Does temperature affect clocks?

[Carey Carlan]

I was running wild sound for a video recently and the audio ran long by
about half a second over 70 minutes. That hasn't happened before. We were
shooting in an outdoor pavilion with temperatures in the 90s. Can heat
affect the speed of a quartz crystal?

[Arny Krueger]

"Carey Carlan" > wrote in message
...

>I was running wild sound for a video recently and the audio ran long by
> about half a second over 70 minutes. That hasn't happened before. We
> were
> shooting in an outdoor pavilion with temperatures in the 90s. Can heat
> affect the speed of a quartz crystal?

Not all audio interfaces use quartz crystals. Some use ceramic resonators.

Either way the temperature coefficient of oscillators like these is strong
enough that using a thermal sensor or oven to compensate for thermal drift
is not uncommon.

Most crystals are +/- 0.005% tolerance or better at 20 degrees C. 0.005%
over 70 minutes is about a quarter of a second.

So, half the problem you observed was *not* due to temperature but due to
random variations in initial value of clock frequency in just one of the two
devices involved.

So, there's two devices and if they happen to be at opposite ends of
tolerance, we've predicted your actual problem, no? ;-)

[Scott Dorsey]

In article >,
Carey Carlan > wrote:
>I was running wild sound for a video recently and the audio ran long by
>about half a second over 70 minutes. That hasn't happened before. We were
>shooting in an outdoor pavilion with temperatures in the 90s. Can heat
>affect the speed of a quartz crystal?

Sure. Also, were you using the same video recorder that you had been using
before?

It never hurts to add a few extra slates here and there.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[William Sommerwerck]

"Arny Krueger" > wrote in message
...

> Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
> 0.005% over 70 minutes is about a quarter of a second.

You forgot the percent sign (an easy mistake). It would actually be a bit
over 0.002 seconds.


> So, half the problem you observed was *not* due to temperature but
> due to random variations in initial value of clock frequency in just one
> of the two devices involved.

> So, there's two devices and if they happen to be at opposite ends of
> tolerance, we've predicted your actual problem, no? ;-)

No. Probably not.

A half-second over 70 minutes is a fairly severe timing error (in terms of
the real-world TB stability possible), though well-within the range that a
DAC can (or should be able to) lock onto. * Whether it constitutes a
/practical/ problem depends on whether you're able to achieve and maintain
sync over the 70 minutes.

* Your average DAC has no trouble locking onto the 44.058kHz sampling rate
from a video-based digital recording system (eg, the PCM-F1).

[Mark]

>
> > Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
> > 0.005% over 70 minutes is about a quarter of a second.
>
> You forgot the percent sign (an easy mistake). It would actually be a bit
> over 0.002 seconds.
>
>
huh???

0.005/100 * 60 seconds/minute * 70 minutes = 0.21 seconds

Mark

[Arny Krueger]

"William Sommerwerck" > wrote in message
...
> "Arny Krueger" > wrote in message
> ...
>
>> Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
>> 0.005% over 70 minutes is about a quarter of a second.
>
> You forgot the percent sign (an easy mistake). It would actually be a bit
> over 0.002 seconds.


Hmm

calculator says:

0.005/100 * 70 * 60 = 0.21

or

70*60 = 4200

4200/100 = 42

42 * 0.005 = 0.210

where did I go wrong?

[William Sommerwerck]

"Arny Krueger" > wrote in message
...
>
> "William Sommerwerck" > wrote in message
> ...
> > "Arny Krueger" > wrote in message
> > ...
> >
> >> Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
> >> 0.005% over 70 minutes is about a quarter of a second.
> >
> > You forgot the percent sign (an easy mistake). It would actually be a
bit
> > over 0.002 seconds.
>
>
> Hmm
> calculator says:

> 0.005/100 * 70 * 60 = 0.21

> or

> 70*60 = 4200
> 4200/100 = 42
> 42 * 0.005 = 0.210

> where did I go wrong?

Duh... Hmmm... Sorry about that.

"My mind is going, Dave. I can feel it."

[Don Pearce[_3_]]

On Thu, 26 May 2011 14:00:46 -0400, "Arny Krueger" >
wrote:

>
>"William Sommerwerck" > wrote in message
...
>> "Arny Krueger" > wrote in message
>> ...
>>
>>> Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
>>> 0.005% over 70 minutes is about a quarter of a second.
>>
>> You forgot the percent sign (an easy mistake). It would actually be a bit
>> over 0.002 seconds.
>
>
>Hmm
>
>calculator says:
>
>0.005/100 * 70 * 60 = 0.21
>
>or
>
>70*60 = 4200
>
>4200/100 = 42
>
>42 * 0.005 = 0.210
>
>where did I go wrong?
>

Google says

http://www.soundthoughts.co.uk/look/percent.png

d

[Ben Bradley[_2_]]

On Thu, 26 May 2011 18:42:58 GMT, (Don Pearce) wrote:

>
>Google says
>
>http://www.soundthoughts.co.uk/look/percent.png
>
>d

You made a screen image of a Google result?

http://www.google.com/search?q=.005%25+*+70+minutes

[garyvee]

On May 26, 6:28*am, Carey Carlan > wrote:
> I was running wild sound for a video recently and the audio ran long by
> about half a second over 70 minutes. *That hasn't happened before. *We were
> shooting in an outdoor pavilion with temperatures in the 90s. *Can heat
> affect the speed of a quartz crystal?

I know this isn't an answer to your question but I highly recommend
the book "Latitude" by Dava Sobel. It's a great historical account
about the quest for a simple and practical method to determine a
ship's latitude in the 18th century. The primary focus is on clocks
designed by John Harrison to determine latitude at sea. Solving the
temperature problem was part of his solution.

Gary V

[Bill Graham]

Carey Carlan wrote:
> I was running wild sound for a video recently and the audio ran long
> by about half a second over 70 minutes. That hasn't happened before.
> We were shooting in an outdoor pavilion with temperatures in the 90s.
> Can heat affect the speed of a quartz crystal?

Yup.

Temperature effects
A crystal's frequency characteristic depends on the shape or 'cut' of the
crystal. A tuning fork crystal is usually cut such that its frequency over
temperature is a parabolic curve centered around 25 °C. This means that a
tuning fork crystal oscillator will resonate close to its target frequency
at room temperature, but will slow down when the temperature either
increases or decreases from room temperature. A common parabolic coefficient
for a 32 kHz tuning fork crystal is -0.04 ppm/°C².


In a real application, this means that a clock built using a regular 32 kHz
tuning fork crystal will keep good time at room temperature, lose 2 minutes
per year at 10 degrees Celsius above (or below) room temperature and lose 8
minutes per year at 20 degrees Celsius above (or below) room temperature due
to the quartz crystal.

(From Wikipedia)

[Don Pearce[_3_]]

On Thu, 26 May 2011 20:09:32 -0400, Ben Bradley
> wrote:

>On Thu, 26 May 2011 18:42:58 GMT, (Don Pearce) wrote:
>
>>
>>Google says
>>
>>http://www.soundthoughts.co.uk/look/percent.png
>>
>>d
>
> You made a screen image of a Google result?
>
>http://www.google.com/search?q=.005%25+*+70+minutes

I did, didn't I?

d

[Arny Krueger]

"Don Pearce" > wrote in message
...
> On Thu, 26 May 2011 14:00:46 -0400, "Arny Krueger" >
> wrote:
>
>>
>>"William Sommerwerck" > wrote in message
...
>>> "Arny Krueger" > wrote in message
>>> ...
>>>
>>>> Most crystals are +/- 0.005% tolerance or better at 20 degrees C.
>>>> 0.005% over 70 minutes is about a quarter of a second.
>>>
>>> You forgot the percent sign (an easy mistake). It would actually be a
>>> bit
>>> over 0.002 seconds.
>>
>>
>>Hmm
>>
>>calculator says:
>>
>>0.005/100 * 70 * 60 = 0.21
>>
>>or
>>
>>70*60 = 4200
>>
>>4200/100 = 42
>>
>>42 * 0.005 = 0.210
>>
>>where did I go wrong?
>>
>
> Google says
>
> http://www.soundthoughts.co.uk/look/percent.png
>

I can live with that~ ;-)

[Carey Carlan]

(Scott Dorsey) wrote in
:

> In article >,
> Carey Carlan > wrote:
>>I was running wild sound for a video recently and the audio ran long
>>by about half a second over 70 minutes. That hasn't happened before.
>>We were shooting in an outdoor pavilion with temperatures in the 90s.
>>Can heat affect the speed of a quartz crystal?
>
> Sure. Also, were you using the same video recorder that you had been
> using before?
>
> It never hurts to add a few extra slates here and there.
> --scott

Same cameras (two of them). Ran an indoor concert just the week before
with no hiccups.

Had a slate at the beginning and the conductor put her hand to her chest at
the end of the program (in a *whew* gesture) and banged the lav. That gave
me my closing slate. Worked OK, but I hadn't seen the problem before.

[Carey Carlan]

garyvee > wrote in news:22146ae1-219a-4db8-aea8-
:

> On May 26, 6:28*am, Carey Carlan > wrote:
>> I was running wild sound for a video recently and the audio ran long by
>> about half a second over 70 minutes. *That hasn't happened before. *W
> e were
>> shooting in an outdoor pavilion with temperatures in the 90s. *Can heat
>> affect the speed of a quartz crystal?
>
> I know this isn't an answer to your question but I highly recommend
> the book "Latitude" by Dava Sobel. It's a great historical account
> about the quest for a simple and practical method to determine a
> ship's latitude in the 18th century. The primary focus is on clocks
> designed by John Harrison to determine latitude at sea. Solving the
> temperature problem was part of his solution.

I'll look for it in the library. Saw a PBS special (Nova?) on that years
ago and was totally unimpressed with Harrison's design. He kept putting
patches on top of other patches rather than refining the original concept.
If I wrote software like that (a) it wouldn't fit, (b) I couldn't maintain
it and (c) it would be slow as Christmas.

[Mark]

>
> The challenge is implementing a TXCO in *products like the Microtrack or
> Zoom H4 as well as the ubiquitous camcorders. *They are doing marvelous
> things with technology, but like the $3 thousand lumen LED light bulb, its a
> few years off.- Hide quoted text -


It's not that much of a challange anymore.

0.005% = 50ppm.

TXCOs that are better then 1 PPM over temperature are routine these
days and they are tiny..

http://www.greenrayindustries.com/tcxo2.html

just for one example


Mark

[Scott Dorsey]

Arny Krueger > wrote:
>
>The challenge is implementing a TXCO in products like the Microtrack or
>Zoom H4 as well as the ubiquitous camcorders. They are doing marvelous
>things with technology, but like the $3 thousand lumen LED light bulb, its a
>few years off.

There's one in the RME ADI-8. And my Prism has an oven.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[William Sommerwerck]

>> I know this isn't an answer to your question but I highly recommend
> the book "Latitude" by Dava Sobel. It's a great historical account
> about the quest for a simple and practical method to determine a
> ship's latitude in the 18th century. The primary focus is on clocks
> designed by John Harrison to determine latitude at sea. Solving the
> temperature problem was part of his solution.

"Longitude", not "Latitude".

[William Sommerwerck]

"Arny Krueger" > wrote in message
...

>> Google says
>> http://www.soundthoughts.co.uk/look/percent.png

> I can live with that~ ;-)

And I apologize for my momentary lapse. (No wisecracks, Arny. And I mean
it.) I figured out why I made the mistake I did.

[Arny Krueger]

"Mark" > wrote in message
...
>
> The challenge is implementing a TXCO in products like the Microtrack or
> Zoom H4 as well as the ubiquitous camcorders. They are doing marvelous
> things with technology, but like the $3 thousand lumen LED light bulb, its
> a
> few years off.- Hide quoted text -


It's not that much of a challange anymore.

0.005% = 50ppm.

TXCOs that are better then 1 PPM over temperature are routine these
days and they are tiny..

http://www.greenrayindustries.com/tcxo2.html

just for one example



Looked at Digikey's pricing and the cheapest was $25. Not going to be in
Microtracks or Zoom H4s any time soon. They might fit in the case, though!
;-)

[Arny Krueger]

"William Sommerwerck" > wrote in message
...
> "Arny Krueger" > wrote in message
> ...
>
>>> Google says
>>> http://www.soundthoughts.co.uk/look/percent.png
>
>> I can live with that~ ;-)
>
> And I apologize for my momentary lapse. (No wisecracks, Arny. And I mean
> it.) I figured out why I made the mistake I did.


No problem, William.

We all try! ;-)

[Sean Conolly]

"Carey Carlan" > wrote in message
...

> I'll look for it in the library. Saw a PBS special (Nova?) on that years
> ago and was totally unimpressed with Harrison's design. He kept putting
> patches on top of other patches rather than refining the original concept.
> If I wrote software like that (a) it wouldn't fit, (b) I couldn't maintain
> it and (c) it would be slow as Christmas.

IOW - a typical C application - I've seen many that fit that description :-)

Sean

[William Sommerwerck]

> I'll look for it in the library. Saw a PBS special (Nova?) on that years
> ago and was totally unimpressed with Harrison's design. He kept putting
> patches on top of other patches rather than refining the original concept.
> If I wrote software like that (a) it wouldn't fit, (b) I couldn't maintain
> it and (c) it would be slow as Christmas.

By the way, the TV version is a dumbing-down of the book. It refuses to get
its hand dirty with the math, science, and engineering involved.

[Bill Graham]

William Sommerwerck wrote:
>>> I know this isn't an answer to your question but I highly recommend
>> the book "Latitude" by Dava Sobel. It's a great historical account
>> about the quest for a simple and practical method to determine a
>> ship's latitude in the 18th century. The primary focus is on clocks
>> designed by John Harrison to determine latitude at sea. Solving the
>> temperature problem was part of his solution.
>
> "Longitude", not "Latitude".

Yes. Latitude is easy to determine. You just shoot the sun at high noon, and
its deviation from straight up is your latitude. Unfortunately, longitude
requires exact knowledge of the time of day, and this is why accurate
navigation depended on accurate time keeping.

[Paul[_13_]]

On May 26, 6:28*am, Carey Carlan > wrote:
> I was running wild sound for a video recently and the audio ran long by
> about half a second over 70 minutes. *That hasn't happened before. *We were
> shooting in an outdoor pavilion with temperatures in the 90s. *Can heat
> affect the speed of a quartz crystal?


Yes, as the other posters have mentioned, it's
parts per million per degree Centigrade, for any
crystal oscillator.

But is there a way for you to synchronize your
audio to your video, by connecting their clocks, so
they use the same clock?