LinuxCNC and DB25-1205 board

I have been building a small 3 axis router for a little while.I now have the build almost complete and need to get the motors working.I bought a NEMA 23 kit from Longs Motor and have been impressed by the quality of the components.The breakout board is a DB25-1205 and I have found very little about people successfully using one of these boards with LinuxCNC.I have laid out the hardware on a bench and made all the connections and there is the correct voltage at every terminal I have tested.I have downloaded the .iso and been running it as a live disc-nothing too hard there as I have been using Linux for several years.The setup for stepconf seems to be OK but when I test the axes-nothing moves.My question is how do I check whether the parallel port is working as I haven't used one for at least 9 years and the old computer that has one doesn't necessarily still have a working port.

Alternatively,is there a function within LinuxCNC that diagnoses whether the port is functioning?I would really like to get this equipment working as its already here,but I do wonder if I should study the possibilities of an Arduino setup. I feel that the whole hobby CNC ecosystem needs to prepare for the demise of parallel port controllers as they are pretty much dependent on an obsolete technology.If no other sensible alternatives exist I would reluctantly go the Mach3/4 route but I am less interested in revisiting old technology than I am in making parts for my projects.

Quote:

---

Originally Posted by routalot

I have been building a small 3 axis router for a little while.I now have the build almost complete and need to get the motors working.I bought a NEMA 23 kit from Longs Motor and have been impressed by the quality of the components.The breakout board is a DB25-1205 and I have found very little about people successfully using one of these boards with LinuxCNC.I have laid out the hardware on a bench and made all the connections and there is the correct voltage at every terminal I have tested.I have downloaded the .iso and been running it as a live disc-nothing too hard there as I have been using Linux for several years.The setup for stepconf seems to be OK but when I test the axes-nothing moves.My question is how do I check whether the parallel port is working as I haven't used one for at least 9 years and the old computer that has one doesn't necessarily still have a working port.

Alternatively,is there a function within LinuxCNC that diagnoses whether the port is functioning?I would really like to get this equipment working as its already here,but I do wonder if I should study the possibilities of an Arduino setup. I feel that the whole hobby CNC ecosystem needs to prepare for the demise of parallel port controllers as they are pretty much dependent on an obsolete technology.If no other sensible alternatives exist I would reluctantly go the Mach3/4 route but I am less interested in revisiting old technology than I am in making parts for my projects.

---

Have you tried the parallel port tester:

LinuxCNC Documentation Wiki: Parallel Port Tester

This allows real time reading of the inputs and toggling the outputs to check your breakout

You may have to edit ptest.hal to change the port address, its default is 0x378
( "0" is a good choice if you have a plug-in parallel port card with unknown address, "0" means the first parallel port found )

also here's a video that shows the DB25-1205 breakouts 5V power setup:

https://www.youtube.com/watch?v=bWVKJUnSLUc

Thanks for that link,I ran the test and the output pins all appear to be good.I will pursue the wiring as shown in the linked video.

Well I spent a bit of time trying some of the preset configurations and getting nowhere.Looks like I will have to select "other" and go through the connections one at a time until things happen.At least I have the consolation that no smoke has escaped so far.

If anybody is interested,I checked out all the pins on the back of the computer and they all passed current when the test utility was used.Incidentally if you do this test remember that Linux is case sensitive in a terminal and don't go looking for a downloads folder because there isn't one.If you cd to Downloads you will find the port test utility you downloaded by following the earlier link.Once I had remembered this step i used a test bulb to verify that the outputs on the breakout board were responding to the test on the same No pin of the port,which reassured me that both the cable and the board were working.Now to return to the control module and motor wiring.

I finally got the steppers to move.It was a combination of inverting the pin functions on the BOB and discovering that either the person compiling the circuit diagram of the one connecting cables to the motor may have been colour blind.The next challenge is to get the machine moving as it is supposed to do.In the stepconf part of the process,there is a test facility and by altering the steps/revolution and micro-stepping values,the movement was correct when tested.

The interesting part comes when running a test program and finding that the movement on all three axes was half that required.Should I be looking at DIP switch settings or experimenting with the stepconf numbers?

I have not used Linux CNC but if it can control the parallel printer port like Mach3
you should be able to use the DB25-1205 breakout board :-

Attachment 400290
Note -
I have only shown 1 of 12 output circuits
and 1 of 5 inputs circuits

provided you connect the +5V to the stepper driver + step & +direction terminals
and the boards opto isolator NPN photo transistors to the - step & - direction terminals

the DB25-1205 outputs can sink more current to ground than the 1K pull up resistors can supply from the +5V supply

at high step rates connecting the breakout boards opto isolator output to the DQ542MA stepper drivers opto isolator inputs will limit the maximum step rate

"817" opto isolator are not the fastest and tend to round of the square edge of the step pulses

John

Thanks for the contribuiton.I am starting to feel that I am making progress as the axes now move.The next challenge is to make them move the distance they should according to the g-code.I have to say the best resourc I have found so far is Brenda E M's video on youtube,but since she and I have different drivers,I am very keen to locate a solution and the huge resource that this site represents has to be the logical place to begin.

I now have the axes moving the correct distance.I believed the default +/- 15 mm travel meant that the axis moved 15mm and then back to it's start point.It seems to mean that it will move 15mm one way and then 15mm the other side of the start point giving 30mm of total travel.Now I have calibrated to this,the basic movements are correct.What concerns me now is that having generated some test programs,the trace on the monitor shows engraving sequences with a radius in some of the corners that should be sharp.I suspect this is a speed/acceleration related issue.

Quote:

---

that having generated some test programs,the trace on the monitor shows engraving sequences with a radius in some of the corners that should be sharp

---

The g-code is describing sharp turns ? No "roll around sharp corners" or cutter compensation ?

By default linuxcnc goes as fast as it can by rounding corners. Look at g64. It allows you to tell linuxcnc how close to follow the programmed path. Sof
G64P.005 says go as fast as it can while not deviating more than .005

Sam

To both the last two posters.The test program was an engraving sequence with sharp corners and thus no cutter compensation.I have now run the machine under power and with fairly slow acceleration it followed the text very well.Then I ran an engraving test with concentric circles and the outcome wasn't concentric.I will try again with a G64 manually inserted in the program and see what happens.For now ,I am just elated to have come so far in a few days after weeks of struggling with the DB25 1205 and the refining of the settings and adding home switches will be the next challenge.This forum is a great resource and I suspect I will be tapping into your collective expertise again as I progress.