Shop OBEX P1 Docs P2 Docs Learn Events
"Prop Dongle" Protoboard 1.5" x 2.75". Available for purchase! - Page 2 — Parallax Forums

"Prop Dongle" Protoboard 1.5" x 2.75". Available for purchase!

24

Comments

  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-08 09:05
    yes please!! I'll PM my details.
  • BeanBean Posts: 8,129
    edited 2007-06-08 15:49
    Graham,
    Any ideas about why I don't get color ? I was running the graphics demo from the library.

    On a CRT I get a stable picture, but no color. If I run it into my PC with a TV card, I get some coloration but the screen is VERY jittery.

    I wouldn't think the resistor values are THAT critical, but maybe they are ?

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • Paul BakerPaul Baker Posts: 6,351
    edited 2007-06-08 17:31
    The most common reason for color not showing is a clock which is a couple percent off it's stated value in code. The resistors shouldn't be super critical, I just successfully tested broadcasting TV with nothing plugged into the holes, granted it's range was a couple inches and it's broadcast, not composite. But if you plan on producing them I would use the correct value since they where chosen for a DAC with near 75 Ohm matching.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • BeanBean Posts: 8,129
    edited 2007-06-08 19:09
    Paul,
    · I have found the problem.
    · If I specify "_xinfreq = 5_003_600" (actually 5_002_800 to 5_004_500 works) then I get good color.
    · So it seems the xtal is not operating at the proper frequency.
    · The xtal is 5MHz 20pF is there any reason it is off frequency this far ?

    · Here is the datasheet http://www.ecsxtal.com/store/pdf/csm-7x.pdf

    [noparse][[/noparse]edit] I looked in the manual and it states 36pF for XTAL1, but I can't see 16pF causing the frequency to be off over 700ppM.
    Also if the xtal is expecting at 20pF load, and is seeing at 36pF load shouldn't it be LOW in frequency ?

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com


    Post Edited (Bean (Hitt Consulting)) : 6/8/2007 7:35:52 PM GMT
  • Paul BakerPaul Baker Posts: 6,351
    edited 2007-06-08 20:41
    Try XTAL2 which is a 20pF circuit and see if that gets the frequency closer to the stated value.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Paul Baker
    Propeller Applications Engineer

    Parallax, Inc.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2007-06-08 20:47
    Bean,

    Your problems may be related to power distribution and bypassing, rather than to the crystal. The Propeller Vss connections could be tightened by using vias inside the QFP footprint to the groundplane on the bottom of the board, rather than to the scattering of grounds on top. You will probably want to get rid of the copper pin labels on the bottom, too, so the groundplane isn't so broken up. I'd also fill the area topside within the QFP footprint with Vdd (except for the ground vias). And, finally, bypass caps: at least two very close to the chip on opposite sides.

    After I've laid out a board, I like to look at just the groundplane and visualize where the ground currents are flowing. It's sort of the "feng shui" of PCB design. If the ground return current from each device can flow freely back to the regulator, without constriction, without switching sides, and without being channeled to comingle with the return currents from other devices, I figure I'm a step ahead. The same goes for Vdd: If each device is fed by an independent path back to the regulator, that's great. It's not always possible, and fat traces can help when it isn't; but it's nice when it can be arranged.

    One final thought: Is the crystal can touching any part of the circuitry?

    -Phil
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-09 00:30
    Phil, how do people get away with running them on breadboards?

    Graham
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2007-06-09 01:57
    Graham,

    Divine intervention?

    -Phil
  • BeanBean Posts: 8,129
    edited 2007-06-09 02:48
    Tried to run the XTAL2 would not do anything, but if I use PLL8X instead of PLL16X then it does run, but no color at either frequency.
    I will try putting my xtal in the Prop Demo board and the prop demo board xtal in my dongle and see if either or both work.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • BeanBean Posts: 8,129
    edited 2007-06-09 12:30
    Okay, I tried my xtals in the Propeller demo board and they work fine.
    So I guess it IS a layout issue of some kind.

    I'll keep working on it and let you guys know what the problem turns out to be.

    I see I don't have any caps near the propeller chip, and I'm not 100% sure about the power from the USB connection...

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-09 12:52
    You could insulate the io vias on the back of the board and then press some copper foil on the back to rule out a groud plane problem.

    Graham
  • BeanBean Posts: 8,129
    edited 2007-06-10 00:36
    I built another unit in it works fine.
    I think I may have gotten the Propeller chip too hot on the first one.
    I will try replacing it and see if that fixes the problem.

    Graham,
    I will send you this working prototype on Monday.

    [noparse][[/noparse]edit] Yep, I replaced the propeller chip, now the first unit works fine too. Yahoo....

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com


    Post Edited (Bean (Hitt Consulting)) : 6/10/2007 1:08:54 AM GMT
  • BeanBean Posts: 8,129
    edited 2007-06-11 02:15
    Graham,
    I hope you weren't in love with those extra pin connection.
    I removed them·for several reasons. It clutered up the back, broke up the ground plane, was just about impossible to solder a wire to them anyway.

    Here is the latest, I also put a 10 Ohm resistor in series with the +5V line after I blew a USB port because I had the +5V line shorted to ground (you would think they would protect them against that ???). But I've read that USB ports can provide up to 500mA, so the 10 Ohms should cover me and still allow over 100mA before the 3.3V regulator drops out.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com


    Post Edited (Bean (Hitt Consulting)) : 6/11/2007 11:29:40 AM GMT
    945 x 704 - 195K
    925 x 521 - 89K
  • rjo_rjo_ Posts: 1,825
    edited 2007-06-11 04:43
    Bean,

    I like the foot print... perfect for my robot weed wacker... which is I am putting together around an HB-25 and used plumbing supplies.
    But I want is an automatous weed wacker... so I want to detach my Dell... and of course that means I would need a power supply.

    What should I do?

    Is there a power supply that has a USB connector?

    Rich

    ps... any chance of talking you into Prob based stereo-video applications?
  • BeanBean Posts: 8,129
    edited 2007-06-11 11:30
    Rick,
    You could just get a USB extension cable and connect it to a +5V regulated power supply.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • BeanBean Posts: 8,129
    edited 2007-06-12 22:06
    Rick,
    What am I thinking, you can just connect +5V to the "5V" and "Gnd" pins in the column of pin connections.

    I just ordered another set of prototype boards. Assuming they check out alright, I should have production boards available around July 9th.
    Will probably be $49.95 $59.95 or $8.00 $10.00 for just the PCB.
    I will make everything available, parts list, schematic, even the ExpressPCB layout. So you can DIY if you want.

    P.S. I went with a PTC resettable fuse instead of the 10 ohm resistor to protect the USB·port.

    [noparse][[/noparse]edit] I had to change the prices, with the·quick turn-around PCB the cost was alot more.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com


    Post Edited (Bean (Hitt Consulting)) : 6/17/2007 2:12:04 PM GMT
  • JamesxJamesx Posts: 132
    edited 2007-06-13 11:50
    Bean:

    These look great. Count me in for a couple of assembled units.

    Jim C
  • Chris MerckChris Merck Posts: 55
    edited 2007-06-13 14:14
    Bean,

    Excellent work! Good call removing the extra copper between the prototyping pads.

    This could be useful for development on the go for sure. Even if you are using the VGA_text object, it could be modified to send screen data out over serial to a terminal emulator on the other end. (this could come in handy for Proguehttp://forums.parallax.com/showthread.php?p=655083 development on the train).
  • Tracy AllenTracy Allen Posts: 6,660
    edited 2007-06-13 16:31
    Count me in for one too! Looks great.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tracy Allen
    www.emesystems.com
  • codemonkeycodemonkey Posts: 38
    edited 2007-06-13 17:06
    Question about the layout, since i know a bit less than zero when it comes to making boards: on the right hand side of the board, there are two holes for ground that chains from top to bottom in the picture and the holes are surrounded by no copper but bridged to the ground (plane?). Is this lack of copper an artifact of the board layout software or does it serve some important function? TIA
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-13 17:31
    I beleive it is to improve soldering by thermally isolating the pad from the ground plane, keeps the joint neat in one place.

    But I'll be interested to hear if my guess is right.

    Graham
  • BeanBean Posts: 8,129
    edited 2007-06-13 17:43
    Graham is correct. This is a thermal via. If the copper was solid it would be really hard to solder to the pad because the copper would "suck" the·heat away from the iron.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • BeanBean Posts: 8,129
    edited 2007-06-15 02:44
    I was looking at the new PropRPM. I noticed that the prototype area is just non-connected pads (except the power and ground at top and bottom).
    Now I setup the Prop Dongle like a breadboard (with 5 pins connected together on either side).

    So my question is: What would you prefer. A) Pads connect in a line like a breadboard (current layout), or B) Pads not connected (similar to PropRPM) ?

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    ·
  • Chris MerckChris Merck Posts: 55
    edited 2007-06-15 02:52
    You could also do a hybrid: pins connected in a line but with a gap in the center for placing ICs across.
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-15 12:20
    Chris, that's how it is, the gap is formed by the power lines.

    Bean,

    I like it the way you have done it for this size of board, I'm not really a fan of the other way anyhow and its easier to break tracks then mess around bridging.

    Graham
  • codemonkeycodemonkey Posts: 38
    edited 2007-06-15 12:32
    Graham,
    Thanks for the answer.

    Bean
    I like this thing and once it's nailed down, it would make a great small board and i want one too. I was going to suggest that it be just pads but graham's suggestion of scraping traces off allows the best of both worlds.
  • BeanBean Posts: 8,129
    edited 2007-06-16 03:33
    Thanks for the comments guys.
    Here is a pic of the (hopefully) finished layout.

    I aligned the power buss holes because the first thing I wanted to do was install 3 pin headers and I would have wasted 3 rows for each one. With the busses aligned I can just install them right across, then use a resistor from the pin hole to the row.

    I hope to have these ready sooner than expected because I'm going to order them for 2 day service instead of 10 day service.

    I will also be including a 3 foot USB extension cable (in case your computer is not handy, or you want to use the dongle on your bench).

    Bean.


    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com
    1129 x 613 - 132K
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-16 11:28
    Hi Bean, the dongle arrived this morning and it is working perfectly!

    When thinking about how I might use it I had a thought.

    What about making it such that a right angled socket could be added once the protoboard area was removed. See the attached picture of my coolrunner demoboard, it comes with some little io modules that can be plugged straight in. These could be of the professional type like these or homemade from a bit of vero/perf board, allowing you to try lots of little experiments with the same dongle.

    As far as I can tell all that would be needed is a 3.3v connection along the edge, that means loosing an io or making th board just a touch wider.

    Might I also suggest that if you can get them a 90 degree socket for the TV out would also look better and produce a flatter dongle that is easier to shrink wrap.

    Cheers,

    Graham
    800 x 600 - 62K
  • Graham StablerGraham Stabler Posts: 2,507
    edited 2007-06-16 11:30
    I'm an idiot, it already has a 3.3v in the line of connections so all I need to do is chop the board and find the connectors.

    Graham
  • BeanBean Posts: 8,129
    edited 2007-06-16 12:04
    I was wondering what you were talking about [noparse];)[/noparse]

    Yes I specifically made it so you could "chop" off the prototype area and still have all the signals.

    BTW I made the TV output a 3 pin header so you can plug a serial LCD into it also. For those using the dongle "on the road".
    (Ground / +5V / Signal) I haven't posted a picture of this latest change yet.

    Bean.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    “The United States is a nation of laws -· poorly written and randomly enforced.” - Frank Zappa

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    www.hittconsulting.com


    Post Edited (Bean (Hitt Consulting)) : 6/16/2007 12:12:42 PM GMT
Sign In or Register to comment.