Fwd: Re: Missing data

[Istvan Retaller]

Sziasztok!
Mit szoltok hozza, hogy Atmelek szerint gyakorlatilag nincs feleledesi
ideje a processzorba epitett komparatornak?
Udv, RI


------- Start of forwarded message -------
From: avr at atmel.com
To: istvan.retaller at evoran.hu
Subject: Fwd: Re: Missing data
Date: 2002.11.13. 14:18:23


Dear customer,

this time is something we have never measured very exactly.
I.e., the best-case is when the signal swing on the AC inputs is large; then
the propagation delay is at its shortest.
Under these circumstances, and even @24MHz, the delay has never been above 1
cycle. And with smaller signals, delays are not far from this.
So the limit of a few hundred ns is VERY generous.

And the AC has always been used in-house like this:
1) switch it on (ACSR)
2) use it immediately.

I.e. no delay has ever been implemented, measurement code follows enabling
immediately.
This has never proved to produce unaccounted variance in the measurements.

Does this answer your question?

Best Regards,
Eivind A. Sivertsen

Please keep the complete email correspondence in any reply!
----------------------------------------------------------------------------
---------------------
AVR support mail        mailto:avr at atmel.com
Atmel AVR page http://www.atmel.com/atmel/products/prod23.htm
Info and software http://www.avrfreaks.net
Discussion forum http://www.avrfreaks.net/phorum/
Third party vendors http://www.atmel.com/atmel/products/prod205.htm
List of distributors http://www.atmel.com/atmel/contacts


----- Original Message -----
From: "Istvan Retaller" <istvan.retaller at evoran.hu>
To: <avr at atmel.com>
Sent: Tuesday, November 12, 2002 1:47 PM
Subject: Re: Missing data


> Dear Mr. Sivertsen,
> I am afraid I did not describe my question correctly.
> There is bit7 of ACSR register to power on/power off the ADC.
> It is kept continously off to spare the 1mA consumption of the ADC.
> Then I power it in each 4th seconds for a short period.
>
> The thing I do not understand:
> Even if I power-on it by clearing bit7 of ACSR register, an
> analogue device needs some time to be ready after powered up.
> Of course I can use it immediately, but there must be some limit
> until it estabilishes and supports _correct_ results. In my guess
> it should be a few times 10 ms range, but I would like to know
> the correct answer.
> Best rgds,
> Istvan Retaller
>
> 2002.11.12. 12:57:58, <avr at atmel.com> wrote:
>
> >
> >Dear customer,
> >
> >the analog comparator has practically no "warm-up time".
> >The propagation delay (page 72 in datasheet) is typically 750 or 500 ns
> >according to Vcc.
> >If you are wondering about any due waiting time after power up etc., in
> >practice the AC will be available when the unit is powered up and the
clock
> >is running.
> >
> >Best Regards,
> >Eivind A. Sivertsen
> >
> >Please keep the complete email correspondence in any reply!
>

----------------------------------------------
>>> Gentlemen, I am working on a battery-operated equipment and to keep
>>> power consumption as low as possible, I keep analog comparator mainly off
>>> and switch back for the minimum time to check battery. But... I
>>> cannot find any data about the warm-up time of the analog comparator
>>> (datasheet: 0839.pdf) Please inform me as soon as possible
>>> Regards, Istvan Retaller

-------- End of forwarded message --------