Spindle control problem

If the "TTL" represents the parallel port output, this circuit will probably not turn all the way off as it should if the parallel port output just swings to 3.3V. A quick fix would be to add another 270 Ohm resistor across the OPTOs LED or add a red LED in series with the 5V

I have had the box out and measured actual voltages on the input to the Triac isolator/driver with both computers now.

The +5v which comes from the power pack that feeds the Xylotex is according to my 'digital' meter actually only 4.4 volts. I may need a larger smoothing capacitor as its fed from a 5v regulator which has +12 going in and plenty of amps so should be 5 but this may be my cheap 'highly accurate' digital meter.

Anyway from the old computer (which switches the motor OK) the figures are 0.41 volts and 3.7 volts As the resistor limiting the LED current is actually 267 ohms I make that a current of 2.5 ma

The new computer outputs are 0.31volts and 3.05 volts which I make about 5ma which happens to be the switch on current.

I increase the resistor to 470 ohms the current should be about 2.5ma which should solve it

I like the idea of putting a LED in series . Many thanks to all who helped.

look at the opto isolator spec and see what the max current is , i would guess a resistor in series with the opto of around 330R to 470R with 5v supply should be fine
so your along the right lines as the opto is not fully switching off , and of course only 4.4v as you say is not helping .

The problem is that an output of 3.05 V will not fully turn off the OPTO (4.4-3.05 = 1.35V, still enough to turn on the OPTOs ~ 1V drop LED) . A larger series resistor is not the right solution as the OPTOs LED will still be slightly on regardless of the resistor value. Dropping your 5V with a series red LED or even a silicon diode will fix this. OPTO couplers are notoriously inaccurate current sensors, often having 3 to 1 or even 5 to 1 ranges of turn on currents, much better to make sure the OPTOs LED is off (<1v) so you truly have 0 input current when its supposed to be off...

Red LED fitted and first EMC metal cut. Really smooth and drives faster than Turbocnc under DOS. Really pleased. Now on Linux machine I need to find a gcode generator.
Many thanks to all who helped

John

John I have downloaded the zip file for ngcui and put it in a dir under emc2.
I found instructions to merge it with EMC2 but want to try it alone first.
The files I have are
ngcgui
ngcgui_app
ngcgui_ttt
tkapp

A simple question how do I run it on its own?