Hi,
Rimmel and I have been sharing a few PM's so that he can understand the requirements of wiring servos to Mach/ESS/BoB..

There is nothing secret about it so I thought I would post so others can see (benefit?) and/or comment.

Note that I use Delta servos and the diagram I have posted is how I have programmed and hard wired the Delta drives and the BoB is my own design.
Having said that Rimmel's Panasonic drives and MB3 breakout board should be sufficiently similar that the same scheme can be applied. Note also that there are dozens
of different ways and/or variations on how this can be done, with and without good reasons. This is just how I have done it, and with any sort of luck will be able to communicate
why I did it this way.

The first question Rimmel asked is where the 24V comes from? Firstly his MB3, and my own BoB have a 24V supply. However my Delta servos have an internal 24V supply
and I use it for the a bias source for the photodiodes. So there are two sources of 24V power. Note that the supplies share a COM. I imagine the Panasonic drives will have a similar
built-in supply.

A brief description is in order.
Not counting the two differential Step/Dir pairs there are two outputs from the BoB to two inputs of the drive. They are an Enable and a Reset. Note that I have used the internal 24V
supply, via the built-in dropper resistors, to the photodiode. If the BoB goes low then current will flow through the photodiode and thus signal the drive. Note also that the transistor in the
BoB can only sink current when the ESS drives the base high. That is to say the ESS and therefore Mach must be active before the transistor can conduct, and is fail-safe. If the ESS or Mach
give up the ghost for any reason then the Enable signal will disappear and all servos will stop.

The Reset signal is to reset the drive after a fault condition, with any sort of luck hardly ever. My Delta servos, if they fault, require either a power cycle OR a Reset in order to clear a fault condition.
I have a button (on screen) that I activate which in turn pulses the Reset output of the BoB, with its transistor switching on and sinking the photodiode current thereby signalling the drive and
resetting it.

Note also that I have just one Enable output and one Reset output and they are sent to all servos. Rimmel has commented that his MBS3 BoB outputs can sink 70mA. Each servo sources about 5 maybe 10mA
so having three servos commoned together would still only require the MBS output sink 15-30mA, well within spec.

The alarm is an output of the drive and therefor an input to the BoB. Ihave used the 24V supply of the BoB via a dropper resistor to source about 5mA. The phototransistor in the drive conducts in a fault
condition and therefore drags the BoB input low, to nearly 0V, which in turn signals (active low) the ESS.

I have one alarm input per servo so Mach and therefore I know which servo faulted. If you are trying to economise on inputs you could have all servo alarms hooked to just the one input, and that one input
would trigger an Estop. Your choice.

There is a risk to the opto-isolators created by paralleling inputs powered by different 24V supplies. If one supply is off its opto-isolators will receive a reverse bias from the other supply. Their cathodes will see about 22V or 23V from the shared signal line ( 24V minus the emitter drop of the other optos) while the anodes will be at zero volts ( the unpowered 24V rail).
The maximum reverse rating of the typical opto-isolator is about 5 volts. In the real world the maximum is usually far higher than the spec but applying 4 times the rating is a risk.

Hi MN300,
there is no risk here. Any one optoisolator is biased by one supply only.

Looking at the diagram the two input optos of the drive (Enable and Reset) are biased by the 24V supply within the drive.
The transistors in the BoB sink current only, they cannot source, ergo the optos are influenced by one supply only.

Looking at the opto output in the drive (Alarm) the photodiode is biased by the 24V supply of the BoB, so again there is no possibility
of that opto being subject to two supplies.

I drew the diagram in symbolic form only, there are biasing and current limiting components I have omitted. In particular each and
every photodiode has an anti-parallel diode such that NO PHOTOIODE EVER experiences more than a few volts reverse bias.

In the real world the maximum is usually far higher than the spec but applying 4 times the rating is a risk.

I agree with your assessment, but it does not in fact happen because of the protection circuits built into the drive. When composing the diagram
I wanted to keep it as simple as possible to convey maximum understanding. Including anti-parallel diodes etc. would detract from that goal.

Craig

Hi,
I've taken these direct from the Delta manual for my servo drive. The first pic is of a DI (digital input), but is a small
diagram and so have included a second pic of the singled ended pulse input circuit, should you want to use it, because the diagram
is much clearer with respect to the biasing/current limiting and in particular the anti-parallel diode to protect against high reverse
bias of the photodiode.

Before anyone starts complaining that the current in the photodiode of the pulse input is 25mA or thereabouts, its because you need
that sort of current to signal at 200kHz, the rated max signaling speed of this circuit. Closer inspection of the DI circuit will reveal that the
bias current limit resistor is 4.7kOhm, and so the photodiode forward current is limited to 5mA or thereabouts. That's perfectly acceptable
in a slower circuit like this.

In any event I did not dream up my wiring scheme un-aided, I used  and relied on the manufacturers manual.

Craig

Hi Craig,
Yes, anti-parallel diodes will protect the opto-isolators. It's rare for manufacturers to publish a schematic with that much detail so I assume the worst.

Any one optoisolator is biased by one supply only.

Your drawing indicates multiple servos with individual power supplies sharing the reset and enable lines. This does create the situation where a dead supply will attempt to reverse bias the optos. The active supply will pull up the shared lines when the open collector output is off.

John

Hi,

It's rare for manufacturers to publish a schematic with that much detail so I assume the worst.

BS, all the good manufacturers publish just exactly this information, if they don't they are junk and I'll have nothing to do
with them. Even the cheap Chinese servo drives publish these diagrams albeit somewhat less detailed and less clear, but even
the worst of them publish this at least.

This does create the situation where a dead supply will attempt to reverse bias the optos.

If one supply in one of the drives failed then the remaining supplies would back-feed the drive with the dead supply, and were it not
for the anti-parallel diode it would reverse bias the opto. This is the classic 'wired OR' situation and is widely used. If the built-in supply
dies then the whole drive is f....d and will have to be replaced, so who cares about the opto?

The brands that we are talking about here, Delta and Panasonic have lifespans of decades, not years or months. The probability of failure is low,
and if a drive were to fail why would a 5W flyback inverter fail, is it not more likely the three phase DC to AC inverter of many hundreds of Watts
fail?

In any event I did not dream up my wiring scheme un-aided, I used  and relied on the manufacturers manual.

This is the scheme suggested by the manufacturer, not me, and to date its worked perfectly.

Craig