Having the board in front of me would help or even a high res image. http://linuxcnc.org/docs/2.6/html/hal/parallel_port.html
"The HAL driver supports the bidirectional port, and allows the user to set the data group as either input or output. If configured as output, a port provides a total of 12 outputs and 5 inputs. If configured as input, it provides 4 outputs and 13 inputs. In some parallel ports, the control group pins are open collectors, which may also be driven low by an external gate. On a board with open collector control pins, the x mode allows a more flexible mode with 8 outputs, and 9 inputs. HAL and Open Collectors HAL cannot automatically determine if the x mode bidirectional pins are actually open collectors (OC). If they are not, they cannot be used as inputs, and attempting to drive them LOW from an external source can damage the hardware. To determine whether your port has open collector pins, load hal_parport in x mode. With no device attached, HAL should read the pin as TRUE. Next, insert a 470 ohm resistor from one of the control pins to GND. If the resulting voltage on the control pin is close to 0V, and HAL now reads the pin as FALSE, then you have an OC port. If the resulting voltage is far from 0V, or HAL does not read the pin as FALSE, then your port cannot be used in x mode." On 10/31/2018 09:13 PM, Chris Albertson wrote: > So it is looking like the external resisters some how become disconnected > when you swap out the operating system? Interesting? > > No, seriously, the hard part about debugging is that all the documentation > describe the INTENDED design and the way things SHOULD work. The first > step in debugging is to realized that "This is Impossible. It can't be > happening" > > The resisters reacting to which OS is running? Impossible, I think. > > Last time I had to debug a 3D printer controler board I looked at the volts > output by a microcontroller pin and said "that's impossible, I set it to > "1" but I see a zero. Worse I set four pins to go up/down/up/down all > together at about 1Hz and only 3 of them moved. "Impossible? Not if > there is a solder defect on the PCB that grounded a pin and cause it to > "blow". > > I think when you find something that is identifiablely impossible you are > 60% of the way fixing it. > > > > On Wed, Oct 31, 2018 at 4:09 PM bari <[email protected]> wrote: > >> I just looked at the manual for the Moschip 9865 and there isn't any >> mention of software select for the LPT signal pullups. It doesn't even >> mention any being internally available. >> >> Looking at the reference schematic they show that external 4.7K ohm >> resistors are to be used on all the LPT pins including 33 ohm series >> resistors. >> >> I didn't see anything in the Moschip Linux driver source at the asix.com >> website about the pullups. >> https://www.asix.com.tw/products.php?op=pItemdetail&PItemID=125;74;111 >> >> >> On 10/31/2018 05:12 PM, Chris Albertson wrote: >>> On Wed, Oct 31, 2018 at 1:46 PM Todd Zuercher <[email protected]> >> wrote: >>>> Gene, >>>> >>>> My problem isn't the breakout board per say. It is that this wonky PCIe >>>> parallel port card isn't pulling up the input pins in Linux (it does in >>>> Windows). >>> If it works in Windows it's obviously a software problem. You need to >>> look a the source code for the paraport driver in Linux. >>> >>> Try looking in the /proc/sys/dev/parport..... file system. If you >> drill >>> down you can see the list of modes the kernal THINKs the port suports. >> It >>> might be the Kernal thinks the one you need is not suported. The fix for >>> this would go the driver. But is your Kernal up to date, you don't want >> to >>> fix something that was allready fixed some months or years ago. >>> >>> Did the card need a special driver install in Windows? If so then it >>> mi=ust be non-standard. >>> >>> >> >> >> _______________________________________________ >> Emc-users mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/emc-users >> > _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
