Shop OBEX P1 Docs P2 Docs Learn Events
Noise questions — Parallax Forums

Noise questions

BTXBTX Posts: 674
edited 2011-06-02 11:23 in Propeller 1
Hi everybody !
I'm designing with propeller chip since 2006 year, for industrials applications. But there is an issue that I never can get full 100% work well.
It is the reset of the pchip randomly, usually when I'm starting an standard contactor near him.
My boards always have optocoupled inputs and outputs, I'm using switching power supplies, always exclusive for my pchip boards, I'm wiring GND's correctly, I'm using PQFP chips, correctly decoupled with .1uf and a 1uf capacitors, my PCB's use all Vcc pins wired together, and all GND pins wired too, wired ,anyway I can avoid some inesperated resets in some ocations.

Could it be, due a isue in the reset pin of my boards? should I put it at Vcc using a resistor to avoid this ? or, what could I do to solve this?
Any ideas ?.
Cheers.
Alberto.

Comments

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-05-12 10:26
    A stiffer pull-up on /RST would certainly help, maybe with a little capacitance thrown it. Don't overdo the capacitance, though, as it could affect the timing of the bootloader operation. BTW, if /BOE is tied low, you don't want to tie /RST directly to Vdd with a removable jumper, say, since the brownout circuit will try to pull it low during a brownout condition.

    -Phil
  • BTXBTX Posts: 674
    edited 2011-05-12 10:37
    Thank you Phil, I will try this, yes my BOE is always tied low, so I'll use a "maybe" 4k7/10K to pull up the /RST pind, and I will test with differents capacitances.
    Should not be this a common practique for all boards so ? Did I missed some document where this is write ?
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2011-05-12 10:40
    I think you could safely go as low as 2.2K without issue.

    -Phil
  • BTXBTX Posts: 674
    edited 2011-05-12 10:42
    Ok Phil, understood, thank you again !!
  • TubularTubular Posts: 4,717
    edited 2011-05-12 16:43
    Alberto,

    I agree with what Phil says, and you should try that first.

    Depending on the layout you may also be getting direct radiation from the arc of the contactor. We use a small RC snubber for supressing the contactor coil, and sometimes also the contactor poles.

    If you do bypass the contactor poles, please be aware of the safety implications. These bleed a few mA of current into the "switched off" load (motor/light). Usually the current is less mA than a RCD safety switch is set to trip at (30mA), but it depends on the capacitor value, frequency (50/60Hz) etc. Please be careful and PM me if anything is not clear.

    DSC01275.jpg
    DSC01275.jpg
    1024 x 680 - 45K
  • BTXBTX Posts: 674
    edited 2011-05-13 09:24
    Hi Tubular.
    Thank you for posting too !, it's not an EMI, all is inside a Faraday system, contactors have their parts too, it's comming by the power supply.
    I will try the solutions in next months, when I'll have an automation to install.
  • madrfskillsmadrfskills Posts: 24
    edited 2011-05-23 06:55
    One thing you may want to try is a lossy pi network on the micro's power rail. I've found the following works very well for nasty environments:

    Filter.PNG


    I usually take input at 3.6V, and size the zeners to protect the caps. Key to performance is cap choice - I like to use the Kemet T496 fail-safe fused tantalums. Having a fuse inline with a tantalum is always nice but the point is that the additional ESR from the fuses is somewhat helpful from a damping perspective. The inductor is a Coilcraft, part number as shown. I have used this with 10uH inductors as well, not much difference in performance. Following all of this is a 3.3V output LDO.

    BTW, having separate faraday cages for differing zones of protection has helped me with very high current designs. Depending on your volume you may be able to justify some customized surface mount shields -- see http://www.gore.com/en_xx/products/electronic/emi/snapshot/index.html.

    V/R
    Mike
    Filter.PNG
    794 x 480 - 16K
  • BTXBTX Posts: 674
    edited 2011-05-24 19:59
    Hi Mike.
    Thank you so much !!, it is an interesting solution to add to the boards. I'll try it too.
  • LeonLeon Posts: 7,620
    edited 2011-05-25 00:27
    TVSs on the regulated supplies should help.
  • BTXBTX Posts: 674
    edited 2011-05-25 04:55
    Thank you so much Leon !! I will try a mix.. of solutions :)
  • Zap-oZap-o Posts: 452
    edited 2011-06-02 11:23
    I simply place a switch in between the reset pin and the transistor used to program the propeller. I never had a propeller randomly reset. Is it possible that your code is causing this? I find that if I dont leave a repeat statement in a given loop the prop will act strange. Perhaps you have an opened ended function somewhere?

    This crude code may cause issues... 
    Pub Main
 dira[somepin]~~
 Do_something

Pub do_something
outa[somepin] := 1

    This crude code will not cause issues... 
    Pub Main
 dira[somepin]~~
 Do_something
 repeat 

Pub do_something
outa[somepin] := 1
Sign In or Register to comment.