Re: Re: Whoooosh..thud..oh bugger

[Martin Howell <martin.howell@xxxxxxxxxxxx>]

Thanks Simon, not as strange a suggestion as it might seem.
I originally planned a fully automated system, using a
controller devised by Rodney Hall (on this list, thanks
Rodney), but decided against it on the grounds that it would
be pretty complex - IR beams to trigger both open and close;
obstruction detection; manual overrides etc - and that it
would need too many fail-safes to stop it from squashing
people, and avoid getting Aunt Edna getting trapped in the
lounge for several days while we worked out what had gone wrong.

I'm sticking with the simple manual approach for now, with a
series of micro switches to slow the door down for the last
6 inches of travel and to shut the power off before the door
hits the stops, together with some cushioning of the cable
drive mechanism.  As a last resort (and with higher WAF), I
could disconnect the drive completely and just add a handle
on both sides, though there's still the problem of small
boys closing the door enthusiastically and forgetting to
move all their body parts out of the way...

Martin

Simon Pawson wrote:
> I tried this reply last night but haven't seen it yet...
>
> Coming totally from a strange direction....
>
> could you do something with a reversing sensor system for a car. The
bar
> graph of LEDS could be converted to opto-isolators driving relays with
> build in resistors to slow the travel.
>
> Might be hard getting a reliable bounce if wall/frame mounted. Easier
> spotting the wall from the door but not so elegant. Could be set up as
a
> safety measure as well. Child in the way stops (or even beeps).
>
> Sorry just an odd thought - back to the merlot.
>
> Simon
>
>
> In message <461A91D0.8000200@xxxxxxx>, Martin Howell
> <martin.howell@xxxxxxx> writes
>> Hmm, this unit from RS gives the 'soft start' I'm looking
>> for, but not the 'soft stop, which is where the real problem
>>  lies.  I agree with your analysis of the drive train
>> issues, but I'm concerned that strengthening the cables and
>> applying more tension will just move the force to the next
>> weakest part of the drive - the cable fails at present
>> because of a joint, inserted partly to prevent breaking a
>> more difficult to fix part of the drive.  I'm looking at a
>> more simplistic system of forcing a slow down at each end of
>> the movement, watch this space to see if it works, or if the
>> whole contraption bursts through the wall into the garden :-)
>>
>> Martin
>>
>> David Cole wrote:
>>> Martin
>>>
>>> Sounds to me like you need a PWM (Pulse Width Modulation)
drive to
>>> efficiently control the speed of the motor.
>>>
>>> Something along the lines of 510-1209 at RS should do it, this
would
>>> give the acceleration and deceleration whilst also limiting
the
>>> current if the door jammed. Alternatively google for PWM DC
drive
>>> circuit
>>>
>>> You want to ensure that your drive train (cables) are as stiff
as
>>> possible, this is why industrial applications usually use rack
and
>>> pinion drives. This may mean increasing the diameter of the
cable to
>>> ensure that you don't have any stretch which causes jerk. It
sounds
>>> like you have a lot of inertia in your system it may help if
you add
>>> additional gearing which will reduce the speed, increase the
torque
>>> but also decrease the amount of reflected inertia seen by the
motor.
>>>
>>> Hope this helps
>>>
>>> Regards
>>>
>>> Dave
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>> --
>>
>>
>> Grizelli Associates Limited
>> IT Project Consultants
>> 2c Nethermoor Road, Middlezoy, Bridgwater, TA7 0PG
>> UK Company registration no. 2468872
>> VAT registration no. 713 1775 49
>>
>>
>>
>>
>>
>>
>>
>>
>

--


Grizelli Associates Limited
IT Project Consultants
2c Nethermoor Road, Middlezoy, Bridgwater, TA7 0PG
UK Company registration no. 2468872
VAT registration no. 713 1775 49