Propeller power issues

Vega256 · 2012-06-04 15:10

Hi guys,

I have my Prop powered through a 3V3 regulator; it along with the Propeller is on a breadboard. I am having issues with the chip power-cycling and randomly powering on and off. I have monitored the voltage during operation and it seems to be steady and quiet. Anyone have a solution or even an idea of what is going on here?

Phil Pilgrim (PhiPi) · 2012-06-04 15:17

Are you sure it's power-cycling, or just resetting? What kind of connections do you have to nRST and nBOE? What else is on the board? USB interface? PropPlug header? Is the board connected to a PC when this happens?

-Phil

Vega256 · 2012-06-04 15:30

I don't think it's resetting; I'm not sure. I have RESn connected to serial and BOEn to ground. Only other thing on the board is an EEPROM and an A/V conn. It does this connected or disconnected to the PC.

Leon · 2012-06-04 15:39

Solderless breadboards are not known for their reliability.

Phil Pilgrim (PhiPi) · 2012-06-04 15:46

Vega256 wrote:

I have RESn connected to serial...

Connected to serial what? And how is it connected?

-Phil

Vega256 · 2012-06-04 15:51

Phil Pilgrim (PhiPi) wrote: »

Connected to serial what? And how is it connected?

-Phil

RS-232

http://www.parallax.com/Portals/0/Downloads/docs/prod/prop/SerialtoPropeller.pdf

turbosupra · 2012-06-04 17:26

Disconnect the rs-232 after adding code to have an LED flashing on one of the pins

And you can also have your code use the code below, if you put this at the top of pub main it will force your application to wait for a keystroke through the parallax serial terminal before it continues through the mail loop. This will allow you to know if the code resets or not as it will not allow the code to progress to the LED flashing code. If the LED continues to flash eternally, it isn't resetting and has something to do with that rs-232 connection.

       tmp:=-1
        repeat while tmp==-1
          pst.str( string( " ",$0d,"== please press key to start ==",$0d ) )
          tmp:=pst.rxcheck
          if tmp==-1
             waitcnt( clkfreq+cnt )

Vega256 · 2012-06-09 09:10

Now, I'm starting to think that the Prop is resetting, and has something to do with BOEn. According to the datasheet, RESn is an input when BOEn is high, and an output when BOEn is low. In the typical connections diagram, BOEn is low; are there other ways it can be connected?

LoopyByteloose · 2012-06-09 10:12

Take a look at what is shown in the Propeller Manual. That is the preferred way.

Phil Pilgrim (PhiPi) · 2012-06-09 19:10

Vega256,

So tell us about the filter and bypass caps you're using, and where they're physically located.

-Phil

Vega256 · 2012-06-09 21:00

Phil Pilgrim (PhiPi) wrote: »

Vega256,

So tell us about the filter and bypass caps you're using, and where they're physically located.

-Phil

I've got caps located at both power pins on the Prop, and in the serial-to-PC connection as indicated in the schematic.

Phil Pilgrim (PhiPi) · 2012-06-09 21:23

Does the voltage regulator have input and filter caps? Which regulator are you using?

-Phil

Vega256 · 2012-06-09 21:51

Phil Pilgrim (PhiPi) wrote: »

Does the voltage regulator have input and filter caps? Which regulator are you using?

-Phil

Yes; LD33V

Toby Seckshund · 2012-06-10 01:45

Do the "resets" happen along with any other local events, such as the fridge kicking in etc?

On my usual setup I use a 9 Volt SM PSU regulated down to 5 Volts, and then again to 3.3 Volts (with all the usual input/output caps). The Prop gets 100nF on each of the sides and usually a 47-100uF Tant cap right underneath it. With all of this I always get a reset when I switch on/off the 'scope (with or without any probe connections) if the serial lead is connected to the PC. The SM PSU has an earth connection all the way through.

I should try a temp setup from a 9 Volt battery to prove if the problem is just one great long earth loop. On some boards I made I have put a switch/link in the path from the serial /RST to the Prop, mostly with the three transistor circuit.

Heater. · 2012-06-10 02:16

Ah, the fridge problem. After years now I still have my Prop resetting when the fridge goes on/off. The USB powered ASC board does not have this problem nor my dev boards for other micros. Only serial connected Prop boards (Triblade or home made) have this issue.
Any suggestions for fixing this?

Phil Pilgrim (PhiPi) · 2012-06-10 09:34

heater wrote:

Any suggestions for fixing this?

Stiffer pull-up on nRES; bigger filter cap (~1000uF) ahead of the 3.3V regulator; diode ahead of the pre-regulator cap.

-Phil

jmg · 2012-06-10 15:31

Heater. wrote: »

Ah, the fridge problem. After years now I still have my Prop resetting when the fridge goes on/off. The USB powered ASC board does not have this problem nor my dev boards for other micros. Only serial connected Prop boards (Triblade or home made) have this issue.
Any suggestions for fixing this?

First, check exactly what triggers by removing the serial connection controls - first the Reset, then the Data lines until only ground remains.
Remote cables + Mains relay switching events can inject quite large current impulses into boards, with very fast edges.
Once you know which pin(s) are the culprits, you can apply remedial action.

Spice can be useful here : Feed a fast, high voltage edge into some fF, coupled into some nH with some nominal damping R/C, and you'll see what is needed to kick 'some volts'. eg a sub-ns 1000V edge with a pathway modeled as 30fF, will kick 4V+ on a 50nH PCB trace.

This is why 'relay control' causes so many switching related issues.

Propeller power issues

Comments