IHSS57-36 with 7i92 don't move

More
13 Sep 2021 00:20 #220382 by markusk
i'm a bit at a loss here. granted, i'm a noob, so what else to expect ¯\_(ツ)_/¯

short intro: i'm trying to retrofit an early 90ies Newing Hall engraving machine (x/y table with pneumatic solenoid tool) with some modern controller, tinkered with a smoothieboard/smoothieware yet but figured i could as well get a bit more knowledge and gain more flexibility by using linuxcnc.

researching over the last year i decided to 'go FPGA' right from the start, and i'm rather confident that i understand the concept enough to do it - but still i lack lot's of basics. also the motors seem to be (judging from search results on this forum) rather exotic which makes it hard for a beginner.

before getting into details of wiring/hal/configs may i just describe the symptoms i get - maybe that tells enough:
everything's powered, connected, starting linuxcnc/axis gui i try to jog a motor. but it doesn't move - it just "spits" what feels like a single step (or not much more) at the end of the commanded movement. and no matter if clicking + or - it's always the same direction.

i'm currently trying to make sure that the wiring really is correct as it somehow looks to me like this could be it - but even trial and error didn't change anything so far.

any idea what i might be missing?

tia,
markus.



 

Please Log in or Create an account to join the conversation.

More
13 Sep 2021 00:54 #220384 by tommylight
Wiring and timing, set timings to at least 5000 both for step and 20000 both for dir.
Those are located in the .ini file with the same name as the config folder.
The following user(s) said Thank You: markusk

Please Log in or Create an account to join the conversation.

More
13 Sep 2021 02:24 #220390 by PCW
Replied by PCW on topic IHSS57-36 with 7i92 don't move
In addition to timing, how is the 7I92 wired to the drives?
The following user(s) said Thank You: markusk

Please Log in or Create an account to join the conversation.

More
13 Sep 2021 08:29 #220412 by markusk

Wiring and timing, set timings to at least 5000 both for step and 20000 both for dir.
 

these were my suspects for the last couple of weeks yet, but trying anything i could think of - i just couldn't get any different result so far.

i'll post my current state when i'm back at the workshop and verified the pinout from fpga-level.

thanks, markus.

Please Log in or Create an account to join the conversation.

More
13 Sep 2021 15:48 #220451 by PCW
Replied by PCW on topic IHSS57-36 with 7i92 don't move
"it just "spits" what feels like a single step (or not much more) at the end of the commanded movement. and no matter if clicking + or - it's always the same direction"

This sounds a lot like step/dir being swapped...

Also if you are using a 7I92 without a breakout board
there is really only one type of connection that makes
sense: +5V to drives STEP+ and DIR+, Drives STEP-
and DIR- to 7I92 step/dir pins. And then set the step
polarity to active low.

Note that I do not recommend connecting motor drives directly
tp the 7I92 but it can be done if you are careful no only connect
the 7I92 IO pins to OPTO isolated signals
The following user(s) said Thank You: markusk

Please Log in or Create an account to join the conversation.

More
14 Sep 2021 11:28 #220577 by markusk

"it just "spits" what feels like a single step (or not much more) at the end of the commanded movement. and no matter if clicking + or - it's always the same direction"

This sounds a lot like step/dir being swapped...


this sir was the simple answer i had hoped for.   seems from all the trials and errors i've been through this somehow slipped my tests.. now that you said it i checked the fpga-pins and lo and behold it says 'Dir' for 1/2 where i had it hooked on the motors 'Step' pins.. we're moving now. 

Also if you are using a 7I92 without a breakout board
there is really only one type of connection that makes
sense: +5V to drives STEP+ and DIR+, Drives STEP-
and DIR- to 7I92 step/dir pins. And then set the step
polarity to active low.


why's that? i think to remember to have read differently in other places for reasons i can't recall - and it's working now with directly connecting as designated. if doing it the way you suggest - where would i set the 'step polarity to active low'?

Note that I do not recommend connecting motor drives directly
tp the 7I92 but it can be done if you are careful no only connect
the 7I92 IO pins to OPTO isolated signals

that's what i came to understand in the meantime - that the 7i92 only seems to be meant as a 'gateway' between computer and other mesa-cards. interpreting your suggestion with my basic electronics knowledge i assume you mean to protect the mesa-card from currents flowing back from the machine/motors, right? not sure how likely this could happen with my rather 'light' machine, but i'll keep it in mind.


another basic question: using these motors with it's integrated controllers this way it isn't 'closed loop' - right? 

thanks,
markus.

Please Log in or Create an account to join the conversation.

More
19 Sep 2021 21:46 - 19 Sep 2021 21:49 #221151 by markusk

Note that I do not recommend connecting motor drives directly
tp the 7I92 but it can be done if you are careful no only connect
the 7I92 IO pins to OPTO isolated signals


well it seems i had to learn it the hard way - after accidently wiring the motor-plug from the wrong side and using it that way it seems like the pins on the mesa went belly up - i'm measuring constant 5V on the pairs which should output step/dir as soon as i start linuxcnc. am i right assuming these pins are unrecoverably busted?

since there's P1 which is currently unconfigured with the default bit-file (and assuming it has the same capabilities like P2) i would be thankful for pointers how to get P1 configured.

lesson learned: found and ordered a cheap breakoutboard with optocouplers (for further use with the 92) and got a 7i96 on the way (which takes a bit longer).

thanks,
markus.
Last edit: 19 Sep 2021 21:49 by markusk.

Please Log in or Create an account to join the conversation.

More
19 Sep 2021 22:43 #221158 by tommylight

i'm measuring constant 5V on the pairs which should output step/dir as soon as i start linuxcnc. am i right assuming these pins are unrecoverably busted?

You are not right, not yet ! :)
7i92 will have all IO at 5V as it has pull up resistors on all IO also used for level shifting to 5V tolerance as the FPGA is a 3.3V part.
Take this with a grain of salt, PCW is the master of Mesa, so waiting for him is always time well spent.

Please Log in or Create an account to join the conversation.

More
20 Sep 2021 15:01 #221204 by PCW
Replied by PCW on topic IHSS57-36 with 7i92 don't move
All I/O high after an error free LinuxCNC start probably means that
one or both of the bus switches has been damaged. This can happen
if voltages > +7V or less than -0.5V have been applied to any IO pin.

Typically damage is more likely from negative inputs. Even touching
and I/O pin to frame ground can damage the card because its easy
to have a more than 0.5V difference between 7I92 5V power ground
and frame ground when high frequency EMI is present.
The following user(s) said Thank You: tommylight, markusk

Please Log in or Create an account to join the conversation.

More
30 Sep 2021 23:27 - 30 Sep 2021 23:28 #221924 by markusk
to rephrase:
not all pins go high - only the 2 pairs i had wrongly wired to the second motor/controller. the first one still works fine.

my question still stands - is there a bit-file to be found that configures the P1 the same way as P2 is with the default firmware? am i assuming correctly that the P1's pins still all should work fine?

thanks or all your wisdom,
markus.
Last edit: 30 Sep 2021 23:28 by markusk.

Please Log in or Create an account to join the conversation.

Moderators: cmorley
Time to create page: 0.088 seconds
Powered by Kunena Forum