OT (?) Laptop as terminal

You don't *touch* the video!

Consider:

- chances are, you're sitting in front of an LCD monitor right now.

- that monitor is able to display a variety of "video resolutions"

- remove the "signal processing" electronics from that monitor

- this PCB has (essentially) three connectors on it: power, video in and LCD panel *out*

- place this inside laptop case

- connect the "panel out" connector to an appropriate "controller in" connector on the "adapter board"

- connect the laptop's LCD panel to a suitable "controller out" connector on the "adapter board"

The point of my post is, besides physical differences, what other

*operational* differences might there be between panels to preclude this?

(note that you may have to tell folks "use the signal processing board out of a Model 289 Fraggletronics LCD monitor to make this work... hopefully, Fraggletronics isn't "Unobtanium")

Yes -- though I usually don't need the mouse.

And, rather than "one off", if I am going to invest the time doing it, I'd like others to be able to follow in my footsteps instead of having to repeat the learning process.

[I.e., I can easily make that "one off" *today*... put my 7" monitor inside *any* laptop case -- with a lexan bezel to cover up the space that the LARGER, original display occupied -- and just wire the keyboard matrix to a keyboard decoder from an old PS/2 keyboard]

That's big, though. E.g., I could just *toss* the 7" monitor and 12" keyboard into a briefcase and I've achieved that level of "portability" :-/

The "beauty" of laptops is that you can throw 6 or 8 of them in a drawer and forget about them.

I curse myself for discarding an old laptop that had a built-in AC power supply feeling it was "too old" to be useful :<

Yes, I stand corrected

I have seen various sizes of foil connectors used, and would not like to guarantee the matrix pinout between models.

Assumes the matrices are all pinned out the same for column and row your row and column lines could well be different between models, meaning having to add a patch panel and different connector placements.

You may be looking on what you think is a row return, but actually is another column.

See later on slack..

The signals to the LCD are digital, but there are many variants. The input signal is most likely analog with nothing to do conversion left as mother board has been removed.

Hope for commonality across LCD panels has been hoped for by many engineers over the last decade.

Then find one or more of the keyboard mounting screws is in the way on this model.

Whilst also ensuring that the connector to the LCD panel will still reach. Many laptops have a keyboard securing screw very near the graphics connector. Most have two screws near the hinges, which is where the existing LCD cable goes.

You may get something to work with keyboard flapping in the breeze.

You are sure they are all pinned in same manner for column and row.

The LCD cable has little slack in most laptops in the base part and any array of suitable connectors each with choice of drive methods

LVDS and TMDS variants parallel data Alternate pixel drives (normlly dual LVDS/TMDI)

Doubt you would be able to make it reach without something like three possible connectors on your board and an adapter cable to go between your drive electronics and the existing cable.

...

NO, the controller is on the MOTHERBOARD you have removed, next to no LCDs in laptops have the scaling electronics in the LCD, it is the responsibility of the graphics controller to scale graphic images to fit the LCD panel. Even when appearing to do lower resolution this is done by the removed graphics controller driving the fixed timing to the LCD panel, with filled/stretched video format.

LCD Panels are very very dumb. Myriad of different signalling and connector methods.

LCD MONITORS have EXTRA electronics to do the conversion from analog multi format to fixed digital size for LCD panel.

If you are going to repackage an exiting monitor you might as well leave it as is and get cheap USB/PS2 keyboard and mouse and plug them in direct.

If you are still using the guts of a laptop you do NOT have the innards of an LCD monitor, just a dumb LCD panel once the motherboard has been removed.

--
Paul Carpenter          | paul@pcserviceselectronics.co.uk
    PC Services
 Timing Diagram Font
  GNU H8 - compiler & Renesas H8/H8S/H8 Tiny
 For those web sites you hate

Why bother it and other electronics and supply rails may not fit inside a laptop. Also see later

A PARTICULAR LCD panel out, possibly supports a few variants by different connectors.

Assuming the panels match there are MANY variants on signalling for LCD panels, your "adapter" may not support the LCD in the laptop.

The cabling is VERY unlikely to match and these connectors can often be a pain to get in small quantities.

Your "adapter" may not support that LCD resolution

Your "adapter" may not support that LCD connector

Your "adapter" may not support that LCD signalling Even to the extent that your LCD pannel is alternate pixel drive and the "adapter" only supports single pixel drive.

Your "adapter" may not support that LCD backlight control or backlight inverter, you do realise the second highest power drain in a laptop is the screen backlight drive.

Your "adapter" is of a strange shape to fit in casing

Your "adapter" has a myriad of power rails to be provided, meaning extra PSU board(s).

These boards change as much as laptops, and you would need to make a matrix up of which good monitor to pull apart to match all the screen types of all the laptop models.

Then you will need further matrices of cabling arrangements, where to get suitable connecors or the like. How to connect the cable up to your PCB.

Alternatively for each LCD panel and laptop combination what the pin out is and which wire to cut and solder to which hole on the PCB.

Having seen too mmany variants on graphics LCD connectors and signalling that when I did a 3D LCD controller from cameras to drive two screens, I developed a multi format converter that went out to TWO standard 40 way IDC headers. Any cabling or signalling conversion for different connectors and signalling was done via cables and/or piggy back boards.

Piggy back boards took TWO x 24 bit parallel data and clocks, then converted drives to LVDS or TMDI or Alternate pixel drive or even did video line inversion as well, depending on panel connected.

If no special drive was need the drive was cabled in direct.

--
Paul Carpenter          | paul@pcserviceselectronics.co.uk
    PC Services
 Timing Diagram Font
  GNU H8 - compiler & Renesas H8/H8S/H8 Tiny
 For those web sites you hate

With that attitude, then Jolitz should *never* have wasted all those "hours and dollars" and I'd be running commercial BSD's instead of "open" ones! :> (i.e., someone has to do it first or everyone falls on this same hurdle)

Consider the number of repurposed and hacked devices that are now commonplace, this seems like it falls in the category of "incredibly obvious" (far moreso and more "useful" than, for example, repurposing an NSLU2 or a WRT54G)

Presumably that talks to "special software" on a PC? I.e., that software reads a captured frame buffer and repaints it onto a window in the PC?

Note that I could also outfit those machines incapable of console redirection (in the BIOS) with "Weasels" fitted to a terminal server (I should contact Jonathan to see if they ever went the logical next step to include a NIC)

Of course, the latter wouldn't help any of the "non-PC" machines that I have...

You missed his at least as widely. The problem is with the PS/2 ports of your headless devices, which you apparently intend to hot-plug to after they've got stuck. It's _their_ PS/2 ports you're eventually going to fry one by one.

Hardly. Trackpads essentially don't exist outside laptops; trackpoints even less so. These things are built expressly and exclusively for laptop usage.

No such thing as "the" LCD panel interface. They're basically similar, but way too different for this to be an interface viable for repurposing.

Congratulation, you've just rediscovered the reason KVM switches _do_ exist.

Here is your (non-LCD) portable monitor:

-- Roberto Waltman

[ Please reply to the group. Return address is invalid ]

OTOH, Software is infinitely hackable. Chips aren't. Laptops are very highly refined to achieving one goal, which is not your goal.

However, if you wanted to attempt something like this, the only sensible path is to do it largely in software, by putting a Linux kernel and terminal emulator in the BIOS ROM. You can ditch the hard drive and other componentry, but re-use the existing CPU and graphics pipeline which already has the right connections to the panel.

Linux in the BIOS ROM has been done. A simple PS2 and/or USB interface could be easily added, probably as a USB peripheral.

I guess I am missing something here... how is the laptop going to capture the video from the headless device? I.e., you still need a frame grabber (like the device mentioned up-thread). In which case, what does Linux-in-ROM buy you that "Windows-on-disk" doesn't?

You don't need the video for those BSD servers, you only need a terminal. But I guess that still doesn't help if you can't boot and need to tweak the BIOS settings. The video rates in that mode is very minimal though, so a classic TV framegrabber could almost do it.

If you have hardware that needs this kind of maintenance often enough to warrant having this hardware diagnostic tool, piff it and buy something more reliable.

I've addressed this issue several times over the last couple of decades. Always came to the same conclusion: "forget it".

Put your existing keyboard and monitor into a briefcase with a battery.

If you have usb, there exist flexible keyboards that roll up pretty tidy.

I find VNC works nicely on a PDA, but you have to have the wireless interface and the system has to be "UP" for it to work. If you can't boot, you don't have much choice but to use a keyboard/monitor with interfaces supported by the BIOS.

Modifying a laptop won't buy you much over what you already have. Still need cables to plug in. And it's an extremely difficult task. Trivial in concept. The devil is in the details...unpublished details.

Why are you hot plugging? The second sentence of my original post: "From time to time, a machine fails to boot to multiuser run level."

Machine fails to boot. You can't *see* anything because there's no display. You can't talk to it because the network I/F is unresponsive. What do you do? Push reset (or cycle power for boxes with no reset button). Is problem repeatable, or a fluke?

If problem persists, you need a keyboard+display:

TURN OFF THE POWER. Drag the machine out to a point where you can access the keyboard and display connectors (since my machines run headless, I "hide them" in out of the way places where they can largely be ignored) Plug in keyboard, mouse, display. Power up machine. Watch boot process to see where things are stuck. Most times, this is pretty obvious: a floppy sitting in a drive; a SCSI cable that has been removed; a removable disk "cartridge" that isn't seated fully and making a poor connection, etc. TURN OFF THE POWER. Fix problem. Unplug keyboard, mouse, display. Slide machine back where it belongs. Reboot.

I learned ~30+ years ago not to plug in keyboards on powered up machines when a friend complained that his keyboard was "dead". He was very grateful that I happened to have an assortment of those little resistor-shaped fuses (because I got stuck in Romney, West Virginia one weekend waiting for an express delivery of one of them to fix a piece of equipment).

Note that old keyboards were power hungry devices. This isn't the case with newer ones, for the most part. And, the touchpads use < 10mA.

In recent years, I've seen more "blown keyboard" problems traced to mechanical failure -- typically, inductors separating from the PCB as it is repeatedly flexed near the keyboard connector.

Note: the touchpad assemblies *for* laptops are *exactly* the same as those packaged standalone. I.e., take a laptop touchpad, solder 4 wires to the flex-wire connector, route them to a PS/2 connector and you have a standalone "touchpad mouse".

Put an FPGA between the outputs of the "controller" and the panel. All you are really looking to do is reroute signals from here to there; you aren't really doing any "processing".

Recall, this doesn't have to work with EVERY laptop made and EVERY "controller". E.g., I have a Neotus LCD "monitor" here that, I'll bet, is too weird and obscure to be worth the effort to make a "compatible" solution! (OTOH, it might be

*exactly* the right device for the job... I should pop the back off and have a look!)

I have four KVM's. All sitting in storage because they are impractical in my situation (all except the Rose are incompatible with older Sun boxen). The most practical tool, to date, has been a KVM *repeater* (extender) as it lets me leave something plugged into a machine and

*later* plug a keyboard/mouse/display into the "remote end".

Far more work than interfacing the controller out of an LCD monitor to another LCD panel. :< Software solutions tend to be "neverending" whereas a bit of hardware is done once it is done. :> (if it doesn't work on *your* particular set of components... "sorry"!)

It's not a question of reliability. I have a *lot* of machines here. Usually set up for specific purposes. So, which machine I am using depends on what I am doing. And, since I tend to end up "doing" something for "a while", it often means that a machine won't see any use for months or longer (I suspect it's been close to a year since I fired up my SB1000; probably almost that long for the SB150; *longer* for the Voyager -- but, it's got a display so that's a non-issue :> )

If, for example, I've traveled since the last time a machine was used, it's probably no longer connected to power (I unplug things when I am going to be away, "just in case"). Invariably, I have no idea what the last thing I used a machine for might have been (which peripherals are connected? which disk is installed? is there a floppy left in the drive? how did I configure openboot? etc.).

So, I power up a machine and *hope*. Most of the time, things work fine. Sometimes, they don't! :-/ E.g., last night, one of my NAS boxes just sat there blinking at me. I figured maybe it was checking the drives so let it stew; this *morning* I pulled the plug for a few seconds (no reset button and the "soft power" switch was unresponsive) and it came right up.

When something doesn't work as expected, the last thing I want is a "major production" to figure out what's gone wrong. I want to be able to *easily* figure out what's happened and fix it -- especially when the "fix" is usually something trivial. E.g., the last machine that failed to boot did so because I had previously been using it to wipe drives and had left a wiped drive in it *as* the boot disk: time to repair, 20 seconds; time to drag out monitor and keyboard to diagnose the problem, 5 minutes.

I suspect this sort of thing is commonplace at server farms. I was asked to design a solution for one installation (~150 servers) because the "obvious" solutions were laughable (console per machine, 16 port KVMs, etc.) at that scale. The most *practical*, in their case, was simply carrying a keyboard+display from machine to machine, as required. (but, then again, their machines are set up to be "accessible")

Wow! That's *cool*! Wouldn't work here, though -- no place to project onto that is within reach of the machines in question. :<

But, I'll bookmark that and add it to my "toy" wish list. Thanks!

(hmmm... since the bulbs in my projectors are often $100+, I wonder what the bulb in *that* must cost -- given it's selling price?)

Despite our warmer temperatures, not a blossom on any of the trees. Nasty winter really beat up on them. I suspect they will spend this season just regrowing

*leaves* -- definitely no fruit will be coming. :< (I'll send along some pix)

I've given up on that sort of thing. It's too frustrating. And, you never are 100% sure that a ball isn't making intermittent contact, etc. I'll stick to gullwing parts and their ilk -- where the only issue is having a big enough magnifying glass! ;-)

Bear in mind that that was not a "big bang" project that emerged out of nothing, but an evolution and combination of what had come before. Even BSD itself emerged as the result of gradually replacing portions of the ancestral Unix code base. And they are notable _because_ they represent a great achievement.

Yes. Windows software at that.

--
Andrew Smallshaw
andrews@sdf.lonestar.org

You're describing the original work at *Berkeley* (from the AT&T sources). I was refering to The Jolitz's' "hobby" (a reference to the unpaid "hours and dollars") to bring it to The Masses... (the original "386BSD" releases)