My project is a remote and manually controlled double door gate, in operation since end 2015.
The sketch includes a counter that terminates an operation if it takes too long (suspected mechanical failure) and current sensing in case of an obstruction. In such events relay power is removed, the display says "ALARM" and there are 100 beeps from the buzzer. After that, the control buttons are enabled again.
Yesterday holiday guests noticed the beeping when closing the gate, probably because an end switch malfunctioned. For 10 minutes they kept pressing buttons and tried to push the doors back open with their full bodyweight. The result is a burnt out current sensor and a severely damaged linear actuator.
What can I possibly do to prevent this? A buzzer and a text on the display clearly aren't enough.
The old saying "it's impossible to make things foolproof ... fools are so ingenious" comes to mind here. There is almost always a way to (try to) do something that the designer never though of. The best you can do is work on changing your design/code to deal with the situations like this.
I happen to know a guy that built a factory conditional access door specified for 10k accesses/day with 10 years warranty. Those workers were tough. The door was (still is) tougher, no matter how hard it was pushed, bumped in, beer splashed or jumped on top of it. I can ask details about what other abuses (hammers, levers, etc..) it had to resist.
Was sold for a 4 digit figure/piece, yet way cheaper than what some swiss-made established company asked.
He kept spare parts for warranty that was never reclaimed but I don't know if warranty expired yet.
Anyone interested?
For-profit companies would not be. Many make most of their money selling repairs and replacement parts for poorly designed and failure prone equipment.
Example: I have a UV sterilizer for a drinking water system. The electronic ballast for the UV tubes fails about once a year, and the company wants $500 to replace it. I now use a standard $20 fluorescent lamp electronic ballast instead.
What did you expect to happen when people who were not familiar with the operation of the gate needed to get through it and it went into that failure mode?
It must either be so durable that it can take it when people who want to get through it when it doesn't want to open, or it must be readily bypassable so they can go around it, or you should not expect untrained users to operate it without damage. How do they know that if they wait for the beeping to stop, the door would magically open again (maybe)?
CDK1:
probably because an end switch malfunctioned.
reed switches or hall sensors + magnets to replace or double the stop command right before the end switches clicks.
The result is a burnt out current sensor and a severely damaged linear actuator.
A trade secret of my friend's factory door was to have very strong rigid springs between actuator arms and door hinges. Anyone trying to bump it could only compress against the spring a couple inches and was pingpong-ed back by the spring. That way he protected the gears. He designed his door at home. It senses the slightest push while it opens or close and reverses the movement, to prevent accidents.
What can I possibly do to prevent this? A buzzer and a text on the display clearly aren't enough.
nope. Even a big red poster "Don't even think to abuse me, I have 100kV in my punishment capacitor" might be tempting for some.
Is it possible for the alarm condition to cause a physical lock to be employed so that all the pushing forces are taken by the lock and not by the actuator?
Alternatively maybe the alarm condition would de-energize everything so the door opens easily by pushing?
I don't understand why the system would return to an operating condition without someone inspecting it and pressing an "OK" switch.
blimpyway:
I happen to know a guy that built a factory conditional access door specified for 10k accesses/day with 10 years warranty. Those workers were tough. The door was (still is) tougher, no matter how hard it was pushed, bumped in, beer splashed or jumped on top of it. I can ask details about what other abuses (hammers, levers, etc..) it had to resist.
Was sold for a 4 digit figure/piece, yet way cheaper than what some swiss-made established company asked.
He kept spare parts for warranty that was never reclaimed but I don't know if warranty expired yet.
Anyone interested?
CDK1:
My project is a remote and manually controlled double door gate, in operation since end 2015.
The sketch includes a counter that terminates an operation if it takes too long (suspected mechanical failure) and current sensing in case of an obstruction. In such events relay power is removed, the display says "ALARM" and there are 100 beeps from the buzzer. After that, the control buttons are enabled again.
Yesterday holiday guests noticed the beeping when closing the gate, probably because an end switch malfunctioned. For 10 minutes they kept pressing buttons and tried to push the doors back open with their full bodyweight. The result is a burnt out current sensor and a severely damaged linear actuator.
What can I possibly do to prevent this? A buzzer and a text on the display clearly aren't enough.
change the drive mechanism. there is no reason in the world that the abuse on the door should damage the motor and I would need to see the schematic to see why it would damage the current sensor.
a manual release might be in order as well. might be a good idea to allow a partially opened gate to be opened fully by some device.
as for the fools, IIRC, didn't they used to pour boiling oil over the rampart walls ?
DrAzzy:
What did you expect to happen when people who were not familiar with the operation of the gate needed to get through it and it went into that failure mode?
That is not what happened. They already had their car outside, then they heard the alarm buzzer and walked back to the gate where they kept pressing buttons and when that didn't help started demolishing things.
There is a leaflet explaining things, with instructions how to disable the drives, but holiday guests never read anything. I can confront them with the security camera pictures but that doesn't repair the damage.
I do not know exactly what caused the malfunction. The limit switch disconnects power to the "close" relay coil causing the electric motor to stop immediately. This was an extremely hot afternoon, maybe the relay contacts were fried and the relay kept conducting. In case the gates cannot fully close because of an obstruction, the current sensor output rises above the limit and the Arduino performs the alarm routine.
There is also a stainless wire string on the edge of the gate, attached to a microswitch. If the switch opens briefly, the alarm routine is performed.
Maybe it is better not to sound the alarm at all, or attach an external siren so the whole neighborhood knows something is wrong.
dave-in-nj:
change the drive mechanism. there is no reason in the world that the abuse on the door should damage the motor and I would need to see the schematic to see why it would damage the current sensor.
At the moment I am so fed up I didn't bother to remove the covers and assess the damage, but at least one spindle is bent.
Current sensing is done with ACS-712s. The sketch takes 3 samples and calculates the average value. It must be that way because the current fluctuates wildly. These devices can handle severe overloads but only for a few milliseconds. Current isn't measured during the first 100 ms to ignore the inrush current, repeatedly starting with a blocked motor causes the semiconductor to overheat and die.
I've tried a lot of limit values and finally used a very high one. The supply voltage can be 11.5V at night or 14.5V in bright sunlight, there is wind pressure to be considered, leaves on the ground, fallen figs from a nearby tree etc. I don't want the alarm to sound without a valid reason.
I'm sure this is not what you want to hear. If you have guests..... Assuming guests are temporary. And you don't want them to destroy the gate while they're there. Remove the gate!
remove the alarm. they probably would not have bothered to go back to close it. Holiday guests do not read and to not close doors. the alarm probalby made them think they did something wrong and your video would show their attempts to fix it.
Use a simple mechansim to allow the gate to close if the motor fails. maybe a magnetic coupling can detach from the motor and allow the gate to close without the motor. making such a connection should not be that hard.
Maybe use a magnetic coupling both ways, but a solenoid type lock when it is closed. a mechanical detent is also not that hard to do. it can be 90% of the strength of the motor/gearing.
the more I think about it, you could put on a rail on the door, then a slide on the rail the slide
without photos or sketchs, we can only guess. but really, you are on an electronics forum asking mechanical questions.
you can put in a simple speaker that issues verbal commands.
'do not be alarmed, the gate is experiencing a minor malfunction. please do not touch the gate while we try to auto correct this error. Please continue on your way, a notification has been sent to a technician to come and re-set. Please leave the area and please do not touch the gate. please do not try to open or close the gate. please allow the gate to self correct. please let the technician come and make simple repairs and not have to fix and catastrophic damage you will do by being stupid and touching the gate at this time. please leave the area and do not touch the gate. warning, if you touch the gate, you can do thousands of dollars of damage. warning, if you touch the gate we will post a video on youtube with your stupid actions and how you cannot follow directions.....
if you can hear this, you have not left the area as requested and we can only assume you want to destroy things. please do not be a jerk and touch the gate and destroy things. are you still here ? LEAVE NOW the loud siren that will hurt your ears will start in 9..8..7..
For the time being I have disconnected the drives, devised a simple mechanism to close and lock the gate in the evening and removed the key from the master switch. We just have to think about it, because there are deer and boar waiting for an opportunity to get it and redecorate the garden.
What seems to have happened is the following:
One of the four "Songle" 30 Amps relay contacts did not close immediately when the Arduino switched it on, causing the left gate to hit the right one. The motor current exceeded the limit and the alarm routine was executed. The same relay also doesn't always open when the coil is removed. It happened on an exceptionally hot day, so I guess the plastic relay casing is a bit deformed.
The gate has a warning sign in English, German and Croatian saying the gate should not be blocked or touched, but in this case the guests choose to ignore that.
I think I have to replace the Chinese relays with larger automotive ones with 12V coil voltage. And let the alarm routine end in loop so it needs a reset to restore operation.
I suspect you are beginning to realize why equipment supplied by a professional firm is more expensive than DIY stuff
Many times I have built something that works perfectly well for me and the first time a friend tries it screws up because he does something perfectly reasonable that I had not thought of.
