(12.10.2016, 14:15)Erwin van der Zwart Wrote: Hi AEK,

We did this Lutron link 2 years ago on a big project.

Keep in mind to send a telegam once every x period otherwise the TCP connection will be closed (after 12 hours i believe) by Lutron.

We included a mechanisme to request date or time every hour from the Lutron server to keep the connection alive.

BR,

Erwin

(02.02.2017, 17:44)Domoticatorino Wrote: Hi Erwin,
is it possible to have some suggestion about Lutron integration? In the next week I will have to integrate it.
Let me know if it could be possible.
Thanks.

(14.11.2022, 16:36)admin Wrote: What kind of status do you want to get? Do you already know the required status request message that should be sent to Lutron?

(21.11.2022, 08:12)admin Wrote: Which commands are you using for control? The status command is similar but it begins with ? instead of # and the last parameter (value) is skipped. This example parses HVAC query: https://forum.logicmachine.net/showthread.php?tid=414&pid=2197#pid2197
You will need a separate resident script for status queries.

(22.11.2022, 12:29)admin Wrote: Try this, no guarantees that it will work because we don't have any Lutron devices to test. Check Logs for any errors if the script does not work.
Adjust each query call as needed. First argument is the query request (without \r\n, it's added automatically), second argument is the status group address where the new value is written.

Code:

require('user.lutron')

sock, err = lutron_login()
if not sock then
  log('no connection to device', err)
  os.sleep(5)
end

function close(err)
  sock:close()
  sock = nil

  log('receive failed', err)
end

function query(req, out)
  if not sock then
    return
  end

  sock:send(req .. '\r\n')
  local res, err = sock:receive()

  if not res then
    return close(err)
  end

  if res:sub(1, 1) ~= '~' then
    return close('invalid reply ' .. res)
  end

  if res:sub(2, #req) ~= req:sub(2) then
    return close('invalid reply ' .. res)
  end

  local val = res:sub(#req + 2)
  val = tonumber(val)

  if val then
    grp.checkupdate(out, val)
  end
end

while sock do
  query('?DEVICE,10,6,3', '1/1/1')
  query('?DEVICE,10,1,2', '1/1/2')
  query('?DEVICE,10,3,4', '1/1/3')

  if sock then
    os.sleep(5) -- delay between each poll cycle
  else
    break
  end
end

(29.11.2022, 10:14)admin Wrote: What is the resident script sleep time set to?
For outputs use this query format:

Code:

query('?OUTPUT,151,1', '1/1/151')

(29.11.2022, 10:35)admin Wrote: Add logging to the query function like this and post what you get in Logs tab:

Code:

function query(req, out)
  if not sock then
    return
  end

  sock:send(req .. '\r\n')
  local res, err = sock:receive()

  if not res then
    return close(err)
  end

  if res:sub(1, 1) ~= '~' then
    return close('invalid reply ' .. res)
  end

  if res:sub(2, #req) ~= req:sub(2) then
    return close('invalid reply ' .. res)
  end

  local val = res:sub(#req + 2)
  log('response', req, out, val)
  val = tonumber(val)

  if val then
    grp.checkupdate(out, val)
  end
end

(29.11.2022, 11:01)admin Wrote: Try this query function:

Code:

function query(req, out)
  if not sock then
    return
  end

  sock:send(req .. '\r\n')
  local res, err = sock:receive()

  if not res then
    return close(err)
  end

  local pos = res:find('~')
  if not pos then
    return close('invalid reply1 ' .. res)
  end

  res = res:sub(pos)

  if res:sub(2, #req) ~= req:sub(2) then
    return close('invalid reply2 ' .. res)
  end

  local val = res:sub(#req + 2)
  log('response', req, out, val)
  val = tonumber(val)

  if val then
    grp.checkupdate(out, val)
  end
end

(29.11.2022, 11:16)admin Wrote: You can lower the sleep time between polling cycles by modifying this line (currently it's 5 seconds):

Code:

    os.sleep(5) -- delay between each poll cycle