Oil the gears and fire up those brains....

[andrewpro]

Update! Three things. 1) all the below still holds true. 2)I dind't really mention it was an 18X I was gonna use, but it's inconsequential because 3) I will soon have a 40X to use isntead.
================================================
Ok, I need input.

Origionally I was jsut designing a data aquisition and control system for a radio telescope.

This then evolved into a "hey, why not make a data aquisition box separate from the radio telescope so I can use it for other things when not observing"

Then it became "Since I have to have some outputs for control, why not make lotsa outputs"

Now it's "Why dont I jsut hang off every accessory I have or can acquire, and make what I have dubbed a "Big Honkin I/O Cube".

I'm calling it a cube, but it's probably going to be more of a big rectanglular box. My idea is to modularize it, each different function on an identically sized board that can be stacked up or removed as needed.

So far I've got the following:

128 bi-directional I/O with interrupt (on two 64 I/O boards)

16 channels of DAC (possibly only 8, and use the other 8 as reference voltages for the first bank).

ADC- this ones still up in the air. I could use analog switches into the picaxe ADC's (giving an easy choice of 8 or 10 bit), or use up 3 I/O's on the picaxe and use a TLV2543. That'll give me a choice between 8 through 12 bits across 11 channels, but be alot slower. oh decisions decisions...=)

a single PWMout through a fixed discrete darlington transistor (as that's all the 18X has so far as I know)

a multiplexed strobe for clocking things in and out of whatever it is that's decided to connect to this frankensteins monster. Perhaps play double duty and use more analog multiplexers for things such as DS18B20 temp sensors.

standard Gammut of MAX6955 features, the "keypad" feature of which will probably be used for the interrupts from the I/O ports for speedier data acquisition (and also onboard keys!)

Onboard memory is an undecided option as of yet, as this is mostly designed to be connected to a computer.

Whew...now that that's outta the way, it comes to needing your input. As an american, I'mgoing to follow the bigger and more is beter stereotype! What else can I shove into this thing? Gimme your ideas, no matter how outrageous, I can probably figure out how to fit it in.

Let your imaginations fly and help my burned out brain!

--Andy P

Edited by - andypro on 7/1/2005 10:22:02 PM

Edited by - andypro on 7/1/2005 10:24:58 PM

 

[Jeremy Leach]

I like the idea of modularisation - powerful. Difficult to think of what to add as you seem to have it covered ! but why not add Infrared input for good measure.

I'm more interested in your telescope control ! - I've got an optical telescope that I haven't really used much, it has manual controls for Alt-azimuth and I was thinking of a PicAxe solution to control it....

 

[andrewpro]

Hmm...well....radio telescopes and optical telescopes are in a sense entirely different animals. The drive mechanisms in thier simple forms really arent that different, as long as they use the same sort of movement.

In your case,the alt-azimuth mount isn't difficult to drive, per say. It all depends on what you want to do with it. (btw, another of my passions is astronomy )

If all you want to do is move it to a specific altitude and azimuth,then take over manual tracking from there, that's relatively easy I suppose. If you actually wanted to track in alt-az mode that gets pretty difficult. I've thought about it (for a drive system for a dobsonian) using a picaxe and really found it non-feasible. You can point it to something and take over tracking, or you can track, but both in an altaz mode involves alot of maths that the picaxe simply cannot do perhaps with the math co-processor it could, but it probably wouldn't eb fast enough and youd get alot of acumulated error.

If you werent going to move the scope from generally the same area, it's possible you could pre-determine the tracks to take and develop an algorithm or lookup tables to do it. It would be alot of work, but not IMPOSSIBLE, just not a very good way to do it.

The other option is to tip your altaz over and turn it into an equitorial mount. All an equitorial mount is, anyways, is an alt-az mount tipped over so it's parallel to the earths axis.

This makes life ALOT easier as all you need to do is drive one axis at sidereal rate once youve "locked on" to your target. locking on to the target is also easy as it's always going to be in the same place relative to the meridian and time, so you could even store your favorite objects in an eeprom and have the picaxe drive the scope there automatically for you. or go full control from a pc.

I'll tell ya though...from my experience (and this is only my opinion) for jsut pure pleasure observing, I actually prefer a non computerized, mostly manually controlled scope. It offers mroe challenge and better satisfaction to me when I find somehting on my own. Tracking is nice though, as turning the slow motion controls all the time while trying to focus or change eye pieces is a pita, especially at high magnification.

For the purely scientific persuits, computerized is the way to go. You cang et alot more done, collect alot more data, and get alot better repeatability in alot less time.

so it all depends on what you want to do as an observer as to which direction to take with controlling your scope. It's not a simple prospect, but combining the two hobbies (electro and astro) is always fun and often results in some ingenious mad scientist type contraptions. I jsut love those


--Andy P

 

[pittuck]

Tracking is dont using mathematical calculations to move 'stepper motors'?

Perhaps you could look at that floating point co-processor to do the math and have a i2c link to talk to it???

Should be feasible... Dunno about the math tho.

 

[andrewpro]

pittuck said:

Tracking is dont using mathematical calculations to move 'stepper motors'?

I says:

Hmm..that bit confused me a little. Did you mean "tracking is 'done' using mathematical calculations to move stepper motors?"

If so, in alt-az mode, yes, it is. You have to mvoe two motors at once, at jsut the right pace, to get the proper arc that the celestial object will follow, and it's not going to be the same for any two points in the sky really. The circles that they trace are different sizes the farther from the pole you are.

In equitorial mounting, you only have to move one axis at sidereal rate to track, which is pretty easy.