Re: [CAD_CAM_EDM_DRO] following error problems - BDI software problems

got_milk_eh wrote:

>Hi,
> After some checking I think my following error problems may lie
>with the BDI itself. The version I have is kernel 2.2.18 The
>stepper board uses 1284 signalling to communicate. 2.2.18 linux
>needs to be patched to support 1284 through the parport module.
>

No. The connection to the parallel port here is through the real time
module, which is NOT
running under Linux. it is essentially on its own machine, in real
mode, and has full access
of ALL hardware, top to bottom.

> I
>don't see this being loaded on startup. So the question I have for
>the BDI people, has this patch been applied to the kernel that goes
>with 2.2.18 on the BDI or is there a compiled module available that
>I can load? Thanks,
>
>

So, there is no connection here. In fact, it is best if the EMC machine
has NO parallel port
modules (parport, parport_pc, lp) loaded at all. These could interfere
if you accidently gave
a print command.

You might as well stop asking all these questions on this list. There
are only a couple of these
boards out in the field at this time (I'm trying to improve that
situation) and there are only 2
or 3 other people in the world who have looked at the code to any
extent. My universal stepper
controller and PPMC boards work in an entirely different manner than any
other EMC setup,
especially stepper EMC systems, and so answers from other people's
experience may be quite
confusing.

I think we are real close to getting it working, but I have to tell yout
that I am the ONLY
person in the world who has actually set up and run the universal
stepper board at this time.
(I hope you won't be mad at me for not laying all this out at the time
you ordered it.)

(Now, splicing in another message ---)

Now, I'm confused. You say you got EMC to jog the axis with the

> feedrate set very slow. Was the axis
> really moving, or just the numbers in the commanded position changing?
>

The axis was not really moving just the numbers when in the commanded mode. In actual mode, nothing moved, numbers or axis.

>> Let's start at the beginning. Do you have a green light on the board,
>> labeled Estop OK?
>
>

yes

>> change. If not, then the encoder counter is not counting the step
>> pulses going out. If the axis is actually
>> moving, then the encoder counter must be set for encoder (closed loop)
>> rather than open loop.
>> Switches 1-4 should be open (off) for open-loop, and all other switches
>> should be closed (on).
>
>

Okay, the 4 switches are off, the rest are on. I have reset the ini to the original one you emailed me with the patch for emc. I have updated it for my lead screws so 20tpi screws and 200 rev motor gives 4000. So its all in inches.

I put the resistor in the two sockets that are closest together and are also right near the led for ssr8. When I start emc and turn the machine from estop to on, the led does not light. I'm assuming this means that there is something wrong then with the computer to board connection? Leading to the following errors since no counter feedback? To double check, my bios is set to EPP/SPP mode with 378/IRQ7 as the settings for the parallel port. I've got a 1284 cable. Are there any linux settings I should check? Is there some way to test the signals coming out of the cable end, logic probe, etc? Thanks,





You don't want to look at these signals. They are really a
communications network, and you'd
need a logic analyzer to make any sense of it. But, there is a
diagnostic, which I will send you.
There are no Linux settings that I can think of. What BIOS settings are
available for your parallel
port? EPP/SPP sounds like it should be one or the other, not both at
the same time.
Try disconnecting the board, or just unplug the wall wart power supply.
EMC should not run,
and you can then try this command :

tail -n 20 /var/log/messages

There should be some stuff from EMC telling about the RT buffers that
communicate between
the Linux and RT sides of EMC, and then info from ppmc_encoder,
ppmc_rate and ppmc_dio,
each one telling you it found such a device at IEEE-1284 address xx, and
what version that
device is. With the board off, it will not be able to find these
devices. Since EMC comes
up, it is communicating well enough to find the 3 functions on the board.

So, first, you should run EMC with the board off, and see the messages,
then turn the board on and
try it again. Let me know what the messages say.

What type of computer is this? What model of motherboard is it?

The next thing I'd try, is forget ALL changes to settings, and run the
original .ini file exactly as
supplied. Correlation between motion and values on the screen will be
wrong, but this will eliminate
any foulups due to the metric units or the particular step/units
conversions you need. Once we get
it running in inches at the same settings I use here, it should be
simple to change one thing at a
time until we get it working with the desired units, etc.

The diagnostic program can scan the IEEE-1284 bus and report what
devices it sees, and it can
also read the encoder counters (a really good check if you have an
encoder you can hook up) and
it can also set the rate generator to produce any rate you want (real
number -10 .. +10).
It can also exercise the digital I/O.

Jon