×
Forum Header
Gecko driver timings in stepconf wizard
01 Aug 2013 02:41 #37305
by spike
Gecko driver timings in stepconf wizard was created by spike
Hi,
I recently used the stepconf wizard to set up a configuration for my gecko drives and I noticed that the timings suggested by the gui are significantly different from what gecko specified in their datasheets. I had a look at the datasheets of the other gecko drives and it seems that sometimes the timings were matching the spec while sometimes they were just absurdly off (20µs for direction setup, are you kidding?).
I've only got the 213V and 203V drives here to test (and they are working to the specs), but I trust gecko and their datasheets, so I've attached a patch for the stepconf.py file to match the timings to the specs. I've also taken the liberty to add some missing drives. This could probably be applied to the pncconf.py file, too.
Best regards, Chris
I recently used the stepconf wizard to set up a configuration for my gecko drives and I noticed that the timings suggested by the gui are significantly different from what gecko specified in their datasheets. I had a look at the datasheets of the other gecko drives and it seems that sometimes the timings were matching the spec while sometimes they were just absurdly off (20µs for direction setup, are you kidding?).
I've only got the 213V and 203V drives here to test (and they are working to the specs), but I trust gecko and their datasheets, so I've attached a patch for the stepconf.py file to match the timings to the specs. I've also taken the liberty to add some missing drives. This could probably be applied to the pncconf.py file, too.
Best regards, Chris
Please Log in or Create an account to join the conversation.
01 Aug 2013 02:43 #37306
by spike
Replied by spike on topic Gecko driver timings in stepconf wizard
Hmm... I'm getting a 404 when I try to view the patch, so here's it again in plain text:
--- stepconf.py.orig 2013-07-31 21:23:29.098978237 +0200
+++ stepconf.py 2013-07-31 21:30:12.001000969 +0200
@@ -70,6 +70,10 @@
- ["gecko201", _("Gecko 201"), 500, 4000, 20000, 1000],
- ["gecko202", _("Gecko 202"), 500, 4500, 20000, 1000],
- ["gecko203v", _("Gecko 203v"), 1000, 2000, 200 , 200],
- ["gecko210", _("Gecko 210"), 500, 4000, 20000, 1000],
- ["gecko212", _("Gecko 212"), 500, 4000, 20000, 1000],
- ["gecko320", _("Gecko 320"), 3500, 500, 200, 200],
+ ["gecko201", _("Gecko 201"), 1000, 2000, 200, 200],
+ ["gecko201x", _("Gecko 201X"), 1000, 4500, 200, 200],
+ ["gecko202", _("Gecko 202"), 1000, 2000, 200, 200],
+ ["gecko203v", _("Gecko 203V"), 1000, 2000, 200 , 200],
+ ["gecko210", _("Gecko 210"), 1000, 2000, 200, 200],
+ ["gecko210x", _("Gecko 210X"), 1000, 4500, 200, 200],
+ ["gecko212", _("Gecko 212"), 1000, 2000, 200, 200],
+ ["gecko213v", _("Gecko 212V"), 1000, 2000, 200, 200],
+ ["gecko320", _("Gecko 320"), 1000, 2000, 200, 200],
+ ["gecko340", _("Gecko 340"), 1000, 2000, 200, 200],Please Log in or Create an account to join the conversation.
- Todd Zuercher
-
- Offline
- Platinum Member
-
Less
More
- Posts: 4973
- Thank you received: 1373
01 Aug 2013 08:20 #37317
by Todd Zuercher
Replied by Todd Zuercher on topic Gecko driver timings in stepconf wizard
I am only venturing a guess. (I have no idea of the source of the timings used in stepconf.)
Some of the Gecko drives listed have been around for quite a while, with a number of revisions over the years. It is quite possible the maximum timings for the drives changed with drives revisions, and the default settings were not updated to match them.
Some of the Gecko drives listed have been around for quite a while, with a number of revisions over the years. It is quite possible the maximum timings for the drives changed with drives revisions, and the default settings were not updated to match them.
Please Log in or Create an account to join the conversation.
01 Aug 2013 12:30 #37320
by spike
Replied by spike on topic Gecko driver timings in stepconf wizard
Quite possible. But I assume that the datasheet for a let's say Gecko 201 drive has correct timings for every 201 revision out there if it doesn't state differences in hardware revision. Unlike many Chinese manufacturers where the specs are merely a suggestion. And even if the timings changed with hardware revisions, the dir setup value for the 201 is off by a factor of 100. That's just highly unlikely. On the other hand, I seriously doubt that a 210 would work reliably with a step time of 500ns.
Please Log in or Create an account to join the conversation.
01 Aug 2013 21:15 #37323
by PCW
Replied by PCW on topic Gecko driver timings in stepconf wizard
Actually you dont want to use minimum timings anyway (there's 0 margin in that case) and a 200 ns DIR setup and hold will fail as soon as you use almost any opto-isolated breakout board.
What makes most of this dont-care is that target mode PCI, PCIE or LPC writes on a modern PC typically take 500 nsec to 2 usec, so 200 ns timing is not even possible and the shortest step pulses (1 PCI write cycle time is the minimum possible) are only used in the reset-mode.
Also unless reset mode is used I'm pretty sure the actual direction setup and hold times are quantized (and rounded up) to one base thread period so will always be in the 20 to 100 usec region. Note that long direction setup and hold times have minimal effects on motion so are also pretty much dont-care...
What makes most of this dont-care is that target mode PCI, PCIE or LPC writes on a modern PC typically take 500 nsec to 2 usec, so 200 ns timing is not even possible and the shortest step pulses (1 PCI write cycle time is the minimum possible) are only used in the reset-mode.
Also unless reset mode is used I'm pretty sure the actual direction setup and hold times are quantized (and rounded up) to one base thread period so will always be in the 20 to 100 usec region. Note that long direction setup and hold times have minimal effects on motion so are also pretty much dont-care...
Please Log in or Create an account to join the conversation.
02 Aug 2013 04:20 #37333
by spike
Replied by spike on topic Gecko driver timings in stepconf wizard
But that doesn't explain why the wizard is suggesting the minimal values for some devices (e.g. the 203v which uses the values straight from the datasheet), even shorter than minimal for others (500ns step time for 201 and 202) but ridiculously long values for the setup timings. I'd like it to be a little more consistent. I can understand that most users probably use some kind of opto-isolation that adds a little delay, but that should be the same for every drive. So either leave it up to the user to increase the default settings to work with his breakout board or use some arbitrary factor (like times 2 or 3) for the datasheet values to work with a great varieties of breakout boards.
I'd much rather like the plain value from the specs and add a potential delay of a opto-coupler myself, but that's just a personal preference.
I'd much rather like the plain value from the specs and add a potential delay of a opto-coupler myself, but that's just a personal preference.
Please Log in or Create an account to join the conversation.
Time to create page: 0.168 seconds