Centroid Community CNC Support Forum

Presumably it's driving an optoisolator which requires a specified current to turn the LED on sufficiently. The drop across an LED is typically around 2V, so there will be a resistor within the DMM drive that is specified for use with 5V by default.

If you run the inputs with greater than 5V, you need to limit the current to the same level. If they have used 1200 ohm to drop 19V (=24 - 5), that suggests around 15mA required. I would expect to see around 200 ohm in series with the opto input inside the DMM drive to give the same current from 5V if you took the lid off (200 = (5-2) / 0.015).

These "high speed" optos are often not quite as fast as they claim to be, so I wouldn't be surprised if they are only just about up to the task. This may be why there has been so much messing about with resistor values, DB5 connection etc. I suspect that running them at a lower pulse frequency would almost certainly help.

DMM will refund the cost of the cables. Modified DMM cables have worked well for me. I wish they would just sell them in the known working configuration.

slodat wrote: ↑Wed Jun 24, 2020 9:25 am DMM will refund the cost of the cables. Modified DMM cables have worked well for me. I wish they would just sell them in the known working configuration.

Follow the attached Centroid schematic (included in the Centroid Schematic set) and you should have no problems.

Marty

Marty, I have fully documented on two machines how there were lost steps using the method shown in that schematic and the lost steps are resolved when switching to 5v on the DB25. What’s the benefit of continuing to insist otherwise? I don’t understand. There are a lot of us that have had this problem and it’s resolved by going to the DB25 step/direction signals. The DYN4 is 5v logic. Help me understand?

Edit: looks like the schematic is using +5v and the screw terminal step/direction. Hard to see the whole drawing on my phone. When I tried this, I still lost steps. I’m happy to see the approach has evolved. Is this working for people that previously had lost steps?

slodat wrote: ↑Wed Jun 24, 2020 9:37 am Marty, I have fully documented on two machines how there were lost steps using the method shown in that schematic and the lost steps are resolved when switching to 5v on the DB25. What’s the benefit of continuing to insist otherwise? I don’t understand. There are a lot of us that have had this problem and it’s resolved by going to the DB25 step/direction signals. The DYN4 is 5v logic. Help me understand?

Edit: looks like the schematic is using +5v and the screw terminal step/direction. Hard to see the whole drawing on my phone. When I tried this, I still lost steps. I’m happy to see the approach has evolved. Is this working for people that previously had lost steps?

Nearly all my DYN4 installs followed that schematic. Yep 5vdc logic.

I also used the DMM supplied cables, I had some that worked and some I had trouble with. Clearly they kept revising it. They "say" they had resistor tolerance issues. Resistor drops the voltage from 24vdc to 5vdc.

So, I just follow the schematic and make my own shielded cables when I do a DYN4 install.

Ahh... that’s the difference from what I have done. Thank you, Marty!

martyscncgarage wrote: ↑Wed Jun 24, 2020 9:55 am I also used the DMM supplied cables, I had some that worked and some I had trouble with. Clearly they kept revising it. They "say" they had resistor tolerance issues. Resistor drops the voltage from 24vdc to 5vdc.

Indeed - sounds as if they have been playing about with the resistor values hoping to find a consistent solution. Describing the issue as "resistor tolerance" suggests that the resistors they used were some way away from their nominal value. That just doesn't happen these days - try to find a resistor with a worse tolerance than 1-2% and you will struggle.

A slightly straighter answer would be that the circuit operation is marginal (possibly "submarginal"?), either they aren't driving enough current through the opto or (more likely) the opto's abilty to switch fast enough is marginal. If it were a simple matter of requiring more current, they would just drop the resistor value further. I'd call this a "component choice" issue, rather than "resistor tolerancing". I wonder how this would be affected by temperature and ageing. Perhaps somebody might like to heat and cool the optos to see what happens - in the interests of science....

I installed the small VFD marty suggestioned a month or so ago gotnir hooked up and running. It was supper easy to hook up and took about 5 mins to have it running..

I have a question on the custom lube pump settings in the wizard. First off my machine is set up with metered check valves that give a metered shot of oil every time the pump come on. The original operation had it set up for the pump to turn on every time the machine made a rapid movement then turn back off. Since there's not a setting like that in the wizard I figured it just let it run for 3 mins or so and then pump lube for 3 seconds. My metering valves only work with the first surge in pressure of the pump then they seal back off. So running the lube pump any longer than necessary is just dumping lube on the ball screws since they aren't metered. So when to set up a custom lube cycle it appears that any accumulated time under 10 mins defaults to zero time which in turn leaves the pump on constantly. I can go in and set the time for a few mins and it doesn't work. If its set to over 10 mins it will work. I also went in and changed parameter 179 to a value of 503 (5 mins accumulated time with 3 seconds of pump time) and tried that. It didn't work. It appears that you cannot change parameter 179 to 0503. It will not accept a time of less than 10 mins and you can't use a leading zero in the parameter. Any suggestions?

Does anybody have any input on my issue

You can always edit the lube pump stage in the PLC program to have it do whatever you want. I use my mill in manual mode quite a bit, so the normal mode of operation where the timer only accumulated when in program run did not work for me. I changed mine to accumulate time whenever not in estop.