watch?v=irqlrVUbjuQ>, >> <http://www.youtube.com/watch?v=CCG4daPvuVI>. >> >> It seems to be on a great path for the future. I have been following the >> developers list and all the changes made and proposed and the general >> direction look good. Altium development seems to be more focussed on >> cosmetics and subscription lock-in and extracting yet more money for >> paid add-ons. >> >> Kicad is cross-platform, 64-bit and written in screaming fast >> C++. AFAICT Altium is 32-bit only and I gather it consists of 15 million >> lines of object pascal dating back to the the 80's. Good luck rewriting >> that... There are important bugs going back years and for me it crashes >> several times a day. Sometimes I think they just add more and more stuff >> around a core that they no longer understand. >> >> Kicad Development has been boosted recently with major and ongoing >> contributions from CERN. >> >> <http://home.web.cern.ch/about/updates/2015/02/kicad-software-gets-cern-treatment> >> >> There is a major new release due real soon now (until recently you had >> to build from source to get anything not ancient). > > I'm on the KiCAD user mailing list as I am expecting to convert to it > for my next project. They are all talking about the new release which > seems to be out, but I'm not sure it is the actual "stable" release or > a trial version before the official release.
It is RC1, a "release candidate". I.e. it is a trial version. It is still probably the best version to install rather than the actual previous "released" version which is a few years old and missing a lot of the good stuff. I would expect it to be generally compatible with the to-be-released version when it appears.
> I read about a *lot* of issues people have with the product, but they > all seem to be little things with easy fixes/work arounds. Some of > the posts here seem to say KiCAD is not ready for prime time, but no > one has given any real details that can be evaluated
I did my best but it was last Christmas and my memory is not great.
>, or the problems are rather old and seem to have been fixed. As you >say, CERN is pushing the software and making major contributions to the >development. So today's KiCAD is likely not much like the KiCAD of a >few years ago.
I was evaluating the bleeding edge version as of ~9 months ago so it is close but should be more polished now. It is pretty good as-is. The best open-source solution and better than many commercial offerings including what I was using. There definitely are people using it professionally for non-trivial projects and it is only going to improve with time. -- John Devereux
Robert Loos wrote:

> Am 23.09.2015 um 17:06 schrieb rickman: > >> Why is that so hard? What else is difficult with KiCAD? > > You cannot draw polygons in copper or use them as a footprint. (not even > circles or ellipsoids, I believe).
Wow, that sounds serious. I've done a little poking around with Kicad, and did generate the gerber files for a project that another guy designed in it. How do you do SMT pads, then? Thanks, Jon
On Thu, 24 Sep 2015 09:44:50 +0200, Robert Loos wrote:

> Am 23.09.2015 um 17:06 schrieb rickman: > >> Why is that so hard? What else is difficult with KiCAD? > > You cannot draw polygons in copper or use them as a footprint. (not even > circles or ellipsoids, I believe).
I've generated footprints with circular pads. Mostly by using the python script available here: http://dlharmon.com/geda/footgen.html
> [snip] > I could not figure out how to draw complex board outlines (with slots, > segments of a circle etc) and how to draw exact lines (like 47.7mm to > the left, arc 45&deg;, r=5, 10.2mm up, ...). There seems no way to make an > arc part of a polygon at all.
There's a crude way of doing it, setting the grid and making careful use of the mouse. I completely agree that that's a crappy workaround - the user interface could (and should) enable exact placement of board outlines.
> You cannot trust the library. In my first test, the 7493 had no ground > contact due to an error in the schematic library.
I've never seen a part library yet that could be "trusted", especially when it comes to ground and power connections.
> Kicad has impressive features (length tuning, differential pairs, the > calculator is great!) but also big deficits. Too many to use it for > professional work. > > Also, the installation process for me was lengthy. There are no > instructions. I don't know where the libraries are (footprints seem to > be on the Internet only). The schematic libraries are in the program > directory! No idea how to move them to a network share. I nearly gave up > before I had a running version. > Under 'Settings', I find only three paths to modify. Two of them point > to directories that do not exist on my PC. > Would it have been so much effort to make a clean installer that really > works?
I look forward to improved installation procedures as well. I haven't used their schematic editor, only PCBnew and associated footprint handling sfwr. We tried Altium. Struggled with the license manager issues, which despite their tech support were a serious pain. Struggled with importing old Protel designs. In the end (of the evaluation period) we were unconvinced that it was worth the pain, suffering, huge learning curve, and unknown future uncertainties (i.e. subscription model). Of course, none of us do layout as our primary activity.
Den torsdag den 24. september 2015 kl. 20.48.48 UTC+2 skrev Jon Elson:
> Robert Loos wrote: > > > Am 23.09.2015 um 17:06 schrieb rickman: > > > >> Why is that so hard? What else is difficult with KiCAD? > > > > You cannot draw polygons in copper or use them as a footprint. (not even > > circles or ellipsoids, I believe). > Wow, that sounds serious. I've done a little poking around with Kicad, and > did generate the gerber files for a project that another guy designed in it. > > How do you do SMT pads, then? >
you can't draw pads free hand, but you can place pads that are rectangular,trapezoidal,round or oval of what ever size you like and you can make multiple pads with the same number to get them connected so it is possible to piece together almost any shape -Lasse
Lasse Langwadt Christensen wrote:

> Den torsdag den 24. september 2015 kl. 20.48.48 UTC+2 skrev Jon Elson: >> Robert Loos wrote: >> >> > Am 23.09.2015 um 17:06 schrieb rickman: >> > >> >> Why is that so hard? What else is difficult with KiCAD? >> > >> > You cannot draw polygons in copper or use them as a footprint. (not >> > even circles or ellipsoids, I believe). >> Wow, that sounds serious. I've done a little poking around with Kicad, >> and did generate the gerber files for a project that another guy designed >> in it. >> >> How do you do SMT pads, then? >> > > you can't draw pads free hand, but you can place pads that are > rectangular,trapezoidal,round or oval of what ever size you like and you > can make multiple pads with the same number to get them connected so it is > possible to piece together almost any shape > > -Lasse
Sometimes it is even easier to use a text editor and directly edit the KiCAD files. One big advantage is that this files are pure ASCII text and their format is documented and clear enough by itself without consulting the documentation. I have done such things with old OrCAD which had an text export/import function but with KiCAD it is even simpler. Some things are often done more effective using a text editor than using all that pointy-clicky GUI stuff. Real engineers don't need no f..g GUI. :-) -- Reinhardt