Timing Circuit
in [Design]
|
Prev: DTMF Decoder IC
Next: 4 digit LCD or LED display
From: Winfield Hill on 30 Jan 2006 21:36 Spehro Pefhany wrote... > > There are sub-dollar parts (qty. 100) that can exceed your accuracy > spec with no external parts (internal calibrated RC oscillator) and > (most important) NO manual trimming or accuracy testing. By using an > 8-pin part with an external resonator you could get 0.5% accuracy for > about 65 cents (qty. 100)- eg. ATtiny11. You'd have a few I/Os left to > play with, can add additional functionality cheaply, and the circuit > will become more difficult to copy. Seems like a no-brainer. If the > startup will pay you something for the couple of weeks it will take > you to figure this stuff out (almost all learning curve), you'll both > be ahead. I second that. -- Thanks, - Win
From: Ken Smith on 31 Jan 2006 09:47 In article <mmbst15a3g6rmf8756admfh087ufvp43ok(a)4ax.com>, John Fields <jfields(a)austininstruments.com> wrote: [....] >However, using a low tempco rheostat and cap to set the center >frequency takes out the uncertainty due to the chip's initial >accuracy, leaving only the cap's tolerance and the pot's tempco to >deal with (assuming the cap's tolerance includes its tempco), so you >wind up with, essentially, a 2% machine. Use a thru hole film cap. After it is installed, cycle the whole PCB between the hot box and the freezer a few times befor you adjust the pot. The quick warm up and cool down work the stresses out of the parts and get the initial drift mostly over with before you tweek the pot. If you do 90% of a resistor's value with a fixed resistor and the last 10 with a pot, you can reduce the tempco effects in the pot and make the tweeking easier. -- -- kensmith(a)rahul.net forging knowledge
From: Phil Hobbs on 31 Jan 2006 09:58 Ken Smith wrote: > If you do 90% of a resistor's value with a fixed resistor and the last 10 > with a pot, you can reduce the tempco effects in the pot and make the > tweeking easier. > The wiper connection is liable to become unreliable, so it's usually better to combine pots and fixed resistors in parallel, or better, series-parallel, to confine the adjustment range. Either way, the thing will still work at some level if the wiper opens. Cheers, Phil Hobbs
From: Ken Smith on 31 Jan 2006 10:08 In article <43DF7B1B.7030004(a)SpamMeSenseless.pergamos.net>, Phil Hobbs <pcdh(a)SpamMeSenseless.pergamos.net> wrote: >Ken Smith wrote: > >> If you do 90% of a resistor's value with a fixed resistor and the last 10 >> with a pot, you can reduce the tempco effects in the pot and make the >> tweeking easier. >> > >The wiper connection is liable to become unreliable, so it's usually >better to combine pots and fixed resistors in parallel, or better, >series-parallel, to confine the adjustment range. Either way, the thing >will still work at some level if the wiper opens. It depends on whether a little out of spec is better than completely stopped. You could also use one of those EEProm based digital pots. In 555 like circuits, changing the effective voltage source for the resistor also can be used to trim the frequency. -- -- kensmith(a)rahul.net forging knowledge
From: Jimbo on 31 Jan 2006 22:33
What's the "wiper connection"? |