Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Safe Temperature Lowering rate?

  • Please log in to reply
25 replies to this topic

#1 Rick Evans

Rick Evans

    Lift Off

  • -----
  • topic starter
  • Posts: 21
  • Joined: 17 Jul 2013

Posted 22 August 2018 - 05:31 PM

Is there a recommended safe rate to lower the temperature of a cooled CCD or CMOS camera? I recently acquired an ASI294MC Pro.

 

TIA



#2 bobzeq25

bobzeq25

    Hubble

  • *****
  • Posts: 16380
  • Joined: 27 Oct 2014

Posted 22 August 2018 - 05:50 PM

With my ATIK 460EXM I just select the temperature.  The software selects the rate for cool down and warm up automatically.


  • xiando likes this

#3 Rick J

Rick J

    Cosmos

  • *****
  • In Memoriam
  • Posts: 8376
  • Joined: 01 Mar 2008
  • Loc: Mantrap Lake, MN

Posted 22 August 2018 - 08:50 PM

My cameras date back to 2004 when I bought them used.  I use both nearly every clear night.  One an old ST-7 for photometric work and the other an STL-11000XM for color imaging.  I've posted more than 1000 images from the latter here over the years and taken over 1600 objects.  My software dates back to even before 2004 and it, unlike some found today, has no system for slowly raising or lowering the temperature other than doing it manually.  I just cool at max and warm by turning the camera off.  After many thousands of observations, I've never had an issue with this.  I don't know if today's chips are brittle hence the controlled warming and cooling but the two I use have no issue with max rate warming or cooling.  My end of session is almost always at dawn when I've been in bed for hours.  The computer just disconnects from the camera then turns off the power to it and the rest of the gear.  Closes the roof and turns off all power to the observatory after parking the scope.  By then (a couple minutes at most) the CCD is back to ambient usually 35C to 40C warmer than the imagine temperature.  It warms at 20C a minute without harm.  I see no reason to change my system.

 

Rick 



#4 jgraham

jgraham

    ISS

  • *****
  • Posts: 20118
  • Joined: 02 Dec 2004
  • Loc: Miami Valley Astronomical Society

Posted 22 August 2018 - 09:04 PM

I don't know about a recommended rate, but with my ASI294MC Pro (and ASI071MC Pro) I start at 20C, and then lower the set point 5C at a time, letting it settle for a bit at each temperature. I do this as part of my setup routine and it doesn't take a lot of time.

 

A couple of other notes...

 

I cool the camera to 0C in the summer and -10C in the winter. It can go a lot colder, but I like to keep the load at less than 30%. I also aim for stability more than freezing the bajeebies out of the camera.

 

Before disconnecting my camera I set the temperature to 20C and turn the TEC off. That way the TEC doesn't immediately start cooling with I turn it back on.

 

Enjoy your new camera!



#5 Alex McConahay

Alex McConahay

    Cosmos

  • *****
  • Posts: 7974
  • Joined: 11 Aug 2008
  • Loc: Moreno Valley, CA

Posted 22 August 2018 - 09:19 PM

I notice my camera will turn off completely when a sequence for some reason shuts down, and within a minute is back at ambient or so. Does that not mean it warmed up way too fast? Yet, it has not actually broken yet after seven years.

I am not advocating it is good, or safe to ignore the gradual cool down/warm up......but, really, has anyone ever had a problem?

Alex



#6 entilza

entilza

    Soyuz

  • *****
  • Posts: 3716
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 22 August 2018 - 10:01 PM

I'll chime in, my QHY163M I set to -15C at a timer of 7 minutes in SGP. For both warm up and cool down. Never an issue with frost and I assume less thermal shock if there is such a thing for sensors.

#7 happylimpet

happylimpet

    Soyuz

  • *****
  • Posts: 3812
  • Joined: 29 Sep 2013
  • Loc: Southampton, UK

Posted 23 August 2018 - 06:22 AM

I think the main potential canger is rapid heat up from killing the power....peltier coolers generate a lot of heat and if switched off suddenly all that heat (perhaps 60c? bit of a guess) can propagate back to the cooled chip (perhaps -20). So I always run the fans a while after turning the peltier off.

 

But yeah...interested to know if anyones ever had a problem!!!



#8 spokeshave

spokeshave

    Surveyor 1

  • *****
  • Posts: 1979
  • Joined: 08 Apr 2015

Posted 23 August 2018 - 07:19 AM

The real danger is cooling down too quickly, not heating up too quickly. A TEC can introduce a fairly astounding temperature change when switched on at full power - on the order of 100C in a second or so on its face. This has the potential to introduce significant thermal shock. Warming, on the other hand, is a completely passive process. While it is true that the heat on the hot side of the TEC will migrate to the cold side, that takes time and while the entire system will warm to ambient over the course of just a few minutes, that is nothing like the potential thermal shock of an instant-on 100% power TEC. So, the place to exercise caution is during cooldown. That is not to say that a rapid 100% power cooldown is certain to damage your sensor. As Rick pointed out, some cameras can tolerate this repeatedly without damage. However, that doesn't mean that all cameras can. 

 

Tim


  • bobzeq25 and ks__observer like this

#9 CharlesW

CharlesW

    Long time member

  • ***--
  • Posts: 3659
  • Joined: 01 Nov 2012
  • Loc: Chula Vista, CA

Posted 23 August 2018 - 09:43 AM

I maintain the observatory scope of a prominent person in the telescope industry. He uses an SBIG 16803 camera. That camera dive bombs from ambient to -20 just as fast as it can. Considering that the sensor in that camera costs more than most folks entire image train, I wouldn’t be to concerned with cooling speeds. Thermal shock is a wive’s tail that needs to die. 


  • Ishtim, DS INC, tolgagumus and 1 other like this

#10 spokeshave

spokeshave

    Surveyor 1

  • *****
  • Posts: 1979
  • Joined: 08 Apr 2015

Posted 23 August 2018 - 10:33 AM

I maintain the observatory scope of a prominent person in the telescope industry. He uses an SBIG 16803 camera. That camera dive bombs from ambient to -20 just as fast as it can. Considering that the sensor in that camera costs more than most folks entire image train, I wouldn’t be to concerned with cooling speeds. Thermal shock is a wive’s tail that needs to die. 

This is, quite frankly, terrible advice. Thermal shock is not a wive's tail. I don't understand why people think that experience with one camera is somehow applicable to every camera. The KAF-16803 is a very robust sensor designed to be cooled. It's datasheet says that it can tolerate -60C. A sensor like the KAF-8300, on the other hand, is only rated by the datasheet to -20C, and I know of at least one instance of one of those sensors being destroyed by thermal shock. The sensor die completely delaminated from the package.

 

The datasheet for the IMX294 (asked about in the OP) is not published so we really don't know what kind of environmental conditions it is rated for. That leaves us only with the manufacturer's recommendations, and every astro camera manual I have ever read (and I have read many and even written some) cautions against cooling too rapidly, and many camera firmware codes prevent too-rapid cooling. It is reckless to advise people to ignore the manufacturer's recommendations based solely on your experience with a single, different camera. 

 

I'm not saying that thermal shock will instantly kill every sensor. That's clearly not the case. But TECs can cool very quickly and sensors are made of bonded components with different CTEs. Physics being physics and not a wive's tail, it deems that bonded materials with differing CTEs will experience significant stress on those bonds when exposed to rapid temperature changes. Rapid cooling of those sensors increases the risk of catastrophic failure. That is a risk that is easily avoided. I can't imaging why anyone would argue against it.

 

Tim


Edited by spokeshave, 23 August 2018 - 10:50 AM.

  • pfile, jdupton and Der_Pit like this

#11 Ishtim

Ishtim

    Apollo

  • -----
  • Posts: 1297
  • Joined: 10 Nov 2007
  • Loc: N. Alabama, USA

Posted 23 August 2018 - 10:53 AM

That camera dive bombs from ambient to -20 just as fast as it can. 

My SBIG STL has being doing this for over 10 years.  Warm up however, is somewhat regulated by only applying a percentage of TEC power for a short period of time. 



#12 CharlesW

CharlesW

    Long time member

  • ***--
  • Posts: 3659
  • Joined: 01 Nov 2012
  • Loc: Chula Vista, CA

Posted 23 August 2018 - 11:12 AM

This is, quite frankly, terrible advice. Thermal shock is not a wive's tail. I don't understand why people think that experience with one camera is somehow applicable to every camera. The KAF-16803 is a very robust sensor designed to be cooled. It's datasheet says that it can tolerate -60C. A sensor like the KAF-8300, on the other hand, is only rated by the datasheet to -20C, and I know of at least one instance of one of those sensors being destroyed by thermal shock. The sensor die completely delaminated from the package.

 

The datasheet for the IMX294 (asked about in the OP) is not published so we really don't know what kind of environmental conditions it is rated for. That leaves us only with the manufacturer's recommendations, and every astro camera manual I have ever read (and I have read many and even written some) cautions against cooling too rapidly, and many camera firmware codes prevent too-rapid cooling. It is reckless to advise people to ignore the manufacturer's recommendations based solely on your experience with a single, different camera. 

 

I'm not saying that thermal shock will instantly kill every sensor. That's clearly not the case. But TECs can cool very quickly and sensors are made of bonded components with different CTEs. Physics being physics and not a wive's tail, it deems that bonded materials with differing CTEs will experience significant stress on those bonds when exposed to rapid temperature changes. Rapid cooling of those sensors increases the risk of catastrophic failure. That is a risk that is easily avoided. I can't imaging why anyone would argue against it.

 

Tim

Does anyone here remember the firestorms that erupted over folks receiving their filters with dust on them? You’ve heard of one camera that had a failure. What kind of meltdown would we see if whole swaths of cameras failed after a year because of shock? One failure is an anomaly, not even statistically relevant. 



#13 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23348
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 23 August 2018 - 12:11 PM

Is there a recommended safe rate to lower the temperature of a cooled CCD or CMOS camera? I recently acquired an ASI294MC Pro.

 

TIA

ZWO's ASI cameras have built-in cool rate throttling. It will usually take several minutes for the cameras to reach your setpoint without any software-controlled cooldown scheduling. I would let the camera do it's thing, and not worry about it. It used to take 2-3 minutes for my ASI1600 to cool, however with the latest firmware, both the ASI1600 and ASI183 take about 5 minutes to reach setpoint. And that is without any software control over the cooling. This should be fine to prevent any kind of thermal shock issues. 

 

The only caveat I can think of here is if cooling within 5 minutes causes any dewing problems. Some people on occasion have noted that they can end up with some dew with the built-in rate (which is not really fast, but not particularly slow either). If you experience some dewing, then using a slower cool rate with software support can help avoid it. This was more a problem on the original ASI071 cameras, which seemed to have a bit of a sealing flaw or something. I think that was resolved with the ASI071 Pro. 

 

It is up to you to use a warmup schedule or not. These cameras do not have any built in warmup scheduling as far as I know. If you simply turn off the camera, the mass of the camera and heatsink will absorb a lot of the heat released by the TEC, so I don't think it is a big problem, but it certainly wouldn't hurt to have a warmup schedule configured in your software (i.e. SGP). 


Edited by Jon Rista, 23 August 2018 - 12:13 PM.


#14 *Axel*

*Axel*

    Vostok 1

  • -----
  • Posts: 197
  • Joined: 01 Mar 2016

Posted 23 August 2018 - 01:22 PM

I always warm up the camera (ASI 1600mm c pro) when I'm done to bring it back from - 20c to ambient. I usually target 5mins to 10mins warm up period and during that time I pack all the stuff.

#15 Der_Pit

Der_Pit

    Viking 1

  • -----
  • Posts: 724
  • Joined: 07 Jul 2018
  • Loc: La Palma

Posted 23 August 2018 - 01:40 PM

It used to take 2-3 minutes for my ASI1600 to cool, however with the latest firmware, both the ASI1600 and ASI183 take about 5 minutes to reach setpoint. And that is without any software control over the cooling.

Maybe somewhat OT, but I just searched through ZWOs website and did not find any mention of firmware and/or updates of it for the ASI1600.  Where/how did you get those?



#16 tolgagumus

tolgagumus

    Vendor

  • *****
  • Vendors
  • Posts: 3630
  • Joined: 21 Jul 2014
  • Loc: NJ

Posted 23 August 2018 - 04:11 PM

I agree with Charles above. I have never seen a camera going bad because thermal shock. I know 10 year old cameras where people just shut the power off at the end of the night and they keep working. I think thermal shock is a myth 



#17 Jon Rista

Jon Rista

    ISS

  • *****
  • Posts: 23348
  • Joined: 10 Jan 2014
  • Loc: Colorado

Posted 23 August 2018 - 04:38 PM

I agree with Charles above. I have never seen a camera going bad because thermal shock. I know 10 year old cameras where people just shut the power off at the end of the night and they keep working. I think thermal shock is a myth 

I don't think warming is a problem. It's the instant cooling that can be a problem. If a camera doesn't have built in cool rate throttling, turning a TEC on at maximum power can cool down to setpoint almost immediately. That is what can be a problem. I've read reports of the TEC separating from the sensor back, or even sensors themselves cracking, due to such rapid cooling. The materials often cannot handle it.



#18 entilza

entilza

    Soyuz

  • *****
  • Posts: 3716
  • Joined: 06 Oct 2014
  • Loc: Oakville, ON, Canada

Posted 23 August 2018 - 05:06 PM

If we look at damage to optics, it's usually a shock to cold, as in the case of using compressed air on glass. Now I am not sure how fast the TEC's can cool from say 23 degrees to -15 at 100% as I've never tried it... It's probably not as fast as the instant thermal shock reaction of the compressed air.

Similarly to taking out a hot glass from the dishwasher and putting it in cold water.

Harder for particle order than particle chaos I guess... So if I had to be cautious I'd be cautious of cooling at 100% rather than warming by shutting off.

#19 xiando

xiando

    Fly Me to the Moon

  • *****
  • Posts: 6374
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 23 August 2018 - 05:26 PM

Like Bob, I just let the Atik firm/software handle it. They seem to use a "paced" approach, so I'm not terribly worried about thermal shock with my Atik Camera. Can't say as I would feel the same if it dive-bombed onto the setpoint, but it doesn't.



#20 jdupton

jdupton

    Surveyor 1

  • *****
  • Posts: 1775
  • Joined: 21 Nov 2010
  • Loc: Central Texas, USA

Posted 23 August 2018 - 06:04 PM

All,

 

   I think all sides of this disagreement have merit - not a concern, concern with cooling, and concern with warming. In my opinion, there are considerations for all aspects that can come into play. Consider the following:

 

  • During initial cool-down, the TEC is at ambient. As it operates, the hot side heats up and the cold side cools down obviously. The TEC could be considered to be at peak efficiency when it first starts up. It loses efficiency as it operates since it becomes progressively harder to extract heat from the already cold side and dump heat into the already hot side. This makes it possible to have greater temperature changes at start-up.
     
  • During the warm-up period at the end of the session, the TEC is already quite warm on the hot side and quite cool on the cold side. If you turn off power, the TEC ceases to actively transfer heat. However, once the TEC ceases operating, it just becomes a conduit for thermal gradients in the system. The heat of the hot side and its associated heat sink immediately starts flowing towards the cold side. The thermal mass of the hot side is generally designed to be greater than that of the cold side to improve cooling efficiency. This makes it possible to have considerable temperature changes immediately after turning off the cooler.

   I am in the camp for cooling and warming gradually in all cases even where the camera is known to control cool-down rates. (The camera cannot have any control of warming rates if you simply turn everything off.) On my cameras, I shoot for what I think is a reasonable compromise. I set SGP to cool to 0° C (about 25° C below ambient) in 5 minutes. I also set it to warm to 20° C in 5 minutes. At that point, I turn off cooling to the TEC and let it warm at a rate governed by the current ambient temperature.

 

   Regarding damage to the sensor chip from cooling or warming too rapidly, I think there are two cases depending on the technology involved. For older sensors, I doubt there is much cause for concern. For some newer sensors, I think there are considerations to make for cooling and warming of the sensor. In my opinion, more care should be exercised with newer BSI sensors.

 

   Consider the construction of non-Back-Side-Illuminated sensors (read as 99%+ of what has been on the market for many years). As illustrated in my crude sketch below, the silicon sensor chip is bonded to the chip carrier (most often ceramic) with a "glue-like" bonding agent. Electrical connections between the chip and the carrier are made with a technique known as "wire bonding". Tiny gold wires are welded to on-chip contacts at one end and then welded to on-carrier contacts at the other.

 

WireBondSensor.png

 

   This construction means that the sensor chip has good thermal contact with the carrier (though the back side "glue") and the wire bonds are flexible enough that any minuscule difference in thermal expansion is easily handled. In such a sensor construction, there is a very low risk of damage from any reasonable cooling or warming.

 

   Now, the construction of a sensor which uses Back-Side-Illumination (BSI), is quite different. Here, the chip cannot be "glued" to the carrier since the back side is used for sensing the light. Instead, the sensor uses what is known as Flip-Chip attachment. The electrical contacts on the chip are joined to the electrical contacts of the carrier by means of very tiny "solder balls" placed between the chip and carrier. During the manufacturing process, these solder balls are heated so that a reliable electrical contact is made when the balls start to melt. See the crude illustration below.

 

BSISensor.png

 

   As can be seen, the only actual thermal contact area between the silicon BSI sensor chip and the ceramic carrier is only made through the limited number of solder balls. This bottleneck may (I repeat may) cause some problems if the sensor / carrier assembly is cooled or warmed too rapidly. The silicon chip itself may lag behind thermal changes at the ceramic carrier causing some amount of stress on the solder ball connections. (Think of the ceramic carrier expanding or contracting with temperature faster than the silicon sensor chip even if both have very similar CTEs [coefficient of thermal expansion].) It is my opinion that cameras incorporating BSI sensors should probably be cooled and warmed at a slightly lower rate than normally attached sensors. In addition, I think more "soak time" after reaching a target set-point temperature may be wise because of this thermal lag.

 

   While I still use the same 5 minute cooling / warming period with my Sonly BSI IMX294CJK sensor camera as my older Sony ICX694ALG sensor, I do allow more soak time before starting imaging of a target.

 

 

John


Edited by jdupton, 24 August 2018 - 08:36 AM.

  • happylimpet, Jon Rista, entilza and 1 other like this

#21 John Miele

John Miele

    Gemini

  • *****
  • Posts: 3358
  • Joined: 29 May 2005
  • Loc: North Alabama

Posted 24 August 2018 - 08:46 PM

What do the camera manufacturers say or recommend? If they are silent on the issue then maybe it's not really an issue? My ASI071pro manuals has no warnings about cooling/heating rates. Does anyone with other camera brands have warnings in their manuals?



#22 rkayakr

rkayakr

    Apollo

  • *****
  • Posts: 1487
  • Joined: 27 Oct 2010
  • Loc: Northeast Ohio

Posted 25 August 2018 - 10:28 AM

For events that happen rarely, or not at all, it is difficult to determine if a preventative measure is effective or necessary. For example, I drink beer to keep the Velociraptors away. I say it works well, as we haven't seen any around here. My wife is unconvinced.


  • John Miele likes this

#23 spokeshave

spokeshave

    Surveyor 1

  • *****
  • Posts: 1979
  • Joined: 08 Apr 2015

Posted 25 August 2018 - 11:43 AM

What do the camera manufacturers say or recommend? If they are silent on the issue then maybe it's not really an issue? My ASI071pro manuals has no warnings about cooling/heating rates. Does anyone with other camera brands have warnings in their manuals?


Some do, some don't. Your ASI071 has no warnings because ZWO has built an automatic ramped cool-down routine into its drivers so you can't apply an instant 100% cooling to the sensor. It seems highly unlikely to me that ZWO would go to the trouble to do so just to address a "wive's tale". Most acquisition software has provisions to set the cool-down temperature of the camera. That seems like a bunch of needless coding to address a "wive's tale".

I honestly don't understand the controversy. Thermal shock is a thing. It exists in the world. People have destroyed optics by hitting them with canned air due to thermal shock. I have personally seen a sensor destroyed by thermal shock. So it happens. We astrophotographers tend to be fairly fastidious about caring for our gear. Taking care to properly cool our sensors should be part of that care. For the life of me I can't understand why anyone would recommend otherwise.

Tim
  • Jon Rista likes this

#24 xiando

xiando

    Fly Me to the Moon

  • *****
  • Posts: 6374
  • Joined: 27 May 2015
  • Loc: Cloudy NEOhio

Posted 25 August 2018 - 12:40 PM

Blasting a hot sensor with a jet of nearly liquid propellant is far different than turning on your cooler. In most environments, your peltier would likely fail due to an over-current situation (it itself would suffer from thermal failure due to the thermopile junctions separating) before the sensor IC experiences thermal shock.


Edited by xiando, 25 August 2018 - 12:41 PM.


#25 John Miele

John Miele

    Gemini

  • *****
  • Posts: 3358
  • Joined: 29 May 2005
  • Loc: North Alabama

Posted 25 August 2018 - 06:24 PM

"Your ASI071 has no warnings because ZWO has built an automatic ramped cool-down routine into its drivers so you can't apply an instant 100% cooling to the sensor. It seems highly unlikely to me that ZWO would go to the trouble to do so just to address a "wive's tale". Most acquisition software has provisions to set the cool-down temperature of the camera. That seems like a bunch of needless coding to address a "wive's tale".
 

Ok. Now that reasoning does make sense...John




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics