kbrown

11 hours ago, Ratlet said:

Mid job a bombs gone off, however as I keep telling my kids, the jobs not done till you tidy up.

I can't really do that since I'm such a bad example

12 hours ago, Dr_Ju_ju said:

To keep down on the clutter I've taken to keeping all the smaller stuff, in clear food container boxes, which at least let me see what's inside...

I'm sort of doing something similar. Often when working from home I order delivery for lunch and I prefer shops that deliver in reusable food containers Now I'm probably going to have to pay a visit to the local pound shop as well in order to pack everything up in order to relocate everything...

8 hours ago, Peter Drew said:

I've not seen anything yet that I would class as messy, the first photo describes my desk after tidying.  You may remember that hands down I won the untidiest workshop challenge a few years ago.     🙂 

Haha. Yes, this was also after a (sort of a) tidy up

A little bit off topic but since I'm having to move most of my DIY hobby stuff from one place to another soon, it gives me a chance to do something about the big mess I always seem to end up with not matter what:

I suppose the biggest problem is that it seems nearly impossible to categorise all the little bits and bobs ranging from bolts to batteries, from screws to surface mount components, from tapes to transformers etc etc etc... If I did that and stored each category in a box for example, I'd run out of storage space very quickly. The table in the photo is very messy but I still sort of know where things are. It just sometimes gets annoying when you don't have any space left to work on the next brilliant idea you came up last night

How messy are you? How do you deal with the amount of precious junk you've accumulated?

First encounters with the not so nice British weather revealed a few problems. Moisture was getting inside the dome and no amount of heat could get rid of it.

There was a few places I could identify. First one is visible in the photo above. The neoprene gasket between the box and the dome wasn't working properly as the underside of the dome isn't perfectly flat. Some water might have gone through the screw holes too.

Also I don't think this box was ever meant to be mounted this way as the screws that hold the lid down are in a deep recess that just invites water in.

To remedy these issues I did a few things. I "glued" the neoprene gasket to the underside of the dome (but not to the box) with some silicone. I added rubber washers to the screws holding the box lid down:

I also added rubber plugs onto the screw recesses. Finally I replaced the dome screws with flanged stainless steel ones to both prevent rust and to increase the surface area pressing down on the rubber washers.

Hoping this will do the trick

Been learning and tinkering with InfluxDB time series database and Grafana data visualisation tool. Have both of them running on my NAS in docker containers. Bit of Python and php scripting and I've got a fancy weather dashboard on my website updating every 5-10 minutes:

http://knowhere.myqnapcloud.com/WordPress/weather-dashboard/

Just managed to configure indi-allsky to upload its data to my website. Not sure how useful this is to anyone apart me but why not

http://knowhere.myqnapcloud.com/WordPress/the-eye-on-the-sky/

After a bit of tinkering with the config and python scripting, I managed to get all the sensor data embedded into the images:

Got better white balance too for both day and night images which was a bit more complicated than I would have liked

The biggest challenge with my setup is the limited ram of the RPi3 (1GB). If I turn on certain functions like detection map or logo overlay, then it'll struggle with memory and start swapping and all goes a bit haywire. I have ordered a 4GB RPi4 but I'm not sure if I will actually use it for this as it would run quite a bit hotter and there's no ventilation in the enclosure. The RPi3 together with the NVMe adapter are already keeping the dome temperature like 10C above the ambient. Yesterday was very hot day and the RPi3 CPU temperatures were close to 85C so it might have already been throttling down the speed to keep things cool...

Houston. We have first light:

Temperature reading is not working yet for the image. Not sure why but doesn't matter really as my indi driver does read it correctly...

Here's how it's sitting on the edge of the garage roof:

Nearly there. Decided to add an external 500GB NVMe storage with a USB adapter to both increase the storage capacity and reduce the I/O on the SD card. I've setup indi-allsky to use this drive for its database and image / timelapse storage. It's not the speediest thing with the USB 2,0 connection of the RPi3 but neither is the SD card.  I had it running for a couple of days on my desk and it all seemed to work okay. Added bonus is that the drive generates a little more heat so maybe the need for the dew heater inside the dome is minimal.

Just need to sort out a few more niggles with the LAN connection to the garage and then I can install this to the final location. Happy days

Meant to link this here earlier: https://github.com/aaronwmorris/indi-allsky/wiki

Worth reading through the sections under "Tuning". There's a lot of information on how to reduce the I/O writes to the SD card etc on an RPi.

13 hours ago, myk68 said:

Thanks for the reply, Keith.

I wondered about setting the DL folder somewhere else but there doesn't seem to be an option to select one other than the one it creates when it saves.

Think I'll have to look at this Allskeye.

 

I was able to do this with indi-allsky. Had to modify the setup.py though but luckily all the folder variables were at the top of the file.

Had to jump through a number of hoops but I now have the first working version of the INDI driver for this:

I couldn't find libraries / examples on how to read the sensors connected to the GPIO via C++ so I wrote a python script that uses various python libraries in order to read them. The script dumps all the sensor data into a json file periodically. I've set this to run as a systemd service so it's always running. My INDI driver then reads the json file and updates the UI properties etc. Then in order to control the dew heater with PWM I had to install pigpio and run it as a daemon. The INDI driver connects to it and issues the desired PWM duty cycle commands. Bit clunky but it works

For the actual all sky camera I'm running this: https://github.com/aaronwmorris/indi-allsky

Well it's not the prettiest thing I've ever made inside but it does work

I'm able to read the BMP280 and AHT20 sensors via the i2c interface and the DS18B20 sensor via the 1-Wire interface. I also added an STP55NF06 mosfet in series with the dew heater so I could control it's power with PWM0 (GPIO18). All there's left to do is to write proper software for this. Happy so far

Decided to add a DS18B20 temperature sensor inside the dome so I can control the dew heater better as I'm also adding a outside temperature / humidity / pressure sensor:

Temperature sensor inside the dome

Designed and printed a sensor screen for the temperature / humidity / pressure sensor. Probably an overkill but at least it looks kinda cool

It'll be sitting on the bottom of the box like this.

Out of curiosity I thought I'd check what the effect of the dew heater would look like on a thermal camera.

Here's before turning it on:

And here's after about 10-15 minutes of full power:

6-7 degrees above ambient. Hoping this will be enough

So here's my take on an all sky camera that I've been wanting to build for a long time. I've finally made a start. It's going to be a Raspberry Pi 3B + Raspberry Pi HQ camera based system with a DIY nichrome wire dew heater. Everything will be powered from a PoE splitter which gives me max 5V / 2.4A.

Here's what I've got so far:

RPi 3 and the PoE splitter + a 50g bag of rechargeable silica gel.

Camera and lens are attached to the box with an adapter I 3D printed. Around this is a 3D printed bracket that holds the thin nichrome wire dew heater. It's about 20cm long 0.2mm thick and draws about 400mA at 5V. Seems to warm up nicely but not too hot to melt the plastic.

I'm using one of these pipe shelf brackets to attach it to my garage wall. It also serves the purpose of the cable entry to the box. The box itself is IP65 rated and I've tried to seal everything up as good as I could with neoprene gaskets and silicone. Also what's not shown in the pictures is one of these pressure compensation valves (IP68) I added to the bottom of the box:

More to come soon...

On 20/08/2023 at 01:17, bosun21 said:

I just clean my mirrors with tap water with a few added drops of washing up liquid. It's the final rinse that needs to be distilled water to ensure there is no residue left when dried. I bought 5 litres, have cleaned two scopes and still have half leftover. This is only for mirrors and corrector/meniscus plates and not lenses per se.

That's what I do as well. No point on wasting expensive stuff if you're just trying to get the dirt off the surface to begin with...

Hi,

So I'm finally actually putting together an All Sky Camera which has been on the bottom of my To-Do list for way too long. It's going to be a RPi3 + RPi HQ camera based and powered with a 5V/2.4A PoE splitter so I don't have to run multiple cables. I'll be running my dew heater off of that 5V as well. My question basically is what do you think is the minimum adequate power for the dew heater? I'm thinking of using just nichrome wire ring around the camera inside the dome as I've got plenty of that available. Something like this:

The thin yellow torus represents the dew heater. As a test I cut off about a 20cm long nichrome wire which is about the circumference of the torus and applied 5V from the ends of it. It drew about 350mA (1.75W) and did get warm to the touch but not too hot I would have had to let go. Do you think this will be enough? The diameter of the dome is about 10cm.

I'm trying to keep the power consumption down as in the future I'm hoping to add weather station functionality to this as well and run all the peripherals from the same power supply.

Brilliant. Thank you both! not quite sure which variant I have. It is TO-39 package. Will check when I take it apart. IP65 should be plenty good.

A while back I made a DIY SQM meter and a Cloud Detector using a TSL237 light sensor and a Melexis MLX90614 IR temperature sensor. I had this riding on my telescope pointing the same part of the sky I was imaging and I didn't put anything in front of the sensors since I was only going to use it when the weather was good enough for imaging. However now I'm building an All-Sky Camera and have been thinking of putting the SQM / Cloud Detector in the same enclosure with it as a permanent installation. This presents a problem though. While the TSL237 sensor probably could survive the elements as is, I don't think the MLX90614 is hermetically sealed. I've read a few forum posts and articles of people trying to find something suitable to put in front of it without affecting its performance, but so far I haven't found anything that would work okay. The only thing I can think of is a mechanical cover I could open and close perhaps with a servo to protect the sensors when the weather is not good enough. Any other ideas?

My concern would be that if you tighten it in let's say cold and then it gets exposed to warmer temperatures it might warp due to the tension and plastic getting a bit softer and you suddenly loose the grip...

3 minutes ago, Captain Scarlet said:

My own effort to reduce spikes/help local contrast has seen me buy some 0.2mm thick stainless steel sheet and I’ll get my local metalwork company to guillotine it into 15 or 20mm wide strips. I’ll also reduce the number of vanes on my 8” from 4 to 3. These will replace my current 4-vane spider whose vanes are 1.2mm thick. The idea is that fewer vanes will reduce total length of diffraction edges, and the thinness will make the spikes longer and fainter, throwing most of the diffraction far beyond a planet.

I might be wrong but I thought that 3 (straight) vanes would produce 6 spikes since they're all pointing at different directions?

2 hours ago, vlaiv said:

Too low SNR for effective deconvolution. Also - since it is very specific kernel, regular gaussian PSF won't be of a much help for deconvolution.

I was mainly thinking the mode in PI where you sample the PSF from your image rather than a synthetic Gaussian PSF. Nonetheless, it sounds like I'm going to pass testing this idea.

I don't mind small diffraction spikes anyway. I always keep my CCD rotated about 45 degree angle (apparent) off the spider vanes. Reduces the effect a bit as the pixels don't align with the spikes without sacrificing anything.

2 hours ago, LuckieEddie said:

I tried these a while ago. I found that on my 8" Newtonian they did almost eliminate the diffraction spikes but in doing so spread that extra light evenly around. This meant that the stars went from tight points with sharp spikes to softer and larger without spikes. For me I considered that a failure.

Did you try something like deconvolution in PixInsight on them?

The guy who posted it on FB said it worked great when he was imaging the cat's eye nebula as the centre of it is very bright and the diffraction spikes would have ruined it.

Suppose if you could easily put the mask on and take it off it might be worth it depending on the target.