Gina's Observatory Roll-Off-Roof Automation - Work in Progress

[Gina]

Next stage is the detection of fault conditions.  I think the clearest way of coding this is to use functions to keep the code in the main flow simpler.  Fault detection can just clear the flags but there are a couple of loose ends to tie up. 

I now have the timeout and motor over-current faults detected and the motor turned off.  The over-current test was achieved by reducing the current limit and loading the motor by hand.

This completes three of the four testing stages.  Next is to get the status messages sent back to base tidied up and working properly.  I may also make the code more modular to make it easier to read.

[Gina]

Now looking at the messages sent back to base through MQTT.  So far I have mainly used one status message plus motor current and time running but I think it may be easier to split the status messages, at least for testing.  MQTT can handle a large number of messages so no problem there.

I'm thinking the state of both flags and both limit switches.  Since it doesn't make sense to set the roof both opening and closing at the same time I have provided an interlock to prevent this.  Thus the flags can only have 3 states - clear, open set or close set.  Same limitation applies physically to the limit switches.

I currently have this in the Dashboard but I shall simplify it.

[Gina]

Everything now seems to be working except that the elapsed time when opening or closing the roof isn't cleared.

Next stage is to sort out all the connectionsbetween control box and observatory.

    1. Power in 2
    2. Motor 2
    3. Limit Switches - each 4 wires + screen
    4. Rain sensor - 2 power and 2 contacts

[Gina]

Limit switch connections.

Gnd
NO
NO
NC
NC

[Gina]

Added 5 pin connectors to the stripboard for the Limit switches and connected the logic lines and Gnd.  I may not bother to connect the make contacts to the LEDs.

Took the motor drive unit out to the observatory and screwed it to the wall, threaded the chain through the pulleys and drive unit and connected to west end of roof.  Just need to sort out the rest of the connections and hoping I may have it working tomorrow.

[Gina]

Rain sensor connections.

Extension cable :-
    Orange - Yellow - Power
    Or/Wht - Green  - Gnd
    Blue      - Blue    - RainSensePin
    Blue/W - White  - Gnd

[Gina]

While we have had some decent weather I have concentrated on the hardware, in particular the observatory.  I found a couple of problems.  The roll off roof was not moving as easily as it should and ROR part of the lock down lever caught on the cover for the large chain pulley.  In clearing the jam I broke the woodwork the chain was attached to at the west end.  This actually showed up a fault in my design so was a benefit.  I have now changed the design.

This was the earlier design and the edge of the board broke away with the eyebolt

 Rather than using an eyebolt through the fascia board of the ROR frame, the chain tensioning spring was attached to a main frame construction bolt. I also found that the spring I was using wasn't strong enough so I replaced it with a stronger one - or rather two, to get the right length.

I held the long length of chain level with a temporary support rig while I pulled the chain out and hooked the spring over the bolt end and screwed on a retaining nut,

Most of the sticking problem with the ROR was due to swelling of the wood in the damp weather and I cleared the problem by removing some wood from the ROR frame and a corner post.  It has also freed up quite a lot from use - the roof hasn't been open much in the last few months!  It's still sticking slightly on the north side (I think) and I shall be looking into that.  All the same, I have had the roof moving under simple motor drive and been able to check motor current.  Mainly this has been under 2A but did increase to around 3A towards the end of the opening run.  Closing was much easier (strangely) - 1.5 to 2A all the way.

 

[MarkAR]

Posted by: @Gina

Closing was much easier (strangely)

Could be flex in the roof frame making it easier one way than the other.

[Gina]

Going back to the control box, just designed and now 3D printing a backplate for the box so that I can attach it to the dividing wall in the observatory.  The control box will just clip on.

[Gina]

Posted by: @MarkAR

Posted by: @Gina

Closing was much easier (strangely)

Could be flex in the roof frame making it easier one way than the other.

Yes, it could be.

 

[Gina]

After a bit of a problem with the chain binding I swapped tension springs to reduce the tension and now the roof opens and closes alright.  I had to increase the motor current limit in the ESP32 as it was tripping at too low a current.  There is a bit of a graunching noise from the chain on the smaller pulley due to being slightly out of alignment.

[Gina]

The problem is that the end of the chain is too far back and the line-up is wrong causing the chain to want to ride up on the side of the pulley.  I now have a sort of eye-nut that could move the end of the chain further out and more in line with the pulley,  I have a crude diagram to help explain this.  The blue line shows the present line of the chain and the red line how the triangle ring shaped nut could help.

[MarkAR]

The blue line set up doesn't look good at all. there red is much better. Just wondering if the other eye bolt you pictured before would be longer. Set a nut and washer on it first so that the alignment is better then bolt in place from outside.

[Gina]

Posted by: @MarkAR

The blue line set up doesn't look good at all. there red is much better. Just wondering if the other eye bolt you pictured before would be longer. Set a nut and washer on it first so that the alignment is better then bolt in place from outside.

I can't get the present bolt out though I could drill another hole right the way through and out through the shiplap cladding.  OTOH the thread starts too far from the eye and there wouldn't be room for it.  It has to clear the large pulley.  Anyway, I'll see how things look when I get back to it a bit later.

Thanks for your thoughts Mark ? 

 

[Gina]

Just about had time this morning to check the new chain system when it started raining.  The triangular eye-nut thing is too high - catches on the large pulley.  Might be able to remove the existing nut and use the new thing. Might then clear but not sure.  The existing nut is very rusty and might not come off.  May see if I can get some measurements when it stops raining.  I had just put the spring back on the bolt when the rain came - closed the roof with the motor.  Seemed very slow!!

[MarkAR]

Posted by: @Gina

the thread starts too far from the eye

Might be able to use a die to lengthen the threads.

[Gina]

Checking the threshold level I should theoretically need for the motor current.  The I to V ratio in the driver is 140mv per Amp.  500mV corresponds to 3.57A which seems a suitable limit.  Going by the graph below this suggests a level of 1000.  I've been using 250.  I reckon it's been triggering at around 1.5A - 2A.  The graph would indicate triggering should be less than an amp so I think something doesn't add up.  Think I shall stick with the empirical view and use 500 or a bit more rather than the full 1000.

[Gina]

Back to the chain drive.  Two changes made to get the chain more in line with the small pulley.  Use that triangle eye-nut to bring the end of the chain further from the roof frame and (bigger job) move the small pulley further from the motor unit and from the end of the chain.  I had made the decision not to cut the chain to length until I was absolutely sure of the length I wanted.  The spare was hooked over a nail on the ROR frame.  Now that decision has paid off as I have used all the spare chain to accommodate the move of the small pulley.  I corrected for error in chain length by changing the springs to give the correct overall length.  Here are some photos.

[MarkAR]

Fingers crossed for fair weather tomorrow to give it a good shake down test.

[Gina]

Posted by: @MarkAR

Fingers crossed for fair weather tomorrow to give it a good shake down test.

Thank you Mark. 

There are still a few things to sort out.  There's still some binding on the north side when opening the roof which I think may be the weather proofing.  Also, I have yet to connect and test the rain sensor.

 

[Gina]

Apart from the roof remote control, I also have to sort out the weather proofing where the eye-bolt broke the woodwork and rubber covering and the bottom of the pulley cover.  I'm thinking of something attached to the ROR framework.

 

[Gina]

Arranged weatherproofing by the cover for the big chain pulley.  Crude but adequate I believe.

[Gina]

Also been clearing some more binding today but still more to do.  Set up the limit switches and the roof now stops exactly where it should within a couple of millimetres of the end of travel.  Remote operation is now working fine from indoors - the binding is not so bad as to limit the movement of the roof.  Travel time it around 40s.

I still have the east fold down window to sort out.  If I want to use the cord system running off the roof, I shall need to provide something to limit the movement to 90 degrees ie. horizontal, whereas it now goes right down to touch the wall.

[MarkAR]

Posted by: @Gina

something to limit the movement to 90 degrees ie. horizontal,

How about a couple of large L-brackets bolted outside to stop and support the weight.

[Astrodad]

Looks fantastic, I've told the wife we need to move in order so I can have my own observatory in the back garden, right now I have to keep taking everything outside to set up each time and am getting moaned at for leaving the back door open ?