LCD -- i.MX233 Hand-Held Multimedia Board
Casainho
verify when I have the LCD on my hands.
But looks like that all LCD pins are available on 0.1'' expansion
header. If we want to use the touchscreen, we will need to put the
touchscreen IC on secondary board and solder 2 wires on main board
(the wires for battery temperature and voltage measure).
So, since we will need to make a secondary board for the buttons, I
think we should use that board for the LCD and buttons.
Matt, could you please tell your opinion on this? are any pins missing
on expansion header?
And finally, could you design a board for this? or at least the schematic?
Later I could try mill and drill this board (easier if a board off 1 layer).
Another question is how to connect this secondary board to main. The
headers could be directly soldered... but then we would be blocked to
access main board, which is not good.
But I have females and males headers of 0.1'', the problem is that
they may take to much space on Z axis.
Bob
I have just had delivered some low profile 0.1 headers from Toby
Electronics. The female is 3.5mm and the male is 1.5mm in profile
giving a total minumum board to board gap of 5mm. This could be
reduced slightly by removing the plastic spacer from the male
connector after soldering (DIY step).
http://www.toby.co.uk/content/catalogue/products.aspx?series=B01-xXxx-AG1-G
http://www.toby.co.uk/content/catalogue/products.aspx?series=LHCD-xxS-R-027-xxx
I believe the tallest component on the core board is the audio
connector at 5mm, however if overlap of the audio connector is not
required then I'm not sure what the limit is, I think there is a
ferrite core ~3mm?
Casainho
> I was just thinking about designing such an IO board.
We must work on it now.
> I have just had delivered some low profile 0.1 headers from Toby
> Electronics. The female is 3.5mm and the male is 1.5mm in profile
> giving a total minumum board to board gap of 5mm. This could be
> reduced slightly by removing the plastic spacer from the male
> connector after soldering (DIY step).
I have in stock some of females and males, that I bough on cheap from
local store. I will measure them and verify.
> I believe the tallest component on the core board is the audio
> connector at 5mm, however if overlap of the audio connector is not
> required then I'm not sure what the limit is, I think there is a
> ferrite core ~3mm?
Hmmm, nice that 5mm if we will need that for audio connector.
Since we will buy the LCD and other components from digikey, we could
find there such 0.1'' female and male headers.
Bob
> local store. I will measure them and verify.
~7.5mm for the female.
The low-profile versions Toby supply are made by Valcon but I can't
find a website for them and Digi-key don;t seem to stick them. A quick
search at Digi-key did not reveal any low-profile connectors but they
have a LOT of connectors ;)
Usually I find it too restrictive to use only one supplier.
Bob
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SAM1095-20-ND
female equivalent is the Samtec HLE range, but they don't appear to
be stocked. They are also about twice the cost of my UK source.
john cooper
> So, since we will need to make a secondary board for the buttons, I
> think we should use that board for the LCD and buttons.
I've heard this all somewhere before. ;)
But seriously, I was trying to underscore this need very
early on. A secondary i/o board is a practical concession
in a minimized form-factor device and it needs to be
designed in tandem with the base board. Doing so after the
fact may lead to unwelcome compromises and an exercise in
frustration.
I spent some time investigating the prospect of keeping it all
on one board. But this would require both stand-off LCD mount
(and we're somewhat there with an FFC/FPC cable connection) and
a standoff full SD socket (seem to have become unusually rare).
Both place unwelcome restrictions on a dense single board layout
as their external physical location creates a non-negotiable
land location on the board.
Even maneuvering around that, the prospect of a key/input
area sinks the ship. Aside from the sprawling board
area consumed, thru-vias need to be staggered around voids
in the switch array. And using bare snap-domes almost
certainly will require blind vias.
A second board has its own considerations. 1.6mm laminate
x2 is wasteful of Z-axis budget, and coordinating component
profiles on both boards to minimize board stack offset is
another puzzle. One issue which took me some time to
resolve was finding external wire to board connectors able
to fit (and moreover function) in the restriction of a 5mm
board offset.
Also the common CUI and Keystone 3.5mm audio jack SMD connectors
are of 5mm height which leaves zero clearance but is so close
that it is temping to try. Two less than ideal possibilities here
are milling 0.2mm from the connector body top or (ugh) a 0.2mm
polyimide/kapton tape shim under one board/board SMD connector
housing. I'll likely mill the connector. Another possibility
of addressing the 5mm ceiling is using a 2.5mm SMD jack.
> And finally, could you design a board for this? or at least the schematic?
> Later I could try mill and drill this board (easier if a board off 1 layer).
I found using even 2 layers for the user interface board
in my design to be a little tight in some cases. With a
single layer board you'll likely end up with wire jumpers
consuming a fair amount of board space on the side they
sit and opposite side for protruding wire tails and solder
fillets.
A two layer board will allow direct (adhesive) mount of
the lcd to one side while the opposite side can be used
for lcd support circuitry. The board surface over-mounted
by the lcd can still be freely used for interconnect of
the support components residing on the opposite side. Doing
so worked out fairly well in my case.
> Another question is how to connect this secondary board to main. The
> headers could be directly soldered... but then we would be blocked to
> access main board, which is not good.
> But I have females and males headers of 0.1'', the problem is that
> they may take to much space on Z axis.
Yes that's the decision I had to make quite a while ago.
Its a trade off of either common utility (eg: using 100mil
headers) or maximum compactness using hand-held device
class connectors. Unfortunately there isn't much of a happy
compromise between the two alternatives.
In cases like this I also avoid through-hole connectors as
they claim non-negotiable real estate on both board surfaces.
Given the physical constraints of where lcd/ffc and input
devices are going to land, through-hole connectors may be
unusable.
I seem to recall finding SMD mount 100mil headers and sockets
somewhere in the past. Try Sullins. I know they have 2mm
versions as I have samples of them here.
-john
Casainho
> Here is an equivalent Digi-key part, male
> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SAM1095-20-ND
> female equivalent is the Samtec HLE range, but they don't appear to
> be stocked. They are also about twice the cost of my UK source.
They are expensive :-(
I can buy for 0.5€ the 40 pins equivalent, at local store. Both female
and male connected make 11mm, so the PCB1 and PCB2 would have 11mm
between them.
I think we can win some mm on printing our enclosure becase:
- we can print with less 0.5mm each wall, meaning less 1mm in total;
- there are 5.4mm of space between PCB and case, we could go with
3.2mm only which is the Z of SDCard connector, so, less 1.2mm;
- finally we can solder the headers or use the low profile ones. Until
then, while we are developing, we can just design and print the
enclosure a few Z mm more, for the female/male headers usage.
Audio connectors have 5mm Z, which I think are the biggest size on top
of the board. Under the board, SDCard connector is the bigger with
3.2mm.
The final dimensions of printed enclosure should be: Y 99mm, X 66mm
(for 2mm walls and one AA battery and support, Y 18mm).
Bob
> I can buy for 0.5€ the 40 pins equivalent, at local store. Both female
> and male connected make 11mm, so the PCB1 and PCB2 would have 11mm
> between them.
> (for 2mm walls and one AA battery and support, Y 18mm).
internal plus wall thickness!
The choice is to use a low profile connector, or a thicker case. We
could use the wasted space for a flat battery, but not an AA shape.
Designing the IO board is tricky, for reasons John mentions. Some of
the LCD signals are not on the expansion connector, but I may be able
to work around that. The expansion connector overlaps the LCD
footprint, so with a through hole connector the LCD will need to be
raised slightly off the board. I think that will be OK as we have a mm
and a bit spare. If I use SMT female that may actually be a higher
profile, so we lose some mm and get no gain.
Having the male pins on the JTAG connector is awkward, because a
standard JTAG connector will not mate with a low profile header. It
may be possible to pass the JTAG signals through to a standard
connector on the IO board.
Anyway, I have started on a schematic for an IO board.
Bob
\hardware_design\lyre_v1_io. There are also pdf files in the group
files area.
I have not attempted to place the touch screen controller, I'm not
sure there is much point. I have used the PWM0 signal to drive the
backlight. I have used 8 GPIOs for 8 switches, I think there are
another 2 GPIO available. It would seem like a good idea to have a
power switch on this board but I have not placed one yet.
By removing a couple of unused pins on the expansion connector, the
2.4 LCD FFC can fold under the LCD. The LCD will need some plastic
spacers under it and maybe some spots of glue to hold it in position.
I couldn't work out how to tell pcbnew to place the connectors on the
copper side but reverse the pinout to match the top, so the connectors
are shown on the wrong side, but the pin out is correct I think.