Instructables

Steppers and EMF.?

I have build this Bi-Polar stepper circuit and it works great on the bread board. I etch a board 3 times, 2 with heat transfer and 1 with photo resist. I use 2 completely separate power supplies but with the PCB's i think i am experiencing EMF or feed back. I am using an adjustable schmidt trigger for the clock. What happens is i can hardly get any rpm's, sometimes if the trigger is set just wright it will turn up fairly well, at that time if i try full step the stepper stops turning or will reverse sometimes, if i try ccw it may or may nor reverse and i may stay in what ever direction and any given time . Now at a slow rpm full step normally work's and cw/ccw may or may not work, also it can trigger the half or full step. Also i can disconnect one coil and the stepper will continue to turn, If i try that with the other coil the stepper will stop. So it looks like it is only running on one coil. Is that possible? Keep in mind i have bread boarded it and work's fine. Its just on the 3 pcb's that all this happens. As for the diodes it seams to work better with 1N4001's than 1N4148's I have checked every trace with an DMM and made sure they are going to the correct pins ect. I have tried different IC's L297 and L298 and steppers 2 different kinds and multiples of each. So i am thinking this is EMF, feedback, or a design problem. I have tried to eliminate most every thing i can think of. Comments?     

Picture of Steppers and EMF.?
Check your supply grounding: only ground the motor supply to the 5V supply at a star point. I suspect that the injected current from the motor upsets the logic.

You look to have adequate decoupling in the design.

Show me the actual board layout.
WWC (author)  steveastrouk2 years ago
Hi Steve
I havent for gotten to reply and pick best answer, still in progress. Hopefully fairy soon. WWC
The real fun starts when you try to run two stepper motor drivers at the same time. Getting one to work right is nothing! At least you save on barber bills. You'll pull all of your own hair out.
WWC (author)  steveastrouk2 years ago
The board.
btm.bmptop.bmp
WWC (author)  steveastrouk2 years ago
Again
Screenshot_1.png
WWC (author)  steveastrouk2 years ago
If you can not see it very well tell me and i will try and put on an imaging site.
Screenshot_4.png
pfred22 years ago
The schematic calls for BV27 diodes which are ultra high speed devices. 1N4001 diodes are regular speed. Until you meet spec there I don't hold out much hope for proper operation. Once you've done that if you are still experiencing problems then optically isolate your step and direction lines and your 297/298s should do whatever it is that they're capable of. I doubt any of the other "solutions" you've gotten so far are going to help you any. One more thing, tie off your half/full control line. Full stepping is of no practical use for most applications. Finally for a bench clock to test stepper drivers I use a 555 timer.
nobykobe pfred22 years ago
I am going to use 555 timer as you suggested with l297 as in this circuit (http://www.ohmslawcalculator.com/555_astable.php)
is R1=R2 =5 kohm, c1 = 0.1 mf,giving Cycle Time: 1.039ms,Time Low: 346.500 μ s enough because i red in other posts something about acceleration and i did not get it well.
Lose the big encircling ground line, split the grounds, so that IC31's heavy ground currents don't flow in the same lines are IC 21.

Also, steppers HAVE to be accelerated to high revs - you can't jsut whack a HF clock on them and expect them to catch up.

Steve
WWC (author)  steveastrouk2 years ago
OK i will try it. Will get back later.

Thanks
Wayne
Nice board work though, by the way. Very tidy - its just some of the tracking, whilst it looks nice, could well inject noise
WWC (author)  steveastrouk2 years ago
Thanks. The design is from the website on the board. Its and older German site, has lots of great info . The big encircling ground line is actually the motor Pos line. What i figure you are saying is cut the gnd traces to the IC'S. Bring them out and tie together. All the remaining gnd traces will be on the motor side. The pos and gnd's will be separate for there respective power supply's. Then connect the gnds at the " start " where everything first comes together. Is this a fairly correct description? I tried to design a board in Eagle, but hard to make a one sided board without many jumpers. Even a 2 sided i couldn't get away from 1 jumper. But PCB design is still new to me, not much experience to draw from. I am glad i started using photo resist. Before that i had about a 1% succes rate with heat transfer.
iceng WWC2 years ago
Run separate +5 and gnd (4) wires to the supply and add a bead and
another capacitor....   .   .   .   A
CKT.png
nobykobe iceng2 years ago
Do i connect both of the grounds while running separate +5 volts supplies . ?
iceng nobykobe2 years ago
You need to run separate ground wires in order to avoid large motor ground return currents from destabilizing the L297 IC21.

a
WWC (author)  iceng2 years ago
+1
iceng WWC2 years ago
:-)
Yeah, you've got it pretty well nailed. Basically, make signal currents flow in the signal supply lines, make motor currents flow in the motor supply lines, and DON'T let them mix ;-)

So, as you say, take the supplies directly from their respective supplies, tie the grounds together at a convenient point, but again check carefully that no motor current will flow in the common line. Its called a "star ground"

Moving to photoresist - you'll never move back ! What are you using to etch ? We swtiched to cupric chloride, and its giving fantastic results.

Steve
May you update the above circuit picture as i am kinda lost in all these posts ?!
Thanks