mikroBUS module dimensions

Peter JakackiPeter Jakacki Posts: 9,598
edited 2020-09-12 - 00:18:28 in General Discussion
I've read the spec and created footprints for the socket and modules but I'm a little confused, all due to one single dimension that is missing. The socket is 1125mils long and seems to extend back 150mils from the back of the pin header. That is clear in the drawings but the plug-in module is the same length but seems to extend 100mils back yet there is no dimensions to indicate this, only the chamfer indicates this.

So according to what I can interpret, the module outline would extend up and out of the open end of the socket outline by 50mil? Sure, this might seem a minor point but I know too well the importance of accurate footprints. Just wanted to confirm all this before I hit the make button.

I haven't had another look at this since late last night, so maybe I just need a cup of coffee.

b8adac3cae4e2047d131e33c9e4fec.png
bec0b9b1dfaf1ba07891cbe0c4a1a6.png

Comments

  • Looks right to me, each side has eight pins 0.100" pin spacing, the two rows are 0.900" wide
  • The outer footprint is 1.100"
  • Two units side by side have 0.250" pins spacing between units. Measured from my EasyPic7 board
  • That hieght is 1.150" not 1.125"
  • Like this
    1200 x 1600 - 725K
  • Yes in my experience, while the spec defines S/M/L boards a lot of them don't fit those rules, so you just have to regard it as kind of 'open to infinity' so boards can extend as far as they like

    The sideways pitch of sockets is 1150 mil, they add an extra 50 mil in between what you have shown above
  • Peter JakackiPeter Jakacki Posts: 9,598
    edited 2020-09-12 - 02:49:58
    DigitalBob wrote: »
    The outer footprint is 1.100"

    The socket outline is 1.1" but the module is 1" allowing for a gap.
  • Yes Peter. Looks to me the 1125 long pcb extends 0.050” past the screen outline (also 1125) on the socket. A bit weird!
  • It's better to have a little extra wiggle room, rather than have it too tight.
  • Peter JakackiPeter Jakacki Posts: 9,598
    edited 2020-09-13 - 00:34:40
    @DigitalBob - If you have done it accurately, then you have wiggle room available according to the spec. The main reason why I posted the spec for the socket and the module was that like so many mechanical drawings they are normally incomplete. In this instance the module had the outline correct and then the pin pitch (100) and row spacing (900), but it didn't have any dimensions for the relationship between the pins and the outlines unless you assume that the bottom right 100 vertical refers to the 50mil spacing from the pin. But if you do then your drawing will also look the same, and then applying the 1125 height results in the module sticking out 50mil over the top of the mismatching socket. Don't make any assumptions, just check it yourself as if you were a "robot". So the drawing itself is sloppy.

    btw - I did make one other assumption, that the 900 pitch rows are centered on the 1000 outline. While that is a safe assumption, you shouldn't have to assume.

    microbus%20module.png
    672 x 684 - 45K
  • Is there a measurement you want I'll take a picture and post it. Your pin layouts are correct
  • Imagine you are a robot, you have drawn the outline and also drawn the 100mil pitch 900mil wide connector but where do you place that connector within that outline using only the numbers that are given? If it were a plotter, it would be unplottable.

    I can look and take a guess but the result is the module sticks out 50mil from the top of the matching socket outline which seems strange. Get it?
    Btw, those outlines are the spec itself.
  • So, Peter, are you saying the drawing is not really correct as there is no dimension from any pin to outline in either x or y direction to place the pin array correctly within the outline?
  • Peter JakackiPeter Jakacki Posts: 9,598
    edited 2020-09-13 - 06:30:04
    So, Peter, are you saying the drawing is not really correct as there is no dimension from any pin to outline in either x or y direction to place the pin array correctly within the outline?

    That's right, it's a sloppy drawing and if it were a machine/pcb file it would fail. But the consequence of the guessing is that the top of the plug-in module protrudes 50mil above the top of the socket outline. You would think that the socket outline would contain the module inline completely. If this is the only source of information about the dimensions then no wonder there is dimensional disparity among manufacturers.

    EDIT: Here are the pcb footprint outlines I constructed from the drawings for socket and module. The socket is the broad tan outlines and the module footprint has been dropped over it with the thin brown outline. Notice the top of it protrudes the socket outline on the right side, which is the top. Of course I can live with it, that's not my problem. But if I had an smd component that had ambiguous dimensions on it I would try to use another brand that was clearer or at least have a sample I could measure, which isn't easy. Electrical "dimensions" or voltages are easy, they are always referenced to the common or ground. Mechanical dimensions can have multiple references (often missing) especially when it comes to the contacts part vs the body part.


    1056 x 922 - 34K
  • I don’t know why the electronics industry is so bad with dimensioning.

    Mechanical drawings I’ve seen from other industries usually have a singular reference point from where all dimensions emanate from.

    Connector dimensions often seem to have the pins all relative to each other, and the case dimensions again relative to each other, but what’s often missing is the relationship (measurement) between the two sections. And just to add to the confusion, measurements are often relative to another measurement, rather than back to a reference point. Plus, often there are missing critical measurements missing too, necessitating bringing out the calipers to try and measure them.
  • Yes, it is that bad. Sometimes I'd rather just have a accurate hires drawing with just one x and y dimension for calibration, and then I can measure it whichever way I want.
  • Perhaps they have a gerber file somewhere, either of the footprint only, or including the footprint ?
  • I'm sure I will be fine. It was more of a grumble because I come across this stuff all the time, but this was from a European company that also sets the standard for a module. How hard can it be to get their own drawings right?
  • I think you'll find these 'errors' may actually be essential ingredients in successful platforms
    - Arduino has that 0.160" offset between top headers
    - Mikrobus seem to have this 0.050" unaccounted for
    - pmod have pin 1 as the gpio pin, rather than GND

    So the question is, what will be the P2D2 quirk that makes it a runaway success ?

    I have one of the v8 development boards, they don't bother with the shorter outline, the height of the socket on the silk screen is around 2.4"
  • You can buy a Click proto board for $3. Whenever I design a PCB, I print it out and lay the parts out on it to check the clearance before ordering.
  • DigitalBob wrote: »
    You can buy a Click proto board for $3. Whenever I design a PCB, I print it out and lay the parts out on it to check the clearance before ordering.

    That's easy enough for simpler boards but still not accurate. Even if the printer was a real 1200dpi and I printed on polyester, you couldn't check the mating to the pads under some parts. I might use this method just to check the placement and "feel" of the main connectors, but that's about it as it is a poor substitution for accurate footprints and outlines.

    Have you never created footprints? Have you tried to follow the drawings? Have you tried to work out whether they are showing you the bottom or top view because it is such a mess? What is the recommended pcb pad size. This is sheer frustration for many components and having to check and create footprints for new components can take quite a while.

    The point of this whole thread was not about me being worried or that I couldn't get it right, as I have done many hundreds of tight pcb designs, it is about the discrepancies in mechanical drawings of components. In this case a European company which normally care about small details and because they are setting the standard, they should know and understand the details. Who there is checking what is published?

  • fwiw I've found Mikroe responsive on this kind of thing. its worth contacting them if you haven't already.

    Last month I sent a request for mechanical dimensions late Friday night, they responded within an hour that they'd have the info to me Monday, which they did
  • Tight PCB design is just time consuming, not really super technical, it just comes down to following the schematic and making the most out of your board space, good ground planes, noise, trace lenght, like the distance from a CPU to the I/O, distance equals nanoseconds.
  • Mikroe email support is Ok. Takes a day coming out of Belgium.
  • DigitalBob wrote: »
    Tight PCB design is just time consuming, not really super technical, it just comes down to following the schematic and making the most out of your board space, good ground planes, noise, trace lenght, like the distance from a CPU to the I/O, distance equals nanoseconds.

    I haven't seen any of your tight pcb designs that you have done, or any other ones either so I don't know where you are coming from with those generalities. They used to refer to pcb designs as "artwork" and if done well they could be a work of art, but to get there sure was 'art work :)


  • Most of my PCB's are propriety. But I have posted my Version of the prop flip etc. So if your click design is not propriety want don't you post it so some of the forum minds can assist with some input.
  • I'm ok, you just keep reading that I need assistance when I am just saying that the information is sloppy. I did post my footprints and they are correct as per the drawings. Did you get my point about the 50mil overhang being a little strange as a result????
    None of this is slowing me up at all, just sharing my thoughts.
  • All of your dimensions look accurate,, but on the 1125 height it looks like it could go 1150, if that 25 is not critical use 1150. So getting back to the top drawing the header is 800 then 175 from the header to the top outer and 100 for the chamfer. That's 1075, so that leaves 50 from the bottom of the chamfer to the bottom outline. So the 50 is the only dim. Not listed.
Sign In or Register to comment.