Firefly MV Mono - Is it worth cooling?

[SnakeyJ]

Finally got the missing bits to complete this (bar a more reasonably sized cpu cooler), but should be able to get the camera cleaned and rebuilt properly working through my to do list:

CLS filter for the camera nosepiece to seal and prevent condensation.

Dessicant package.

TIM between CMOS and cold finger.

Smaller/Lighter CPU Cooler pref with variable speed control.

LX control via GPIO - here's hoping.

Darken/paint the upper side of the cold finger around the CMOS.

Clean the CMOS, lens housing, nosepiece and camera internals.

Check that there's no vigneting or obstruction caused by the slightly increased focal length and the copper heat sink.

Investigate shielding/passive cooling for other pcb components to reduce read noise.

Also just got a set of Baader LRGB CCD filters so hoping to get out and find a hole in the cloud for an attempt at Jupiter and Saturn over the weekend.

[Cath]

Still at it then

Cooling definitely helps, which was to be expected really. Shame they didn't use the Q package version of the image sensor chip and have a hole in the PCB to allow a heatsink attachment directly to the back of the image sensor chi, but they didn't.

[long_arms]

I'll be interested to see results of the LX control, could be very handy.

Did you get these filters http://www.firstlightoptics.com/rgb-filters-filter-sets/baader-rgb-filter-set-with-ir-cut.html?

I've been tempted before so will look forward to seeing your results!

Dan

[SnakeyJ]

I'll be interested to see results of the LX control, could be very handy.

Did you get these filters http://www.firstligh...th-ir-cut.html?

Just waiting for my GPIO harness to arrive - not 100% sure it will work as it may just trigger to the max exposure then cut out automatically - the tech reference guide covers the fireflys and fleas and is a bit vague. However here's hoping!

Re the filters the basic Baader RGB IR-Cut set does not have IR-CUT coatings on the colour filters, so you need to pair each one with the IR-CUT. Way too fiddly for my liking, though you could leave the IR Cut on the nosepiece then add each of the colours in a filter wheel. However I want to use IR-PASS with my luminance to enhance the seeing and get the best detail. Also this set is not parfocal so you will need to play around with the settings rather more for getting a good colour balance - that said the sensor itself is not truely parfocal so not sure how this pans.

I got the Baader CCD LRGB 1.25" Set from Bernard @ Modern Astronomy - http://www.modernastronomy.com/filtersBaader.htm#bLrgb though FLO also stock if you want to source from them - much quoted review of these at http://panther-observatory.com/Baader_LRGB.htm The astronomik ones worked out more expensive as I couldn't find a combined set price and the QHY ones which look cheap are also not IR-Block coated.

Hopefully this set will last me until I need the bigger 2" filters for widefield DSO stuff, but should work well with the ZWO ASI120MM or QHY5L-II (whichever one I end up buying)

[SnakeyJ]

Here are some shots of the heatsink & mods to the camera housing taken during dissasembly/cleaning last night - the HS itself is cut from 3mm thick copper sheet and is insulted with foam and electrical tape.

1. Camera & HS, Cooler and Peltier components during dissasembly:

2. HS and camera with housing removed:

3. HS Removed from PCB:

4. Slots to Camera and Lens Housing:

5. HS Tip - Rear with Thermal Pad, Front Painted with Black Matt:

6. Adhesive Thermal Pad - does not stick/do not buy this product!:

As the adhesive thermal tape/pads does not work (certainly not sticking to the ceramic surface of the peltiers, despite pre cleaning with IPA), I have gone back to using lots of electrical tape to secure the heatsink, peltier and fan. I have ordered a tube of Arctic Silvers thermal epoxy http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htm - which I will use once I source a suitably smaller and more asthetic heatsink/fan and it is marketed as a permanent solution.

Once the PCB and heatsink/lens housing were reassembled I masked up the GPIO, USB socket and lens housing before liberally spraying the pcb with plastic/insulating coating to protect the board components from condensation.

No time to test light night as under strict orders to turn in before midnight, but I think the improved thermal junction between the HS tip and CMOS should yield some improvement. Still not the pretty but functionally complete.

I have just found and ordered a low profile CPU Cooler from overclockers which is 150gms lighter than my existing @ 205gms and cheap £ 9.98

I can trim the clips off and put a grill over the front.

[Cath]

Shame the stick pads wouldn't stick .. I guess the cooler is made from ceramic type stuff ?

Good to see your still at it though Jake

[SnakeyJ]

Shame the stick pads wouldn't stick .. I guess the cooler is made from ceramic type stuff ?

Yep the ceramic plates on the peltier cooler - hoping I have better luck with the thermal expoxy.

[SnakeyJ]

So glad I didn't get the epoxy early, as I thought'd test this tonight but no cooling - after much faffing around I got the multimeter out and realised I'd broken or blown the peltier - I do have a spare, but no heart to do it all tonight.

[Cath]

Oh dear Jake

Any idea what caused it do you know?

[Cath]

I just fort. Why not try bolting your heatsink to the camera heat plate with the peltier sandwitched as you did before with the metal screws but instead use nylon screws? .. I'm assuming nylon screws/bolts won't transfer heat like metal ones.

Can't give up after all you've done so far Jake!

[SnakeyJ]

Hi Cath - not really sure on the how and why, I'm not sure if perhaps I bent or stressed a lead during the dissasembly. Pretty sure I didn't drop or Knock it and there was no obvious sign that I'd blown or overheated it - just bad luck I guess, but it has definitely put me off the idea of epoxying the unit between the cpu cooler and heatsink.

I'm not sure if nylon screws/bolts would be strong enough especially at M5/M6 threads when warmed and cooled at different ends - I could sleeve/washer steel bolts in nylon/plastic for insultation at the cold end

I think I favour something I can bolt or clamp together, but need to keep the weight down and include an insulating layer, alternatively I can just bind with electrical tape at least for the short term.

At least with these I can use a decent TIM compound (like the Artic Silver/MX4) which has produced the best cooling results so far - not so worried about a little delay on the cold side, but want the heat to dissipate as qucikly as possible from the hot side.

Certainly not giving up I've got the spare to hand, but have ordered up another two of the peltiers from China, which I think is around £ 6.40 delivered to UK. If I can get a decent result from the Firefly I shall move on to the xbox live and possibly something a little better.

BTW - My GPIO harness turned up in todays post. Have to resist the urge to tinker tonight as I have to get ready for the IOW Star Party (wet weather gear advised but will go whatever). I will take the firefly just in case the clouds part.

[Cath]

Cooled:

Found out what the problem is causing the image to look like that.

Sorry about that Jake, Fixed now.

[SnakeyJ]

Found out what the problem is causing the image to look like that.

Sorry about that Jake, Fixed now.

That sounds like great news - is it settings or code and are you willing to post or PM?

I got back in from the IOW star party today, though had no chance to use the new cooled camera for real - though I had it mounted and running for four hours, at a nice steady set point of -5C. Unfortunately Saturn decided not to put in an appearance!

I've been running with a low profile CPU cooler - but this is not nearly as good as the big heat pipe cooler at dissipating the heat and runs at about +12C above ambient on the warm side - giving me about 22C warm to -5 cool side, or delta T of 27C. I will go back to using the old cooler as this runs at +3.5C above Ambient and should get me down to around -15C quite comfortably. I really need a good clear night to test this on Jupiter and Saturn so hoping for a break in the weather over the next week.

Mothers day tomorrow so will probably be in/busy for the day but hope to get the chance to play with the GPIO harness in the evening.

[Cath]

That sounds like great news - is it settings or code and are you willing to post or PM?

It was a coding problem, well not a problem but I was doing a gamma thing when saving images, which is what I've removed.

I got back in from the IOW star party today, though had no chance to use the new cooled camera for real - though I had it mounted and running for four hours, at a nice steady set point of -5C. Unfortunately Saturn decided not to put in an appearance!

I've been running with a low profile CPU cooler - but this is not nearly as good as the big heat pipe cooler at dissipating the heat and runs at about +12C above ambient on the warm side - giving me about 22C warm to -5 cool side, or delta T of 27C. I will go back to using the old cooler as this runs at +3.5C above Ambient and should get me down to around -15C quite comfortably. I really need a good clear night to test this on Jupiter and Saturn so hoping for a break in the weather over the next week.

Mothers day tomorrow so will probably be in/busy for the day but hope to get the chance to play with the GPIO harness in the evening.

Lots to do

[SnakeyJ]

First tests back on GPIO bulb exposure using the GPIO harness - the camera only claims to support trigger mode 0 and 3. By grounding ping 3 of the GPIO cable I am able to send the trigger function and get a frame this functions in both Flycap and Firecapture, at least as long as I set the camera registers usings flycap first. Trigger Mode 1 is what I need to get the a bulb LX exposure - although this definitely does not work in firecap, I did get some longer exposures using Flycap, which allowed me to set the firefly to trigger mode 1 although this should not be supported.

I'm a little short on time tonight, but will rig a switch in between pin 3 and pin 1 on the GPIO harness and test this again tomorrow with a board lens on the firefly. This will allow me to test some manually timed exposures and produce some images.

If this is successful I shall check options for a variable time switch or perhaps arduino or computer control via RS232.

[SnakeyJ]

LX not as easy as expected. The firefly mv seems to only support two trigger modes from http://www.ptgrey.com/support/kb/index.asp?a=4&q=239

Mode 0 - Standard Trigger Mode

Mode 3 - Skip Frames Mode

For both of these the exposure length is determined by 'Shutter' value/key in the registry - which after a quick reg search appears to be absent on my machine.

What I would like is Mode 1 - Bulb Shutter Mode, though this does not appear to be supported on this camera.

However, all may not be lost. In the SDK examples there is an application called extended shutter - this appears to interface the camera directly and is set at default at a 3000ms shutter/exposure. This programe appears to run fine (see attached screen shot):

There are source code examples for this, though I do not have any VB/Visual C programming tools to examine more closely - though looking at this in a text editor it looks pretty simple.

Question is can I modify this to be triggered by the shutter button, choose the output file type/location and change the exposure times with relative ease?

[SnakeyJ]

I've also checked the specs and they do say:

Shutter

Automatic/Manual Shutter modes

0.03 ms to 512 ms (extended shutter mode)

No limits set here on the extended/manual shutter mode so should be possible...

[Cath]

It doesn't appear to have BULB mode

But it does have software trigger - which means no external trigger would be needed.

I would have added BLUB mode to the software I'm doing but as their is no BULB mode I can't

[SnakeyJ]

Makes me want to BLUB too

Desipte the lack of firmware support for Bulb function - it looks as though longer intergrations/exposures are possible via the SDK examples, the software control/trigger is more useful (especially for warm room/conservatory control - must get a view finder cam and focuser control sorted!).

I'm working for the next 24-36hrs straight, but will see if i can get a copy of the Visual Studio Express downloaded over the weekend to tinker with the ExtendedShutter code to see if it can produce what I want.

Thanks again Cath for input/support on this - much appreciated as always!

[Cath]

No problem at all Jake, glad if I can help in any way.

The extended shutter code was one of the first example sources I went through, it only works with certain cameras, and guess what - the MV isn't one of them

But sure download MS's express compiler and have a play! .. The good thing about doing your own code is that you can write your software to do exactly what you want, look however you want etc rather than trying to get other peoples software doing what it is you want.

[SnakeyJ]

Ok - it has been a while, so time for a quick update on the project.

Cooling definitely worth the effort, particularly on recent attempts to image Saturn, where this has enabled me to really push the gain up using direct register controls in Firecapture 2.2 and the previous 22 beta 37. Using these controls and my 8" Newt at F15, I have been able to push the gain up to 24dB and then use the x1.2 signal boost for 28.8dB for the Blue channel. Using the cameras ROI mode @ 300x300 this has yielded frame rates of about 32L, 24IR, 22R, 22G and 15B.

Without the cooling running I can push the same settings, but the resulting images are way too noisy and not worth processing - even with dark frame subtraction (which leaves impressive holes). I'll try and post up some comparitive images later this evening as an example.

The main shortcommings that I have found with the cooling is the output cooling of the peltier - this needs quite a large heavy CPU cooler to be effective (at least with this TEC), and the lightest that I have been happy with has added 370gms to the weight of the camera. Although when running there is little to no apparent vibration caused by the fan (at least not detectable above the effects of the seeing), when the fan is turned on the torque at startup does cause some shudder. This is a nuiscance if you are hovering around the set point temprature and the fan is going on/off regularly. This could of course be negated with a proper PWM TEC controller, with variable control of TEC current and fan speedl). The other issue is that of heat pipes within the CPU cooler - by design these work far better when the cooler is horizontally mounted, but are much less effective when approaching the vertical (or pipes tilted down from the heatsink base). This results in cooling variation between best deltaT of 25C below Ambient and worst 10C. Although I can mitigate the camera cooling to a certain extent by rotating my scope within the tube rings, I frequently end up in positions where the finders are inaccesible/difficult to use.

Having just brought one of the new QHY5L-II mono's, I'm not sure that that the firefly is going to receive a lot more attention (though I will certainly be hanging on to it for some comparitives with Jupiter next January). However, there are a number of things that I would still like to try, though perhaps on a higher resolution, less noisy, more sensitve sensor, preferably something offering some decent driver/firmware support for long exposure (30-120 secs would be useful).

The main modification would be to watercool the TEC, with a 6mm silicon tube cooling loop running to a radiator/resevoir and pump hung below the tripod, I have sourced some light aluminium water blocks, which whilst not quite as thermally efficient as copper offer a significant weight advantage. This would give significant weight advantage to the camera and should virtually eliminate any vibration from cooling fans. Watercooling should also be unaffected by orientation of the camera, though there may be a small reduction of flow due to variation in water head. This should deliver a consistant and more reliable delta T between -25 to -30C. Additionally the weight saving at the camera and increased cooling capapcity should allow the use of a two stage peltier cooler if a larger delta T is required, -50C? Though the lower tempratures would also require an anti-dew heater and sealed enclosure with dessicants.

With regard to the PWM TEC control I will be looking to make an arduino controlled unit, following some of the many excellent guides and sketches that are available online - It will be designed to run as a closed box (with LCD display and control buttons), but I'm also hoping to use the arduino's usb port to be able to control and monitor this from my imaging laptop.

Four months ago all of this would have seemed far too daunting, given my then severely limited electrical and practicle skills (notably soldering). Whilst I can't claim to be skilled in either regard, the path ahead looks considerably less steep and more achievable now

[Cath]

Hi Jake

Thank you for the write up

Just have to bare in mind that the water could end up freezing in the cold months I guess in the pipes, not sure. But It's still got to be much better than trying to hang a heavy heatsink etc on the camera. Makes more sense to me.

[SnakeyJ]

Hi Jake

Thank you for the write up

Just have to bare in mind that the water could end up freezing in the cold months I guess in the pipes, not sure. But It's still got to be much better than trying to hang a heavy heatsink etc on the camera. Makes more sense to me.

Thanks Cath - hopefully the freezing nights are behind us until next winter, though I will use a cooling fluid that offers some anti-freeze, corrossion and microbial inhibitors. Although these are not quite as efficient as cooling with pure water, the benefits outweight disadvantages. I'll probably go with the Feser F1 fluid, which does all of this, has a freezing point of -5C and is electrically non conductive if there is any spill or leakage. Even on a very cold night the thermal output of the peltier should keep this well above freezing (From memory my heatsink temprature was running at 6-8C above ambient with the firefly at 24dB sustained on my last outing.

Just jogged my memory, but can recall using cheap vodka as anti-freeze back in Germany in the late 80s - I think it was around 4DM a litre and certainly cheaper than the anti-freeze in the NAAFI (the exchange rate was around 4DM to the pound) at the time!

[Cath]

The sensor certainly does get warm.

I found out how to fix the problem I've had with the camera only running for a minute or 2 and disappearing from the hardware list. If I insert one of my 5M USB repeater leads between the camera and the PC it's fine, it runs forever with no problem at all. So I think FireFly's USB circuit/PCB design on the board is not fully matched properly, it's the only USB hardware I have this problem with.

I too would like a QHY5 ! .. maybe one day.

[SnakeyJ]

I found out how to fix the problem I've had with the camera only running for a minute or 2 and disappearing from the hardware list. If I insert one of my 5M USB repeater leads between the camera and the PC it's fine, it runs forever with no problem at all. So I think FireFly's USB circuit/PCB design on the board is not fully matched properly, it's the only USB hardware I have this problem with.

I too would like a QHY5 ! .. maybe one day.

I'm glad you sorted this issue out - I was thinking you might have a bad/faulty camera as mine has been very stable since fixing the cable at the micro-usb port, so this is good news. I haven't run it over anything longer than a 2m USB lead, which to be honest is a little restrictive - so I may lash out and get a 5m powered lead for this (I have numerous reports of people having problems with USB hubs and imaging so think its better to keep the camera fed directly.

Noise is the main reason for the upgrade to the QHY - though increased resolution, 74% QE, 14bit ADC, and 'out of the box' exposures between 20ms and 600 seconds are all contributing factors Cost I think is pretty good at £ 229, certainly cheaper than the £ 299+ for the ZWO ASI120MM (which uses the same sensor and has very similar performance spec).

However, at this price it will be a long time before I consider dissasembling to fit my own cooling (though I am considering a cooled block/collar to fit around the body).