How to use one crystal to clock multiple props?
MarcG
Posts: 14
I was wondering how I can have one crystal, clock multiple props.
Will this work... Can I daisy chain the XO pin of one prop to the XI pin of another and continue this way for a total of 10 props spread out across a 17 inch span. (See attached PDF.) If it will work, then would I set the slave props OSCMx register to XINPUT?
If above idea won't work, or if there is a better way. please let me know. Thanks
Will this work... Can I daisy chain the XO pin of one prop to the XI pin of another and continue this way for a total of 10 props spread out across a 17 inch span. (See attached PDF.) If it will work, then would I set the slave props OSCMx register to XINPUT?
If above idea won't work, or if there is a better way. please let me know. Thanks
Comments
-Phil
1) Have a separate crystal for each Propeller
2) Designate one Propeller as master and use an I/O pin for a clock. This requires that the program in that Propeller initialize the I/O pin and one of the cog counters in the "main" cog so that the I/O pin outputs the necessary clock.
3) Have a single external crystal oscillator that's connected to each of the Propellers' XI and set up all the Propellers with XINPUT clock mode.
Note that the Propellers initialize themselves using the fast internal RC clock. Once this is done and a program is loaded from EEPROM or an attached PC, the Prop switches over to whatever clock is specified in the program.
On common powering up, and simutaneous resets the two of then got on fine.
(this was with the Master being a Blade2 and the Slave being a PropCMD/VT100)
'Interesting observation! When this happened, was the slave clock configured for PLL and/or did you have other PLL counters running? Without any PLLs running, I would think that the Prop would behave in a totally static manner. Of course, having PLLs with a 4-8 MHz capture range changes all of that. So it's probably not a good idea, in general, to drive one Prop from another's counter output.
-Phil
A common reset button would have probably worked.
There is also a 74LVC1104 IIRC.
The Prop starts up using its internal RC clocks and then, code permitting, fires up the external xtal and PLL bits (powersaving I suppose). Even if the Prop is static, clock wise, the interuption did crash the second (VDU) Prop every time. If I had gone even cheaper and not put on the reset switches I may have never noticed.