[EE]: Trying to ID a connector
Philip Pemberton
I picked up a couple of "slightly used" EL displays the other day,
apparently out of a Sun server of some description. They're made by
Planar, have 3 LEDs and a 160x80 graphical EL panel with a touch screen.
Planar part numbers are 996-0277-01 and EL160.80.38, Rev E. Sun P/N is
370-2311-02 Rev 50. All the juicy details are in the following forum
thread (most of the good stuff is on page two, the last post includes a
pinout):
http://forum.lcdinfo.com/viewtopic.php?t=1051
Problem is, while I have the pinout, the connector is an odd little
thing that looks like a shrunken version of a Centronics printer connector:
http://www.skippari.net/lcd/forumstuff/EL160.80.38-SM1_2.jpg
It has 32 pins, is surface-mounted, and is made from white plastic. No
manufacturer ID or branding that I could see.
Does anyone recognise this connector? While I can desolder it with the
hot-air gun and replace it with a bit of ribbon cable, I'd rather use
the "proper" connector if it's something reasonably easy to get...
Thanks,
--
Phil.
pic...@philpem.me.uk
http://www.philpem.me.uk/
--
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Mike Harrison
>Hi guys,
>
>I picked up a couple of "slightly used" EL displays the other day,
>apparently out of a Sun server of some description. They're made by
>Planar, have 3 LEDs and a 160x80 graphical EL panel with a touch screen.
>
>Planar part numbers are 996-0277-01 and EL160.80.38, Rev E. Sun P/N is
>370-2311-02 Rev 50. All the juicy details are in the following forum
>thread (most of the good stuff is on page two, the last post includes a
>pinout):
> http://forum.lcdinfo.com/viewtopic.php?t=1051
>
>Problem is, while I have the pinout, the connector is an odd little
>thing that looks like a shrunken version of a Centronics printer connector:
> http://www.skippari.net/lcd/forumstuff/EL160.80.38-SM1_2.jpg
>
>It has 32 pins, is surface-mounted, and is made from white plastic. No
>manufacturer ID or branding that I could see.
>
>Does anyone recognise this connector? While I can desolder it with the
>hot-air gun and replace it with a bit of ribbon cable, I'd rather use
>the "proper" connector if it's something reasonably easy to get...
>
>Thanks,
>--
>Phil.
>pic...@philpem.me.uk
>http://www.philpem.me.uk/
Looks similar to the family used on older TTL interface TFT panels - pretty sure Digikey had them -
maybe worth a parametric search starting with pin count and pitch.
Probably made by someone like Hirose or JST
Philip Pemberton
> Looks similar to the family used on older TTL interface TFT panels - pretty sure Digikey had them -
> maybe worth a parametric search starting with pin count and pitch.
> Probably made by someone like Hirose or JST
That's what I thought -- I've got some NEC 640x480 panels on order which
use a Hirose DF9. The DF9 looks similar, but isn't quite right (not
available in 32-pin, and too small pitch). Also doesn't carry their
logo; IIRC Hirose stamp their logo into almost every connector they make.
I forgot about JST though -- thanks for the lead. Though it's fairly
unusual to see Asian connectors (Hirose, JST, et al) on Western kit;
this display and most of its PCBs bear the legend "MADE IN USA BY
PLANAR". Hmm.
It does look fairly simple to reverse engineer. The SED1355 has one
address pin (A0), but there are two on the connector pinout (A0 and A1).
There's some form of PLD on the board which appears to run the touch
screen and possibly the addressing/chip select for the LCD controller.
I'm willing to wager that if A1=0, the 1335 is selected, which leaves
two DIO ports for the touch screen.
Given that the touchscreen is a "fixed pad" one -- i.e. a 6x3 matrix of
transparent switches, I suspect one port is the "scan" output and the
other is the "response" input. Should be fun to reverse engineer (and
probably take about half an hour to do so).
If I can't figure it out, I'll hack up a PIC to generate some bus
cycles, solder some "spy wires" onto the chips and connectors, and watch
the thing with a scope or LA ;)
Mike Harrison
>On 14/06/10 14:42, Mike Harrison wrote:
>> Looks similar to the family used on older TTL interface TFT panels - pretty sure Digikey had them -
>> maybe worth a parametric search starting with pin count and pitch.
>> Probably made by someone like Hirose or JST
>
>That's what I thought -- I've got some NEC 640x480 panels on order which
>use a Hirose DF9. The DF9 looks similar, but isn't quite right (not
>available in 32-pin, and too small pitch). Also doesn't carry their
>logo; IIRC Hirose stamp their logo into almost every connector they make.
>
>I forgot about JST though -- thanks for the lead. Though it's fairly
>unusual to see Asian connectors (Hirose, JST, et al) on Western kit;
>this display and most of its PCBs bear the legend "MADE IN USA BY
>PLANAR". Hmm.
...In which case check Molex and Tyco/Amp
Philip Pemberton
> ...In which case check Molex and Tyco/Amp
Nope, it's a Hirose. FX2 series, 1.27mm pitch to be precise. The board
connector is an FX2-32P-1.27SV(71), which means the mating connector is
an FX2-32S-1.27SV(71).
I just spotted the "HRS" logo stamp on the *inside* of the connector
moulding. Madness. Not stocked by Farnell, RS or Rapid, but is stocked
by Digikey at a fairly high price (£3.51 for either gender). Ugh. Looks
like I'm replacing the connector with ribbon cable... unless somebody
happens to have a couple of spare Hirose FX2-32S plugs kicking around?
I just buzzed out some of the LCD wiring -- A0 is wired straight to the
SED controller, while A1 is wired to nCS. That means the address map is
something like this:
A1 A0 Function
0 0 SED1335 status flag read / display data and parameter write
0 1 SED1335 display data read / command write
1 0 PLD?
1 1 PLD?
Given that there's an interrupt pin on the connector (and that the SED
doesn't have an interrupt output) I suspect the PLD scans the
touchscreen continuously and keeps track of which switches have been
pushed. That would imply that to read the touchscreen state, you write a
row address to one port and read the pin states for that row from the
other. There may well be an interrupt flag / enable as well...
--