Troubleshooting Advice
Shuman (OSURC)
Posts: 12
Background:
We're using the Quadrover for a mars rover challenge in early June. The controls are handled with the propeller protoboard. During a rehearsal we managed to break the propeller. After brining it back to the lab we found that the propeller was not executing any instructions. We tried to use the internal RC oscillator and it executed instructions again. Our latest test was to see if the XTAL1 source would work if the PLL was bypassed, and it still executed.
Here is the code that we used to test the propeller. The clock is set correctly if P17 toggles:
CON
_clkmode = xtal1 '' This line sets the clock to a valid source
''_clkmode = xtal1 + pll16x '' This line sets the clock to an invalid source
_xinfreq = 5_000_000
PUB Main | i
dira[noparse][[/noparse]17] := 1
repeat
outa[noparse][[/noparse]17] := !outa[noparse][[/noparse]17]
waitcnt(clkfreq/20 + cnt)
I'm looking for some ideas on how the PLL could be broken so we can find and prevent future problems.
Thank you,
Matt Shuman
OSURC[url=http://][/url][url=http://][/url][noparse][[/noparse]url]
We're using the Quadrover for a mars rover challenge in early June. The controls are handled with the propeller protoboard. During a rehearsal we managed to break the propeller. After brining it back to the lab we found that the propeller was not executing any instructions. We tried to use the internal RC oscillator and it executed instructions again. Our latest test was to see if the XTAL1 source would work if the PLL was bypassed, and it still executed.
Here is the code that we used to test the propeller. The clock is set correctly if P17 toggles:
CON
_clkmode = xtal1 '' This line sets the clock to a valid source
''_clkmode = xtal1 + pll16x '' This line sets the clock to an invalid source
_xinfreq = 5_000_000
PUB Main | i
dira[noparse][[/noparse]17] := 1
repeat
outa[noparse][[/noparse]17] := !outa[noparse][[/noparse]17]
waitcnt(clkfreq/20 + cnt)
I'm looking for some ideas on how the PLL could be broken so we can find and prevent future problems.
Thank you,
Matt Shuman
OSURC[url=http://][/url][url=http://][/url][noparse][[/noparse]url]
Comments
This may help you:
Paul Baker (Parallax)
Applications Engineer
Click to send Paul Baker (Parallax) email.Personal Homepage Not AvailableSend a Private Message to Paul Baker (Parallax)AIM Not AvailableICQ Not AvailableY! Not AvailableMSN Not Available
Date Joined Jul 2004
Total Posts : 5692
Posted 3/21/2008 9:27 PM (GMT -4) Quote This PostIgnore Posts From Paul Baker (Parallax).Alert An Admin About This Post.
Sounds like you fried your PLL, to test this load a program with the clock mode set to RCFAST, if it runs, the PLL is gone.
We have determined the leading cause for this are ground loops and inductive kickback.
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
Source
http://forums.parallax.com/showthread.php?p=717270
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Aka: CosmicBob
Conversely I can try to find the post that I go into great detail about board do's and don'ts wrt making the Propeller less susceptable to PLL failures. This problem has happened enough that I will be writing an application note on how to minimize this, but it will be many months from now before I will have enough time to devot to it.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.