Relay timing - MAHO MH400E - gearbox control

[RotarySMP]

I am working with a friend who is a professional programmer to create a LinuxCNC comp to control the 18 speed gearbox of my MAHO 400E CNC mill:

I have a question about the timing of motor relays (please refer to the attached schematic), and I would like your opinions:
Sometimes I need to power a relay to reverse the direction of travel of a 24VDC motor, and power the motor relay it at the same time. We could implement a delay in the software, but it would be added complexity.

Is it okay to set 11K7 "reverse direction" and also 11K4 to power the motor at the same time (1ms) ? Or would it be better to give some delay for the reverse relay to close and settle before putting on the power. If both relays are are powered similtaneously and if the motor relay closes faster, then the motor circuit will start to energise in one direction, and then the reverse relay will have to break that rising EMI of this inductive load, before reversing it, causing possible sparking of the contacts? Or is this not an a valid consideration with only 24VDC?

If anyone is interested in the larger project, the retrofit is documented in the LinuxCNC forum:
https://forum.linuxcnc.org/12-milling/33035-retrofitting-a-1986-maho-mh400e?start=450

While the gearbox control software project is in github:
https://github.com/jin-eld/mh400e-linuxcnc/wiki
Mark

[awemawson]

It is usual to have a contact on the 'forward' relay that disables 'reverse' and a contact on the 'reverse' relay that disables 'forward' to ensure that they CANNOT both be energised.

[RotarySMP]

Thanks for looking at this Andrew.

These is not separate forward and reverse relays, but a "reverse direction" and a "motor on" relay. Motor on always powers the motor, CW if 1K7 is de-energised, CCW if 1K7 is energised.

At the start of a gear change both are de-engerised. If I need to turn CW, no big deal, only the "motor on" relay is energised.

The questions is the case I need to turn CCW. So I command "reverse direction" and "motor on". Does it matter if I energise both at the same time, with the risk that the "Motor On" is a little quicker, and starts building EMF in the forward dierection, before 1K7 closes, and reverses it?

[PK]

I guess it depends on how far you want to go. Andrews approach of going from ON+DIRECTION to FWD+REV is sensible and reflects his experience with motor control hardware.
Staying in that space, you can have FWD+REV+BRAKE where the brake relay kicks in when the other two aren't on using simple relay logic. The brake relay just connects a resistor across the motor and pulls some energy out of the mechanism before the next relay engages.

As a software guy I say put a prox tach on the motor and, for a few lines of code control the motor in every situation.  Now you could do this in electronics, but it's nowhere near as easy as doing it in software.

Just my 2c worth.
PK

[RotarySMP]

Thanks PK.

THe relays are in the mahcine and work, I am not changing them, just activating them with LinuxCNC instead of the original Phillips controller.

My question is limited to whether delay is needed between switching with 24VDC relays.

[hanermo]

Everything electrical has delays, spikes and sparks related to voltage, current and coils in the circuit.

Some old stuff was very costly, very well made, and can resist some forms of abuse for a long time, or forever.
Most older expensive high end electrical stuff for milling machines was like that, at that time.

I would think the Maho relays are similar to that.
Overbuilt, robust, etc.
But they were almost certainly not expecting to be switched at 1 ms intervals or less via modern sw.

Since it´s a gearbox and the usage is likely to be very sporadic and sparse ..
I would very much prefer a robust enforced delay of some kind.

There could be inductive kicks, emf driven spikes, whatever, otherwise.
E.
My servo drives live and work for quite a long time on power-out, about 1 second as a tlar.
I think due to emf and caps and servo motor coils.

My wag is about 0.5 secs delay would be a good start.
Comparing it to VFDs and other std power electronics.

My servo drives can do 20 ms stops at 7.5 kW / 3 secs (300%) + whatever they dump (I suspect over 20 kW momentarily) but they are specifically built for that and I think they dump into the grid.

If you are going to switch fwd/reverse, fast, I would suggest some caution.
The spindle motor + hw has huge power and momentum.

I used to sell Haas machines.
The acceleration took about 30 kW to get to top speed, in 1-2 secs.
Cutting took mostly 3 kW or so.

I think that running a machine at speed, and reversing it electrically, fast, will blow the relays in a colourful way.
And probably all sorts of other stuff.

My servo spindle refit 2.5 kW has about 30 kW sustained power belts, and the mounts are over 50 kg in mass.
My WAG is that stops are over 20 kW in peak power.

[RotarySMP]

Thanks for the inputs. It was not the 1.5kW main spindle motor I was talking about sequencing, but just the little 24V 0.8A gearbox shifter motors.

We decided to implement 200ms delay between energising relays. It seems to work really well:

Mark

[hanermo]

Nicely done !

[RotarySMP]

Thanks. Now I am troubleshooting two Heidenhain LS-403 linear encoders with are making trouble.
Mark

[RotarySMP]

Turns out the Y axis linear encoder was fine. The PSU I added to power the Mesa 7i77 and the encoders was only set at 4.7V, whereas the Heidenhain manual specifies 5V +/-5%. Turned up to 5.1V it is stable.

The Z encoder was also okay. It was the third channel of the EX which is dead. I just bought a single channel 602D EXE off ebay and installed that. Now it works fine.

Today I made the first chips.

there is still a fair bit of work to finish wiring up all the buttons and functions of the user interface panel. The bellows covers on X are scrap, so they will need to be replaced before the enclosure goes back on. As it is it flings chips everywhere, so I don't want to use it too much before that enclosure is on there.
Mark