Homebrew_PCBs

Hello Harvey,

I'm interested in getting more information on your drilling setup,
mostly the optical stuff - camera, software, monitor. What supplies
the reticule image for the crosshairs?

Thanks.

Mark

On Fri, 15 Aug 2014 12:52:18 -0400, you wrote:

>Hello Harvey,
>
>I'm interested in getting more information on your drilling setup,
>mostly the optical stuff - camera, software, monitor. What supplies
>the reticule image for the crosshairs?

OK, in short. The optical stuff is a 12 volt/24vac camera with almost
any lens that works well. You want to move it out away from the CCD
so that you have the equivalent of a closeup extender.

That goes to a standard color monitor. This is an LCD monitor that I
found (runs off 12 volts DC), and has an overlay with a graticule on
it. I've been tempted to make it more complicated by adding a
crosshair generator, but that'd be another PC board and some video
work. The graticule is easier. Just make the pattern on a
transparency and put it over the monitor itself. Almost any old TV
would work.

The software is custom, and is a simple Z axis setup. There are two
pushbuttons (green/red) and a rotary encoder with a push switch
(Electronics goldmine sold them). There's an end travel (and top
travel) led sensor.

Boot up, the display says to push green to home. That homes the drill
and runs it up to the rest position. pushing the footswitch (or the
green pushbutton) starts a drill cycle, three speeds possible, up to
the drill top position. At that point the drill retreats back to the
rest position and turns off the Proxxon tool (solid state relay, works
better with the 12 volt model and transformer).

Pushing the red button puts the drill at the top position, which is
where I can change the bit from the top.

Pushing in the rotary switch/encoder put it into an adjust mode, where
I can set the rest and top positions (stored in the processor's
eprom).

Basically, the software is not all that complicated, but does rely on
an infrastructure of display drivers (it's a smart display that I
built), a modified keyboard driver package, and a few other bits and
pieces.

Physically, the camera is mounted on a PACE hot air station stand (I
managed to find two for not so much money), which gives me an
overarching arm for the camera. I can move it up and down to focus
the camera and move it out of the way for drill change. I suppose
another enhancement would be to put a stepper drive on that.

You will have to align the camera center point to the tip of the
drill. In my case, the adjustment is entirely mechanical, with the
camera swinging left/right on the mounting screw, but tilted (spring
used on screw) forwards/backwards as needed. That's a trifle awkward,
and the electronic crosshair generator would be better. That would
require a small CPLD, a sync separator, a synchronized pixel clock,
and some counters. Ideally, the graticule has ticks on both axis
lines, since that helps align the center of large holes.

The smallest hole I drill is 0.021 for a via, and I estimate that I
can repeat that hole to within 0.002 inches roughly. The size of that
hole on the monitor is the critical part.

You'll want to backlight the board. Red light may be most useful, but
I haven't experimented with that yet.

Hope that this helps.

Harvey

>
>Thanks.
>
>Mark

Thanks, Harvey, for the detailed description - Very interesting
setup. One of these days I'll work on something similar, though I
rarely drill a lot of holes. I use almost exclusively smt and usually
make single sided pcbs using zero ohm resistors as jumpers.

Mark


At 06:50 PM 8/15/2014, you wrote:

>On Fri, 15 Aug 2014 12:52:18 -0400, you wrote:
>
> >Hello Harvey,
> >
> >I'm interested in getting more information on your drilling setup,
> >mostly the optical stuff - camera, software, monitor. What supplies
> >the reticule image for the crosshairs?
>
>OK, in short. The optical stuff is a 12 volt/24vac camera with almost
>any lens that works well. You want to move it out away from the CCD
>so that you have the equivalent of a closeup extender.
>
>That goes to a standard color monitor. This is an LCD monitor that I
>found (runs off 12 volts DC), and has an overlay with a graticule on
>it. I've been tempted to make it more complicated by adding a
>crosshair generator, but that'd be another PC board and some video
>work. The graticule is easier. Just make the pattern on a
>transparency and put it over the monitor itself. Almost any old TV
>would work.
>
>The software is custom, and is a simple Z axis setup. There are two
>pushbuttons (green/red) and a rotary encoder with a push switch
>(Electronics goldmine sold them). There's an end travel (and top
>travel) led sensor.
>
>Boot up, the display says to push green to home. That homes the drill
>and runs it up to the rest position. pushing the footswitch (or the
>green pushbutton) starts a drill cycle, three speeds possible, up to
>the drill top position. At that point the drill retreats back to the
>rest position and turns off the Proxxon tool (solid state relay, works
>better with the 12 volt model and transformer).
>
>Pushing the red button puts the drill at the top position, which is
>where I can change the bit from the top.
>
>Pushing in the rotary switch/encoder put it into an adjust mode, where
>I can set the rest and top positions (stored in the processor's
>eprom).
>
>Basically, the software is not all that complicated, but does rely on
>an infrastructure of display drivers (it's a smart display that I
>built), a modified keyboard driver package, and a few other bits and
>pieces.
>
>Physically, the camera is mounted on a PACE hot air station stand (I
>managed to find two for not so much money), which gives me an
>overarching arm for the camera. I can move it up and down to focus
>the camera and move it out of the way for drill change. I suppose
>another enhancement would be to put a stepper drive on that.
>
>You will have to align the camera center point to the tip of the
>drill. In my case, the adjustment is entirely mechanical, with the
>camera swinging left/right on the mounting screw, but tilted (spring
>used on screw) forwards/backwards as needed. That's a trifle awkward,
>and the electronic crosshair generator would be better. That would
>require a small CPLD, a sync separator, a synchronized pixel clock,
>and some counters. Ideally, the graticule has ticks on both axis
>lines, since that helps align the center of large holes.
>
>The smallest hole I drill is 0.021 for a via, and I estimate that I
>can repeat that hole to within 0.002 inches roughly. The size of that
>hole on the monitor is the critical part.
>
>You'll want to backlight the board. Red light may be most useful, but
>I haven't experimented with that yet.
>
>Hope that this helps.
>
>Harvey
>
>
> >
> >Thanks.
> >
> >Mark
>
>
>
>------------------------------------
>Posted by: Harvey White <madyn@...>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBs
>------------------------------------
>
>Yahoo Groups Links
>
>
>

On Sat, 16 Aug 2014 08:51:49 -0400, you wrote:

>Thanks, Harvey, for the detailed description - Very interesting
>setup. One of these days I'll work on something similar, though I
>rarely drill a lot of holes. I use almost exclusively smt and usually
>make single sided pcbs using zero ohm resistors as jumpers.
>

I do double sided boards, with the boards made from two .021 boards
epoxied together (use 1 hour epoxy!). Alignment is very critical and
the drilling setup (finally!) makes that work a lot better. Since I'm
doing 64/100/144 pin TQFP chips, I can't live with the luxury of
exclusively resistive jumpers.

Latest project involves a graphics panel driver with an S1D13781
graphics chip (Epson, anyone wanting to use it please talk to me,
there are some non-obvious things you need to do to make it work). an
Xmega, a touch panel controller, -24 volt supply, I2C communications
(dual channel).

Harvey

>Mark
>
>
>At 06:50 PM 8/15/2014, you wrote:
>>On Fri, 15 Aug 2014 12:52:18 -0400, you wrote:
>>
>> >Hello Harvey,
>> >
>> >I'm interested in getting more information on your drilling setup,
>> >mostly the optical stuff - camera, software, monitor. What supplies
>> >the reticule image for the crosshairs?
>>
>>OK, in short. The optical stuff is a 12 volt/24vac camera with almost
>>any lens that works well. You want to move it out away from the CCD
>>so that you have the equivalent of a closeup extender.
>>
>>That goes to a standard color monitor. This is an LCD monitor that I
>>found (runs off 12 volts DC), and has an overlay with a graticule on
>>it. I've been tempted to make it more complicated by adding a
>>crosshair generator, but that'd be another PC board and some video
>>work. The graticule is easier. Just make the pattern on a
>>transparency and put it over the monitor itself. Almost any old TV
>>would work.
>>
>>The software is custom, and is a simple Z axis setup. There are two
>>pushbuttons (green/red) and a rotary encoder with a push switch
>>(Electronics goldmine sold them). There's an end travel (and top
>>travel) led sensor.
>>
>>Boot up, the display says to push green to home. That homes the drill
>>and runs it up to the rest position. pushing the footswitch (or the
>>green pushbutton) starts a drill cycle, three speeds possible, up to
>>the drill top position. At that point the drill retreats back to the
>>rest position and turns off the Proxxon tool (solid state relay, works
>>better with the 12 volt model and transformer).
>>
>>Pushing the red button puts the drill at the top position, which is
>>where I can change the bit from the top.
>>
>>Pushing in the rotary switch/encoder put it into an adjust mode, where
>>I can set the rest and top positions (stored in the processor's
>>eprom).
>>
>>Basically, the software is not all that complicated, but does rely on
>>an infrastructure of display drivers (it's a smart display that I
>>built), a modified keyboard driver package, and a few other bits and
>>pieces.
>>
>>Physically, the camera is mounted on a PACE hot air station stand (I
>>managed to find two for not so much money), which gives me an
>>overarching arm for the camera. I can move it up and down to focus
>>the camera and move it out of the way for drill change. I suppose
>>another enhancement would be to put a stepper drive on that.
>>
>>You will have to align the camera center point to the tip of the
>>drill. In my case, the adjustment is entirely mechanical, with the
>>camera swinging left/right on the mounting screw, but tilted (spring
>>used on screw) forwards/backwards as needed. That's a trifle awkward,
>>and the electronic crosshair generator would be better. That would
>>require a small CPLD, a sync separator, a synchronized pixel clock,
>>and some counters. Ideally, the graticule has ticks on both axis
>>lines, since that helps align the center of large holes.
>>
>>The smallest hole I drill is 0.021 for a via, and I estimate that I
>>can repeat that hole to within 0.002 inches roughly. The size of that
>>hole on the monitor is the critical part.
>>
>>You'll want to backlight the board. Red light may be most useful, but
>>I haven't experimented with that yet.
>>
>>Hope that this helps.
>>
>>Harvey
>>
>>
>> >
>> >Thanks.
>> >
>> >Mark
>>
>>
>>
>>------------------------------------
>>Posted by: Harvey White <madyn@...>
>>------------------------------------
>>
>>Be sure to visit the group home and check for new Links, Files, and Photos:
>>http://groups.yahoo.com/group/Homebrew_PCBs
>>------------------------------------
>>
>>Yahoo Groups Links
>>
>>
>>

A little out of my league <smile>. My needs are simpler, though I
often do use TQFP packages. Those, however, often need vias under the
chip, which I can't do. To make alignment less critical I would use
larger vias so you can't "miss" as easily.

Mark


At 09:38 AM 8/16/2014, you wrote:

>On Sat, 16 Aug 2014 08:51:49 -0400, you wrote:
>
> >Thanks, Harvey, for the detailed description - Very interesting
> >setup. One of these days I'll work on something similar, though I
> >rarely drill a lot of holes. I use almost exclusively smt and usually
> >make single sided pcbs using zero ohm resistors as jumpers.
> >
>
>I do double sided boards, with the boards made from two .021 boards
>epoxied together (use 1 hour epoxy!). Alignment is very critical and
>the drilling setup (finally!) makes that work a lot better. Since I'm
>doing 64/100/144 pin TQFP chips, I can't live with the luxury of
>exclusively resistive jumpers.
>
>Latest project involves a graphics panel driver with an S1D13781
>graphics chip (Epson, anyone wanting to use it please talk to me,
>there are some non-obvious things you need to do to make it work). an
>Xmega, a touch panel controller, -24 volt supply, I2C communications
>(dual channel).
>
>Harvey
>
> >Mark
> >
> >
> >At 06:50 PM 8/15/2014, you wrote:
> >>On Fri, 15 Aug 2014 12:52:18 -0400, you wrote:
> >>
> >> >Hello Harvey,
> >> >
> >> >I'm interested in getting more information on your drilling setup,
> >> >mostly the optical stuff - camera, software, monitor. What supplies
> >> >the reticule image for the crosshairs?
> >>
> >>OK, in short. The optical stuff is a 12 volt/24vac camera with almost
> >>any lens that works well. You want to move it out away from the CCD
> >>so that you have the equivalent of a closeup extender.
> >>
> >>That goes to a standard color monitor. This is an LCD monitor that I
> >>found (runs off 12 volts DC), and has an overlay with a graticule on
> >>it. I've been tempted to make it more complicated by adding a
> >>crosshair generator, but that'd be another PC board and some video
> >>work. The graticule is easier. Just make the pattern on a
> >>transparency and put it over the monitor itself. Almost any old TV
> >>would work.
> >>
> >>The software is custom, and is a simple Z axis setup. There are two
> >>pushbuttons (green/red) and a rotary encoder with a push switch
> >>(Electronics goldmine sold them). There's an end travel (and top
> >>travel) led sensor.
> >>
> >>Boot up, the display says to push green to home. That homes the drill
> >>and runs it up to the rest position. pushing the footswitch (or the
> >>green pushbutton) starts a drill cycle, three speeds possible, up to
> >>the drill top position. At that point the drill retreats back to the
> >>rest position and turns off the Proxxon tool (solid state relay, works
> >>better with the 12 volt model and transformer).
> >>
> >>Pushing the red button puts the drill at the top position, which is
> >>where I can change the bit from the top.
> >>
> >>Pushing in the rotary switch/encoder put it into an adjust mode, where
> >>I can set the rest and top positions (stored in the processor's
> >>eprom).
> >>
> >>Basically, the software is not all that complicated, but does rely on
> >>an infrastructure of display drivers (it's a smart display that I
> >>built), a modified keyboard driver package, and a few other bits and
> >>pieces.
> >>
> >>Physically, the camera is mounted on a PACE hot air station stand (I
> >>managed to find two for not so much money), which gives me an
> >>overarching arm for the camera. I can move it up and down to focus
> >>the camera and move it out of the way for drill change. I suppose
> >>another enhancement would be to put a stepper drive on that.
> >>
> >>You will have to align the camera center point to the tip of the
> >>drill. In my case, the adjustment is entirely mechanical, with the
> >>camera swinging left/right on the mounting screw, but tilted (spring
> >>used on screw) forwards/backwards as needed. That's a trifle awkward,
> >>and the electronic crosshair generator would be better. That would
> >>require a small CPLD, a sync separator, a synchronized pixel clock,
> >>and some counters. Ideally, the graticule has ticks on both axis
> >>lines, since that helps align the center of large holes.
> >>
> >>The smallest hole I drill is 0.021 for a via, and I estimate that I
> >>can repeat that hole to within 0.002 inches roughly. The size of that
> >>hole on the monitor is the critical part.
> >>
> >>You'll want to backlight the board. Red light may be most useful, but
> >>I haven't experimented with that yet.
> >>
> >>Hope that this helps.
> >>
> >>Harvey
> >>
> >>
> >> >
> >> >Thanks.
> >> >
> >> >Mark
> >>
> >>
> >>
> >>------------------------------------
> >>Posted by: Harvey White <madyn@...>
> >>------------------------------------
> >>
> >>Be sure to visit the group home and check for new Links, Files, and Photos:
> >>http://groups.yahoo.com/group/Homebrew_PCBs
> >>------------------------------------
> >>
> >>Yahoo Groups Links
> >>
> >>
> >>
>
>
>
>------------------------------------
>Posted by: Harvey White <madyn@...>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBs
>------------------------------------
>
>Yahoo Groups Links
>
>
>

I think the thing that would ease alignment of vias isn't just via size, but via landing size.

David Bobb -- N5MUX

On Aug 16, 2014 10:12 AM, "Mark Lerman mlerman@... [Homebrew_PCBs]" <Homebrew_PCBs@yahoogroups.com> wrote:

 

A little out of my league <smile>. My needs are simpler, though I
often do use TQFP packages. Those, however, often need vias under the
chip, which I can't do. To make alignment less critical I would use
larger vias so you can't "miss" as easily.

Mark

At 09:38 AM 8/16/2014, you wrote:
>On Sat, 16 Aug 2014 08:51:49 -0400, you wrote:
>
> >Thanks, Harvey, for the detailed description - Very interesting
> >setup. One of these days I'll work on something similar, though I
> >rarely drill a lot of holes. I use almost exclusively smt and usually
> >make single sided pcbs using zero ohm resistors as jumpers.
> >
>
>I do double sided boards, with the boards made from two .021 boards
>epoxied together (use 1 hour epoxy!). Alignment is very critical and
>the drilling setup (finally!) makes that work a lot better. Since I'm
>doing 64/100/144 pin TQFP chips, I can't live with the luxury of
>exclusively resistive jumpers.
>
>Latest project involves a graphics panel driver with an S1D13781
>graphics chip (Epson, anyone wanting to use it please talk to me,
>there are some non-obvious things you need to do to make it work). an
>Xmega, a touch panel controller, -24 volt supply, I2C communications
>(dual channel).
>
>Harvey
>
> >Mark
> >
> >
> >At 06:50 PM 8/15/2014, you wrote:
> >>On Fri, 15 Aug 2014 12:52:18 -0400, you wrote:
> >>
> >> >Hello Harvey,
> >> >
> >> >I'm interested in getting more information on your drilling setup,
> >> >mostly the optical stuff - camera, software, monitor. What supplies
> >> >the reticule image for the crosshairs?
> >>
> >>OK, in short. The optical stuff is a 12 volt/24vac camera with almost
> >>any lens that works well. You want to move it out away from the CCD
> >>so that you have the equivalent of a closeup extender.
> >>
> >>That goes to a standard color monitor. This is an LCD monitor that I
> >>found (runs off 12 volts DC), and has an overlay with a graticule on
> >>it. I've been tempted to make it more complicated by adding a
> >>crosshair generator, but that'd be another PC board and some video
> >>work. The graticule is easier. Just make the pattern on a
> >>transparency and put it over the monitor itself. Almost any old TV
> >>would work.
> >>
> >>The software is custom, and is a simple Z axis setup. There are two
> >>pushbuttons (green/red) and a rotary encoder with a push switch
> >>(Electronics goldmine sold them). There's an end travel (and top
> >>travel) led sensor.
> >>
> >>Boot up, the display says to push green to home. That homes the drill
> >>and runs it up to the rest position. pushing the footswitch (or the
> >>green pushbutton) starts a drill cycle, three speeds possible, up to
> >>the drill top position. At that point the drill retreats back to the
> >>rest position and turns off the Proxxon tool (solid state relay, works
> >>better with the 12 volt model and transformer).
> >>
> >>Pushing the red button puts the drill at the top position, which is
> >>where I can change the bit from the top.
> >>
> >>Pushing in the rotary switch/encoder put it into an adjust mode, where
> >>I can set the rest and top positions (stored in the processor's
> >>eprom).
> >>
> >>Basically, the software is not all that complicated, but does rely on
> >>an infrastructure of display drivers (it's a smart display that I
> >>built), a modified keyboard driver package, and a few other bits and
> >>pieces.
> >>
> >>Physically, the camera is mounted on a PACE hot air station stand (I
> >>managed to find two for not so much money), which gives me an
> >>overarching arm for the camera. I can move it up and down to focus
> >>the camera and move it out of the way for drill change. I suppose
> >>another enhancement would be to put a stepper drive on that.
> >>
> >>You will have to align the camera center point to the tip of the
> >>drill. In my case, the adjustment is entirely mechanical, with the
> >>camera swinging left/right on the mounting screw, but tilted (spring
> >>used on screw) forwards/backwards as needed. That's a trifle awkward,
> >>and the electronic crosshair generator would be better. That would
> >>require a small CPLD, a sync separator, a synchronized pixel clock,
> >>and some counters. Ideally, the graticule has ticks on both axis
> >>lines, since that helps align the center of large holes.
> >>
> >>The smallest hole I drill is 0.021 for a via, and I estimate that I
> >>can repeat that hole to within 0.002 inches roughly. The size of that
> >>hole on the monitor is the critical part.
> >>
> >>You'll want to backlight the board. Red light may be most useful, but
> >>I haven't experimented with that yet.
> >>
> >>Hope that this helps.
> >>
> >>Harvey
> >>
> >>
> >> >
> >> >Thanks.
> >> >
> >> >Mark
> >>
> >>
> >>
> >>------------------------------------
> >>Posted by: Harvey White <madyn@...>
> >>------------------------------------
> >>
> >>Be sure to visit the group home and check for new Links, Files, and Photos:
> >>http://groups.yahoo.com/group/Homebrew_PCBs
> >>------------------------------------
> >>
> >>Yahoo Groups Links
> >>
> >>
> >>
>
>
>
>------------------------------------
>Posted by: Harvey White <madyn@...>
>------------------------------------
>
>Be sure to visit the group home and check for new Links, Files, and Photos:
>http://groups.yahoo.com/group/Homebrew_PCBs
>------------------------------------
>
>Yahoo Groups Links
>
>
>

On Sat, 16 Aug 2014 11:12:01 -0400, you wrote:

>
>A little out of my league <smile>. My needs are simpler, though I
>often do use TQFP packages. Those, however, often need vias under the
>chip, which I can't do. To make alignment less critical I would use
>larger vias so you can't "miss" as easily.

Well....

I don't do vias under the chip at all. I've recently hit on a design
that takes all the supply voltages, spiders them out a little, then
goes to the bottom layer. Under the chip are all of the bypass caps,
which have a big common ground plane under the chip itself. The chip
ground leads themselves go to a common ground plane under the chip on
the top layer. The ground leads from the chip are also spidered out
and run under the chip to the ground plane there. Seems a slight bit
complicated, but other than the vias required to go top to bottom, you
have pretty decent access to all the pins.

The latest vias have a 0.021 hole, and are about 0.043 in diameter, so
alignment can be a bit picky here.

Why so small (relatively, I guess)? 0.5 mm spacing TQFP 144 pin
packages is why...

Harvey

>
>Mark
>
>
>At 09:38 AM 8/16/2014, you wrote:
>>On Sat, 16 Aug 2014 08:51:49 -0400, you wrote:
>>
>> >Thanks, Harvey, for the detailed description - Very interesting
>> >setup. One of these days I'll work on something similar, though I
>> >rarely drill a lot of holes. I use almost exclusively smt and usually
>> >make single sided pcbs using zero ohm resistors as jumpers.
>> >
>>
>>I do double sided boards, with the boards made from two .021 boards
>>epoxied together (use 1 hour epoxy!). Alignment is very critical and
>>the drilling setup (finally!) makes that work a lot better. Since I'm
>>doing 64/100/144 pin TQFP chips, I can't live with the luxury of
>>exclusively resistive jumpers.
>>
>>Latest project involves a graphics panel driver with an S1D13781
>>graphics chip (Epson, anyone wanting to use it please talk to me,
>>there are some non-obvious things you need to do to make it work). an
>>Xmega, a touch panel controller, -24 volt supply, I2C communications
>>(dual channel).
>>
>>Harvey
>>
>> >Mark