Shop OBEX P1 Docs P2 Docs Learn Events
Propeller Project Board USB - VGA issues — Parallax Forums

Propeller Project Board USB - VGA issues

MoskogMoskog Posts: 554
edited 2013-10-07 18:03 in Propeller 1
LA6WNA and I are working on a project that will include several Spinnerets and Propeller boards and a Visual Basic application. The idea is to monitor the state of several ventilation systems placed on different facillities. Older systems that was made before the idea on remote controlling by the Internet. Soon we will have a test system ready and that will be placed on a vent system at the nearby school. The test system will be monitoring only but we plan to make a two-way communication later to be able to make a full remote control of the systems.

We plan to use several Propeller Project Boards USB (#32810) because they are nice and clean and simple to connect to the main PCB's. But we ran into a VGA issue when transferring the project from the PPD-board to the new project board. We plan to be able to use a standard VGA-screen to monitor values on location in addition to the Spinneret but during the tests the VGA screen randomly gets black like it turns off. Usually it gets black after a short, blue blink after reset but sometimes it gets black later in the program code and won't turn on again. We are using the standard VGA_text.spin and VGA.spin without any special effects, just a blue screen with white chars.

The PPB USB board is fed by a 12V 1A supply and the VGA connector is the slim type bought from Parallax. The only thing different,unlike the Professional Board, is that this new board are using different resistor values at the VGA output.

We have tested the board with a VGA-safe code too but with same result, screen turning black. LA6WNA also faced the same issue recently with another PPD USB board, on a different project.

So now we wonder if this is a problem related to the board itself as we can't find anything that could possibly be the reason, are there other people here that has used the same combination, the new board and VGA?
«1

Comments

  • Ken GraceyKen Gracey Posts: 7,392
    edited 2013-03-25 09:47
    Moskog, I've alerted our Tech Support staff to this post for a reply.
  • RaymanRayman Posts: 14,662
    edited 2013-03-25 10:29
    I think I might see why... Pin9 of the VGA connector is sometimes sensed by the monitor and the monitor will go to sleep if it's not there...

    Looking at the schematic, I don't see anything going to pin 9... Maybe running a jumper wire from +5VDC to pin9 on the vga connector will help?
  • MoskogMoskog Posts: 554
    edited 2013-03-25 13:11
    Rayman!
    I took a look at the schematics of the Propeller Professional Development Board and it seems like pin 9 is connected to 5V there. I also took a look at the Propeller Proto Board (#32212) which I've been used for several projects without any kinds of VGA trouble. Pin 9 is connected to +5V there too. So maybe we have something here.
    So, beside the different resistor values on this new board we also have this missing +5V on VGA pin 9.

    Ken!
    Look forward to hear from the Tech Support team.

    Won't be able to test with a 5V jumper until tomorrow, but I will post the result!
  • MoskogMoskog Posts: 554
    edited 2013-03-26 00:48
    OK, testing the VGA with connecting +5V to pin 9. The monitor still gets black.
    It seems that the monitor remains awake, not sleeping, as the green button remains green during blackout.

    The monitor is a Dell model nr 1905FP.
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 02:48
    In that case, it sounds like a software issue. Unless... Are you saying the exact same code works on these other boards but not this one?
  • MoskogMoskog Posts: 554
    edited 2013-03-26 04:17
    Rayman wrote: »
    ...Are you saying the exact same code works on these other boards but not this one?

    Yes, the project was made and tested on the professional board without any problem.

    We also tested some other files thats been used with the older Propeller Proto Board (#32212) without any problems, and faced the same issue.

    LA6WNA did mention this problem in an earlier thread where he used this new board with his heat-pump project. (In the Project forum)
  • Toby SeckshundToby Seckshund Posts: 2,027
    edited 2013-03-26 05:20
    Pin 9 on the 15 way D type is there to supply the EEPROM in the monitor, so the "PC" can read what it is even if the monitor is off ( pins 15 & 12 carry the I2C signals). Sometimes Pin 9 is simply grounded with all the others in the middle row and this would short out the +5V rail, a lot of the older VGA cables have that pin missing.

    Does the monitor accept the signalling for that display mode, given on the sync pulses timings and their polarities. Most modern monitors do their signal level clamping on the tail end of the syncs so if the pic content is too early then that could cause problems.

    One of my "favorites" is getting the H & V syncs the wrong way around!

    Alan
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 05:36
    After looking at that other thread, maybe what you have is a power problem...
    Could it be that your regulator is overheating and shutting down after some amount of time?

    Is it just the VGA output that stops? Or maybe the whole system shuts down?

    Maybe if you add an LED to the setup, maybe make it blink or something, you could tell...

    Or, just measure the 3.3 volt rail when the monitor goes black...
  • MoskogMoskog Posts: 554
    edited 2013-03-26 06:14
    @Toby Seckshund
    Does the monitor accept the signalling for that display mode, given on the sync pulses timings and their polarities. Most modern monitors do their signal level clamping on the tail end of the syncs so if the pic content is too early then that could cause problems.

    Same .spin-files have been tested on the older boards using same monitor without any problems.
    One of my "favorites" is getting the H & V syncs the wrong way around!

    I guess if that happend the display would never act normal at all, now it usually start up at reset and get black after one second, or stays normal for a time or doesn't start up at all.

    @Rayman
    Is it just the VGA output that stops? Or maybe the whole system shuts down?
    Maybe if you add an LED to the setup, maybe make it blink or something, you could tell...

    It is just the VGA, the main loop is up and running like normal. We do have a loop-LED that blinks and the Propeller keeps feeding the spinneret.
  • kuronekokuroneko Posts: 3,623
    edited 2013-03-26 06:21
    Moskog wrote: »
    It is just the VGA, the main loop is up and running like normal. We do have a loop-LED that blinks and the Propeller keeps feeding the spinneret.
    IIRC the VGA driver used by VGA_Text uses a live parameter array in the VAR section (and keeps updating itself). Could it be that something else is corrupting that memory? Have you tried a fire-and-forget VGA driver? It's unlikely to be an issue but you never know until you check ...
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 06:44
    What happens when the screen is black and you restart the program?

    What happens if you turn the monitor off when it is black and then turn it back on?

    Do you have another monitor you can try? Do you have a scope or multimeter?
  • MoskogMoskog Posts: 554
    edited 2013-03-26 09:10
    vga_issue.jpg


    Pic showing the bottom side of the Propeller board how the VGA connector is connected to the Propeller.
    Rayman, if the screen is black and we restart the program it usually blinks blue or stay blue for a short time before getting black. Sometimes it remains black and sometimes it turns on getting blue and stay blue. Second question, not sure we tested turning off/on the monitor during program-run today. Third question, sorry we haven't tried another monitor today, but LA6WNA used another monitor on his heat-pump project, different but still a Dell. Facing the same problem.
    And kuroneko, its hard to believe this is a software problem as the same code ran pretty well on the Professional Developing board using the same monitor.
    864 x 619 - 546K
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 09:24
    I'm curious, if there's no 5V on the board, how did you connect +5 VDC to pin 9?
    Sure it wasn't 3.3V you put there?
  • MoskogMoskog Posts: 554
    edited 2013-03-26 09:32
    vga_issue2.jpg

    Good question, we took 5V from the main board. We have a 1,5A 7805 there for other stuff.
    837 x 654 - 446K
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 09:45
    Try loading a simple VGA code, like VGA Text Demo or something and see if the same thing happens...

    You may also want to double check your wiring and check continuity with a multimeter...
  • LA6WNALA6WNA Posts: 138
    edited 2013-03-26 10:34
    What happens if you turn the monitor off when it is black and then turn it back on?
    Yes, I tried to turn off and on the monitor with the power button. When monitor is in blackout state, it remains blacked when turning on the monitor again. When it is in normal state, it remains normal afther a power off and on again.

    One thing that maybee not come clearly out of this tread is that when monitor turns up running normal (it sometimes do so), it can stay normal for a long time.
    In my heat pump project, VGA blackouts could happen when a relay powered on or off during normal program run. We do not have any relay in this project yet, but it comes.:smile:
    To be 100% sure of this is not a software problem, we`ll try to load up only a VGA demo to see what`s happening. Have`nt tried that yet. We`ll post the result of the test.
    Another thing; the new PPBoard USB do have 130 ohm "pull-downs" at the RGB lines to the VGA output. Cant see them at older boards or at the proffesional develop board. Could this be a reason for older VGA monitors to not respond normal? Also 510ohm serial resistors instead of 470ohm at the older boards. Just a shot in the blind......... Feel we`ve tried many possiblities without finding the real bug yet, but I`m sure we`ll find the "burried dog" at last....
    Thanks a lot everyone for good comments and for trying to help "Two good neighbours".:smile:
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 10:49
    I don't think the resistors matter, the 130 Ohms is probably just a better way of doing what was done before.

    I have seen a bad solder connection on a sync signal giving similar effects as you describe.
    I think a bad ground connection can also do it.
  • LA6WNALA6WNA Posts: 138
    edited 2013-03-26 13:12
    Just tested a third board of same type (PPB usb new edition). Loaded up the a vga demo spincode and got the same problems. Also tried with a second vga demo, but same results. This board has NOTHING else connected but a vga output. I`ve also measured pin for pin from vga connector to pin 16 to 23 on propeller. All connected with the right pins and I got the right R-values. No shortcuts etc. My idea-bank is empty for the moment...
    Also noticed that pin 11 on vga connector is grounded. This don`t happens at the old boards. Could this affect the monitor performance..???
    I can either not see that my connections should be wrong.??
    Just like this:
    VGA PROP
    1----240R----23 'forms R signal
    1----510R----22 'forms R signal
    2----240R----21 'forms G signal
    2----510R----20 'forms G signal
    3----240R----19 'forms B signal
    3----510R----18 'forms B signal
    13--240R----17 'H-sync
    14--240R----16 'V-sync

    At all, 8 wires. See the picture from Moskog. Is this as simple that I`ve done something wrong here???

    I`ve also tried interchange H and V wires, but then the monitor did`nt start at all. I put them back again.
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 13:31
    How about ground? Do you have a ground connection between PPB and the Propeller??
  • LA6WNALA6WNA Posts: 138
    edited 2013-03-26 13:44
    Not sure what you mean..?? Propeller is grounded to the board from factory. VGA connector ground connections also made from factory. We`re now talking about one single PPB of the new type, without any other connections or daughterboards. Specially hooked up for this test. Board is brand new.
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 13:54
    Ok, I thought you were plugged into that bigger board... I have to say, I think I'm out of ideas...
    If the regulator isn't overheating, not sure what else could do it...

    Think I'd try 5 or 3.3 V on Pin 9 again...

    If you have a scope, it might help to look at the sync signals and make sure they are there...
    You could just look at the DC level with a multimeter too, that might tell you something if it changes when the screen goes black...
  • MoskogMoskog Posts: 554
    edited 2013-03-26 13:56
    How about ground? Do you have a ground connection between PPB and the Propeller??

    Guess you mean ground between the PPB and the main PCB, yes, the grounding is taken good care of.
    I tried an LC monitor tonight to see if there were any difference from the Dell, but with same result.

    As LA6WNA now has tried the third Propeller Project Board and with plain VGA demo and with nothing else connected and still face the problem I think there is nothing more for us to do.
    Nothing else then await for the Tech Support Team to arrive.
  • jmgjmg Posts: 15,173
    edited 2013-03-26 14:19
    LA6WNA wrote: »
    One thing that maybee not come clearly out of this tread is that when monitor turns up running normal (it sometimes do so), it can stay normal for a long time.
    In my heat pump project, VGA blackouts could happen when a relay powered on or off during normal program run. We do not have any relay in this project yet, but it comes.:smile:

    Are all cut-to-black events related to something like a Load switching ?
    Does it usually always start ok from a cold start ?

    If the system is largely stable, and only rarely goes out to lunch, you need to focus on 'how rarely' ?

    If you have a Multimeter with Frequency Counter mode, leave that connected to the H sync, or better still, a scope.

    I'd try load nothing but a VGA test pattern into a board, and see what your longest Up Time is ?
  • Kevin CookKevin Cook Posts: 159
    edited 2013-03-26 15:20
    Hello All,

    I have good news! “Propeller Project Board REV B” has a fix for this issue. I’d be more than happy to send you a REV B once they arrive next week. Just send me an email (kcook@parallax.com) with your information and I’ll be sure you get the REV B at no cost.

    We discovered a trace issue that made one of the Propeller decoupling capacitors to be ineffective and therefore, on higher definition and some standard VGA video, the reliability was not there. If you were to look at the Cogs running the VGA signals you would see an “out of phase” effect due to this issue.

    You should be able to run some simple VGA with the current board. I’m very sorry about the problem!

    -Kevin Cook
  • LA6WNALA6WNA Posts: 138
    edited 2013-03-26 16:00
    BINGO!!
    Thanks a lot, Kevin. Now I can stop spending my time banging my head in the wall...:lol:
    I`ve scoped the sync pulses tonight, and it seems like the V-sync is getting crazy when the phenomen arrives. I also tried connect an other VGA connector with 470 and 240 ohm resistors and no pull-downs, just like the old boards, but negative results...
    So, when your post arrived here I tought; YIPPI.....
    We`re really grateful for your support. I`ll send you an email directly tomorrow. Its midnight here, so I guess I go seeing into my head for a while.
    :lol::lol:
    Again: THANK YOU ALL FOR NICE REPLIES AND SUPPORT.
  • jmgjmg Posts: 15,173
    edited 2013-03-26 16:01
    Kevin Cook wrote: »
    We discovered a trace issue that made one of the Propeller decoupling capacitors to be ineffective and therefore, on higher definition and some standard VGA video, the reliability was not there.

    Is there a user-patch solution to this problem ?
  • Ken GraceyKen Gracey Posts: 7,392
    edited 2013-03-26 17:13
    Excellent work, Kevin. Thanks for providing a solution we can all accept (a free replacement!).
  • RaymanRayman Posts: 14,662
    edited 2013-03-26 17:41
    Glad to hear the problem is found... Maybe adding a 10uF or so tantalum cap between Vdd and GND would fix the existing boards?
  • MoskogMoskog Posts: 554
    edited 2013-03-26 23:08
    Ken Gracey wrote: »
    Excellent work, Kevin. Thanks for providing a solution we can all accept (a free replacement!).

    Excellent work, excellent forum and excellent company!
  • Kevin CookKevin Cook Posts: 159
    edited 2013-09-30 14:54
    There is another update coming out for the Project Board, to help resolve some intermittent VGA issues that still existed; even on the REV B boards. You would notice this issue when running any of the "high resolution" VGA objects.

    The crystals' "XI" (input) runs down along the board to the XI on the Quickstart Header. Well this run was inducing some interference on some of the IO.

    The "fix" for any existing REV B users is to cut the XI trace that runs down from the crystal. The only way to do this is to drill a hole, since that trace is on an internal layer of the board. See my attached photos for details.

    As always, we will replace any Project boards with the upcoming REV C. The update will re-route that trace.

    Note, this problem on effects high resolution drivers.

    drill2.PNG
    drill1.jpg
    509 x 489 - 38K
    1024 x 768 - 108K
Sign In or Register to comment.