ZWO Peltier Cooler

[Gina]

Yes, that aluminium block will certainly slow the cooling.  That is a nice low temperature though (if you mean -17.5°C).  Do you need it to cool down faster?  Also, I find cooling down to -10°C or so adequate for removing noise and hot pixels with the ASI185MC camera at 30s exposures (going much longer produces star trails).

[1CM69]

You’ve mounted the small box on the fan side of the main mounting plate, it should be underneath as I have done. 

I know the sounds odd but it initially mounted it as you have only to find it restricted the drawing of cool air into the fan heatsink. 

Simple to test, you should notice an immediate benefit. 

Cheers

[1CM69]

9 hours ago, cjdawson said:

It's been running for a while now, but it's not cooling very fast, it's down to about 17.5 now.

I wonder if this is something to do with the TES1-12704 peltier module, or maybe the thicker aluminium block.

Maybe it's because my peltier module is only 30mm across.  If I can find one that's 40mm maybe that'll cool the alu block better.

Can’t vouch for the 30mm TEC, I specifically used a 40mm to match the alu block dimensions fully. 

[cjdawson]

5 hours ago, Gina said:

Yes, that aluminium block will certainly slow the cooling.  That is a nice low temperature though (if you mean -17.5°C).  Do you need it to cool down faster?  Also, I find cooling down to -10°C or so adequate for removing noise and hot pixels with the ASI185MC camera at 30s exposures (going much longer produces star trails).

 

Nope it’s +17.5 not very good at all.   Hence why I’ll be changing the peltier module

 

5 hours ago, 1CM69 said:

You’ve mounted the small box on the fan side of the main mounting plate, it should be underneath as I have done. 

I know the sounds odd but it initially mounted it as you have only to find it restricted the drawing of cool air into the fan heatsink. 

Simple to test, you should notice an immediate benefit. 

Cheers

That makes sense.   I should be able to do that without much effort.

5 hours ago, 1CM69 said:

Can’t vouch for the 30mm TEC, I specifically used a 40mm to match the alu block dimensions fully. 

Yep, that’s why I decided to order the 40mm module, figured it’s wasted potential cooling surface

[Stub Mandrel]

Personally, I would mount the TEC directly on the camera, not essential, but it seems an extra thermal step for no obvious gain.

If not, you need to insulate the block between the TEC and Camera all around, even a thin layer of foam. If air can get to the block it will defeat the cooling.

Also are you using heatsink compound at the TEC-block and camera -block interfaces.

[Gina]

18 minutes ago, cjdawson said:

Nope it’s +17.5 not very good at all.   Hence why I’ll be changing the peltier module

OH!  I think you need to cool the camera casing directly.  Anything in between adds thermal resistance and worsens the cooling effect.  The Peltier TEC also makes a big difference.  I've found Chinese models to be pretty useless - I bought mine from Farnell Element14 - the model number is in my ASC blog I believe.

[Gina]

4 minutes ago, Stub Mandrel said:

Personally, I would mount the TEC directly on the camera, not essential, but it seems an extra thermal step for no obvious gain.

If not, you need to insulate the block between the TEC and Camera all around, even a thin layer of foam. If air can get to the block it will defeat the cooling.

Also are you using heatsink compound at the TEC-block and camera -block interfaces.

Crossed posts ?  My ASC uses similar kit and works well.  At least until something went wrong recently - I need to bring it indoors and investigate.  It looks like a camera problem - all white image.  I have been very careful with thermal insulation.  This is essential as you say Neil as is thermal paste.

[Gina]

On 06/07/2018 at 14:02, Gina said:

I'll have to see if I can find the model number.  It's 30mm square and about 5mm thick.  Rated at 15v and about 20W so relatively low power.  I bought it from Farnell element14 and I'm pretty sure it's this model :- MCPE-127-10-25 - Peltier Element, Thermoelectric Cooler, 19.6 W, 6.9 ohm, 2 A, 15.7 V, 75 °C

I tried various Chinese Peltier TECs but found they were very inefficient - a lot of power input for little actual cooling.  Cheap but not very cheerful!!

[cjdawson]

I’ve just remounted the box away from the cpu cooler.   That’s made a difference, I’m seeing 15.9 and still dropping.    Going to see how far it goes

[1CM69]

This is the TEC I use had no problems at all: HALJIA TEC1-12703 Heatsink Thermoelectric Cooler Cooling Peltier Plate Module 12V 27W https://www.amazon.co.uk/dp/B079DN16KH/ref=cm_sw_r_cp_api_.4MNBbWVZ7VKC

Looking at the Farnell link, it doesn’t specify the manufacturer other than Multicomp is their own value brand & they source from anywhere. 

Peopke are knocking ‘Cheap’ Chinese but more than likely these days, it all comes from China. 

You can get a bargain by cutting out the middle man, IMO, just saying. 

Glad moving the box has helped ??

[cjdawson]

Down to 13.6, that change made a huge difference.

 

to answer the other questioms, yes there’s heatsink compound everywhere.   I’m still thinking about moving the peltier closer to the camera.   My thought here is to get a couple of 3mm Alu plates. Or maybe a 2mm and a 4mm.    The 2 would be cold side.

[Gina]

But why put anything between Peltier TEC and camera?  The back of the camera is flat.

[1CM69]

I guess you have already made sure if this but it is super important that the main mounting plate is completely insulated from coming in to contact with any of the other metal parts or the Peltier will be either trying to cool or heat this huge area. 

[Gina]

There has to be some fundamental reason why cjd is not managing to cool below freezing whereas I can get down to -15°C with mine.  BTW - I'm not crowing - just trying to help.

[1CM69]

6 minutes ago, Gina said:

There has to be some fundamental reason why cjd is not managing to cool below freezing whereas I can get down to -15°C with mine.  BTW - I'm not crowing - just trying to help.

Mine gets below freezing too, everything depends on a temp drop from ambient. 

When used outside, as per normal use, the lower the starting temp the lower the final working temp. 

You would not get -15° if the ambient was 30° to start for instance. 

Not crowing either ?

[cjdawson]

20 hours ago, 1CM69 said:

This is the TEC I use had no problems at all: HALJIA TEC1-12703 Heatsink Thermoelectric Cooler Cooling Peltier Plate Module 12V 27W https://www.amazon.co.uk/dp/B079DN16KH/ref=cm_sw_r_cp_api_.4MNBbWVZ7VKC

Looking at the Farnell link, it doesn’t specify the manufacturer other than Multicomp is their own value brand & they source from anywhere. 

Peopke are knocking ‘Cheap’ Chinese but more than likely these days, it all comes from China. 

You can get a bargain by cutting out the middle man, IMO, just saying. 

Glad moving the box has helped ??

 

My new TEC1-12703 has arrived.  when I clicked on your link I was pleasently surprised to see that it said that's the one I purchased.  

 

20 hours ago, Gina said:

There has to be some fundamental reason why cjd is not managing to cool below freezing whereas I can get down to -15°C with mine.  BTW - I'm not crowing - just trying to help.

20 hours ago, Gina said:

But why put anything between Peltier TEC and camera?  The back of the camera is flat.

 

I think the fundimental reason is that the TES1 module that I was using is not good enough.   It does work, and I get several degreess of temp drop.  So it's working, just not working as well as has been seen by others.

The reason for the plate between the TEC and the camera is that the plate provides a place for the temp sensor to be plugged in. It also provided a place for the cold finger to screw into.

hmmm, this does give me a thought, maybe that cold finger - the copper threaded rod is causing a slight air gap or something if the threaded didn't line up 100%      just thinking out loud, it's something to add to the items to check list.

 

20 hours ago, 1CM69 said:

I guess you have already made sure if this but it is super important that the main mounting plate is completely insulated from coming in to contact with any of the other metal parts or the Peltier will be either trying to cool or heat this huge area. 

I used the templates that you provided and got them laser cut       When I made the gasket, I did notice that there is a gap between the mounting plate and the bigger plate for the cpu cooler to attach.      I figured that this gap was important as it's what lets the cold from the TEC go through to the camera.   I did make sure that this is centralised and doesn't touch the main plate at all.   The EDPM rubber keeps it away without any problems.

I'm thinking of adding some neoprene or maybe some more rubber to build up the insulation around the cold block,   It wouldn't surprise me if some cold is escaping to the air instead of being transferred through to the camera.

Actually this is one of the reasons why I'm changing from a 30mm TES to a 40mm TEC.    The extra 10mm should help to insulate that block from the air by lowering the exposed surface area.

 

 

[Gina]

You don't need a temperature sensor, the camera provides that information and from the sensor where it's more important.  You will find the sensor temperature is higher than the case even though it's thermally connected to the case for cooling.  There is a certain amount of aluminium in the cold finger that connects the sensor to the case.  The 1/4" UNC/Whitworth threaded hole is just in the case and only meant for mounting on a tripod.  I wouldn't bother filling it with copper - I don't think it will help much if at all.

[cjdawson]

2 minutes ago, Gina said:

You don't need a temperature sensor, the camera provides that information and from the sensor where it's more important.  You will find the sensor temperature is higher than the case even though it's thermally connected to the case for cooling.  There is a certain amount of aluminium in the cold finger that connects the sensor to the case.  The 1/4" UNC/Whitworth threaded hole is just in the case and only meant for mounting on a tripod.  I wouldn't bother filling it with copper - I don't think it will help much if at all.

The temperature sensor is there to provide a mechanism for turning the cooler on an off.  If it's not there, there will be no control of the peltier at all, and whilst this may be what you desire, I want to cool to a given temp, which is both achievable and repeatable.   I'm not looking to get -20C what I'm looking to do is get something that I can readily achieve, so that I can create a lower the thermal currents.   Gain some control over the camera temp allowing me to create a reusable dark library.

[Gina]

But you won't get that from the case temperature let alone that block as the temperature difference between sensor and cooler will vary.  In any case you don't need to control the temperature - just go low enough (or lower) to stop noise and hot pixels.  IMO you are overthinking this - you don't need a darks library for an ASC, particularly with the good quality camera and lens you and I are using.

[cjdawson]

1 minute ago, Gina said:

But you won't get that from the case temperature let alone that block as the temperature difference between sensor and cooler will vary.  In any case you don't need to control the temperature - just go low enough (or lower) to stop noise and hot pixels.  IMO you are overthinking this - you don't need a darks library for an ASC, particularly with the good quality camera and lens you and I are using.

So if I don't need to control the temperature, why bother with any of it at all?   If I didn't need to control it, it wouldn't need any modification.     I'm not overthinking things at all.    In fact. if you want to argue with anyone about the design, you are better off starting a new thread and discussing the overall design aspects there.    I'm not interested in redesigning this project.   I simply want to get it working well enough to not be a complete waste of time, effort and energy (both in contruction and usage)

So far. I've managed to get a 7 degree drop from ambient temp.      What I'm looking to achieve is getting the temp down from 20C to 0C.   So far I've managed about 13.1C    So things are going in the right direction.

I'm very sure that the difference between what I've been getting and what I'm desiring (which should be possible as the demo video shows a drop from 23.8C to 1.5C)  My testing is starting at around 20C, and I'm looking to get to the same 1.5C or 0C.

Mind you there is a difference in the testing as I have my camera mounted on the OTA at the moment.  To remove variables, I think I'll try removing the camera from the scope and proping it on a jar.  Whilst this may not help in a real world application it does allow me to replicate the test conditions better.  In the Real world application, the starting temp is going to be lower than 20C for most of the year anyways.

[1CM69]

2 hours ago, Gina said:

But you won't get that from the case temperature let alone that block as the temperature difference between sensor and cooler will vary.  In any case you don't need to control the temperature - just go low enough (or lower) to stop noise and hot pixels.  IMO you are overthinking this - you don't need a darks library for an ASC, particularly with the good quality camera and lens you and I are using.

I pretty sure that CJD intends using his camera for DSO not as an ASC, as I have used my 224, so a Dark library & repeatable Darks is indeed necessary. 

Yes you are correct that the camera itself will give a temp readout but this is not controllable whereas using a temp probe give control. 

Although this temp probe is only giving the case temp & not sensor temp it bears a relationship and given using a selected exposure time is repeatable. 

This is how I use mine and it works. 

No doubt there are tweaks that can be made here and there, my current published design is a culmination of around 6 other iterations, taking the best from all of them to achieve a finished project that’ll produce results on par with the more expensive cooled versions of these or similar cameras at a vastly reduced cost. 

Cheers 

[cjdawson]

Here’s a progress update.    When I got home from work, I switched out the peltier module with the 40mm one.   Also I double checked that the paste below the block was ok - I wasn’t happy with it, so added some more.   When I reinserted, I double checked that the block wasn’t touching the main part.  Hooked everything up and put it on the telescope for a quick test.    Set it going starting temp was 23.xC.    Went to feed my cat.  When I came back the temp was dropping much much faster and was already down to about 15C.

 

 

I had to go out this evening, so stopped the test at 12.8C    Looks like I have a result.   Will do a full test when I get home to see just how far this will go.    Looks very very promising at first quick 5 min ish test.

 

With any luck it’ll hit 0C and maintain a stable temp.

 

 

I am planning on using this as a Deep Sky setup.    So I’m not fussed about getting to -20, just that can I get to a reasonable stable repeatable temperature.  This will make sure that the camera gives a consistent performance.   -3C is more than enough for my need.

 

I’ll do a full test later this evening.

[cjdawson]

I'm down to 7.5C now, that's down from 20.2C.   That's a good improvement.   I'm thinking of adding another EPDM rubber sheet, to add to help with the insulation.

[John78]

If you use sharpcap, I wrote a script which uses the temperature reading from the on-board sensor to control the PWM power to the TEC - this is what you need to do to get the right results - or add another temperature sensor onto the sensor board.  Measuring the temperature of the body/case is going to be a level of abstraction too far for calibration frame libraries.

You can find the relevant info in this thread and at the end of the thread is what you really should do if you are serious

[cjdawson]

I have sharpcap, but I'm using Sequence Generator Pro for my deep sky work.  

 

Some of your comments are very useful to me...

Quote

 

1.  The whole camera gets soaking wet (the insulation is WIP)

2.  How to get colder (ideally with less power) - I had hoped to be achieving -20c

3.  How to fabricate a thermal break between the camera and the filter wheel

 

 

Here's my thoughts.

1. I've not checked this out yet.  Just been doing some starting tests, on this to make sure that it's good to go.

2. Still working on this bit.  I'm trying to tweak my setup to make it work better.

3. I'd thought about this as well.  Would be great to be able to control where the cold is better to stop it going through to the filter wheel.  I'm sure that will help to make the setup work better overall.