Ok, guys, another case where I have to ditch the uC idea and go semi-discrete. Need a couple long timers so I am looking at the old CD4060, hoping to find some in TSSOP stock.
For 5V it states the frequency as 23kHz +/-10%. But the usual formula is f=1/(2.2*Rx*Cx) and that comes to about 45kHz. I can't imagine these things to have more than a few ten pF in internal capacitances so this doesn't compute. Look at the 4th page (page 3-161):
Any ideas why?
-- Regards, Joerg http://www.analogconsultants.com
-- If you\'re looking for a timer, then you shouldn\'t be concerned about frequency, you should be concerned about how long it takes for the counter\'s outputs to get to where you want them to be after a reset, and T ~ 2.2RxCx will give you that, I think.
Guess I'll just hook one up and check it out.
-- Regards, Joerg http://www.analogconsultants.com
Looking at the datasheet on page 3-160, the chart says that the input capacitance on any input is 5pf (typ) to 7.5pf (max). So that answers the question about internal capacitances. However, the propagation delays vary over about 2:1 range, which corresponds to the freq variation you questioned. Would seem that the delays are the culprits here.
Cheers!!
-- Dave M MasonDG44 at comcast dot net (Just substitute the appropriate characters in the address) Some days you\'re the dog, some days the hydrant.
-- Excellent! Reality is the best test, no?
Why? What's wrong with an AtTiny13 running from its internal 128kHz oscillator?
-- John B
"John B" schrieb im Newsbeitrag news:45b15cf3$0$759$ snipped-for-privacy@master.news.zetnet.net... | On 19/01/2007 Joerg wrote: | | > Ok, guys, another case where I have to ditch the uC idea and go | | | Why? What's wrong with an AtTiny13 running from its internal 128kHz | oscillator?
Needs extra programming = extra cost?!
- Henry
--
Sure is. But there is always that weird aftertaste when you test a part and the results corroborate what you are thinking while pointing out that the data sheet ain't quite right. OTOH this would certainly not be the first time that happens. Last time was a month ago on a TEC controller chip. From Analog Devices no less.
-- Regards, Joerg http://www.analogconsultants.com
over
Would
It would take quite some delay to scoot an oscillator from 23kHz to
45kHz. Something like 20usec. That would be a stunning prop delay ;-)-- Regards, Joerg http://www.analogconsultants.com
Nothing wrong, except that this circuit needs to operate between 4V and
9.5V for VCC. Plus the two timers must be independant for safety reasons. And probably we won't have a power switch, meaning Iq must be a uA or so. [...]-- Regards, Joerg http://www.analogconsultants.com
I know you might not like NXP..... but, for me that was quite painless.
Page 3-160) Figure 12) on the TI Dirty Sheet shows your equation for the 'oscillator' frequency.
Page 5) Figure 4) on the Philips/NXP Data Sheet gives a similar equation. They also give other useful information........ I like Philips and Siemens for Data Sheets....
Anyway,
The 'Logic' diagram on the Philips Data Sheet, Page 3) Figure 3) might make more sense. Notice that there are four flop flips before you get to O3. Flop Flip 1 divides the 'oscillator' frequency by two before the rest of the stuff sees it.
The TI Dirty Sheet is referring to a Clock Signal after Flop Flip 1. Fair enough, the Philips Data Sheet might not specifically make that point but the TI/Harris logic diagram hides it deeper.
Of course if I knew f*ck about logic I would have been able to work it out from the TI Dirty Sheet. But I don't.......
However, having looked at the two.... the TI Dirty Shit might show Flop Flip
1 connected as a Toggle (JK with inverted inputs which could be a D-type) before the rest of the counters get to see it.DNA
Oh, I like them. What I don't like is their marketing "effort" and IMHO their web site is crummy. Their specsmanship was always superb though.
One has to be really, really careful with that stuff. For example Moto/ON often calls the first output Q4 while others call it Q3. That can really throw people a curve when they discover that after the layout is done. We used to have an "anger pad" for such situations. It hung on a wall and read: In case of an upcoming temper trantrum take one sheet, crumple and throw into a corner.
They are basically all used as deviders feeding each other. Not the most modern concept but for a slow timer it's often good enough.
-- Regards, Joerg http://www.analogconsultants.com
Oh, alright then.
You asked..... the difference between about 45KHz and about 23KHz is about a factor of two. It's quite possible that 2 is a digital divide by two so..........
You absolute old patronising bastard you!
Your job interviews must last at least four hours whilst you blither on about the good old days and a few terminal 'war stories'. Then, having bored yourself, you'll forget my name as you ask your secretary to dial me a taxi.
Assuming I get through that load of shit and get the job you might just give me a broken chair and back order a customised bercostat for delivery about the time your Ronny James manages to lock you up a cupboard so they can sort out the power supply pins on an LM324.
Your design reviews will be three hours of nodding at your extreme ability to focus on where to place potentiostat1 on the PCB with a screaming desire to get out, ignore your input and get the job sorted.
Then you'll probably visit three times a day to ask if any of the resistors are a slight shade of blue because something went wrong when you used blue resistors in 1965.
DNA
Ah, that's what you meant. Could, of course, be but what puzzled me was the mention of both on the same datasheet. And they mentioned the oscillator (sans dividers).
Actually, they could last an hour but I asked the people about their stories. Mine I told later after I hired them :-)))
Nah, that was ten years later and what happened was that RF turned a capacitor from white ceramic into green glass. It was amazing.
So, dinner time and then I'll pour us a cold one...
-- Regards, Joerg http://www.analogconsultants.com
OK. So where did I go wrong?
DNA
OK, so where did I go wrong?
DNA
:)
I figure they put the die in rotated by 90 degrees :)
having the power pins between the inputs should help reduce crostalk
Bye. Jasen
On a sunny day (Fri, 19 Jan 2007 20:48:27 GMT) it happened Joerg wrote in :
Maybe you do not like micro controllers, but long precise delays with micros are much easier to make then with discrete or some logic. I assume you have other reasons to dump the micro? Else even the smallest PIC will give you years delay if must be.
Either there is dividing by two, or it uses the RC ramp on both the charge and discharge parts of the cycle..
You didn't. T'was me, I must have misunderstood.
-- Regards, Joerg http://www.analogconsultants.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.