Motherboard ?

Had a bit of a bang come from the psu a couple of weeks back and its caused no end of problems/damage, but I'm on the way back. I've eventually got my machine to run and home as it should but not after having to re-map the wiring. I have been chasing a fault that I thought was the parallel port on the MoBo by trying a couple of pci cards and dis-abling the on board port, but no matter what I did I could not get the machine to respond properly.

All the pci port cards I have tried, have tested OK on the Big John Hal tester, with the led's lighting as they should with the corresponding button press. When I hooked up to the machine I could never get the z axis, pin 7 to reverse direction, untill I re-mapped the pins, even even my logic probe showed that the pin was not changing state when I pressed to change direction in the GUI. So after all this I am assuming that something on the MoBo is not right but I dont know what as every test I have tried has shown the parallel ports to be working properly except the onboard port.

Its probably a mute point as a new Mobo will fix it but has anyone any idea what it might be?

Bill

Hard to tell based on what you say, but unlikely that the onboard port AND PCI port cards are both defective

If your sole fault is that Z will not reverse direction, I would suspect the driver or the wiring, most likely the later.

Swapping over drivers and wiring should quickly establish the area the fault is in, if the Z axis works fine attached to the X axis stuff for instance.

regards

I would tend to agree with you that it would be strange for all the parallel ports to be defective but before I bought my last pci port I did swop drivers around and the drivers worked as they should except the driver connected to the Z axis, so I went for the new pci parallel port which exhibits the same problem pin 7 does not reverse the Z axis direction. I re-wired the Z axis to pins 8 and 9 and it works as it should.

As I said iin my first post its a mute point as replacing the motherboard would definitley cure the problem, but what on the motherboard would cause pin 7 not to work correctly for both the onboard post and pci cards.

Bill

but what on the motherboard would cause pin 7 not to work correctly for both the onboard post and pci cards.

Nothing.

Far more likely to be a break in the parport cable or similar, try testing continuity from D25 plug which connects to the computer to the other end of the cable and thence onto the BOB if that is OK

regards

Yep done that, I dont know how any times, both continuity and using a logic probe to check the pins were changing state when they should as well as any parallel port test programs that I could get to run. I might have put a bit of confusion into this matter as the onboard parallel port does not work well at all. After the bang a couple of the outputs will not work, no leds will light when conected or any sign of voltage where there should be. Thats why I was wondering whether there was anything that would be common to the motherboard and pci slots.

Bill

Thats why I was wondering whether there was anything that would be common to the motherboard and pci slots.

At some point there must be a convergence between ports on the PCI bus and the built in IO ports, but my knowledge of the architecture gets very sketchy at this point.

They will be addressed and implemented by the same low level driver (parport) with each port having a mapped memory area.
Because each pin ends up being a memory address, it seems difficult to imagine a common fault between the two, when they are using different memory addresses,
but what do I know

regards

Onboard parallel ports are almost exclusively LPC* bus devices on motherboards of the last 10 years or so.
There is basically no commonality in the data path between LPC devices and PCI devices that would not have
global (machine crashing) consequences.

An external problem seems much more likely

* LPC bus = Low Pin Count bus, a simplified serial version of the the old ISA bus

I agree with you also PCW as I did with ArcEye but everything that is external to the computer is the same as it was before the big bang, except for new chips in the drivers and a couple of resistors. I even went to the lengths of re-installing Linuxcnc to rule out something corrupt in the OS but still nothing on that pin 7. I think I might just put it down to one of lifes mysterys and get on with setting this machine up, at least I can move 3 axis just need to get them to move where I want them to, god I hate calibrating things.

Bill

PCW wrote:

Onboard parallel ports are almost exclusively LPC* bus devices on motherboards of the last 10 years or so.

Is the LPC bus inferior compared to an add-in card for Linux CNC?

LPC is about the same speed as the old ISA bus so LPC parallel ports are likely to be
a bit slower than PCI/PCIE add-in parallel port cards. This probably doesn't make much
practical difference as these times are small (full range is probably about 0.5 to 3 uSec for the fastest PCI to slowest LPC port) and only a few I/O cycles are used