Advanced Search

A friend an I now spend several hours on finding an old PicKit3 programmer and getting it to run under Win11 using programmer version 8.92.
Unfortunately the 7i73 still does not work.
After flashing the hex file from 7i73r16.zip, the leds are dark and nothing happens. Luckily my friend read out the original hex file before we wrote the new one. We could recover it by flashing the original read hex back on it.
Are you sure the file in the Zip contains all necessary parts? The original file included data from Address 0x0000 and the hex in the zip starts at 0x1000 when checked in the programmer.
This starts being some pita - please help!
best regards,
Alex

Looks like no 5V to the 7I76U

(7I92T W4 needs to be "up")

Hi Cmorley

I like the MacroTab Widget, you can easily make your own simple macros, but I think it needs some fixes.
I can save the current parameters to a file, but there is a problem with restoring the settings from the file.
The "set_style" function also does not work. I tried to fix it, but so far without success.
Could you look at this problem?

I do have ethernet, and I do have a reason for needing wifi in addition. The camera I'm adding to the spindle sends video via wifi.

Below is dmesg output with the wifi card installed. Without it, I think everything is the same except the second card is addressed 0xdcc0 instead of 0xacc0. I can try that later if needed.

Tommylight, I do have the means to change the MB parallel port address. Sounds like something to experiment with, thanks. I have swapped slots for the wifi card and the second parallel port card, same result.

I did verify that the wifi card is known to work with Linux. I don't see why LinuxCNC would care (or know anything) about the wifi card. Any issue here?

$ dmesg | grep parport
[ 7.560447] parport_pc 00:03: reported by Plug and Play ACPI
[ 7.560518] parport0: PC-style at 0x378 (0x778), irq 7 [PCSPP,TRISTATE,EPP]
[ 7.665553] lp0: using parport0 (interrupt-driven).
[ 8.274698] parport1: PC-style at 0xbcc0 (0xbcc8), irq 30 [PCSPP,TRISTATE,EPP]
[ 8.381897] lp1: using parport1 (interrupt-driven).
[ 8.390482] parport2: PC-style at 0xacc0 (0xacc8), irq 16 [PCSPP,TRISTATE,EPP]
[ 8.493589] lp2: using parport2 (interrupt-driven).

Hi All
I added one Smart Serial (instead of previously none) , which did indeed restore the missing 7i76u TB5 and TB6 tabs.
I tried configuring it with pncconf (hal, ini attached) and with mesact2 (hal, ini and .io attached)
I tried two different sets of physical inputs on TB6.
first using (on TB6, where pin 1 is at the bottom edge of the 7i76u board)
input4  (pin5)  x limit and home
input5 (pin6)   y limit and home
input6 (pin7)   y1 limit and home
input7 (pin8)  z limit and home
which linuxcnc did not like,  and it exited with an error (see attached reports)
I then moved the four physical inputs
and configured on TB6 using
input0 pin 1 x limit and home
input1 pin2  y limit and home
input2 pin3  y1 limit and home
input3 pin4  z x limit and home
which linuxcnc did not like,  and it exited with an error.
'hm2_7i92.0.7i76.0.0.input-00' does not exist

two of the inputs (x and z) are normally closed switches providing 24 volts , and two (y and y1) are proximity sensors providing 24 volts .

Using a multimeter all four inputs measure 24v field io (when measured between 7i76 Field IO Power FG, and the relevant TB pin)

I added a photo so that you can see where the input wiring is going.

On power up the 7i92t has PWR led on yellow, CR1 led on green  , the 7i76u has CR3 led on yellow, and the watchdog CR6 led on red (it does not get reset, because the config fails to load)
If Linuxcnc doesn't like the config and aborts , then there is nothing to see using halshow.

its got to be something obvious , just it's not obvious to me
 

Which pulse/rev is best for NVEM RT1052?  


 

This line in halAdd base period after the EMCMOT.

tommylight wrote:

Yes that is the correct place to put it.
While testing this, no stress test, no glxgears, no video, no youtube.

next insert lane give error:

Never said it wouldn't

Only mentioned on board serial & parallel ports on the ISA bus. Which is still there only not physically.

"Asian knockoff", how the fuck would that be read as compatible ?
And where would one find one ?

And once again your stupid senseless post show up and add nothing.

Sometimes you can get less expensive shipping via a forwarder, might take a little longer to arrive, but may work out a little cheaper. But yeah it's been somewhat expensive for shipping Down Under...and no jokes about Qld being 20 years and 1 hour behind....it's non daylight savings time now hahaha

Oh those prices are going to be about 1.6 higher in Aussie beer tokens. So yeah it does end up costing heaps.

If you want to move away from Python, try "Ruby on Rails".
Going from Python to Java is absolutely moving backwards.

You can jam in as many parallel port cards as you have slots. They can be at any free address you have. If the on board one is buggered leaving it enabled will only affect the ISA compatible addresses, IE those hanging off the LPC bus. For the addresses allocated on the PCIe bus different realm, different interface, different bus. Hence the reason on board lpt ports don't show up via lspci, cos they are technically on the ISA bus but physically the LPC bus.