Shop OBEX P1 Docs P2 Docs Learn Events
Propeller resetting when I don't want it. — Parallax Forums

Propeller resetting when I don't want it.

Bobb FwedBobb Fwed Posts: 1,119
edited 2009-06-16 22:35 in Propeller 1
I have had a problem for a little bit now. My propeller project works around a lot of ambient frequencies and random induction. I have noticed that when certain electrical things change suddenly near the PCB, the propeller tends to reset. It seems the programming traces (or wire when I have it plugged in) act as an antenna, and the (somewhat) weak pull up is being over powered by the change in the electrical field.

I am hoping I can fix this by placing a resistor to my 3.3V source from the reset pin on the propeller. Is this sound practice? Will I still be able to program the chip? I figure adding a 2.2K- or 5.1K-ohm resistor to augment the 5K internal resistor should do the trick (I may need even less resistance, but i won't know until I test it).

Are there other solutions?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
April, 2008: when I discovered the answers to all my micro-computational-botherations!

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-16 18:32
    A stiffer pullup might help. But tell us more about your board, particularly about the programming port.

    -Phil
  • Beau SchwabeBeau Schwabe Posts: 6,568
    edited 2009-06-16 18:33
    Bobb Fwed,

    What does your power supply directly to the Propeller look like?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Beau Schwabe

    IC Layout Engineer
    Parallax, Inc.
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 18:38
    Well initially I had fairly short traces (1cm to pin 30/31, 2cm to reset), the programming "port" was right next to the chip. I had the resetting problem but fairly rarely.
    The next version of the PCB I placed the programming "port" at the edge of PCB so the port could accessed from outside of a plastic enclosure. So the traces are now 8cm or so.
    For both, the traces are 0.01"-width with a simple 4-pin 0.1"-spacing header to plug the Prop Plug in.

    The newest version of the board (haven't yet ordered it) has 0.006" wide traces for the programming traces, but the length is about the same. I have also added a optional placement of a resistor as described above.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!

    Post Edited (Bobb Fwed) : 6/16/2009 6:51:02 PM GMT
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 18:44
    The power supply has not had an effect. I have three three different types of sources: a variable-dc-regulated power supply by Tenma (very handy-- up to 16V and up to 20A). I have also used a 7A isolated power supply @ 14V. And finally a 7.4V Li-Ion battery.

    On the PCB itself I have a 1000µF cap on the incoming power, NTE1904 3.3V regulator with a 1F (30-ohm ESR) cap and a 470µF (0.45-ohm ESR) cap for stabilization (the datasheet says the 1904 only requires 27µF or something like that). I have 1F cap so the chip can continue to operate for a short period of time after power has been removed, and I have the 470 for the somewhat low ESR (for stabilization).

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
  • RaymanRayman Posts: 14,833
    edited 2009-06-16 18:50
    You don't have PropPlug and USB cable connected when this happens do you? I've found you have to remove these when there's EMI around...

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    My Prop Info&Apps: ·http://www.rayslogic.com/propeller/propeller.htm
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 18:56
    I have noticed that as well, but it happens with it plugged in and not. But definitely moreso when it is plugged in.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-16 18:56
    That's the reason I was asking about the programming port. What I was hoping to find out is whether the USB circuitry is onboard or whether a PropPlug is required.

    -Phil
  • LeonLeon Posts: 7,620
    edited 2009-06-16 18:58
    Try a TVS on the 3.3V supply. Have you decoupled both supply pins?

    Leon

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Amateur radio callsign: G1HSM
    Suzuki SV1000S motorcycle
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 19:04
    It's a QFP chip. I have the 470 cap as close as I can to the chip as possible, but I do not have ceramic caps on the 4 power connections (like I have the room...).
    I just remembered though. an early prototype did have a Polyester Metalized cap right next to the power. That had the same problems though. I don't think it is a power supply problem.

    With my oscilloscope, there is essentially nothing but a flat line with it turned all the way up (5mV div) on the 3.3V power plane.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-16 19:11
    Bob,

    You really need bypass caps on at least two of the Vdd pins. The 470uF cap is a filter cap and does not provide the bypass function.

    -Phil
  • W9GFOW9GFO Posts: 4,010
    edited 2009-06-16 19:26
    Phil Pilgrim (PhiPi) said...
    Bob,

    You really need bypass caps on at least two of the Vdd pins. The 470uF cap is a filter cap and does not provide the bypass function.

    -Phil

    Are bypass caps the same as decoupling caps, across VDD and VSS? Are 102s appropriate?

    Mine is resetting too once in a while...

    Rich H
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2009-06-16 19:29
    Bypass == decoupling: different words for the same thing. By "102s" do you mean 0.001uF? No. I'd consider 0.1uF (104) an absolute minimum, and they need to be ceramic.

    -Phil
  • SapiehaSapieha Posts: 2,964
    edited 2009-06-16 19:39
    Hi W9GFO.

    To block high frequensies (parasitic frequensies)·to and from Propeller use 10 to 100 nF ceramic capasitors (decoupling) parallel with 10 to 47 uF Tantalum caps for stablity of power to Propeller.
    All near Propeller if You have more IC's on PCB.
    Ceramic caps will be on every power pair to Propeller. Tantalum on pair nearest Crystal.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Nothing is impossible, there are only different degrees of difficulty.
    For every stupid question there is at least one intelligent answer.
    Don't guess - ask instead.
    If you don't ask you won't know.
    If your gonna construct something, make it·as simple as·possible yet as versatile as posible.


    Sapieha
  • W9GFOW9GFO Posts: 4,010
    edited 2009-06-16 20:03
    Thanks, I've got plenty of ceramic 104s but no tantalums.

    Rich H
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 20:21
    Well, I have room to add one capacitor. I have a couple 0.47µF ceramic caps on the way. View the attachment: C2 (1F), C18 (470µF), C21 (0.47µF), R43 (pull up resistor on reset pin).

    I still don't think the prop resetting is a power problem. When I attach an "extension cord" to my programming port the problem is much more frequent. Just that change isn't affecting the power, it is affecting pin 30/31 and reset. The reset pin of which is the only potential problem.

    Regardless of the need to decouple. My original question stands: is it ok to pull the reset pin high? At what point do I stand a chance of damaging the prop plug when I attempt to program with an additional pull-up resistor?

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
    PCB.JPG
    997 x 830 - 176K
    PCB.JPG 175.8K
  • SapiehaSapieha Posts: 2,964
    edited 2009-06-16 20:29
    Hi Bobb Fwed.

    PropPlug drive Reset by transistor (open collector) and that wil not colision with FTD chip pins

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Nothing is impossible, there are only different degrees of difficulty.
    For every stupid question there is at least one intelligent answer.
    Don't guess - ask instead.
    If you don't ask you won't know.
    If your gonna construct something, make it·as simple as·possible yet as versatile as posible.


    Sapieha
  • LawsonLawson Posts: 870
    edited 2009-06-16 20:44
    I've got a 10k pull-up on the reset line of a project of mine. It's been working and programming just fine. (no massive inductive or electrostatic surges near it yet though)

    My other thought would be to add a ground plane. Hm, on second look it would appear that a ground and power plane exist on Bobb's layout, but are hidden. In that case I'd try to minimize the fragmentation of the ground plane. The propeller is very nice it this regard, because it's easy to juggle pins around to untangle a layout. (i.e. almost no fixed function pins) Have you tried adding a metal shield between the prop board and the worst of the interference sources?

    My 2 bits

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Lunch cures all problems! have you had lunch?
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 20:49
    Yes, the ground and power planes are hidden...but they do exist.
    As for shielding it, it would essentially have to be a complete enclosure of metal. The interference is coming from almost 360-degrees on all axis (ie. it's in the middle of the problem). But due to other design and assembly issues (not to mention the bulk pricing) I cannot go to a metal enclosure.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
  • Clock LoopClock Loop Posts: 2,069
    edited 2009-06-16 21:09
    Bobb Fwed said...

    Regardless of the need to decouple. My original question stands: is it ok to pull the reset pin high? At what point do I stand a chance of damaging the prop plug when I attempt to program with an additional pull-up resistor?

    What is your BOEn attached to?

    using a strong pullup resistor on the RES line shouldn't hurt anything.

    If you calculate the current at 3.3v through a 1k, you will see its only 3.3 ma of current. Most LED's pull that much, eaisly.

    But if you use a 1k, it might mess up high speed communications, I use a 1k and its fine for me.

    You can try 10k also. Or variation in between 1k and 10k.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    TERMS OF USE: MIT License

    "Permission is hereby granted, free of charge, to any pers...........................
    ..............................OMITTED FOR FORUM...............................................
    .................. OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. "
  • Bobb FwedBobb Fwed Posts: 1,119
    edited 2009-06-16 22:35
    BOEn is tied to ground.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    April, 2008: when I discovered the answers to all my micro-computational-botherations!
Sign In or Register to comment.