DIY External RC Clock?

Humanoido

It is possible to make a cheap simple low-accuracy slow prop external RC clock circuit? ..a poor man's crystal made from a fed resistor-capacitor circuit etc..

(for academic pursuit, making and testing ultimately low power props, and possibly experimenting with cheap "wide band" home-made clocks)

Thanks for your reply.

JonnyMac

There are two internal RC modes; will neither work for you?

Mike Green

If you want a cheap simple low-accuracy clock, just use the RCSLOW clock source. It's built into the Propeller. When running at around 32KHz, the Prop draws very very little power particularly if you shut down 7 of the cogs and have the 8th waiting most of the time in a WAITCNT. If you calibrate the clock to adjust for chip to chip variations, it should be fairly stable once the Propeller stabilizes at room temperature (assuming that room temperature is fairly stable).

By calibrate, I mean that you have some way to run the clock for a period of time (several hours at least) and enter in the start and end time as measured by some other (more accurate) clock. The Prop would use that to set a more accurate CLKFREQ value than the assumed value.

Humanoido

Mike Green wrote: »

By calibrate, I mean that you have some way to run the clock for a period of time (several hours at least) and enter in the start and end time as measured by some other (more accurate) clock. The Prop would use that to set a more accurate CLKFREQ value than the assumed value.

Excellent. I have noted, that individual props are relatively stable in selecting approximately same/ similar clock values at different startup times. However from prop chip to prop chip, a variance can result. Bean has provided an interesting benchmark that could be used/ adapted for testing the results. http://forums.parallax.com/showthread.php?123678-PE-Basic-Version-0.16-July-11-2011&p=1017831&viewfull=1#post1017831

Mark_T

Humanoido wrote: »

It is possible to make a cheap simple low-accuracy slow prop external RC clock circuit? ..a poor man's crystal made from a fed resistor-capacitor circuit etc..

(for academic pursuit, making and testing ultimately low power props, and possibly experimenting with cheap "wide band" home-made clocks)

Thanks for your reply.

Well I would suggest using a CMOS schmidt-trigger inverter such as 74HC14 plus an RC filter, perhaps 100k - 1M resistor and a few hundred pF. The datasheet (see link) has a relaxation oscillator circuit in figure 14.

Note that its power consumption (20uA?) won't match RCSLOW (about 7uA) as far as I can see - there is a graph of current consumption v. input voltage in the datasheet. For better performance two simple CMOS inverters with multiple feedback would be needed I think.

http://www.nxp.com/documents/data_sheet/74HC_HCT14.pdf