You could use a 24V no-volt release start / E-Stop module. These a re cheap enough. Arrange the guard switch(es) to drop out the no volt coil and one contact to the VFD.

Robert G8RPI.

I had a quick look at the Mitsubishi manual for the FR E729S inverter. The wiring diagram shows details of the safety circuits which show dual inputs S1 and S2 for the safety stop. It is intended that each device in the safety system should be provided with two electrically independent contacts ( closed when safe and open when unsafe ). Each circuit ( PC to S1 and PC to S2 ) should contain one of the dual contacts in each circuit. The intention is that there is two identical safety circuits which normally operate together hence giving two chances of a safe stop – operating only one circuit ( such as in the case of a mechanical failure ) causes a lock-out until both circuits are opened.

PLEASE, PLEASE make sure that the principles are adhered to – they are there to keep us safe ( even if they are a little more difficult to understand and implement ).

No ,

Meant a 24V NVR, powered from the PC terminal and contacts to the safety circuit.

Addressing Bobs comment, while using dual circuit stops is best practice, it depends on the machine. A Failure Modes & Effects Analysis (FMEA) would be carried out and severity of failures addressed. The level of protection will depend on the hazards and use. For a non CNC lathe or mill used by a skilled operator a dual circuit protection system would not normally be required. If you did a safety analysis on a chain saw as a new invention it would never be allowed!.

One particular failure mode of most E-Stops is that the switch module fits on the back of the button assembly and it's actuator has to be pressed to open the contacts and stop the machine. If the switch block is not fitted or falls off for some reason the E-stop wont work, dual circuit or not. The only mitigation for this is testing the E-Stop, but the switch block could drop off as you reset it. I've looked for a fail-safe E-Stop but never found one.

Robert G8RPI.

Inverters are a complexed, software controlled motor drives. As such, we must understand that the drive may fail either without any power to the motor or with full power to the motor ( or anywhere in between ). Since the inverter is under software control, either failure mode is possible and the standard Start, Stop, Forward an Reverse controls which also pass via the software may become inoperable. To prevent unsafe situations, safe shutdown inputs are sometimes provided, generally preventing the output stage of the inverter from driving the motor, preventing power transfer by disabling the output devices.

So, lets look at the safety issues surrounding inverters on machinery :

• Your lathe motor is directly connected to the lathe chuck via a system of belts, pulleys, gears and shafts, therefore any unexpected starting of the motor will result in potentially dangerous rotation of the chuck – imagine should the chuck start turning whilst using the chuck key – potential injuries could result in loss of fingers, hands or even arms.
• Equally, un-safe opening of change-wheel guards etc. also exposes the user to similar hazards should the motor start without warning.

According to the diagram posted by William, he is expecting to find door switches and an external Emergency Stop fitted to his machine.

The reason for the provision of two parallel safety circuits ( PC to S1 and PC to S2 ) is to reduce the affects of potential failures in the safety circuits – i.e. there are two chances of detecting a failure rather than one therefore the door switches and Emergency Stop should form part of these twin parallel circuits.

OK, so lets look at the requirement for safety switches on machinery :

• Safety switches should have two independent electrical circuits ( one for each safety loop ),
• Safety switches should be forcibly disconnected when the guard is opened ( or Emergency Stop pressed ). This ensures that the contacts of the safety switch are forcibly opened even if the switch contacts were welded closed due to a fault.
• Doors should be fitted with a 'tongue in the slot' type safety switch or a suitable mechanical cam to force operate a roller type limit switch. Note that the limit switch must be a safety type switch and that micro-switches generally do not conform to the safety requirements.
• Emergency stop switches manufactured and CE marked ( yes we know the Europe issues and the comments about China Export markings ) by reputable manufacturers will provide dual safe contacts.
• Safe operated switches providing two electrical circuits may still fail because their fixings may come loose allowing the switch to fail to operate. This is the subject of another whole massive set of legislation.

Finally, one should provide two identical loops from PC to S1 and PC to S2 ( this is the reason that the drive manufacturer provides two terminals ) – looping the two safety inputs together defeats the purpose.

One final note regarding the safety circuits, the inverter should not re-start after the safety circuits are completed following a safety stop but must be provided with a manual method of re-starting i.e. a Reset pushbutton or maybe the Start button to be pressed again.

Finally, to address Robert Atkinson 2's comment regarding the integrity of Emergency Stop switches, it is generally accepted that designers of industrial control systems will incorporate a suitable monitoring relay which is specifically designed to detect such faults ( and many others ). The Mitsubishi manual details such a system a few pages on from the normal connection drawing.

Sorry to drag this out, but our safety is of the utmost importance.

Bob

A circuit matching William's requirement would be helpful. For example I'm confused about the practical implementation of: 'one should provide two identical loops from PC to S1 and PC to S2 ( this is the reason that the drive manufacturer provides two terminals ) – looping the two safety inputs together defeats the purpose.'

Also, which rules are quoted and who do they apply to? I note repeated use of the word 'should', which is advisory, rather than the far stronger 'must', which is mandatory. In a safety regime, 'should' allows judgement calls whenever the hazard level doesn't justify the need for full-on mitigation. Health and safety rarely requires 100% compliance with everything that could possibly be done to cover all eventualities. Few home workshops have written policies covering training, procedures, roles, responsibilities, and record keeping even though they 'should'! Not because we're gung-ho, but because those measures expensively address risks model engineers effectively manage in other ways. We aren't employers.

Anyone aware of a VFD starting a machine unexpectedly due to a fault? Certainly possible, but I've never seen a report. I suggest the risk is low, especially compared with an old lathe switched at the mains socket with no NVR or safety interlocks at all. Even they don't cause much trouble!

Amateur radio is much more likely to kill people than Model Engineering. Hams fall off towers and roof-tops, touch power cables with Aluminium ladders and dangling antenna, aerials collapse on them, and they get RF burns, high-voltage shocks mending big amplifiers, and are occasionally struck by lightning. Been a few murders and lost at sea disappearances too. Even so, the risk is statistically tiny compared with riding motorbikes!

Dave