So I'm starting a bit of an electronics project. I am going to call it "The All Things to All People" (TATAP) controller. My intention is to make this 100% open including software, circuit, PCB artwork and 3D model of the case if I end up going down the 3D printer route.
Simple criteria for my design:
1. Build it at home, parts of eBay or common electrical distributors and a PCB simple enough that it can etched at home.
2. As much as possible all in one box design.
3. Modular design. If someone wants to build it but doesn't need feature X then don't populate it.
Anyway this all came about after dragging to astrofest a power hub, usb hub, power monitor, dew heater, and losing power as well. Additionally I want to add a focuser at some point. All of these things were individually strapped to the tripod and there were cables everywhere.
Anyway currently I am planning the following feature list:
12V power input
12V power monitoring (voltage and current)
Piezo speaker
Alarm for low voltage (for when running from batteries)
12V backup battery supply
Trickle charger for backup battery
Alarm for loss of primary supply (for running from 12V and backup battery)
Multiple 12V outputs, relay switched.
Multiple Dew heater controllers
Temperature Sensor
Humidity Sensor
Autofocus stepper driver with temperature reporting (not compensation)
External display and keypad for autofocus controller
Everything fused with blade fuses
All 12V inputs and outputs using secure connectors (leaning towards Anderson connectors instead of those crappy 12v plugs)
A display and keypad to control everything
Computer software to record and log everything
Focuser will be compatible with the sgl_observatory open source ASCOM driver.
USB hub will be in the case and powered from the device, but will be a separate product.
So any ideas of what else can fit in there or features you'd like to see if you want to build one too?
This will likely take me until Christmas or even after. Spare time is rare at the moment and the goal is to have it done well before the next winter season.
Additional Ideas + a dump of my brain to remember what to do:
Osc Prox switches for detecting mount position.
Alarm inhibit via software (don't annoy the neighbours).
USB master reset (via cutting data lines or via asserting reset on USB hub) More thinking needed since this approach would cause controller to drop out as well.
Potential need to switch to ATMEGA328 as 8kb program memory may be too limiting.
Dew controller to run autonomously or via computer, with computer allowing fancier dew control algorithms.
Implement graphing / logging of temperature vs time (may help with dark library on non-temp-controlled CCDs)
Labelling of I/O on computer display ala mounthubpro
GPIO expansion header for unused pins. (low firmware memory cost)
All in one is an idea I had as well, but I more than likely do not have the time to make anything like this anymore. Nice project and good luck with it.
I'd suggest splitting it into two parts so you have a bunch of USB and power connectors mountable on the scope, the heavy bits (like the backup battery) on the ground and a small number of cables between them.
Yes Rick for sure. The battery was only an input connection, no way I want something that heavy mounted on the equipment. By the way already ordered Anderson connectors for all this. Your power rail was my inspiration
Also having a chat to Peter today he mentioned some devices (like his camera) crapping themselves if connected before the power and that makes switchable power risky to remote observatories. Right now I'm looking into making the USB hub itself switchable from this device too. Not quite sure how that will work yet. I'm hoping cutting the 5V from the computer and the power controller at the same time will hard reset all devices, but I may need to cut the datalines to the computer too to force the reset. This will need some experimentation starting with some cheap donor USB devices
Interesting idea... good luck with it! Here are a few other ideas:
* More than one 12V power input port - so you can hot switch between batteries or AC power supply without power cycling everything.
* Multiple temperature sensor probes on the scope side associated with individual dew heaters.
* Ultrasound proximity sensors (e.g. that detects < 3 cm range) that can be added as-needed to avoid OTA vs tripod crashes.
* Following on Rick's idea of a separate ground-based box, perhaps you could integrate a weather-proof area for incorporating an all sky camera? It'd be useful for nomadic imagers - just plonk the battery box onto the ground adjacent the scope and you have an ASC view ready to go.
* Similarly near the box, having an integrated IR temperature sensor and/or an IR rain sensor (that will result in an alarm played via speaker).
* Lots of spare analog inputs and digital outputs that can be easily extended for additional devices (like a Phidgets I/O board).
I'm hoping cutting the 5V from the computer and the power controller at the same time will hard reset all devices, but I may need to cut the datalines to the computer too to force the reset. This will need some experimentation starting with some cheap donor USB devices
I'd be very careful with this approach... I've heard of several people accidentally destroying USB devices when disconnecting pins via multiple relays, and the ground pins being cut momentarily before the power pin - resulting in lots of current where you don't want it
* More than one 12V power input port - so you can hot switch between batteries or AC power supply without power cycling everything.
Already taken into account. The battery backup input will essentially be the same as an unmetered 12V input. Allowing hot plugging of the main powersupply in the presence of the battery, or if you're happy without voltage or current indication just run entirely from the battery backup.
Quote:
Originally Posted by naskies
* Multiple temperature sensor probes on the scope side associated with individual dew heaters.
This may get complicated from an algorithm point of view. I was intending to go down the route on >80% humidity switch on all dew heaters to a pre-set point.
Quote:
Originally Posted by naskies
* Ultrasound proximity sensors (e.g. that detects < 3 cm range) that can be added as-needed to avoid OTA vs tripod crashes.
Brilliant. I like it! Though I don't like the idea of ultrasound but mag prox switches on the other hand would be a great idea. Could even be combined with the power switches to not just give an alarm but actually cut the power to the mount.
Quote:
Originally Posted by naskies
* Following on Rick's idea of a separate ground-based box, perhaps you could integrate a weather-proof area for incorporating an all sky camera? It'd be useful for nomadic imagers - just plonk the battery box onto the ground adjacent the scope and you have an ASC view ready to go.
I think I'm slowly getting out of my depth here.
Quote:
Originally Posted by naskies
* Similarly near the box, having an integrated IR temperature sensor and/or an IR rain sensor (that will result in an alarm played via speaker).
IR rain sensor sounds interesting. Why IR temp sensor? What advantage over the existing humidity and temperature sensor?
Quote:
Originally Posted by naskies
* Lots of spare analog inputs and digital outputs that can be easily extended for additional devices (like a Phidgets I/O board).
Not sure what the application here is but analogue inputs are very computationally intensive. I'd want a pretty good reason to include this because memory is a premium and I'll be struggling to get all this into 8k as it is.
Quote:
Originally Posted by naskies
* 5V power supply for USB hubs.
Already have an internal hub included but I could provide an external 5V line. May not be a bad idea.
Quote:
Originally Posted by naskies
I'd be very careful with this approach... I've heard of several people accidentally destroying USB devices when disconnecting pins via multiple relays, and the ground pins being cut momentarily before the power pin - resulting in lots of current where you don't want it
Yeah I was originally intending to cut power+/- first and then power with GND remaining in contact at all times. Especially these days with switchmode supplies charges can build up where you don't want them and fry stuff quite easily.
Though looking at my USB hub I am now thinking I can just assert the reset pin on the USB hub chips. We'll see.
This may get complicated from an algorithm point of view. I was intending to go down the route on >80% humidity switch on all dew heaters to a pre-set point.
Fair enough. I was mainly thinking that for optics, you want as little heating as possible - e.g. on my RC8 and manually controlled heaters, I get better FWHM numbers without heaters, provided that it doesn't dew up. However, I usually err on the side of too much heat / poorer FWHMs because it's less risky than losing a whole night's worth of subs.
Quote:
Why IR temp sensor? What advantage over the existing humidity and temperature sensor?
Cloud detection - i.e. use the IR sensor to read the sky temperature. If it's much lower than the ambient / ground temperature, then assume clear skies. Otherwise, if it's similar to the ambient temperature, there's a good chance of cloud. Perhaps something like this:
Now that I think of it... a precision light sensor might be handy too as a poor man's sky quality meter, i.e. record ambient light levels as part of the log data. I've noticed that the sky can vary a lot between consecutive nights at a dark site, so this would be handy to record.
A couple of suggestions
To be all things to all people you'll definitely need to add an ASCOM compliant rotator as well as the planned focusser, that would ideally mean a second display line for rotational position !
You have only 8k to work with you say !
All 12v switchable output lines IP addressable.
For observatory control :
Usually achieved by Rain sensor, Day/Night light sensor, Cloud IR sensor, Anemometer and can be configured with external software, parameter settings and hardware. Optionally an external Ambient Temp sensor
Anemometer input - say for digital hall effect trannies and magnets
Anemometer and Rain sensor inputs linked via hardware and software interface to a dome closure relay output for observatory shut down - allowing you to set the values.
May as well have some optional GPIO output relays - to control lighting, motors and other devices
Maybe a flats lighting controller - either for LED panel or better yet an EL panel - so variable output value - not sure if there is an ASCOM standard for Flat lighting control.
Keep things as standard (to other existing devices) as possible so it can work with CCDAP, ACP etc
Variable power Dew controller output - not just ON, but variable with temp feedback. Once again not sure if there is an ASCOM standard for this
Cannon connectors FTW, which i better do soon lol. Sounds like a great idea if you get it off the ground, the ascom focuser would be a real winner Temp sensor another winner having a non regulated ccd it makes choosing dark sets more on the ball. Really i like the whole thing apart from dew heaters which i never use, also i don't think i would risk powering my ccd through it.
Fair enough. I was mainly thinking that for optics, you want as little heating as possible - e.g. on my RC8 and manually controlled heaters, I get better FWHM numbers without heaters, provided that it doesn't dew up. However, I usually err on the side of too much heat / poorer FWHMs because it's less risky than losing a whole night's worth of subs.
Very true. Maybe a global control then. Say to get your dewstrap working on a small scope bias it to 33% and a large scope bias to 100% and then use the global temperature and humidity to control this and start stepping up the heaters as required i.e. start at 3%/10% then 6%/20% 12%/40% as the temps drop etc.
This kind of complicated stuff I can offload to the computer saving valuable program memory on the microcontroller.
Quote:
Originally Posted by naskies
Cloud detection - i.e. use the IR sensor to read the sky temperature. If it's much lower than the ambient / ground temperature, then assume clear skies. Otherwise, if it's similar to the ambient temperature, there's a good chance of cloud.
Neat. Will research into this. If the detector is simple enough to read then the code to trigger an alarm is trivial. If I can find one that runs on an I2C bus then it's not expensive from a software / hardware availability point of view either.
Quote:
Originally Posted by rally
To be all things to all people you'll definitely need to add an ASCOM compliant rotator as well as the planned focusser, that would ideally mean a second display line for rotational position !
You have only 8k to work with you say !
All 12v switchable output lines IP addressable.
For observatory control :
Usually achieved by Rain sensor, Day/Night light sensor, Cloud IR sensor, Anemometer and can be configured with external software, parameter settings and hardware. Optionally an external Ambient Temp sensor
Anemometer input - say for digital hall effect trannies and magnets
Anemometer and Rain sensor inputs linked via hardware and software interface to a dome closure relay output for observatory shut down - allowing you to set the values.
May as well have some optional GPIO output relays - to control lighting, motors and other devices
Maybe a flats lighting controller - either for LED panel or better yet an EL panel - so variable output value - not sure if there is an ASCOM standard for Flat lighting control.
Keep things as standard (to other existing devices) as possible so it can work with CCDAP, ACP etc
Variable power Dew controller output - not just ON, but variable with temp feedback. Once again not sure if there is an ASCOM standard for this
Thanks for the ideas. To be clear that's 8 kilobytes of program memory. The more I read these posts the more I realise there's really two different target applications here: Telescope control and Observatory control. May split some of the ideas like cloud sensor etc into a separate project.
What do you mean by ASCOM rotator? I'm relatively new to this hobby (about 1 year), you talking about observatory roof control or is this something to do with focusing?
As for relays / GPIO I'm not sure I see the point of dedicating contact outputs like that as providing 12V switchable outputs effectively allow you to connect any 12V relay anyway thus making it a relay output.
Variable dew controllers integrating with temperature and humidity sensor is definitely part of what I was going for.
A few things different from that project:
1) I hate the barrel connectors. They are flimsy, easily unplugged and the manufacturing is so lose on tolerances that I actually have one cable which outright doesn't work on one device yet is very tight to plug in and remove on another despite being the same physical size.
2) I wanted a battery backup input and voltage / current monitoring to ride through power outages at the next astrofest and to ensure I don't over discharge batteries when not running from batteries.
3) I hate the barrel connectors.
4) I at the cost.
5) Did I mention I hate the barrel connectors.
Quote:
Originally Posted by 2stroke
Cannon connectors FTW, which i better do soon lol. Sounds like a great idea if you get it off the ground, the ascom focuser would be a real winner Temp sensor another winner having a non regulated ccd it makes choosing dark sets more on the ball. Really i like the whole thing apart from dew heaters which i never use, also i don't think i would risk powering my ccd through it.
XLR connectors are expensive! But otherwise a great alternative. Actually that's unjustified. Quality XLR connectors are expensive. I'm sure ebay I could get them cheap too.
Why would you "risk" powering a CCD through it? Effectively the CCD will be powered by whatever good supply you give it, the only difference being this box will actually have voltage and current monitoring + battery backup. What part of this project worries you for your CCD? (interested incase I need to address something in the design).
Well partially anyway. I received a red LCD panel from ebay. The feature list sounded good. I2C bus compatible ... but no driver. Well CRAP. I spent the entire weekend writing a driver for it. In the end I got it working.
It's currently displaying the Temperature and Humidity from the DHT11 and the Temperature from the DS18B20 (the DHT11 doesn't go below 0 degC which I'm sure some people's scopes do).
Anyway so far I've got these 3 devices working and I've used some 38% of my available program memory. I think I'm going to need to work on my coding efficiency this weekend.
Ask my girlfriend about the state of my desk and she'd say electronics are quite at home with me.
Electrical engineer, working in the oil industry now but during my uni times I worked on a few medical instruments and my degree was mostly microelectronics based.
Well you know your electronics in that case.
Just make sure you don't have too much fun with this....!
Seriously, watching with interest. Even half of what you have outlined would be of interest to many, I'm sure.
I have just built an automatic 2 ch dew heater, 4 outputs, and it's really easy to implement. Although, with all you want to do, perhaps the 328 would be more flexible. I had built most of it and I found the link below which gave me the idea of getting the dew point and do a small calculation based on your conditions, and the guy has done a great job of his dew heater.
As for the operation, it's easy to automatically do multiple sensor under dew strap readings and compare with your dew point. I use a DHT22 to get ambient temp and the humidity, then calculate dew point, compare with temp sensors under the dew straps by x degrees for your conditions. Then pwm to mosfets.
I wrote a small app for logging the data on the PC via USB and I'm sure it will come in handy for all sorts of things. Easy to get the data into Excel and do pretty graphs.
I also have a feature that if you startup the unit with switch 'A' on, it boots into a manual mode and bypasses the auto mode.
I'm using a 20 x 4 red chars on black background LCD in normal 4bit data mode, which is great to show lots of info. It's not the smallest box as it's the 'prototype', however I have a seperate small box that has the temp sensor inputs and the dew heater inputs which can sit on the scope. (very lightweight). I use midi connector for this - 2 pins for power and 3 for data. This small box and midi lead can be disconnected and the main unit used in stand alone manual mode without the sensors but still 2 ch with 4 outs.
Are you using mosfets for your output with pwm? I used the freetronics N-Drive mosfets and on a 10" dew strap, I can't feel them get warm at all. I'm sure though there are much better options though in the sea of fets.
This link is from another Ice In Space member with their dew heater.
Thanks redbeard I may have to quiz you on a few things :-)
I'm still leaning towards the ATMEGA88/168 to keep cost low. I'm doing a lot but the reality is I need to split the project across a few microcontrollers anyway for modularity and to get around that lovely "feature" of serial communications that only one piece of software may access a comms port at any given time. This means Focuser will be on it's own micro freeing up a lot of spare code.
The DHT22 looks like a much better option than the DHT11. I'll look into that. The biggest gripe I had with the DHT11 is its inability to get below 0 degC, a situation which many people may find themselves in.
I haven't decided on a MOSFET yet but I have buckets of them at home so I'll pick any suitable one. On a slow switching circuit like a dew heater the critical part is a low Rds_on value which limits you to any one of about 10000 different ones . As switching frequency increases then the rise/fall time become critical to reducing power loss as well, but this won't be the case here.
I actually do have a question not so much on the "how" but rather the "what". Coding up some algorithm is easy when you know what you want to do.
With temperature controlled dew straps what is the control scheme you're aiming for? Are you trying to keep the tube at a certain temperature compared to ambient? Are you trying to keep the dew strap at a certain temperature? How do you compensate for the location of the temperature sensor (i.e. the tube is nice and warm but the middle of your 12" corrector is dewing up because the heat isn't getting all the way to it.
This uncertainty and the assumption that all devices will dew up equally lead me to the theory that you bias the dew straps as a percentage and then control them globally based on the results of a single temperature / humidity measurement. i.e. when the humidity goes above 95% then put all the dew heaters at their certain biased starting positions.
Also which temperature did you use at the dew strap? A Dallas 1-wire sensor as well?
at the cost.
