× Forum Header

LinuxCNC, DQ542MA driver and DB25-1205 BOB

More
15 Aug 2013 19:54 #37714 by awes
Hope this can help somebody in the future...

LinuxCNC, DQ542MA drivers and DB25-1205 BoB (from Wantai Motors)

When setting this up I had problems to find examples so I thought I might share what worked for me.

Based on this image I found while searching I did the wiring (not mine picture).
img195.imageshack.us/img195/9451/sitemanuevo.jpg
The third red from the top that goes to ENBL + is connected to pin 16.

Dip switches on the driver card are set for 200 steps/rev and 1/8 microsteps

In LinuxCNC config I set:

Driver timing: According to what is documented here wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing

Parallel port setup:
Pin 2 - XStep
Pin 3 - XDirection
.......
Pin 16 - Amplifier Enable (I will connect all drivers "ENBL+" to this pin. This will be a three axis setup for a mill)

X Axis Setup.
Motor steps revolution : 200
Drivers Microstepping: 8

Anders
SpÄnga, Sweden
The following user(s) said Thank You: BigJohnT, ruffle, bottleworks

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

More
23 Sep 2013 04:01 #39114 by gburbeck
Firstly I would like to thank Anders for his post, it proved enormously helpful in giving me the confidence to try to convert my MF70 mill to cnc. This reply is not an attempt to hijack his post but to hopefully offer an alternative option.

I discovered today that following Anders' set up did not work with a similar 3 axis kit that I purchased from Wantai - which included three 57BYGH627 stepper motors.

I found that I could not get the motors to rotate with an active high configuration, as an oscilloscope showed that the high pulse voltage (PUL+) on all three of the DQ542MAs was only just above 2v as opposed to 5v - seemingly insufficient to trigger a pulse. I do not know if this is a design flaw in the particular drivers I have been sent but the approach below seems to have given me a working configuration.

Changing to an active low configuration resolved the problem.

In an active low configuration the PUL+ and DIR+ (for each DQ542MA driver) are tied to the 5v input on the DB25-1205 and the PUL- and DIR- to their respective pins (for instance default X axis linuxcnc pins 2 and 3). The pins should be inverted in the Parallel Port Set-up screen within the Stepconf Wizard .

I have used all other aspects of Anders configuration.

I hope this helps anyone that has encountered a similar problem.

Grant

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

More
25 Sep 2013 23:53 #39185 by fannlow
Hi there,
I'm using DM542A driver board, from longs-motor, but have not had idea of how to retrieve the

Step Time
Step Space
Dir. Hold
Dir. Setup

I'm hoping I can use the DQ542MA board timings?

or if anyone can direct me to the right place?

also, how or where on the driver board, did you get the
steps per rev?

I only see pulse per rev? is this the same?

also for microsteping, my board says its capable of 128, but I saw you wrote 8, so where do I get this from also?

hope you may show me the light, which prevents me from trying to trial and error for almost 4 days now... :(

thanks in advance

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

More
26 Sep 2013 00:57 #39187 by fannlow
am attaching nema 23HS9430 spec sheet
am attaching driver DM542A spec sheet

my current setup : driver DIP switch.

( sw 1, sw 2, sw 3 ) = OFF, OFF, OFF
= Peak (A) 4.20 & RMS (A) 3.00

( sw 4 ) = OFF
= not sure what this means? is there reduction to 70%? or what settings should this be?

( sw 5, sw 6, sw 7, sw 8 ) = OFF, ON, ON, OFF
= 2000 pulse/rev
is more pulse/rev = steps/rev?
if that is the case, is more pulse/rev better or less pulse/rev better? or should i ask if faster?

thanks for helping me sort out or point me in the direction of linking them together =)

many thanks
Attachments:

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

More
26 Sep 2013 17:52 #39213 by andypugh

is more pulse/rev = steps/rev?
if that is the case, is more pulse/rev better or less pulse/rev better? or should i ask if faster?


Pulses per rev is exactly the same as steps per rev.

Al LinuxCNC cares about is how many step pulses it needs to make to move the machine one distance or angle unit. (Typically inches or mm or degrees, but you could use any units).

The motor steps per rev is fixed (200) but the drive can "microstep" between these real steps by sending intermediate current levels to the coils. More microsteps will normally be smoother and will give higher resolution, but eventually you will hit the limit of how many steps per second the software can produce. It rarely makes sense to go above 4x microstepping. (What the drive manual calls 800 pulse per rev)

Set the drive to the motor rated current (3A). Ignore the motor rated voltage. The drives can handle 50V, and you will get more motor speed the higher you go up to that point.

Follow what the drive manual calls the "Yang" wiring scheme. Parallel ports are much better at sinking current than sourcing it. That means connecting 5V to the + terminal of each input to the drive. The USB port is a very useful source of 5V for this purpose. (the signal voltage needs to be referenced to the PC ground, and ideally the motor power should be completely separate).
If using that wiring scheme (active low) then you need to select the "invert" checkboxes for the step pins in stepconf.

None of the manuals has anything to say about the step timing parameters. However if you look at wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing and the Wantai DQ542MA you will see that that has exactly the same manual as your drives, and you should use the numbers for that in stepconf.
The following user(s) said Thank You: fannlow

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

More
27 Sep 2013 01:11 #39227 by fannlow
Yup, did just that.
How do you check from the board or software if that is the case, if I have active low or high.

My current setup has invert on step only at this time. I will upload some screen shots tomorrow for references.

Many thanks

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

More
27 Sep 2013 01:21 #39230 by andypugh

How do you check from the board or software if that is the case, if I have active low or high.

It's a wiring choice you can make, more than anything else. The Parallel port standard is that "1" = 5V. But they are more useful to us generally in current-sinking mode.

wiki.linuxcnc.org/cgi-bin/wiki.pl?Parallel_Port_Tester

Might be useful.
The following user(s) said Thank You: fannlow

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

More
27 Sep 2013 01:38 #39231 by fannlow
I did a few trial and error.

my original planned data to be used for stepConf

200 steps/rev (driver DM542A DIP to 800 pulse/rev for all X/Y/Z)
4 microsteps
1:1 (motor:lead)
7mm (leadpitch, measured the distance traveled by the NUT on leadscrew, by running velocity 1mm/s at 150mm/s accel, for 1 full turn of the leadscrew)
homing to 0 (zero)
travel -800 to 800 (Y-axis)
travel -350 to 350 (X-axis) 2.5mm leadpitch
travel -300 to 300 (Z-axis) 2.5mm leadpitch

however when I try to run the machine test for each axis, I just couldn't get the right travel distance.
eg: if I run +/- 700 on Y-axis using 7mm pitch, my distance traveled is way too short

i couldn't figure what went wrong?
steplen 5050ns
stepSpace 5050ns
Dir. Hold 5000ns
Dir. Setup 5000ns
Not completely sure if these has to do with the NEMA23HS settings?
which is 1us high, 1us low, Dir hold is 2us, which is quite different from the DM542A settings?

forgot to screenshot, will upload tomorrow...

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

More
27 Sep 2013 01:47 #39232 by andypugh
7mm pitch would be _really_ unusual.
Also, very coarse.
Ballscrews are typically 5mm pitch. Trapezoidal leadscrews on eBay are 3mm pitch up to 14mm then 4mm pitch 16mm and over.
The following user(s) said Thank You: fannlow

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

More
27 Sep 2013 01:55 #39234 by fannlow
yea, it wasn't a real leadscrew I would say... I an actual ball screw from Korea, PMI brand, which will arrive mid Oct 2013... along with the bundles, bearings blocks, fixed/open support, etc...

I didn't know if I turned the screw 1 complete rev = distance traveled by NUT = leadpitch?
but I did saw on the stepConf, that it says mm/rev, which is what I did...
maybe I didn't understand it completely and made a mistake which I need to learn? =)

on the new ballscrew which I ordered, it is 25mm dia, with lead 10, which I assume will be 10mm/rev?



I have started a new topic for this discussion, as I didn't want to disturb this topic, since it was already "settled", so I moved over to

www.linuxcnc.org/index.php/english/forum...42-step-timespacedir

I did attached some images for my morning testing, and I used andypugh's suggestions for my evening testing, which will upload tomorrow.
just to compare and learn where my mistakes are from and correct them ;)

this forum is a lot of great help!

really appreciate all these kindness

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

Time to create page: 0.183 seconds
Powered by Kunena Forum