I-Appliance BBS
The Official Source for Internet Appliance Upgrades and Mods

Click Here!
BBS Main List | Sign In | Sign Up | Search | Help | Linux-Hacker.netReply to Thread | Printer |

Home / I-Opener Areas / I-Opener Technical Stuff
TFT Screen Part III
TFT Screen Part 3

New MessageTFT Screen Part III (modified 0 times) turbowsr
Profile
Ok, Part 2 was getting long. Please read Part 1 and 2 completly, thank you.

Part I

http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2485&Idle=0&Sort=0&Order=Descend&Page=0&Session=

Part II
http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2492&Idle=0&Sort=0&Order=Descend&Page=0&Session=

Our goal is to help get TFT LCD screens working on the I-Opener.

04-26-2002 15:07:33

New MessageThe Key - Part #1 (modified 0 times) KneuB_2
Profile | Email
Friday's Lunch

I got one good shot of a small area that shows the MA3 to MA5 pins. NOTE: This small area shot is a half a meg.

each of the pins traceout to a different 8 pin chip labeled D220. What are those D220 chips anyway?


Monday's Lunch

Remove tall pieces to prep the board for another attempt at a full frontal high definition scan.

04-28-2002 01:57:26

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
MA3 goes to 3rd pad from top left of RN11. These are labeled 22R (atleast thats whats on mine). I believe these are resistor packs, the pin on top corrisponds with the pin on bottom. Please correct me if I am wrong. I took some High res. shots of the pack. It seems that no traces come from the opposite pin.

MA4 eventually traces to R169 top pad (Thru the board, comes out beside R162 follows paint to R170 between pads then to R169). Continuity between these two points also.

MA5 eventually traces to RN14 top pin 1 from right. Does not appear to go any where from here.

This is as far as I got. When Knueb gets all the parts removed and the paint off, it will be more clear.

Thanks Knueb.

04-28-2002 09:18:48

New MessageRE:TFT Screen Part III (modified 0 times) KneuB_2
Profile | Email
Update

I got the links to work and added a couple more photos.

The MA3 trace to the third pin on the resistor pack (I hope you are right on that Turbowsr) does not appear to have a trace off the opposing pin as Turbowsr noted. However, even if it did the white paint stripe in that area would conceal it on this board. The absence of a trace from there would correspond to a 800 X 600 DSTN mapping per the Cyberblade spec.

The MA4 trace pops up to pin 1 of the next resistor pack inline(Upper left pin when the cyberblade pad is at the top of the picture). This second resistor pack is the one that looks like I knocked it off it's pads when I removed the Cyberblade chip. If there had been a visible trace connected to the opposing pin on that resistor pack prior to my inadvertent moving of said resistor pack, it's hidden now .... oops

Note that just below the second resistor pack is a 4.7K ohm resistor and on the right of that resistor is a triangle pattern of via's. the via closest to the 4.7K ohm resistor is where MA5 comes back to the top side of the board and ties to pin 1 of a third resistor pack. The opposing pin on that resistor pack does not seem to go anywhere either. Are these "Resistor Packs" really resistor packs? This is not making sense, so what am I missing here.

Board Version

I do not see the yamaha chip on this board (that must be on your new board GWIZAH) but it does have the dual dip switch (SW4) just above the RAM slot, and the netpliance info on the back is;

IA1 M/B REV: J
P/N: DAOI1MB4J2

Would that make this a version 5?

04-28-2002 18:27:32

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Knueb,
Sounds like you followed all the same paths :) Even the triangle pattern. I also noticed that on the MA3 path to RN11 just before it connects, under the paint is a thru hole. It is hard to see, but with Knueb's pic, it is visible. Thru hole leads to a pad.
I took his pick, put in Paint Shop Pro, made it into a greyscale, then worked with the brightness/contrast until I get a good ratio to identify traces. Then take it to 2 color B/W. It makes it easier to follow traces, just color them Red, Blue, etc. to differetiate from other traces and follow point to point.
I ohmed out the Resistor packs, there seems to be little to no resistance from side to corrisponging pin on opposite side. I'm with Knueb on this, it doesn't make since to just end them there.
hmmmm.....
04-28-2002 21:39:18

New MessageWild Guess (modified 0 times) KneuB_2
Profile | Email
I do not know what a resistor array looks like but from diagramsI've glanced at you would not necessarily use the next pin. Could these be resistor arrays or am I way off base here?
04-28-2002 22:05:27

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Looked them up in Digikey (pg. 606, in the book not online). These are lebeled RNxxx on the mobo, for Resistor Network.

http://info.digikey.com/T021/V5/600.pdf
Bottom of the page.
These are not the exact same ones. Different manufactures, manufacturing process change, etc.

04-29-2002 15:07:42

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
D220 or 220
D= +/- 20% (A= +/- 5%, B= +/- 10%, C= +/- 15%, D= +/- 20%) No letter equals max of +/- 20%
22= Ohms
0= the power of ten (0x10=0) (ie 221 = 220 ohms, 222 = 2200 ohms, 223= 22000 ohms, etc.)

So these are Resistor Networks of 22 Ohms at +/- 20% (17.6 Ohms to 26.4 Ohms resistance)

04-29-2002 15:31:58

New MessageRE:TFT Screen Part III (modified 0 times) KneuB_2
Profile | Email
The resistor network diagram shown on the digikey page is what I had mentally associated with resistor arrays. Since all the pins are tied together internally, and two 22 ohm in series resistors is nothing in series with a 4.7K ohm resistor, then the question becomes; Does one of the other pins from each of the resistor networks go to a 4.7K ohm resistor tied to a + voltage trace? (in the case of MA3, the 4.7K ohm resistor pad should be empty) If so, then I would suspect that you have found the high strap resistors.
04-29-2002 18:17:54

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Should have included this thread a while back. This is Programmers thread on trying to get a TFT to work on the IO.

http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2013&Idle=0&Sort=0&Order=Descend&Page=10&Session=

Programmer, with help, did a lot of work trying to get a TFT screen to work. But the end result never seemed to take shape. Codeman supposedly got a TFT to work, but searching the BBS never confirmed this, nor were specifics listed for this.
Thanks to all that worked on this hard before to get us this far.

04-29-2002 20:39:30

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
I guess we aren't the only ones trying to Hack the Trident Cyberblade. Take a look at this:

http://www.acl.lanl.gov/linuxbios/wanted/index.html

The last wanted item!

04-30-2002 17:59:12

New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
Profile
I didn't look at the pdf file, but I thought I could clear up the Resistor Network question. These are exactly like the ones we had to use for the CF modification, so I'm rather familiar with them now. They have 8 pins, the pins accross from each other are the individual resistors. They're "Isolated" resistor networks, meaning the 4 individual resistors in the package have nothing to do with each other. Their pins are not connected. 220 is 22ohms, right. I asked a lot of questions in other forums regarding the purpose of these and ultimately was told that they are necessary in high-speed traces to overcome the capacitence between the traces which could cause erratic signals. They aren't (typically) used as pullup or pulldown resistors.

If it makes it simpler to visualize and understand, just imagine that the resistor networks aren't even there, that the traces just continue in a straight line under the RN "D220".

KneuB_2 this is great!

04-30-2002 18:09:09

New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
Profile
oh...
About picture formats... I scan at high quality (1200dpi or so) then store it in jpg format. I was concerned at first about lost quality, but it still looks great. There is a LOT of "same-color space" on these high-res scans which makes it perfect for jpg format. However, when changing these to B/W like Turbowsr does, they best be in bmp or tiff or similar format. That's because changing the trace colors later on using floodfill will store the clean pure colors that you floodfilled to "jpg-colors" where even a huge space of only one color will be filled with speckles of different (but similar) colors. This is hardly noticeable, but when you floodfill again they won't fill completely because the single-color traces are really made up of dozens of similar colors. I suppose if it's all B/W originally anyways, and no to similar colored traces are near each other, then you could set the floodfill setting to 10% or so and it would still work in jpg. Another option is to put it in 16 color bmp format (before changing trace colors, and after changing to B/W), since you probably only have a few different colors for traces. That will make the file a ton smaller, as opposed to 256 (or 65536!) color.

Hope that makes sense.
Eric

04-30-2002 18:18:55

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Thanks Eric,
Good info on those resistors.
I do save the B&W under bitmap. Then go in and color fill each trace.
Here's my steps under Paint Shop Pro V5 (I don't use any others, so I don't know)
Open Pic
Colors==>grey scale
Colors==>negative image
Colors==>Adjust==>Highlight/Midtone/Shadow==>set Highlight up all the way, both Midtone and Shadow down all the way
Colors==>Decrease Color Depth==>2 Colors
Image==>Mirror (so it can set under topside and line up with thru holes)Believe me, this makes it easier to follow traces.

This worked for me. Now it should look like a schematic. If you know how to do this in any other program let others know. Paint Shop Pro is as powerful and as good for the home user as Adobe (only other REAL competitor) but cost about $300 less. You can also download it and try it out for 30 days full functioning (at least it was that way) from http://www.jasc.com

05-01-2002 16:41:15

New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
Profile
Sounds like you've made it a science.... Did you ever think about buying some inkjet transparancies to print it on? Then you could align the traces on top of the real ones ;)

You could also use the Gimp, which is free and very powerful. I'm not sure if there's a Windows version or not... (I use Linux...)

Eric

05-01-2002 18:47:16

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Eric,
Already working on transparency, and yes there is a Windoze version of Gimp. A search on Yahoo will get the address.

Am in talks now with lead of LinuxBIOS (or FreeBIOS) to help get MVP4/Cyberblade i7 working with LinuxBIOS. Are the Cyberblade specs available around here somewhere :), Legal? The LinuxBIOS team want to know, they want to keep it legit.

Open Source, User Configurable BIOS, faster boot into Linux (or with some work xBSD). I use Mandrake. I also have Debian Linux, FreeBSD, and BeOS. Oh, and this crappy OS. I REFUSE TO UPGRADE TO XP, Win98 works just fine. I only have to reinstall ONCE a year ;)

TFT Screen, Harddrive, CPU upgrade, Open Source BIOS, all on a cheap laptop clone.

05-02-2002 15:13:47

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Ah Ha, found MA5. It comes from UNDER RN14, Under paint, between R179's pads, to R178 top pad. So MA4(R169) is 0 (no 4.7k pullup), MA5(R178) is 0. MA3 goes to R168, with the 4.7k pull up. Thats MA5-3, 001, or 1280x1024x18+18bit TFT. This is assuming 1 is pull up. If we assume pull up is 0, then we have 110, which is 800x600x16bit DTSN. 110 sounds more correct.
05-02-2002 16:08:35

New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
Profile
Just need someone to remove R168, and add 4.7k to R169, R178. Get Wild_Pencil's BIOS (mentioned in Part 1), and a TFT to test.
I am trying to get a TFT, I have spare 4.7K's around, BUT is WP available.
05-02-2002 16:14:05

New MessageStripping the Board (modified 0 times) KneuB_2
Profile | Email
This Past Week

I have spent all lunch hours and a couple of evenings stripping the board. At this point I now have all the tall pieces off and it’s now ready for the next max resolution scan.

Turbowsr, it’s good to see that you took the time to find the high strap resistor pads, I hope you get an opportunity to find the pins (or maybe I should say test points) for the two missing blue signals. I shall continue to concentrate my efforts on the stripping and scanning of the MB. (my TFT has not arrived yet either) I shall also try to post my progress a little more frequently.

The Plan of Attack

  • Get a max resolution scan of the top side of the board with just the little pieces on it for use as a comparison reference after all the little pieces are removed.

  • Remove the little pieces and sand off the rectangular white epoxy ink pads (I’ll leave the lines and text in place for reference if I can).

  • Get a max resolution scan of the top side without any components.

  • Evaluate the scans to determine if it will be advantages to sand off all the markings so that the top side is void of everything except traces.

  • Figure out some way to make this available to everyone.


    Lessons Learned, Observations, & Theories


    Surface mount canister capacitors: Those itty bitty solder points on opposing sides of the capacitor actually extend about 1/3 or the way under the capacitor from each direction. One solder iron can easily liquefy the solder on one pin, but even the slightest tilt of the canister will cause the opposing side to tear up the trace. To remove the capacitors without tearing up a trace requires melting both solder points at the same time. I had to track down a second soldering iron to remove the capacitors.


    Removing multi pin through hole components with a paint stripper heat gun; Making such an attempting in an office environment is frowned on. Doing it on the loading dock solves the office environment concerns but any little breeze significantly reduces the component removal rate. I was averaging about 20 minutes per part and certain pieces such as the parallel connector and the processor socket I simply was not able to remove in such an environment. It would take a lot of lunch hours to remove even just the tall components at that rate. Constructing some sort of wind break is a must. Having a significant portion of the engineering department huddling around trying to figure out what you are doing is a close second.

    The up side of using an 11 AMP B&D paint stripper in this manner was that it doesn’t burn the board even when left in extremely close proximity for an extended period of time. In one case I actually let it “sit” “On the Board” for about 15 minutes with one edge propped up about an 1/8 inch (so the air could flow) without burning the board. OK, maybe this particular paint stripper is a little marginal to be doing this in a breeze, but it did work.

    Post Removal: As slow as the paint stripper was at removing through hole components I decided to try another approach for the posts. The approach that worked well was to use one of those butane micro torches (cigarette lighter on steroids) that Rat Shack dumped for 99 cents about a year ago (it was in the Deals Deals Deals thread, I hope you got one) and a piece of a pop can with a hole poked through it for a heat shield. I put the flame tip directly on the post from the back side of the board, and as long as I did not slide the heat shield off center it did not burn the board. The posts came out very quickly.

    NOTE: The application of the torch without a heat shield caused the surrounding board area to burn almost instantly (this was determined in a practice test conducted on a separate board, “before” using this method on the i-o). Also, when applying the torch flame through a hole in a flattened pop can the heat causes the aluminum to curl, at that time you must be very careful to hold the edges of the hole down, because, if one edge lifts, the flame will burn the board under the edge almost instantly (this was determined on the i-o #$&^%&). Fortunately, the back side of the board had already been photoed.

    Suggestion: To use a pop can as a heat shield, use the bottom of it. Put the hole through the center of the bottom and trim the side of the can down to the point that the edge of the hole and the edge of the can both sit on the board. Because of the shape of the can bottom it resists curling and is a lot easier to work with.

    BTW: The inside for the micro torch is actually a refillable cigarette lighter (looks very similar to a Bic) but the nozzle in the micro torch controls and concentrates the flame a lot better than a cigarette lighter by itself could ever do.

    Removing Components In Your Oven This was definitely the most productive source of last weeks “lessons learned”, “observations” and “theories”. But I’ll have to write about it later.

  • 05-06-2002 00:34:57

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    Whew, I got tired just reading that, KnueB :) I'll be looking for those, just as soon as my new multimeter gets here. Mine is flakey, they don't take kindly to being dropped very often :)
    This has now become 1 of 3 projects. They are all hobby projects, so it doesn't matter in what order I work on them.
    Personally, I found the 20 min's in the oven was easier than what I ever had to do with a heat gun. You just have to keep an eye on the board (my "test" board blistered, but not very bad, you could still use it to follow traces). I used a cookie sheet with a wire rack for cooling bread, etc. with aluminum foil on the rack to catch parts so they would not burn. The wire rack had four prongs that held the board up so that parts fell off.
    Wow, almost 100 posts just to figure out 3 resistors :)
    05-06-2002 15:06:52

    New MessageStatus (modified 0 times) KneuB_2
    Profile | Email
    Status

    A friend has taken the board home tonight to do the max resolution scan.

    turbowsr Your point is well taken. All "Lessons Learned" missives will be placed elsewhere. I will keep this thread short and, hopefully, sweet.

    Tomorrow

    Review The scans.

    05-06-2002 18:26:00

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    I am taking all pertinent information from the TFT Screen's threads and putting it up on my website. If you would like to add anything, send me the html, and I'll give it a page. I will not put anything legally questionable tho. The page will be centered around TFT screens, initially, then other working "upgrades" will be added.
    To add info email me at yahoo dot com.

    \i-opener.turbowsr.com

    Nothing there yet, but will post images and intructions to MA5-3 soon.

    05-07-2002 20:22:29

    New MessageReference SCAN (modified 0 times) KneuB_2
    Profile | Email
    Today's Lunch Hour

    Reviewed the max resolution scans. They're great, resistors are easily readable and everything is in sharp focus, YES !. The smallest file size is 90 Meg however.

    Turbowsr; Is your web site big enough to host one or two of these by any chance?

    Up Comming

    Stripping the remainder of the parts off, is next.

    05-07-2002 20:59:34

    New MessageRE:TFT Screen Part III (modified 0 times) JAK1
    Profile
    Back in January, I was working with Wildfrog on identifying hurdles for TFT functionality. A new job and new baby pretty much put my work on hold. Attempts to contact Wild frog have gone unanswered. Wildfrog, you still out there?

    My apologies for not making public this info sooner. From an email I sent Wildfrog:

    "I haven't discovered what works, but I have deduced what does not. Although there are 3 locations for the pull-up resistors on the board, none are populated, which means it is defaulted to TFT 1024 x 768 x 18-bit. If these locations were used for configuring the display, 2 locations would need 4.7k resistors for DSTN 800 x 600 x 16-bit. In case you are curious the locations are R178(A5), R169(A4), and R170(A3). These pins are shared with the address lines of the SDRAM. I didn't post this because I wanted someone else to confirm this info, was hoping you could do this. By looking at http://www.jedec.org/download/pub21/4_5_6-R9.PDF I was able to discover the locations of A5-3, A5 : pin-34, A4 : pin-32, A3 : pin-30. With this info I started ohm'ing out locations near the relevant corner of the northbridge."

    My info on A3 conflicts with other findings; I'm saying it goes to R170, turbowsr says its R168. Wildfrog said he confirmed my findings but now I have my doubts. Can someone confirm this?

    With the locations unpopulated, we assumed that the bios set the registers and focused our attention there.

    JAK

    05-07-2002 23:54:00

    New MessageRE:TFT Screen Part III (modified 0 times) GWIZAH
    Profile | Email
    Wow, you guys are really taking this to the next level...Here I sit with a TFT grafted onto an I-O's dstn screen trying to connect it directly to the DSTN connector on the mobo, boy do I feel like a putz now :( I just want to commend all of you for your hard work once again. KneuB-2, I *may* be able to host some of those files on my site. 90 MB might be strectchin it a bit. Email me in my profile.
    -wheee!-
    05-08-2002 08:25:52

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    KnueB,
    90 Meg is too big for me to host, I only have 25MB. I'll only be able to show pics, but not be able to host large files. Getting the site up and running will be slower than I expected because....

    I got hit by lightening, well my modem did, took 2 ports out on my switch, an NIC card and the Modem/Lan card I was using. The lightening went thru the switch and hit a dorm fridge it was setting next to. Left a black spot, on both the fridge and Switch. D@mn !t.....:( This just happened to be my proxy. Soo now I have to reconfigure my whole net work (I have 4 PC's connected here). Also took a PC Card slot on my Laptop.
    I know, should have had a surge protector. It was there, but the RJ11 was blown from a previous lightening storm. Like I said before, these lines are crappy around here. I got too far 'Away from it all'

    Jak1,
    I will recheck my findings just as soon as my new multimeter gets here. I checked for continuity with my old one before it had an unexpected fall. When they are checked, there should be a very small difference between the MAx pad and the Rxxx pad. Only a 22ohm resistor between them.

    Since the pullup resistors have been found. Next project is to find the missing 2 blue bits, then I'll work on finding the other 6 for 24 bit TFT.

    Anyone interested in putting a vga adapter on? If you do a search on Ebay for PCCHIPS VGA, you'll find an adapter ment for a PCChips AT board that relocates the VGA where ever you'd like. They are running for about $8. I am going to try to put on a DVI adapter with 24 bit, then hook up from there.

    05-08-2002 16:50:10

    New MessageRE:TFT Screen Part III (modified 0 times) JAK1
    Profile
    turbowsr,

    I think you are missing my point. If all three config. resistors are missing then the bios must be setting the registers.

    During power-on and system reset, pins MA5-3 are used as config lines, but afterwards they are data lines for the SDRAM, page 6-7 in data sheet. By looking at http://www.jedec.org/download/pub21/4_5_6-R9.PDF you will see that address lines 5-3 correspond to pins 34, 32, 30 respectively (the Cyberblade spec sheet shows 168-pin SDRAM so that schematic can not be used). I was able to test for 0 ohms between the SODIMM pins and resistor pads.

    JAK

    05-08-2002 21:21:01

    New MessageRE:TFT Screen Part III (modified 0 times) GWIZAH
    Profile | Email
    turbowsr, you mean adding a VGA-out? Hasn't this alreay been accomplished? I have an I-O currently wired with a VGA adapter from a PCchips mobo. As a matter of fact, that's the onlw way its working since the DSTN has no cable.
    05-09-2002 07:34:41

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    GWIZAH,
    Na, I was just letting anyone that wanted VGA parts know I found some on ebay. Thats all :)
    05-09-2002 15:24:45

    New MessageConfirmation (modified 0 times) KneuB_2
    Profile | Email
    This Weeks Status

    JAK1 is correct, the high strap resistor pads ARE: R178(A5), R169(A4), and R170(A3), and they are all un-populated. The lines comming off the resistor nets all pass over vias before reaching the high strap resistor pads. The VIA's all head off toward the RAM, but I did not get the opportunity to verify which pins.

    Also, a continuity check on this stripped board verifies that the "isolated" ends of each of the high strap resistor pad pair (say that one fast five times) are actually tied together internal to the board.

    Question

    Since the BIOS is being used to setup the registers to drive the DSTN is it possible to use the high strap resistor pads to setup the registers to drive a TFT or will the BIOS over ride?

    BTW: Plucking these little resistors etc. leaves a major solder mess no matter what I've tried. Is there an easy way to clean all this solder up? It will take a lot of solder wick.

    05-10-2002 18:19:28

    New MessageRE:TFT Screen Part III (modified 0 times) JAK1
    Profile
    KneuB_2,

    The bios is probably over riding the config resistors. I believe software could be written to reprogram the registers after the bios has done its thing. The problem for me is that I'm not much of a software guy; hardware hacking and reading data sheets is extent of my talent. If there is anyone out there who knows how to access and change these registers, we could use your help.

    Back when I was working in the circuit board industry we would occasionally need to strip a board. After spraying some flux onto the board, we would dip the entire board into the solder pot of a wave solder machine. After a few seconds, we would remove it and bang the edge of the board on the edge of a table. A good majority of the parts and solder would fly off and hit the floor. Minor cleanup was done with solder wick. When the tech. first told me about this method I was a little skeptical but in the hands of a "trained" professional it worked well.

    The key is fluxing the entire board and then heating the entire board to about 210-220 C, and, of course, wacking it.

    JAK

    05-11-2002 09:18:01

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    Correct. I confirmed that MA3 IS R170 NOT R168.
    05-11-2002 19:03:57

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    Knueb,
    I looked at pics you posted. I don't understand the two missing blue pic you have. All the pins needed are in the square area of G1:M6 (this gives 24 bit). The pic you have does not show this area, but the are designated for CPU. Plus being a lot dark too.
    Can you post the area of G1:M6?
    Thanks.
    Wild_Pencil built a bios with TFT support, but he moved, and apparently has not come back here yet.

    turbowsr

    05-11-2002 19:12:26

    New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
    Profile
    I must be losing it... I know KneuB2 posted some links to pictures he took somewhere above, but I can't seem to find the links now. You never did manage to find a home for the big files, did you?

    I have a webserver sitting in my room which they could be put on, but there are some issues with this:
    A) It's my personal computer, so only on during the day (usually from 6:30AM to ~1AM)
    B) Sometimes I'll have to restart my computer or will have it off...
    C) People with @Home are screwed. For some stupid reason @Home now blocks communication between computers on the @home network (at least, between my and my friends' computers)
    D) Since I'm on @Home, I can NOT have more than one person download it at a time or else things will get very slow and they might notice that I'm sending alot of data...

    I have a few friends who have servers at the UW who might be willing to host... I haven't asked them yet.

    Did you try saving them as jpg files? I'm serious, I took 1200dpi scans which were 20+MB in bmp format and saved them in jpg format as only 1MB. Of course the quality isn't as accurate and the colors may be slightly off, but at that resolution there is a ton of "wasted" pixels making it perfect for jpg format. As I said before, we should probably convert them back to bmp before modifying them as Turbowsr suggests...

    Eric

    05-12-2002 23:45:55

    New MessageRE:TFT Screen Part III (modified 0 times) KneuB_2
    Profile | Email
    Jak1

    I'll give it a try tomorrow evening, I hope burnt PAM will wash off the board.

    Turbowsr

    I'll bring the CD home from work tomorrow night and post some more of the photos.

    Finatronics

    I'm still looking for somewhere to host them, and yours is by far the best offer I've heard so far.

    The scans I have not tried to save in JPG yet, but the photos from the camera I saved in four different formats. The jpg format did indeed make a difference but not as much as the pict format did.

    Sorry for the rushed response tonight, but that's just the way it is tonight.

    05-13-2002 22:01:19

    New MessageProgress (modified 0 times) KneuB_2
    Profile | Email
    Last Knight

    Sprayed the board with PAM to try and keep the parts from re-adhearing, baked it and sent the parts flying ... it worked great. The Pam didn't burn but it did bake on, it's a pain to scrub off.

    Tryed to post some more photos of the cyberblade area but had difficulty establishing an ftp connection, so I e-mailed a couple really small pic's to turbowsr until I can get the ftp problem resolved. I hope that works for you for now turbowsr.

    Experimented with saving TIF files as jpg files. Neither MS photo editor nor MS VISIO could open a 90 meg TIF file (VISIO surprised me on that one). MS Photo draw could do it but it trashed the quality so bad that traces could not be followed at any magnification. The real pleasant surprise was a little 500K software application called Kodakimg.exe It seems to get installed by default with Win 2000 pro, but you have to search for it to find it since it doesn't show up under "Programs". Anyway it not only provides far .... FAR .... better viewing capbilities of TIF and jpg images than any of the microsoft software, it also maintains nearly all the clarity when converting a TIF to a jpg. In addition the 90 meg TIF was reduced to a little over 6 megs using the "MINIMUM" compression setting, now that size is postable.

    Today's Lunch

    Noticed a couple dozen tiny pieces had simply re-located themselves, including two that had obviously bounced around the inside of the oven first since they had re-located onto the back side of the board. Spent the lunch hour removing them with a solding iron.

    Tomorow

    Start sanding off all the white epoxy ink pads to expose the traces. If I can get them all done by the end of lunch my co-worker will have the final max resolution scans on Friday.

    05-15-2002 22:56:30

    New MessageRE:TFT Screen Part III (modified 0 times) 02U2
    Profile
    I learned something interesting today...
    Just passing on this tidbit for whatever it may be worth?

    There is an EEPROM in the inverter which stores its supported LCD type ID code.
    If you replace a LCD with one of a different brand or use a new inverter, the ID information in the inverter EEPROM should be updated.

    http://www.gruftie.net/ibm/servicepac.de.ibm.com/eprmhtml/epr2f/h7084.htm

    05-17-2002 16:35:51

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    Thanks U2O2,
    We are not using the inverter board on the IO, (At least not as I understand it)
    We are connecting a TFT with its own inverter board to the digital video output of the IO board.
    This is how I plan on doing it. At least this is how it should be done. I would not advise using the IO's inverter for anything other than the same DTSN screen.
    Good information non the less.

    turbowsr

    05-17-2002 18:25:14

    New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
    Profile
    Whoops! I completely forgot about this thread... it's been a hectic week. KneuB_2, if you're still interested in using my computer as a host, I can give you an ftp account and a place to upload to. Email me for more info... a3453453452@yahoo.com

    Yeah, the compression setting for JPGs really affects the size. If I use the fine setting on my digital camera the pictures are 600-800k, but if I use the normal setting (same resolution) they're only 300-400k. I can't tell the difference between the way they look.

    For future reference, my site is at http://finatronics.dyndns.org:8123/
    As I said before, there might be a huge chunk of the population who can't go there because @Home has put some sort of block on the routers so @Home members can't talk to each other... It's also my home computer, so it isn't on after I go to bed. Your best bet when downloading or uploading is to start early in the morning a little after 6:30AM Pacific time. That way, if it takes all day, it should probably be done before I go to bed around 12-1AM. I don't have any way to monitor or limit how many people are downloading (that I'm aware of), but if it seems to go extremely slow (less than 5KB/s on a 56k modem, or less than 10Kb/s on a fast connection), it's probably because too many people are using it! Wait until later... Also, I recommend using a download manager so you can resume a download if I have to restart or turn it off for the night.

    OK, that's all I can think of to put in the disclaimer.
    Eric

    05-18-2002 14:20:56

    New MessageRE:TFT Screen Part III (modified 0 times) Finatronics
    Profile
    Oh, I meant to ask about the inverter question... I've never heard of an LCD that has the inverter built in! CCFT tubes are CCFT tubes. Most LCDs just have two leads coming directly from the tube and you plug it into the inverter built into the laptop or whatever. I think that IBM thing is referencing their specific laptop LCD display modules including everything that's built into the housing of the display... which probably includes the inverter. If the display is the same size as the IO display, I wouldn't see anything wrong with plugging the TFT's CCFT into the IO. If the display is larger, then the tube will also be larger which would draw more current... I wouldn't recommend to someone else to use the built in inverter to drive a bigger fluorescent tube, but I probably would do it, and I doubt you'd run into trouble.
    No promises ;p
    Eric
    05-18-2002 14:26:03

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    Well I didn't exactly mean built in to the LCD. I ment built in to most LCD Units. Not too often can you get just an LCD. You usually get LCD, backlight, inverter (sometimes not inverter). I would recommend buying the whole kit 'N kabootle, that way it should be a straight shot from mobo to screen.
    Its a good idea to keep the inverter board in mind when doing this upgrade.
    05-18-2002 22:11:53

    New MessageRE:TFT Screen Part III (modified 0 times) GWIZAH
    Profile | Email
    My TFT is the whole kit n' Kaboodle, meaning it has a built in invereter and two cathode lights. I was under the assumption that most people who are interested in this hack would have the exact config. Though, when we make assumptions...

    Good job so far guys, still lurking, learning, watching... If you guys need anything, let me know.


    -wheee!-
    05-18-2002 22:28:04

    New MessageRE:TFT Screen Part III (modified 0 times) turbowsr
    Profile
    No More Post's here.
    Please continue on to Part IV

    http://www.linux-hacker.net/cgi-bin/UltraBoard/UltraBoard.pl?Action=ShowPost&Board=technical&Post=2516&Idle=0&Sort=0&Order=Descend&Page=0&Session=

    Thank you to all that have brought us this far.

    05-19-2002 17:41:39

    Reply to Thread | Printer |
    All times are PSTPowered by UltraBoard v1.62



    Copyright © 2000, Netmake Inc. All Rights Reserved.
    See Terms and Conditions for more information.




    i-opener opener laptop notebook computer help drivers dll free windows dos repair fix linux mac macintosh 2000 95 98 nt pc configure hardware software sound video netscape explorer network networking lan wan software cmos fat bios printer card mouse modem ide scsi cd rom controllers scanner tape hard drive cgi scripts source code mp3