Post by thread:[EE]: Anyway to create stable 115.2khz pulse

William wrote: > > Hi! > > Now, I am connecting a crystal (14.7456Mhz) to CD74HC4060 (binary > counter), and > get a signal of 115.2khz for 4060 pin 6. > > And I like to know is there anyway to use a 555 timer to generate a > stable, clean (sharp > waveform) of 115.2khz. No! 555 are very unstable even one cycle to the next. For simple apps you can run the 555 into a divide by 16 count, and with 16 or more division per-cycle stability is ok (I have device that uses this system). Then you still get temperature/freq drift, maybe you can work with that though?? Look for ceramic resonators, there are a lot of cheap plastic box types in the 100+ kHz range, maybe you can get one? -Roman -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Hi, Roman. Thanks for your reply and advice. I just look thru RS contalogue, and those resonators are not suitable for my 115.2khz. Actually, I'm going to use 115.2khz for serial communication, so it should be quite precious and waveform needs to be clean and uniform. Late, I think I need to stay to use a crystal and binary counter to get my required frequency! Thank you. Roman Black wrote: {Quote hidden}> William wrote: > > > > Hi! > > > > Now, I am connecting a crystal (14.7456Mhz) to CD74HC4060 (binary > > counter), and > > get a signal of 115.2khz for 4060 pin 6. > > > > And I like to know is there anyway to use a 555 timer to generate a > > stable, clean (sharp > > waveform) of 115.2khz. > > No! 555 are very unstable even one cycle to the > next. For simple apps you can run the 555 into a > divide by 16 count, and with 16 or more division > per-cycle stability is ok (I have device that uses > this system). > > Then you still get temperature/freq drift, maybe > you can work with that though?? Look for ceramic > resonators, there are a lot of cheap plastic > box types in the 100+ kHz range, maybe you can > get one? > -Roman > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> And I like to know is there anyway to use a 555 timer to generate a > stable, clean (sharp > waveform) of 115.2khz. Yes, but accuracy will be an issue if you want to use this for RS-232 timing. RS-232 can only tolerate a 5.8% mismatch between the sending and receiving baud rates at 8 bits per characters. Of course the mismatch must be below that in practise because this is the "guaranteed to fail" value. I try to stick to half that, or 2.9%. It will be nearly impossible to maintain that kind of accuracy with a 555 timer and the associated R/C components. ***************************************************************** Olin Lathrop, embedded systems consultant in Devens Massachusetts (978) 772-3129, spam_OUTolinTakeThisOuTembedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

It will be nearly impossible to maintain that kind of accuracy (2-5%) with a 555 timer and the associated R/C components. Why do you say that? Aren't the primary variations introduced by temperature variations of the RC component values? How much temperature change is necessary for a 5% value shift in a typical component, anyway? Doesn't it tend to related to absolute temperature (ie room temp = 300K, 5% of which is 15 degrees, which is a LOT for any sort of temperature controlled environment (say, indoors where people work), assuming your circuit doesn't use enough power to changes its own temperature much. FWIW, I made a rs232 format/speed converter (6bit 75bps to 8 bit 300bps) for a newswire system some years ago from a couple uarts using a 555 as the clock source. I just stuck an LED on the "framing error" output of the uart, and adjusted a pot in the 555 circuit to the middle of the range where the LED didn't light. In theory, this was "to be replaced with a more stable source if necessary", but it was never necessary. (This DID run inside a temperature controller mainfram machine room...) BillW -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

> It will be nearly impossible to maintain that kind of accuracy (2-5%) > with a 555 timer and the associated R/C components. > > Why do you say that? Aren't the primary variations introduced by > temperature variations of the RC component values? How much temperature > change is necessary for a 5% value shift in a typical component, anyway? > Doesn't it tend to related to absolute temperature (ie room temp = 300K, > 5% of which is 15 degrees, which is a LOT for any sort of temperature > controlled environment (say, indoors where people work), assuming your > circuit doesn't use enough power to changes its own temperature much. First of all, I was talking about 2.9%. You added the 2-5% comment in parenthesis. I wouldn't want to run RS-232 with 5% clock mismatch. The problem is not as much stability as initial accuracy. It will be very difficult to find a resistor and capacitor that together have no more than a 3% error. Most of this error will come from initial part tolerances, but you should probably allow 1% or so for drift due to time and temperature. This leaves 2% for the resistor plus the capacitor tolerance. Yes, you could use a pot and tweak it at production time, but that will be MUCH more expensive and less reliable than just using a crystal or ceramic resonator in the first place. ***************************************************************** Olin Lathrop, embedded systems consultant in Devens Massachusetts (978) 772-3129, .....olinKILLspam@spam@embedinc.com, http://www.embedinc.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

If 555 timer is unstable for generating uniform pulse, then is there any other timer that can preform well, which can generate 115.2khz? I hope I can get an alternative cheaper way to generate the pulse. Thanks Olin Lathrop wrote: {Quote hidden}> > It will be nearly impossible to maintain that kind of accuracy (2-5%) > > with a 555 timer and the associated R/C components. > > > > Why do you say that? Aren't the primary variations introduced by > > temperature variations of the RC component values? How much temperature > > change is necessary for a 5% value shift in a typical component, anyway? > > Doesn't it tend to related to absolute temperature (ie room temp = 300K, > > 5% of which is 15 degrees, which is a LOT for any sort of temperature > > controlled environment (say, indoors where people work), assuming your > > circuit doesn't use enough power to changes its own temperature much. > > First of all, I was talking about 2.9%. You added the 2-5% comment in > parenthesis. I wouldn't want to run RS-232 with 5% clock mismatch. > > The problem is not as much stability as initial accuracy. It will be very > difficult to find a resistor and capacitor that together have no more than a > 3% error. Most of this error will come from initial part tolerances, but > you should probably allow 1% or so for drift due to time and temperature. > This leaves 2% for the resistor plus the capacitor tolerance. > > Yes, you could use a pot and tweak it at production time, but that will be > MUCH more expensive and less reliable than just using a crystal or ceramic > resonator in the first place. > > ***************************************************************** > Olin Lathrop, embedded systems consultant in Devens Massachusetts > (978) 772-3129, olinKILLspamembedinc.com, http://www.embedinc.com > > -- > http://www.piclist.com hint: The list server can filter out subtopics > (like ads or off topics) for you. See http://www.piclist.com/#topics -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

>Hi! > >Now, I am connecting a crystal (14.7456Mhz) to CD74HC4060 (binary >counter), and >get a signal of 115.2khz for 4060 pin 6. > >And I like to know is there anyway to use a 555 timer to generate a >stable, clean (sharp >waveform) of 115.2khz. > No. A 555 isn't stable enough, and doesn't give full swing unless you use a CMOS version. If you use that same crystal with a PIC, and count exactly 32 cycles you'll have 115,200 Hz though. J. Nagy Elm Electronics ICs for Experimenters http://www.elmelectronics.com/ -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics