BS2p40 PAUSE smaller than 1 ms???

Pressus Limited

Hi Guys,

linked to my previous post on stepper motor control, I need to introduce a·PAUSE smaller than the 1ms you get with the PAUSE command to allow me to increase the motor speed.· the A4983 datasheet says it needs a minimum 1uS HIGH step signal with a minimum 1uS LOW signal.··I've put in: FOR·c=1 TO 3:·NEXT·which seems to work ok, my question is, am I right in thinking that this will introduce a delay of approximately 25uS ( (1 sec /12000 instructions per sec) * 3)???

Is there a better or·more accurate way of doing it????

thanks

Jon

allanlane5

If you have an unused pin, you can use a "PULSOUT" statement of the appropriate length.

Pressus Limited

Thanks, but unfortunately I don't have any spare pins :-(

Phil Pilgrim (PhiPi)

You can send a string of positive-going pulses using SEROUT. Set it for inverted mode and send a string of $FF values using the REP modifier. You will get a pulse train whose period is governed by the baud rate. To double the width of each pulse, send $FE instead of $FF.

-Phil

Tracy Allen

Anything you do with program loops take 100s of microseconds, even on the BS2P40. It is an interpreted language. The figure of 12000 instructions per second (83 microseconds per instruction) is an average, and some instructions are faster, some a lot slower. The <FOR c=1 to 3:NEXT> would take more like 650 microseconds.

Another way to get a string of pulses out fast is via the PWM command. For example,
PWM 1,41,1
or
PWM 1,1,41
either put out a string of 41 pulses, but the first form does it in less than 1 millisecond while the second form spreads it out over 41 times as long. The second argument (duty) should be less than 128 and the third argument (frames) can be up to 255.
The individual pulses are a little over 1 microsecond long. (A PWM frame consists of 256 time slots that can be either low or high, and duty=128 gives half and half evenly spaced).

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Beau Schwabe

Pressus Limited,

Here is where a scope really comes in handy. Below are a few different tests. I didn't have a BS2p40, but I used a standard BS2. The idea should still be the same however.

Tracy Allen,

Yes, the code is interpreted and it's going to take awhile to load, but you can create 'do nothing' functions using the LOOKUP or LOOKDOWN to act as a predictable hard coded timing loop, but again a SCOPE really helps to debug and determine what the timing is going to be.



Note: The times below are measuring how long the test pulse was HIGH

Test1:
  HIGH 0
  PAUSE 1             '1360us
  LOW 0

Test2:
  HIGH 0
  FOR B0=1 TO 3:NEXT  '2440us
  LOW 0

Test3:
  LOW 0
  HIGH 0    ' 140us
  HIGH 0    ' 280us
  HIGH 0    ' 420us
  HIGH 0    ' 560us
  HIGH 0    ' 700us
  HIGH 0    ' 840us
  HIGH 0    ' 980us
  HIGH 0    '1120us
  HIGH 0    '1260us
  HIGH 0    '1400us
  LOW 0

Test4:
  HIGH 0
' LOOKUP 0,[noparse][[/noparse]0],W0                     ' 460us
' LOOKUP 0,[noparse][[/noparse]0,0],W0                   ' 570us
' LOOKUP 0,[noparse][[/noparse]0,0,0],W0                 ' 680us
' LOOKUP 0,[noparse][[/noparse]0,0,0,0],W0               ' 790us
' LOOKUP 0,[noparse][[/noparse]0,0,0,0,0],W0             ' 900us
' LOOKUP 0,[noparse][[/noparse]0,0,0,0,0,0],W0           '1010us
' LOOKUP 0,[noparse][[/noparse]0,0,0,0,0,0,0],W0         '1120us
' LOOKUP 0,[noparse][[/noparse]0,0,0,0,0,0,0,0],W0       '1230us
  LOOKUP 0,[noparse][[/noparse]0,0,0,0,0,0,0,0,0],W0     '1340us
  LOW 0

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 6/9/2010 8:57:45 AM GMT

Tracy Allen

Hi Beau,
The OP is asking about the BS2p40, and I surmise that the times you listed above are for the BS2 or BS2e. Right? Surmise based on the 140 µs time for LOW 0: HIGH 0. BS2p is a little over 3x faster. The LOW 0:HIGH 0 sequence takes close to 41 µs on the BS2p.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Beau Schwabe

Tracy Allen,

"The OP is asking about the BS2p40, and I surmise that the times you listed above are for the BS2 or BS2e. Right?" - Yes, that is correct, I did not have a BS2p40 to check, and I indicated that in my post. I was just showing a few different methods to get a deterministic amount of timing that was smaller than the 1ms intervals.

A Scope is essential though when timing like this plays a big part of the design.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

allanlane5

And I have used the "Ant-8" as a very nice USB-attached logic analyzer for these kinds of timing studies.