Limit switches not working

Hi all.
I'm starting a new post, but actually I should keep going on with the old one (here the link forum.linuxcnc.org/16-stepconf-wizard/34...nput-troubles#106943), but as I decided to change the BOB, maybe is more right to start a new one.
Well, briefly:
about 2 week ago I finished the build of my CNC, so started to assemble the electronic.
As in the above post, it seemed my problems where fixed, so I connected everything to my ZP5A-INT BOB, but nothing was working.....
As I was having even some trouble with the old pc I was using to drive all (it keeps resetting), decided to get a new one, so I got an Intel quad core J1900, 8Gb RAM and MB AsRock Q1900M.
Couple of day ago, I got the new PC, installed only LinuxCNC (the new PC has only to drive the CNC), then started the problems....

- After the latency test, I got a value of 24000. Should be a good one, but time to time I'm getting the message about RTAPI ERROR unexpected realtime delay etc etc. I disabled even the onboard audio (i don't need it), but no changes. I cannot avoid to ask myself "what kind of not-of-this-earth pc is needed to drive LinuxCNC??????"

Aside the latency problem, the limit switches don't work again.

The BOB was correctly powered with both sources (USB and external), but nothing to do.
As it seems that the most of the people has good result with the "standard chinese BOB" (to not get confused, the one attached), and as I have one of this kind (it was working - see the post above), I decided to disassemble all the electronics and start from the beginning.
The chinese BOB is powered by USB and by 12V (I'm using the ATX PS of the old PC - ok, 11.8 V but I don't think is a big issue, isn't?)
The limit switches (2 each axis, wired in series, NC), are wired to pin 12-13-15 (X-Y-Z) with common GND.
After run stepconfwizard then LinuxCNC, they don't work....the HAL monitor reports no response when switching them.
The signal goes to the BOB normally, but the PC doesn't read it.
I tried by modifying the BIOS setup for the parport (bi-directional, ECP and EPP 1.9 mode, ECP and EPP 1.7 mode availables): with bi-directional it doesn't work at all, with the others LinuxCNC "see" the motors but the switches....

I found a guy in the forum (sorry, I don't find the post right now) having my same MB, BOB and problem, and he fixed the problem by adding a pull-up resistor (2.2 Kohm) between every optocoupler and the switches' common GND.

Before modify the BOB with the resistors, I would like to have your opinion (and help)

I attach even my HAL file (only the Z limit in it)

I know I've a lot of "road to walk" before getting some nice object from my CNC, but at least I would like the have the machine working.....

Thank you all

Andrea

If parport.0.pin-15-in does not change when you activate you Z limit switch, that suggests a electrical problem with the BOB. Some people have reported that the optocoupler input common (12/24V) is unconnected on at least some of these BOBs

I would measure the Z limit switch input at the BOB and see if the voltage changes there when you activate the switch
Its should be ~10V when the switch is open and ~0V when the switch is closed (with a 12V supply)

Thanks for your answer.
I left tonight for 1 week because of job,but when i'll back i'll try your suggestion.
Anyway,the problem is not only with z limit (this is the one reported in the hal file,but just because i was making several test): all the limit switches give me the same behaviour.
Moreover,with the old zp5a-int (before i switched to the chinese) i got the same.
But before keep annoying you,i'll do the test you suggest me,as soon i'll back home next we.
Only a question (please forgive me,i'm not an engineer): i suppose to check the voltage i've to put the multimeter terminals on pin 15 and on gnd pin and misure there the voltage,right?
Thank you

I tried to check the voltage. Here there are the values (when switched - the switches are set as NC > 0 V):

Z: 10.3 V
Y: 11.3 - 11.9 V (not stable)
X: 10.3 V

Jocman wrote:

I tried to check the voltage. Here there are the values (when switched - the switches are set as NC > 0 V):

Z: 10.3 V
Y: 11.3 - 11.9 V (not stable)
X: 10.3 V

when you operate the switch (say the Z one) does the voltage on the BOB pin 15 change from 10.3 to 0 volts. if so that is correct.

You might have to invert the pin in the parport.0.pin-15-in to parport.0.pin-15-in-not
depending on which version of Lcnc you are using.

well, actually when I operate them, they all switch from 0 to the value
By the way, I'll try inverting the pin as you suggest me

I'm using LinuxCNC 2.7.11 wheezy

Thanks

Hi.
Finally I fixed it (hopefully).
I found this post in the forum

forum.linuxcnc.org/18-computer/28929-no-...ard-parport?start=10

As the guy owns the same MB and the same BOB, i decided to try this solution, so got Yesterday from China 2 more BOBs (you never know, maybe they get fried.....), soldered the 5 resistors and started the machine.
Everything works fine: finally the CNC do the self homing procedure!!!!

Thanks all for the support and the advices

Andrea