Advice Please: Using 2 12v DC power supplies in parallel...

[PerryGunn]

Background - 'cos it always helps

I have some electric driveway gates with an electric lock and sometimes the lock doesn't disengage properly. I've troubleshooted the issue and it appears that sometimes the lock binds slightly on opening and doesn't spring back. As soon as the gates start to open (about 1 sec after the lock should have disengaged), the binding pressure on the lock is released and, if I'm standing there next to it and press the manual release button, the lock will spring open and the gates can operate as intended. If the lock binds and the manual release is not pressed, the gate motors are pulling against the lock and I've already experienced some gate damage due to this.

Mechanical fixes have already been explored and while some effect a temporary solution either the problem reoccurs or they come with 'worse' side-effects.

The binding is due to the pressure that the gate rams apply when closed - with help from the suppliers they've been fine tuned as much as possible but, at times, they still press the gates too firmly against the top/bottom stops. It's a pair of wooden swing gates so the reverse pressure against the lock changes due to the dimensional instability of the wood as the temperature/moisture content changes. The only way to avoid any back pressure is to set the rams up so that they barely close but this leaves too much play and they rattle/bang against the stops when it's windy - which will annoy both the neighbours and, more importantly, my wife

I've spoken to the lock manufacturer and they sell an add-on module which extends the solenoid latching time and suggested that it was applicable to situations such as mine. Great, I thought, I'll take one - however their module requires a minimum of a 0.5sec pulse to function and my control panel only supplies a 0.2sec pulse which they thought was too far outside their requirements for it to work reliably - hence why I've decided to make one myself. It's either that or replace a quite expensive electric lock that has the occasional glitch...

The gate control board (in the garage) sends a 0.2sec 12v DC pulse to energise the solenoid and disengage the lock, this means that, when the lock binds, the solenoid is no longer energised at the point the gates try to open so even though the binding pressure on the lock is released it is in the 'locked' position. If the solenoid activation pulse was longer, say 2-3sec, the lock would release as soon as the gate starts to move and the binding pressure is released.

What I'd like to do

Extend the lock solenoid activation period to around 3sec using the 0.2sec pulse from the Lock Control Module as a trigger. I'd also like to take a 'belt & braces' approach and connect both the Lock Control Module and the timed power to the lock so that if there is a problem with the relay board the lock will still receive the pulse from the Lock Control Module - this is shown by the blue and red lines in the diagram below.

How far have I got

It's been a long time since I've done much in the way of practical circuit work but I've created an extended 12v pulse by using a timed relay taking power from an independent 12v power supply (thinking of a 12v 5A laptop supply when/if it's permanantly installed), and using the 0.2sec 12v pulse from the Lock Control Module as the trigger. I've tested the timed relay hookup using one old car battery as the trigger and another as the power supply and I can generate timed 12v outputs that hold the lock solenoid open for the specified time.


Where I could use some advice...

- In order to utilise both 12v pulses I think i require a diode on the +12v line to isolate the two power lines. Is this correct?

- Do I require one on the -ve line as well?

- What size/rating diode would be required?

- Any issues using a laptop power supply?

- Am I barking up the wrong tree and this won't work?

Thanks in advance

[ Specified attachment is not available ]

[bob91343]

I spent a few minutes rereading and looking at the diagram and frankly cannot make sense out of what you are trying to do or how you are trying to make it work.

[TheMG]

Going back to the basics, you didn't mention if you had measured the voltage getting to the solenoid when it is activated?

It sounds like the power supply is located remotely (not near the gate), so there is probably a non-negligible length of wire between the existing power supply and the solenoid. How long are these wires? What size (gauge) are they? How much current does the solenoid draw when activated?

If there is too much voltage drop over the wires supplying power to solenoid, it will have reduced activation force and may be the root cause of the binding.

If you haven't already investigated this possibility, I would look into it before doing anything else.

[PerryGunn]

Going back to the basics, you didn't mention if you had measured the voltage getting to the solenoid when it is activated?

It sounds like the power supply is located remotely (not near the gate), so there is probably a non-negligible length of wire between the existing power supply and the solenoid. How long are these wires? What size (gauge) are they? How much current does the solenoid draw when activated?

If there is too much voltage drop over the wires supplying power to solenoid, it will have reduced activation force and may be the root cause of the binding.

If you haven't already investigated this possibility, I would look into it before doing anything else.

The lock control module on the main board in the garage and the specs say it's 12v 15W so, to me, that's 1.25A - there is approx. 12m of cable between the control pane and the lock (1.5mm SWA for all but the final 1.5-2m, then it's 0.75mm HR05xxx flexible cable (it needs to be flexible as the lock is in the middle of swing gates). Given its application this length of cable is not unusual and the recommendation was for 0.75mm cable.

I haven't been able to measure the voltage at the lock as the standard pulse is only 0.2sec and I haven't got anything that will measure a transient that short, but you can hear the solenoid operate. The solenoid doesn't pull back the lock into the open position, the lock has a rotating spring-loaded catch and the solenoid just releases the mechanism that stops the lock catch rotating into the open position. The solenoid operates but the rotating catch is what is being friction bound by the gates - when they are closed and the lock is bound it will not rotate into the open position even if I press the manual release button (which releases the same mechanism as the solenoid) or apply 12v directly to the lock terminals (using about 60 cm of 1.5mm solid core T&E mains cable for that test)

With my breadboarded setup the solenoid voltage is applied for 4 secs and the catch rotates freely as soon as the gate rams release the pressure on the gates.

I'm happy that the longer-timed pulse will solve the problem, I'd just like to know if I can safely supply both the original 12v 0.2sec pulse and the timed pulse to the lock without the cross-connection of the two 12v supplies damaging each other - I think I can but I think I need a diode in the position indicated in my diagram

[Ian.M]

Yes, you need a diode there if you have two sources feeding the solonoid load.  However you may also need one between your timer relay output and the load to protect the laptop PSU.  If you do need diodes you can salvage the dual common cathode Schottky diode used as the 12V rectifier in just about any (dead) PC PSU.

However I wouldn't faff around with a laptop PSU and trying to combine the two drive signals as that certainly will decrease reliability.  Find the 12V rail inside or feeding the original lock control module and tap off that to feed your time delay relay.   You must also check whether the original controller switches the positive or negative side of the solonoid output - don't assume the negative side is always connected to ground, so you may need to find its real ground rail to feed the time delay relay!

[PerryGunn]

I spent a few minutes rereading and looking at the diagram and frankly cannot make sense out of what you are trying to do or how you are trying to make it work.

Perhaps a step-by-step will help

- The Electric Lock has a rotating catch which rotates around a 'floating' (has several mm of play in each direction) catch bar on the other gate

In the normal installation
- The Lock Control Module is connected directly to the Electric Lock
- The Lock Control Module provides a 12v 0.2 sec pulse to the Electric Lock
- The pulse energises a solenoid in the lock which releases the locking mechanism and allows the spring loaded locking catch to rotate into the open position releasing the catch bar and allowing the gates to open.
The problem:
- The rotating catch sometimes 'binds' due to the pressure from the electric rams that open/close the gates
- When the catch binds, it is still in the closed position when the electric rams start to operate and the rams try to pull the lock and catch bar apart. The lock is rated at over 2000kg so, when this happens, the rams try to pull the lock and catch bar apart, causing damage to the gates.

If I supply a 12v pulse of a few seconds in length, the solenoid stays energised and the catch will rotate into the open position as soon as the rams start to open the gates and the binding pressure is reduced

In order to do this
- I purchased https://www.amazon.co.uk/gp/product/B06XCT39X4 from Amazon
- Connected a permanent 12v supply (currently a car battery but planning to use a 12v 5A laptop power supply)
- Connected the electric lock to the output terminals
- Connected the Lock Control Module to the trigger terminal
- Set the board up to provide a timed 12v supply (currently set as 4secs) to the output terminals when it receives the trigger pulse

This all works as expected
- Before the gates open the Lock Control Module sends a 0.2sec pulse
- The 0.2sec pulse triggers the timer module and it supplies the lock solenoid with 12v for 4 secs
- The electric lock is free to unlatch as long as the solenoid is energised
- As soon as the gate rams start to open the gates (around 1 sec after the initial unlock pulse from the Lock Control Module), the binding pressure on the lock catch is released and it's free to rotate into the open position (which it does)
- the gates open properly and there's no 'fighting' with the lock

What I want to know is if I can supply the electric lock with both the original 0.2sec pulse from the Electric Lock Module and the timed 12v pulse from the timer board without causing issues by joining the +/- of two different 12v supplies. I would like to do this in case there is a problem with the timer board as the gates will still receive the original 0.2sec pulse - they may still bind occasionally but the majority of the time the lock will work and this will reduce the chance of damage to the wooden gates should the timer board have issues.

I think I can do this but believe that  I need to add a diode on the +12v line from the Electric Lock Module


I hope that explains things clearly enough - if not, I'll attempt to clarify as I'm sure the regulars on here know a lot more about these things than I do and I may have been confusing matters with incorrect terminology. 

[PerryGunn]

However I wouldn't faff around with a laptop PSU and trying to combine the two drive signals as that certainly will decrease reliability.  Find the 12V rail inside or feeding the original lock control module and tap off that to feed your time delay relay.   You must also check whether the original controller switches the positive or negative side of the solonoid output - don't assume the negative side is always connected to ground, so you may need to find its real ground rail to feed the time delay relay!

That was my original thought, and a 12v from the main board would have been preferable, but the main controller board runs on 24v - the Lock Control Module is a tiny clip-in board that seems to be little more than 24v to 12v DC converter but it appears that it's only live for the duration of the unlock pulse.

Given the choice between another 'cheapo' board that does the 24v to 12v DC conversion and a separate power supply that will be running well within its capacity (it's rated at 5A 60W), I thought that the separate power supply was a better bet. The potential for unreliability in the timer board/external PSU combo was what got me started on the possibility of providing both the original 0.2sec pulse and the timer board output to the lock as a 'belt & braces' solution.

I'm planning on mounting the 'lash-up' in a wall-mounted ABS case with the laptop power supply fixed on top (for heat management) and was thinking of buying a 12v panel mount LED that can be mounted on the front/side of the case to show that the laptop PSU is working.
 

[bob91343]

When I read about metal parts binding, my first thought is to lubricate the surfaces to reduce the force needed.  Also polish the surfaces.  Reducing friction may be all you need to do, since it appears that things are marginal.

If there is a spring involved, perhaps making it stronger or weaker would help.  Or adding one to bias the latch into the open position.

Maybe a mechanical solution would do the job.

[NiHaoMike]

If you want to make sure that failure of the unit doesn't cause the lock to fail to work altogether, use a relay that passes through the signal by default, but switches it over to the separate power supply if activated by the 555 circuit. Don't forget the inverse parallel diode across the output.

[BMK]

Some gate controllers have a feature to power one or both of the motors briefly towards the 'closed' position stops during unlocking. Intended to solve this exact problem.
Mine is ELBE 909B has this feature, it is called 'reverse push' and is enabled by DIP switch
Maybe yours has a similar trick.

[PerryGunn]

Some gate controllers have a feature to power one or both of the motors briefly towards the 'closed' position stops during unlocking. Intended to solve this exact problem.
Mine is ELBE 909B has this feature, it is called 'reverse push' and is enabled by DIP switch
Maybe yours has a similar trick.

Yes, mine has a function called 'Kick Back' which is available at the end of the closing phase and/or the start of the opening phase - it was one of the first things I tried. Unfortunately the push towards the closed position is the opposite of what's required as it's the ram pressure that's causing the binding, if they had a function that worked the other way around and relieved the ram pressure briefly before sending the unlock pulse it would have been the ideal solution