Stepper software for prop chip
mikediv
Posts: 825
Hi guys does anyone have a decent program for driving a 6 wire unipolar stepper motor I have a design I am working on and wired up a ULN2803 to a 6 wire Unipolar stepper motor but I can not find any code for the prop that I can use to test this or try and learn how to do it myself if anyone has anything they are willing to share I would be very grateful.
Comments
Rich H
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
The Servo Boss, a 12 channel servo tester kit from Gadget Gangster.
Running a unipolar is very easy as you just need a step sequence a bit like this: (assume 4 coils labeled A,B,C,D)
Forward: Drive A, wait, drive B, wait, release A, wait, drive C, wait, release B, wait, drive C, wait, release B, wait, drive D ..... and so on
Reverse: Drive D, wait, drive C, wait, release D, ....... and so on
This is the half-step method that pulls the armature to the half-step position when both coils are driven. Of course when you forward or reverse you simply drive the adjacent coil of the one that happens to be driven. Use an index of 0 to 7 that wraps around for this. The wait period can be made variable to allow for acceleration and deceleration.
*Peter*
P.S. index has to 8 steps which can lookup a table to decide which coils to drive.
I found this animated schematics
that show the sequence how the coils have to be switched on/off
this one seems to be a 6-wire unipolar on the page it is called "unipolar halfstep"
to drive the steppermotor you can define an array of bytes that contains the bittpatterns
according to the bitpattern shown in this animation I post some lines of code that show the principle
(edit if you don't add blanks between square-brackets this "highly sophisticated" forum-software takes them for changing the fontsize)
Var byte BitPattern[noparse][[/noparse] 8 ] '8 elements accessed by index 0 to 7 NOT index 1 to 8 long count Pub Demo BitPattern[noparse][[/noparse] 0 ] := %1000 BitPattern[noparse][[/noparse] 1 ] := %1010 BitPattern[noparse][[/noparse] 2 ] := %0010 BitPattern[noparse][[/noparse] 3 ] := %0110 BitPattern[noparse][[/noparse] 4 ] := %0100 BitPattern[noparse][[/noparse] 5 ] := %0101 BitPattern[noparse][[/noparse] 6 ] := %0001 BitPattern[noparse][[/noparse] 7 ] := %1001 DirA[noparse][[/noparse] 0..3 ] := %1111 repeat repeat count from 0 to 7 OutA[noparse][[/noparse] 0..3 ] := BitPattern[noparse][[/noparse]count] waitcnt(ClkFreq + cnt)Steppermotors and steppermotor cicruits are much more variable than f.e. the I2C-Bus
You have 4,6,8 coils, 4,6,8 wires, unipolar and bipolar types
This means there can't be one universial object "steppermotordriver" that suits to them all
Some code just puts out step and direction to a controller who does the details of switching the wires
Some other code does the switching of the wires like shown in the codesnippet above
here is an article stepping out with spin
that shows some code how to do it on a level a little above what I have shown in the code snippet
best regards
Stefan
Post Edited (StefanL38) : 9/11/2009 7:22:56 PM GMT
The controller logic needed for steppers is rather simple pattern based sequencing, similar to what is posted above.· You just need to keep repeating the patterns in the desired direction to move the motor.· Where it becomes more complicated is controlling the current to the motor itself.· As steppers are obviously magnetic devices, the primary factor in their performance is current.· It is quite typical to drive steppers with voltages much higher than what is printed on the motor, and instead control the current using a pwm (aka chopper) technique.· This will significantly improve stepper performance, especially at higher speeds where low voltages can cause the motor to get 'out of step'.
Here is·a good reference on steppers and control - refer to the 'Current Limiting' section for the topic I mention above.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Post Edited (Agent420) : 9/10/2009 1:43:50 PM GMT
Var long BitPattern Pub Demo BitPattern := %1000_1010_0010_0110_0100_0101_0001_1001 DirA[noparse][[/noparse] 0..3 ] := %1111 repeat OutA[noparse][[/noparse] 0..3 ] := BitPattern & %1111 BitPattern ->= 4 ' Forward (<-= for reverse) waitcnt(ClkFreq + cnt)▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
I found this code from Ray Allen he is one of the resident genius's here but as hard as I have tired to modify this to work my stepper only hums or buzz but does not move. The steeper motor I am using are 1 amp but by doubling up the ULN2803 outputs it should be able to handle the no load power for enough time for me to see if they will at least step. When I refer to they I am using/trying different steppers but only one at a time
Thanks. StefanL38 that's a very interesting animation thank you for taking the time to find it as well as the rest of you.
Peter Jakaci do you mean I should build a logic table ? If yes how many bits should it be I read Kwinn post on 4 bits but this is where I get confused.
For starters the ULN2803 is not really able to handle 1 or 2 amps total, there is an absolute maximum power rating of 1.47W for the DIP package. What happens is that each darlington will only saturate to around 1.5V @ 500ma each which means just one coil will push this past the limit. BTW, the recommended rating is 0.76W. Bear this in mind when the lid of the chip goes pop and lets all the magic smoke out. It's not really good enough to say that's all you have but if you really really must then piggyback another chip on top. Far easier (in this case) just to use little TO92 transistors (low-sat) that can handle a few amps easy.
The code examples that others have given should help you along, I was trying not to do it for you as the sequence itself is very simple, you just need to understand it. You have your Spin compiler ready and waiting for you to type in a little test sequence, even as primitive as the sequence that I outlined, so what's stopping you?
*Peter*
a square wave signal and I am sure I have the pins right what could I be missing I am obviously missing somehting simple , Oh Peter according to the data sheet the way I have the pins doubled up it states it can handle 1 amp I agree with you this is not the way to go forward but I just want to get my code working to some degree
The code you posted shows you trying to do microstep so just comment this one out and uncomment the full or half step. Make sure your coils and code correspond so that the coils are driven in adjacent sequence. If in doubt, swap a couple around [noparse]:)[/noparse]
*Peter*
maybe if you try it in some kind of a single step mode
with LEDs in parallel to see which coil is switched on/off
I mean a testprogram where you can step through the sequence by pressing a key
pressing the key ONE time and take a look how the shaft was turning for some degrees
pressing the key ONCE again and see if the shaft is turning into the same direction as before or the other
if so change the sequence
or a testprogram to switch on/off IO-PIN 0-3 by different buttons to test what happens if a switch on pin 0 and then pin 1 etc. etc.
best regards
Stefan
but I have these really big steppers I even have a couple 20 amp jobbers its the timing I am having the biggest problem with I am just learning how to program period and I am starting with Spin I know this more advanced than where I am at , I am good at blinking Led's and reading toggle switches but the stuff I want to do (build robots) is taking me forever to master spin,,, no knock against spin its awesome but other than some basic literally I am trying to self teach myself and I do appreciate all the help and tips guys.
I am used to programming in Forth where you can communicate and compile interactively with the system and I can testify that this is a very powerful development environment. In a similar fashion though not quite so interactively you can write a line or two of Spin, press F10, yep it looks good, and build on it. So just proceed a step at a time, just as if you are walking for the first time.
*Peter*
I was doing the exact thing you are with a unipolar stepper just a couple weeks ago. I have some spin code that will step and run the motor forward and reverse for testing.. I will post if you're interested.
Steve
they just Buzz I am starting to think I am doing somehting wrong with the steppers directly let me explain. I got these out of a large photo enlarger they are Vexta 2 Phase PH 2610 6V 1.8 A
6 Wire with the help from some of the guys here I learned how to Ohm out the wires So taking 3 of the six I ohm between them and find the common so 10 Ohms between 2 wires and then 4.8 Ohms from the common to either end wire I end up with 2 common wires and 4 coil wires
I can Ohm them out so I have each Coil you can see from schematic I tied the common wires to pin 10 of the ULN2803 to 6V and each of the other coil wires to an output of the ULN
I also have a few Shibaura S4H56B06K-06 6V 1.1 A 1.8Degree Stepping motor same thing 6 wire I did the same thing to figure out wiring found common and the associated wires to go with the common making up the coil I took them out of working machines and saw them
run with my own eyes but I have never been able to build a circuit and get them to step using my Stamps or Props . I am wondering if I am wrong and these are not even unipolar the reason I said they are is from what I found on internet and one place said 6 wire steppers were unipolar I have to ask is there any way I can connect one to a 5 volt power supply and by hand switch the power on and off just to see them step????
Steve I would love to see your code and how you connected your stepper if you would not mind
One more thing I have the stepper kit from Parallax its a 5 wire stepper that works with the Basic stamp2 or homework board I have gotten that to work just fine with Parallax code and my OEM Stamp2 however it is 5 wires and no matter how I tried to wire up one of the 6 wire steppers they would not work with this example
I very much appreciate being able to come here and ask you guys for help and everyone taking the time to try and help
I can post a schematic if you wish.
..Steve
I'm getting a server error
{{ StepperTest.spin Stepper motor test by Steven R. Stuart 14-Aug-2009 This program was tested on a 1.8° full step motor, 200 steps/revolution uni-polar stepper motor Low power sequence Winding X1 100010001000100010001000 X and Y center-tapped winding 1 Winding Y1 001000100010001000100010 Winding X2 010001000100010001000100 X and Y center-tapped winding 2 Winding Y2 000100010001000100010001 time ---> This sequence will generate about 1.4 times torque Winding X1 1100110011001100110011001 Winding Y1 0011001100110011001100110 Winding X2 0110011001100110011001100 Winding Y2 1001100110011001100110011 time ---> }} CON _XINFREQ = 5_000_000 'Frequency on XIN pin is 5 MHz _CLKMODE = XTAL1 + PLL16X 'System clock = 16x PLL = 80MHz pinX1 = 4 'The pins controlling the motor coils pinY1 = 5 pinX2 = 6 pinY2 = 7 speed = 200 'Higher is faster. 200 = 1 revolution per second VAR byte seq PUB StepperMotor | index seq := @SEQ1 'Select step sequence: @SEQ1 or @SEQ2 dira[noparse][[/noparse]pinX1..pinY2]~~ 'Set stepper pins as output repeat 'Run forever repeat 50 'One full circle (50 steps) repeat index from 0 to 3 'Step thru the four SEQ pattern bytes outa[noparse][[/noparse]pinX1..pinY2] := byte[noparse][[/noparse]seq][noparse][[/noparse]index] 'Output a pattern waitcnt(clkfreq / speed + cnt) 'Run SEQ in reverse repeat 50 repeat index from 3 to 0 outa[noparse][[/noparse]pinX1..pinY2] := byte[noparse][[/noparse]seq][noparse][[/noparse]index] waitcnt(clkfreq / speed + cnt) DAT ' Y1┐┌X2 ' X1┐││┌Y2 SEQ1 byte %1000, %0010, %0100, %0001 'lower power SEQ2 byte %1001, %1010, %0110, %0101 'higher torqueI use (or I am trying) to use the ends of the stepper coils to provide the steps Via the prop outputs buffered through the ULN 2803 I am still not having any sucess getting the steppers to do anything more than Buzz
what frequency are you using ?
did you try changing the A, A' wires ?
did you try changing the B, B' wires ?
steppers often just buzz if one of the wires is connected vice versa than how they should be connected
if you mount something on the axle of the steppers maybe 10 cm long to see a bigger movement of a single step
with a step-frequency as low as 1 Hz and you switch just between two bitpattern of the sequence always for and back does the stepper do that ?
If you add a third bitpattern 1,2,3 and always switch them 1-2-3-2-1-2-3-2-1... does the stepper do that ?
try experimenting with different bitpattern sequencies
best regards
Stefan
I have tried changing the wiring with no luck and I did put a very long wire on the shaft to see if maybe they are moving and I am no seeing it but that is not the case. Great idea though and thank you for it.
I am hoping the L293 will make the difference if you look at my schematic I am using the common of both stepper coils and driving it at +12 so I really do not have a ground for the stepper coils then I was using the prop outputs to send the steps buffered through the ULN2803
I can not thank everyone enough for all the help so I will let this rest until parts come in I have in the meantime tried using LEDS as was suggested here and even my quad scope so I can clearly see the code examples are sending pulses out on the right prop pins so it has to be the way I am driving/wiring the steppers as you have suggested.
A on, B on, A off, C on, B off ...etc
If that doesn't work then it's your motor that's the problem.
The 2803s should be working although individual NPNs (with diodes) would be less problematic and deliver the full voltage and current. Take the ZTX849 in a little plastic TO92 style package, it handles up to 5A continuous with a maximum of 0.22V drop and still the gain is around 100 to 200. Of course MOSFETs have an internal diode and can easily handle the current too, but the NPN is a standard through-hole part and very small.
You don't need a transistor this good and there are plenty to choose from so why bother with special drivers?
*Peter*
I wish I could draw this better but here goes coil 1,,,,, 1
middle leg is common
2 other side of coil so" 1 is Blue" (middle is white and common) and "2 is the other end of this coil red wire" Now the second coil
1
middle leg is common
2 the other side of this coil " 1 is green first wire"(middle is white common same as other coil) and the other end of the coil wire end is "black"
So I have 2 coils oh by the way this is the Vexta 2 phase 2610--- each coil is divided in half by the common wire if you look at my schematic I tied both commons to +12 V and either end of the other coils to the outputs of the ULN2803
To simplify this I will run to Radio Shack tomorrow and pick up a handful of Transistors what do you recommend Pete? Thanks again for all the help
I have no trouble with servos
I couldn't find the 2610 on the Vexta website. Do you have a link?
There is a diagram here of your 6 wire unipolar. (The 2 commons are tied together to the supply voltage for your application)
www.probotix.com/stepper_motors/unipolar_bipolar/
*Peter*
It could be you are confusing the leads, you should be able to confirm the coil wiring·using an ohmmeter.· For each side, the pair with the greater value between them are the leads you want to control by switching to ground, and the wire that results in the lessor reading from the other two is the common which should be connected to the positive rail.
edit - oops, missed your post indicating you have already down this^
Also, I would slow your frequency down to a crawl in your code (maybe 1 pps) so you can visually observe what is happening.· The 'buzz' is a result of your code changing outputs so quickly.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Post Edited (Agent420) : 9/17/2009 2:58:19 PM GMT
here's the datasheet for the L293D buffer/driver: www.datasheetcatalog.org/datasheet/texasinstruments/l293d.pdf.
I've attached a schematic showing how I connected my motor. I don't know if the motor wire colors are of any standard values, but I included them.
To view movement, I just placed a 1 inch piece of black electrical tape to the shaft of my stepper.
Good luck
i have dug up this post because i am a little lost. I have a bipolar stepper. that i would like to use with your code. I am using the PPDB that has the L293 on-board.
The stepper motor has the following wire colors...orange, white, green, yellow, black, brown, red and blue.
i have used a continuity tester and determined that the coil pairs are as follows (orange/white), (green, yellow), (black/brown), (red, blue)
I realize that i need to connect the two opposing coils together making a set of two coils called A and B. So i would have A,*A,B,*B with two sets that are tied together separately. (make sense?)
How do i know what coils oppose each other and belong in A group and in the B group?
The PPDB has a terminal block that starts at GND, OUT0, OUT1, OUT2, OUT3, V+
I have connected a supply voltage of 13.8V to the GND and the V+.
What wires connect to the remainder terminals?
Will the code that was supplied above work with this motor?
Thanks for any help that anyone can render.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
It's Only A Stupid Question If You Have Not Googled It First!!
nnnnnn
nnnnnn
nnnnn
nnnnn
| | | | | |
| | | | | |
A *A,A *A B *B,B *B
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
It's Only A Stupid Question If You Have Not Googled It First!!
I will try to find another way to illustrate.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
It's Only A Stupid Question If You Have Not Googled It First!!
Do you know the brand name and model number?
I don't think unipolar code could ever work with a bipolar motor. [noparse][[/noparse]EDIT: oops, I think I'm wrong about that. What I should say is that you can't run a bipolar motor in bipolar mode using unipolar code. Sorry about that.] I've recently been using a chip for controlling bipolar motors, the L6208. If you look at the attached pdf, you'll find it has a diagram of how it switches the voltage inside the chip to energize the coils for bipolar mode.
Here are some links that help explain a little about different kinds of stepper motor controllers:
dkc1.digikey.com/us/en/tod/STMicroelectronics/StepperFundamentals_NoAudio/Stepper_Fundamentals_NoAudio.swf
www.cs.uiowa.edu/~jones/step/physics.html
I hope that helps a bit,
Mark
Post Edited (ElectricAye) : 4/27/2010 1:11:09 PM GMT