E.stop wiring

Mike,

I asked the forum for advice on the above, some months ago. One reply gave me Axminsters wiring which did not work. I did not buy mine from Axminster, got it from Hong Kong, identical appearance and eve same part no. Another reply gave me a different circuit which did work and has done ever since. Description follows.

Mains Live in to terminal 1 on emergency button. Live out terminal 2 from button to terminal 13 on stop /start. Live out terminal 14 to motor.

Mains Neutral in to start/ stop terminal23. Terminal 24 out to motor.

Whilst it works, have I got it wired correctly?

John

The reason I connected the switched live via the E-stop switch to A1 is

An accidental short between A1 and neutral or earth will blow the mains fuse

If you had live to terminal 13 and neutral to terminal 23

an accidental short between A1 and neutral or earth would energise the coil when the E-stop switch is open

assuming the mains supply is not protected by a residual current circuit breaker

john

Hi, I agree with John Swift1 diagram. The E-stops should only control the coil of the KJD17 and if you want more than one E-stop, then they should all be connected in series with the one shown. Protection of the KJD17 should be via a fuse or circuit breaker which feeds the supply to it, as suggested and non of the load current should go through the E-stops as Mike Poole said. One advantage of this is that the only way the motor will start is by pressing the start button and unless the KJD17 is or has been seriously overloaded, it is unlikely to fail and so if you get a power cut during use, every thing after the KJD17 will be in a safe mode when power is restored.

Regards Nick.

Edited By Nicholas Farr on 10/05/2019 07:17:57

This is not strictly correct. For the sort of loads we are talking about here a decent E-Stop switch can handle the full load and still stop the motor if the no-volt release fails. 16A rated 2 pole E-stop switches are readily available. This primary E-stop is close to the NVR. Additional, remote E-stop switches can be wired in series with the NVR coil. The switch side of the coils should be in the neutral side so the wiring is not connected to a high current live. A short to ground will disable the remote e-stop (test them once a month if you are worried) but will not cause a fire as a live feed might. If you do put the remote switch connection of the holding coil in the live a fuse should be fitted at the source end. There is no advantage in fitting the E-stop local to the NVR in the holding coil circuit rather than the power. The suggested dual power and holding coil wiring costs no more and gives redundancy. Personally I would not trust a cheap Chinese NVR as my final safety device.

I have designed E-stop, guarding and interlock systems for robotic machines to meet European and North American safety standards so do know a bit about this.

Robert G8RPI.

If you put the POWER switching E-stop before the NVR (KJD 17) then the NVR will release when the E-stop is pressed so the motor will not re-start when the E-stop is reset.
If the KDJ 17 is so reliable why are there so many being sold? There are dozens just on ebay, they are not all for new build machines.

The bigger problem is the NVR contact is not failsafe. The only force opening it is a spring. If the spring breaks, coil armature or sear sticks or the contacts weld there is no force to break the contact. A properly designed E-stop has forced break contacts. The force on the button is directly transferred to the contacts in the opening direction. Short of the structural strength of the switch being exceeded the contact will open. With the E-stop in the NVR coil circuit all the NVR failure modes or the E-stop failure modes will cause it not to work.

Robert G8RPI

One further consideration is to fuse the appliance correctly, this formula applies to AC and DC

Power (Watts) = Volts * Current (Amps)

@Grant - you should put a 2A fuse in the appliance plug, always put the closest commonly available fuse to match the power load. If an overload situation occurs, the fuse protects the rest of the components

0.37kW = 220V * 1.68A

Now there are variations to the rule, startup load may exceed minimum fuse rating, but in general, the principle holds true and slow-blow fuses are available too

Whatever you do, don't leave 13A fuses supplied in (almost) all domestic appliances, calculate the appropriate fuse value first

Hi Robert, I accept your credentials, however I've never taken one of these apart, but When you press the stop button, you can feel you are pushing the contactor away, so any spring only holds it away or pushes away during a power cut. Even industrial contactors have springs in them to push the contacts apart and yes I have even seen these contactors welded together due a very large overload. Nothing is without the possibility of failure, but I've only ever heard of the KJD17 and the like to fail to start and that is normally due to the coil going open circuit and that even happens in industrial contactors. In all my working industrial experience, I've never known E-stops to carry the load current, they have always controlled the coil. An E-stop can be subject to an accidental short just as much as anything else, wherever is placed in a circuit, only good design and good wiring practice will reduce such accidental shorts to a minimum in any system. One other point is that a lot of people don't seem to realise is that both the live and neutral carry the same power and have the same potential from each other, it is just that the neutral is grounded at the supply and there is a measurable difference between earth and neutral at the consumer end albeit very small.

Regards Nick.

Bit of false logic in 'If the KDJ 17 is so reliable why are there so many being sold?' : Robert draws the conclusion that they must be unreliable because 'so many are being sold'. That might be true, but there's no evidence for it. We don't know what the size of the market is for these switches, or their failure rate. But the real killer is that we don't know how many are being sold, we only know that they are being offered for sale. If Robert's logic is correct, then it also applies to the entire RS catalogue!

I agree though that these NVR switches aren't fail-safe. But for exactly the same reasons neither are the contactor circuits used in most machine tools.

Problem with fail-safe systems is that covering every contingency, however unlikely, rapidly gets complicated and error prone. In this case, we are reducing a workshop risk, not preventing a second Chernobyl. I suggest there's a hierarchy of measures that could be taken:

1. A fused machine is safer than an unfused machine. (People do replace fuses with paperclips...)
2. The correct type of fuse (or circuit breaker) is safer than the wrong type of fuse (fast-blow, slow blow etc.
3. A switched machine is safer than an unswitched machine because it can be turned off in an emergency. I suggest steps 1-3 are the absolute minimum.
4. An NVR switched machine is safer than a plain switched machine because the machine won't restart automatically after a power cut. I suggest this is well worthwhile.
5. A machine fitted with a prominent easily reached Emergency Stop switch wired to disconnect the entire supply is safer because the machine can be stopped quickly, even by someone who knows nothing about the machine. I suggest this is highly desirable.
6. A machine fitted with safety interlocks on guards, end-stops, and doors etc is safer because the machine stops when the operator does something unwise. Whether or not this one is applied depends on the risk. However, it's not usually desirable, or safe, for small interlock switches to interrupt the full power applied to the machine. Instead, they are wired to control a contactor, or the NVR switch.

For what Grant is doing, Andy's circuit seems entirely appropriate to me:

And if Grant has secondary requirements to cut power, like protecting a door, putting extra switches in the A1 to 24 link wouldn't be insane. However, I wouldn't use a KDJ17 like that if the secondary switching was frequent because it's unlikely that a cheaply made NVR will take a lot of on/off cycles. For multi-repeat switching a contactor is a better bet.

My experience of safety interlocks is in software, where I learned my second favourite adage: 'Fail-safe systems fail by failing to fail safely'. The best advice is 'Never meddle with gunpowder by candlelight'.

Dave

This is my recommended circuit for a KDJ17

It is fine for anything up to a few kW. 0.5mm "twin red" lighting cable would suit the wiring to the secondary E-stop(s)

Dave, I think it is quite reasonable to assume that mot of the KDJ17 NVR's sold on ebay are for replacement purposes rather than for manufacturing. For 3 phase you do have to use contactors etc but for safety critical circuits on production equipment you have to use suitably safety rated components. https://uk.rs-online.com/web/c/automation-control-gear/machine-guarding-safety/safety-contactors/

Robert G8RPI.

Edited By Robert Atkinson 2 on 10/05/2019 19:54:04

I am with Nick in that all my years in industrial electrical maintenance I have not encountered an estop carrying load current but your proposal will meet the minimum category of safety and this will probably be satisfactory for the risk involved. I think most of the safety devices that switch full load current have been ultimate switches on hoists and such like equipment, but this is a thing of the past on most large installations. Still find them on stand alone things like electric chain hoists. Care will be needed in the selection of the stop button to switch load current but most decent quality buttons will be capable. I would still put the stops in the control rather than load circuit as this concept works for all loads. Not come across the idea of a primary and secondary estop, in my world they should all stop the machine without any doubt.

Mike

Hi Mike,

I agree most industrial installations would have the E-Stop in the control line, but most industril machines will be 3 phase and higher current. If we were talking about using industrial quality contactors and switches from reputable manufacturers and suppliers I would be happier with just control line E-stops. However we are talking cheap all in one NVRs made in the far east and I have little faith in their reliability. There is virtually no cost or difficulty impact from putting one E-stop that is close to the NVR in the two phase power input before the NVR. Any additional E-stop(s) can be in the control (holding coil) circuit to simplify wiring and not have power on the wiring. My reasoning is if the NVR does not release when toy press the button you will go for the nearst E-stop which will work 100%. If a remote E-stop does not work you will again go back to the NVR and the primary E-stop. This may seem like overkill, but why not when it is very little cost in time or effort that may save a work piece, machine or bodily harm some time in the future.
One thing is not clear on my drawing, I say the input comes from a CB. Obviously ther shoyld be some kind of isolator between the power source and first E-stop. For single phase machines this would normally be a 13A plug. For larger machines either an industrial plug and socket ot a wall mounted isolator switch.

Robert G8RPI.

I think we broadly agree Robert, I tend to favour industrial control gear even at home but it does cost if you need to buy it new. I doubt this meets any standard for a safety circuit.

Mike

That looks like the original nuclear reactor SCRAM (shutdown) system. Allegedly the first reactor CP-1 had a control rod hung from a rope and the plan was to cut the rope with an axe if the reaction ran away. They still call a emergency reactor shutdown a SCRAM but there is no longer a Safety Control Rod Axe Man

Robert G8RPI.