How inaccurate is a 555 or 7555 REALLY?

[[email protected]]

"mri_bob"


** True - the actual temp range figure is +/- 3 degrees C for +/- 0.1%
frequency variation.

So totally impractical in the real world.

Unless you find yourself a crystal oven and keep it at a steady 70C, or
build yourself a proper temperature controller around a Peltier
junction - the one I published managed
+/-0.001C, but would have been a tight fit in a 3/4 inch package.

** He does that kinda stuff a lot.

Sometimes. I don't get too worried about such allegations until they
put down their own numbers.

** Nice to see you read my explanation.

No temp change = no frequency drift.

Not always true. Inductors have a nasty tendency to shift inductance as
the DC current through the coil changes, even if you keep the
temperature constant (whichcan be a little difficult if you are
changing the heat dissipated in the coil).

 

[[email protected]]

i did not make up the .005% spec for the 555. It is on the first page of
National's data sheet. the poster said he wanted .1% and i said the 555
would not be the controlling factor.

It isn't the whole story, like most manufacturer's specifications.

as for prototype/production issues, the guy said he wanted to make one, he
didn't say 1000. your experience in a production environment is
irrelevant.

If he isn't in a production environment, how many capacitors does he
have to buy to find one that drifts the right way to the right extent?
Bearing in mind that capacitors from a single batch are usualy a lot
more similar that parts from successive batches.

Selecting parts is a mugs games.

i said crystals were good enough. i don't know why you are arguing that
point? you seem to be misstating my position on that one just so you can
disagree with me.

10MHz crystals are a lot better than 32768Hz crystals. My Farnell
catalogue is back i Nijmegen, which made it difficult to be
quantitative.

i said .1% is a minute and a half a DAY and a $2 watch is better than
that,

Mainly because it is temperature controlled by your wrist.

then you rant about needing to go to a lot of trouble to get .5
seconds per day and say that is .01% Well, there are 86,400 seconds in a
day so .01% is 8.6 seconds. You are off by over an order of magnitude and
you are berating me saying i don't know what i'm talking about?

That was a typo - it should have been 0.001% or +/-5ppm. I used to
worry about that sort of error - now I just try and cross-check.
Everybody makes an error like that about once every 30 minutes when
sober ...

and you do not need automatic frequency control in a radio to keep it from
drifting 100 kHz.

Once it has warmed up

nor do you need a PLL. a cheap simple fm ham receiver

Which has an operator to provide AFC via a slow PLL

with no AFC will not drift that much by an order of magnitude.

Regular FM radios have a PLL to provide AFC, lacking an operator
prepared to act as a dumb servo ina slow loop.

To quote you: "Try to understand the technology you pontificate about before you
claim to be getting real"

I do, you don't. Pity about that.

 

[Jim Thompson]

On Sun, 03 Dec 2006 06:44:33 GMT, Robert Baer

[snip]

Ignoring the insufficent supply voltage, the basic inaccuracy of the
555 design rests on the offset voltage of the comparitors combined with
the mismatch of the internal resistor divider ratios.

And once you choose an R-C, what does those irregularities have to do
with stability?

Most of the temperature variation in a 555 timer would come from the
delay time through the comparators to the output.

Crudely speaking, 0.1% cannot be achieved with one 555.

Probably not.

If you want a relatively accurate frequency, look to crystal or
tuning fork oscillators using discrete SMT parts.

Agreed.

...Jim Thompson

 

[mri_bob]

To quote you: "Try to understand the technology you pontificate about
before you claim to be getting real"
I do, you don't. Pity about that.



i guess i will have to turn in my engineering diploma then. i have run
into clowns on newsgroups before who attacked other peoples ideas just so
they could tell their own stories about how much they know, it's nothing
new.

i was attempting to help a novice learn why it would be difficult but
maybe not impossible to do it with a 555 if temperature was tightly
controlled but that a microcomputer circuit would be better. bill has his
own agenda which is different from mine. i'm finished with this topic.
people who need to end their posts with insults have nothing to offer to
me.

-bob

 

[WildIrish]

Thanks for all the great and useful data and information everyone,
I really started a firestorm of argument on this topic. I'm happy
to switch to the dividing circuits, but i'm stuck with using AA's... is
there
any packages that anyone can think of which will divide at that
low of voltage that doesn't have minimum frequency limitations
yet still pretty small?

Programmable solutions are acceptable, however I've gone
crosseyed looking at datasheets that all show minium 3v with
min Fout of xxxkHz instead of low hz ratings.

~Brendan

PS to Phil - not everyone is as worldly as you, I humbly described
myself
as a novice so as not to apear rude, or to get answers that are
rediculously
over my head. Everyone has to learn by asking questions, that is the
notion
behind discussion forums.... not for certain individuals to show off
how
'amazing' they are. 'Novice' is is a relative term... My engineering
degree is
still novice compared to a NASA engineer. I never said 'no experience'
so
your statement of "Then it is time to realise that YOU are not up to
the task"
is rude and uncalled for.... thanks for the positivity.

 

[Robert Baer]

"mri_bob"




** True - the actual temp range figure is +/- 3 degrees C for +/- 0.1%
frequency variation.

So totally impractical in the real world.






** He does that kinda stuff a lot.







** Nice to see you read my explanation.

No temp change = no frequency drift.







** No offence taken here.

Frequency stability of oscillators is a non trivial topic.



..... Phil

But making an oven to keep temperature within 3C is not difficult, so
that part is not "totally impractical in the real world".

 

[Robert Baer]

It isn't the whole story, like most manufacturer's specifications.




If he isn't in a production environment, how many capacitors does he
have to buy to find one that drifts the right way to the right extent?
Bearing in mind that capacitors from a single batch are usualy a lot
more similar that parts from successive batches.

Selecting parts is a mugs games.




10MHz crystals are a lot better than 32768Hz crystals. My Farnell
catalogue is back i Nijmegen, which made it difficult to be
quantitative.




Mainly because it is temperature controlled by your wrist.




That was a typo - it should have been 0.001% or +/-5ppm. I used to
worry about that sort of error - now I just try and cross-check.
Everybody makes an error like that about once every 30 minutes when
sober ...




Once it has warmed up





Which has an operator to provide AFC via a slow PLL




Regular FM radios have a PLL to provide AFC, lacking an operator
prepared to act as a dumb servo ina slow loop.




I do, you don't. Pity about that.

? "Nijmegen" ?
*What* is that?

 

[Robert Baer]

On Sun, 03 Dec 2006 06:44:33 GMT, Robert Baer

[snip]

Ignoring the insufficent supply voltage, the basic inaccuracy of the
555 design rests on the offset voltage of the comparitors combined with
the mismatch of the internal resistor divider ratios.

And once you choose an R-C, what does those irregularities have to do
with stability?

Most of the temperature variation in a 555 timer would come from the
delay time through the comparators to the output.

Crudely speaking, 0.1% cannot be achieved with one 555.

Probably not.

If you want a relatively accurate frequency, look to crystal or
tuning fork oscillators using discrete SMT parts.

Agreed.

...Jim Thompson

That ratio is temperature sensitive...

 

[Jim Thompson]

On Sun, 03 Dec 2006 06:44:33 GMT, Robert Baer

[snip]

Ignoring the insufficent supply voltage, the basic inaccuracy of the
555 design rests on the offset voltage of the comparitors combined with
the mismatch of the internal resistor divider ratios.

And once you choose an R-C, what does those irregularities have to do
with stability?

Most of the temperature variation in a 555 timer would come from the
delay time through the comparators to the output.

Crudely speaking, 0.1% cannot be achieved with one 555.

Probably not.

If you want a relatively accurate frequency, look to crystal or
tuning fork oscillators using discrete SMT parts.

Agreed.

...Jim Thompson

That ratio is temperature sensitive...

What "ratio" is that?

...Jim Thompson

 

[mri_bob]

in passing, i would like to address bill's ridiculous claim that a digital
watch was only accurate to better than a minute or two a day because it
was temperature regulated by the wearer's wrist. ridiculous nonsense that
shows an ignorance of thermodynamics and of the stability of digital
timepieces in general. How do you suppose the hundreds of millions of
cheap digital wall clocks are thermally regulated?

-bob

 

[Spehro Pefhany]

Thanks for all the great and useful data and information everyone,
I really started a firestorm of argument on this topic. I'm happy
to switch to the dividing circuits, but i'm stuck with using AA's... is
there
any packages that anyone can think of which will divide at that
low of voltage that doesn't have minimum frequency limitations
yet still pretty small?

Programmable solutions are acceptable, however I've gone
crosseyed looking at datasheets that all show minium 3v with
min Fout of xxxkHz instead of low hz ratings.

~Brendan

Are you actually trying to generate a sine wave output?


Best regards,
Spehro Pefhany

 

[Kiviranta, Mikko]

? "Nijmegen" ?
*What* is that?

A town in the Netherlands, whose roots lie in the Novio Magnus
military camp by the ancient Romans. Also the home of the Katholieke
Universiteit where my wife went to the medical school.

Regards,
Mikko

 

[[email protected]]

before you claim to be getting real"
I do, you don't. Pity about that.



i guess i will have to turn in my engineering diploma then.

I wouldn't bother making the effort. Diploma's don't count for much.
The IEEE admitted me as member 26 years ago without any evidence of any
academic training in engineering (I do have a Ph.D. in Physical
Chemistry) on the basis of a declaration from one of the IEEE members
at EMI Central Research that I knew what I was talking about. I'd
gotten two patents i the three years I'd worked there, so they may have
had an exaggerated idea of my competence.

i have run into clowns on newsgroups before who attacked other peoples ideas just so
they could tell their own stories about how much they know, it's nothing
new.

And it does tend to stop people sounding off about stuff they don't
know enough about.

i was attempting to help a novice learn why it would be difficult but
maybe not impossible to do it with a 555 if temperature was tightly
controlled but that a microcomputer circuit would be better. Bill has his
own agenda which is different from mine.

Based on a fairly wearing month or so back in 1974 trying to do the
same job for an over-optimistic boss. We ended up getting fairly close
with an emitter-coupled multi-vibrator, but the temperature dependence
of the Early effect was bad enough that even that circuit didn't make
it.

i'm finished with this topic.
people who need to end their posts with insults have nothing to offer to
me.

Then you are condemned out of your own mouth.

Your second message in the thread - message 6 in the lst ends " let's
get real, huh? "
which is the kind of discourtesy I repay in kind. My own idea of my
attitude on this news group is "polite until provoked" but John Larkin
does keep on finding insults in my posts that I could have sworn
weren't there when I composed them (and don't look much like insults to
me when I reread them). Phil Allison would blame this on autism, but he
blames a lot of stuff on autism. He's not an expert psychologist,
unlike my wife's academic colleagues (who won't use me as a test
subject any more, because I know too much).

 

[Robert Baer]

Jim Thompson wrote:

On Sun, 03 Dec 2006 06:44:33 GMT, Robert Baer

[snip]


Ignoring the insufficent supply voltage, the basic inaccuracy of the
555 design rests on the offset voltage of the comparitors combined with
the mismatch of the internal resistor divider ratios.


And once you choose an R-C, what does those irregularities have to do
with stability?

Most of the temperature variation in a 555 timer would come from the
delay time through the comparators to the output.



Crudely speaking, 0.1% cannot be achieved with one 555.


Probably not.



If you want a relatively accurate frequency, look to crystal or
tuning fork oscillators using discrete SMT parts.


Agreed.

...Jim Thompson

That ratio is temperature sensitive...

What "ratio" is that?

...Jim Thompson

The ratio of the ersistor calues in the (comparitor) divider.

 

[Robert Baer]

I wouldn't bother making the effort. Diploma's don't count for much.
The IEEE admitted me as member 26 years ago without any evidence of any
academic training in engineering (I do have a Ph.D. in Physical
Chemistry) on the basis of a declaration from one of the IEEE members
at EMI Central Research that I knew what I was talking about. I'd
gotten two patents i the three years I'd worked there, so they may have
had an exaggerated idea of my competence.




And it does tend to stop people sounding off about stuff they don't
know enough about.




Based on a fairly wearing month or so back in 1974 trying to do the
same job for an over-optimistic boss. We ended up getting fairly close
with an emitter-coupled multi-vibrator, but the temperature dependence
of the Early effect was bad enough that even that circuit didn't make
it.

** And *nobody* thought of using cascode?

 

[Jim Thompson]

Jim Thompson wrote:


On Sun, 03 Dec 2006 06:44:33 GMT, Robert Baer

[snip]


Ignoring the insufficent supply voltage, the basic inaccuracy of the
555 design rests on the offset voltage of the comparitors combined with
the mismatch of the internal resistor divider ratios.


And once you choose an R-C, what does those irregularities have to do
with stability?

Most of the temperature variation in a 555 timer would come from the
delay time through the comparators to the output.



Crudely speaking, 0.1% cannot be achieved with one 555.


Probably not.



If you want a relatively accurate frequency, look to crystal or
tuning fork oscillators using discrete SMT parts.


Agreed.

...Jim Thompson

That ratio is temperature sensitive...

What "ratio" is that?

...Jim Thompson

The ratio of the ersistor calues in the (comparitor) divider.

What makes you think resistor ratios have a TC?

Solid ratios is the only reason we can build analog IC's at all.

...Jim Thompson

 

[Phil Allison]

PS to Phil - not everyone is as worldly as you, I humbly described
myself
as a novice so as not to apear rude, or to get answers that are
rediculously
over my head. Everyone has to learn by asking questions, that is the
notion
behind discussion forums.... not for certain individuals to show off
how
'amazing' they are. 'Novice' is is a relative term... My engineering
degree is
still novice compared to a NASA engineer. I never said 'no experience'
so
your statement of "Then it is time to realise that YOU are not up to
the task"
is rude and uncalled for.... thanks for the positivity.

** YOU called yourself a: " novice with minor experience "

- was I supposed to think otherwise ?

Your posts show a total lack of practical electronics knowledge.

Your alleged "engineering degree " is a worthless scrap of paper.

I absolutely stand by my comment:

" The size and performance criteria you have come up with make it a job for
an
expert. "


You still have a lot of growing to do - pal.




........ Phil

 

[Phil Allison]

"mri_bob"

in passing, i would like to address bill's ridiculous claim that a digital
watch was only accurate to better than a minute or two a day because it
was temperature regulated by the wearer's wrist. ridiculous nonsense that
shows an ignorance of thermodynamics and of the stability of digital
timepieces in general. How do you suppose the hundreds of millions of
cheap digital wall clocks are thermally regulated?

** Jeeze - this desperate Bob wanker only ever opens his fat mouth to
change feet.

Must be a close relative of " Joe Blow"

Obviously, a worn watch undergoes a much smaller range of temperature
variation than one that it not - hence the stability of its crystal based
time keeping is much enhanced.

Battery operated clocks rely on the INDOOR room temp having a *long term *
average that is not too far from 25C - their makers claim an accuracy of
about 30 seconds a month, which is nothing wonderful.

Bob - do you have a diploma in utter stupidity or what ?




....... Phil

 

[David L. Jones]

Thanks for all the great and useful data and information everyone,
I really started a firestorm of argument on this topic. I'm happy
to switch to the dividing circuits, but i'm stuck with using AA's... is
there
any packages that anyone can think of which will divide at that
low of voltage that doesn't have minimum frequency limitations
yet still pretty small?

Programmable solutions are acceptable, however I've gone
crosseyed looking at datasheets that all show minium 3v with
min Fout of xxxkHz instead of low hz ratings.

~Brendan

Hi Brendan
Standard 74HC logic can work down to 2V, and you can get low voltage
crystal oscillator modules. Some micros like the PIC and Atmel AVR
series can work down to 2V too. That's enough headroom for a 2xAA
battery supply.

You can also make your life a bit easier if needed by using a DC-DC
converter to give you 5V from however many batteries you have.

Dave
P.S. Ignore Phil, that's just the way he is unfortunately.

 

[WildIrish]

Thanks for your help. I can potentially bump up to 3v.
Is there any 74HC or similar products that can perform
divisions high enough to divide down a 10mhz oscillator in only ONE
package?
Someone early on suggested using a pair or 74HC's to do the job, is
there
any off the shelf dividers in on package capable of this assuming I
switch
to a 3V supply?

What dividers are used by wristwatches? they
must take a 32768hz oscillator and divide it WAY down until it gets one
second
bursts to advance the stepper motors? What dividers do they use, they
run at 1.5v right?

P.S. Ignore Phil, that's just the way he is unfortunately.

Thanks. That's unfortunate, he seems to know his stuff but exibits
social
issues.... i'd be afraid to work with a guy like him. Phil if you read
this,
don't waste your time on this thread anymore, we are all tired of you
it appears,
and I don't need or want your help. HOWEVER, I appreciate everyone
else's
time and energy in helping me to bridge gaps in my knowledge.

~Brendan