7i96s custom bitfile

Hi all,  long story so please forgive me!

Being as Tormach have updated their software to allow the use of 7i92T boards and they use the same fpga as the 7i96s I thought there must be a way for 7i96s to mimick a 7i92T and allow me to use the 7i96s with pathpilot and not have the firmware (gtag) issue that prevents it from loading.

I assumed it was only the ssr part that is that causes pathpilot to fail to start, so i modified the xml and some vhd files and saved them in the 7i92T directory, synthesised it and to mu surprise spat out a hex file...Yippee I thought as I flashed it via pathpilot, now all it does is the led run at start up and blink both the rx and tx leds on the ethernet port..no communication!.

Whilst I wait for usb to spi to program via jtag (fallback does not seem to work now either) I thought I'd ask the experts if it would be possible for someone who is better at this than stupid old me to generate a bitfile for the 7i96s that is for all intents and purposes a 7i92T with the appropriate pin differences to match the connectors and pcb of the 7i96s ?.

Hope that makes sense...and thanks for any help...

Ian
 

Hey, I just did the same thing, (someone on the forum asked for a 7I92th 7II76_7I85S
configuration so I made one and proceeded to absent-mindedly write it to a 7I96s )

If you have a relatively recent 7I96S firmware, you should be able to recover
by setting both IP address select jumpers "UP" and cycling to power.

Getting back to your original problem, you don't want to write 7I92t firmware to a 7I96s
(as we both found out) what you really want to do is change the card name from 7I96S to 7I92T...



 

Lol...it's not just me then...?

Thank you for your very promt reply. I did the jumper thing but no joy so i think I'll have to use jtag when the parts arrive.

The other issue, for me at least, is the ssr firmware that Pathpilot does not like, hence why i started this adventure, i would like my 7i96s to have pp compatible firmware. I thought i could mix and match the files to acheive this....as far as i can tell there is only one pin that is different on the ethernet data side of things, i thought once i get comm's I can then figure out the i/o but i killed it in the process...

my reasoning was that the 7i92T is now supported by Tormach so that was a place for me to start. My first attemp was simply to change the 7i96s vhd pin file
from
IOPortTag & x"00" & XfrmrOutTag & XfrmrOut0Pin, -- I/O
to
IOPortTag & x"00" & NullTag & x"00", -- I/O

but that still left the incompatible part and Pathpilot complains about it..

so i started modifying the 7i92t....now i'm stuffed......

one more thing if you will?...what do i do with and how do i use the setname files?...

thank you...

setname allows you to change the board name
(its a linux 86x64 executable) but of course
requires communication.

It turns out that the 7I92T source is broken (the interface designer xml file
was corrupted and generated the wrong clocks)
I have fixed this in the distribution.

So a working 7I92T _will_ run in a 7I96s (run meaning Ethernet works) but
you would not want to because all the user I/O would be on random pins,
likely creating conflicts.

If the IP address jumper UP/UP setting does not give you a blinking
/INIT LED. That means you have a fairly old 7I96S which should have its
fallback configuration updated.

Ahh, that makes sense,. i thought it should have worked but I wasn't convinced about the ethernet pins...

efxpt:gpio name="ENAIO" gpio_def="GPIOL_50"   for the 7i96s
vs
efxpt:gpio name="ENAIO" gpio_def="GPIOL_02"  for the 7i92t

currently my 7i96s looks like it starts ok, I see the run of leds sequence but i cannot communicate with it 192.168.1.121
and the rx and tx leds on the ethernet socket blink once every two seconds or so....

Once I get my hands on a programmer I can start experimenting..!

I don't think the random i/o is a problem, to get the device to run is just a starting point.

I really appreciate the time and effort you commit to your support, I realise how much of your own time this must service consume.

thank you!

Ian

Sooo......

after some configuring and csv manipulating I have managed to get most of my 7i96s working with pathpilot, all the stepgens now function, as do the inputs and two of the outputs.

The other four outputs require an alternating signal due to the isolating transformers.

Could I ask PCW if there is a schematic available, if this would help me?.. or is there another way I can simply create a suitable sgnal without the Xfrmout module (that Pathpilot doesn't like)

I could remove the xformers and jerry-rig opto couplers or other devices but I'd rather not if there was a less invasive option...

For now the two remaining operational outputs will suffice, and the expansion port of course..

Many thanks to PCW for his help and support so far, and if there is anything else I can do in return, I'll try my hardest.

cheers all...

Honestly, the best way would be to rebuild pathpilots
hostmot2 driver to support newer modules (like xfrmrs)

There's no way I can think of that would generate the synchronous ~1MHz
signals needed for the transformer drive.

Thank you PCW,

I was afraid of that!... and I presume the sequencing of the outputs correspond to XfrmrRef signal too?...

I might just splice optos into the circuit to gate the mosfets,....ugly but should work as the signals are getting to the buffer ic's as expected.

I must also point out,...the 7i96s is a great piece of kit.....I'm just trying to make it do things it was not designed for!....

wish me luck!.........

Ian

If you are not running Tormach specific hardware, you could probably
just replace the hostmot2 and hm2_eth binaries with current ones from
LinuxCNC and get support for all the new modules (hostmot2) and
cards (hm2_eth).

An off topic question.Why so many people like pathpilot ? Probebasic is very similar and can be customized ?