Picstep problems continue

[randyf1965]

I had 2 boards working with my 2 good pic's.. replaced the bad one. Tried to put my breakout board (non OPTO) in place. NOTHING works. I can see the step and direction signal come from the breakout board to the picstep. If I short S1 on the breakout board I get 2 of the led (2 led wired opposite with a resistor to limit current)to dimly light. tried 3 board 4 pic's same issues.

I have also removed the breakout board and wired
parallel pin 2 to picstep pin 7
parallel pin 3 to picstep pin 8
5v to picstep pin 1
gnd to picstep pin 10
gnd parallel pin 22

same thing is seen on output.

thought I was on a roll after the laser printer pc board etching worked so well.... but all has gone rapidly downhill from there,

[jerber]

When I wired my 1st board to my parallel port, I found out that you have to connect pin 3 (enable/brake) to the +5V.

Are you also sure that you programmed your pic's the right way ?

[randyf1965]

I found I have to GROUND pin 3 on th epicstep board... All 3 of my boards work GREAT hard wired to the parallel port.... my breakout board had MAJOR issues... only works 1 direction (and it stutters that direction) BUT it turned out to be the 5v line!

[randyf1965]

Man... doesn't take long for your jaws to start hurting watching 3 steppers spin as you run a test.nc file thru turbo cnc while grinning like a madman!

WOOHOOO!!!!!!

[jerber]

Originally Posted by randyf1965 I found I have to GROUND pin 3 on th epicstep board... All 3 of my boards work GREAT hard wired to the parallel port.... my breakout board had MAJOR issues... only works 1 direction (and it stutters that direction) BUT it turned out to be the 5v line!

You're right, you have to connect enable (pin 3) to the ground.
I also had problems with my homebuild (I actualy build it at my work) opto-break out board. The picstep runned fine without it but wouldn't move when connected to my board. After using a scope I found out that you have to use (very) fast optocouplers for the step signals. For the other pins the optocouplers don't have to be fast because they just act as swithes. If you're using PWM for spindlespeeed control (Mach2/3) then that pin needs a fast(er) optocoupler too.

[da21]

Originally Posted by jerber You're right, you have to connect enable (pin 3) to the ground. I also had problems with my homebuild (I actualy build it at my work) opto-break out board. The picstep runned fine without it but wouldn't move when connected to my board. After using a scope I found out that you have to use (very) fast optocouplers for the step signals. For the other pins the optocouplers don't have to be fast because they just act as swithes. If you're using PWM for spindlespeeed control (Mach2/3) then that pin needs a fast(er) optocoupler too.

which opto's did you use

[jerber]

Originally Posted by da21 which opto's did you use

The opto's I use are PC910's, from sharp, for the three step signals and ISP521's for the other pins.

[10bulls]

hey jerber!
What pulse width can you go down to? (in Mach2?)...My Fairchild optos are good for down to 4us, but any lower than that and they 'fail to proceed'.

[jerber]

Originally Posted by 10bulls hey jerber! What pulse width can you go down to? (in Mach2?)...My Fairchild optos are good for down to 4us, but any lower than that and they 'fail to proceed'.

According to the datasheet they can handle a pulsewidth of approximately 200 nS, that's fast enough for the 45 kHz Mach2 can pulse out. However I can't seem to drive my steppers, I tested 4 different models so far, any faster than 6 kHz from Mach2, but that's ok for me.

I also played with a hcpl053l (a dual 050l) wich is a SMD 3.3V type but worked nevertheless.

[frog]

I have a question does anyone have experience with 16f84a as a stepper controller I have built my own design and when I up the voltage to 40 volts with 4.1 volt 1.1amp steppers my micros scramble and I have to reprogram them I have diodes out the gazu to stop this I have been working on this night and day for 1 month built the whole thing from scratch it works ok at 5 volts but looses sync on y axis it has to be something simple any help would be appreciated do I need to couple the power pins on the micro with pico caps help me sombeody loosing my hair. are these motors big enough for an 18'' x 18" mill when they are working you cant stop them then they start to stutter then I have to reprogram the chip.I am using a hex buffer in from of tip 122 unipolar motors.

[10bulls]

Originally Posted by jerber According to the datasheet they can handle a pulsewidth of approximately 200 nS,...

I think the datasheet is times are when the input signal is a nice buffered 5V signal.

The signal from the parallel port will typically be a little over 3V, with a lot of
capacitance from the parallel port leads. Alan's circuit was designed so a separate PC side power supply was not required, so he couldn't buffer the inputs. I have attached a trace of an opto output during stepping.
I think this was a 7 or 8uS pulse width.
As you can see, the signal is still pretty low at 2-3uS.

[abasir]

Originally Posted by frog I have a question does anyone have experience with 16f84a as a stepper controller I have built my own design and when I up the voltage to 40 volts with 4.1 volt 1.1amp steppers my micros scramble and I have to reprogram them I have diodes out the gazu to stop this I have been working on this night and day for 1 month built the whole thing from scratch it works ok at 5 volts but looses sync on y axis it has to be something simple any help would be appreciated do I need to couple the power pins on the micro with pico caps help me sombeody loosing my hair. are these motors big enough for an 18'' x 18" mill when they are working you cant stop them then they start to stutter then I have to reprogram the chip.I am using a hex buffer in from of tip 122 unipolar motors.

Switching 40V@1.1A will cause a lot of electrical noise. Apart from bypass caps on the PIC supply (about 100nF), you may need a larger filtering capacitor on your power supply. Also the PCB layout is also crucial. Wire the supply in star configuration.

Maybe start a new thread to avoid hijacking the original thread on PICStep