[sp00k1e]

My Hans Wedemeyer L297/8 board just blew up. The magic electronic smoke escaped from the 298 when I turned on the power. So it is maybe time for a change. I have had no problems with this chipset even though I built it on veroboard and was planning on upgrading it to a prober milled pcb. I use a bungard CCD pcb Mill at work.
My question is what are the advantages over the 297/8 chipset, if any, of the other chipsets designs around. I fancy Alans LMd18245 PICSTEP board, but in the UK the chips are about £11 each (unless anyone knows a cheaper source) and will my machine run faster with more power or be about the same. I have looked at pminmos circuits maybe you could answer seeing as you have seemingly built every combination available? I was going to build your 3977 board but the cost of importing the chips from the US was prohibitative, as i cant find them over here.
I have 6 lead 2A steppers to control

Thanks Paul

[MrBean]

You can get the 3977 in the UK from Farnell.

http://uk.farnell.com/jsp/endecaSear...KU=4126075&N=0

£3.72 ea. Not sure if tat includes VAT though.



Regards Terry.....

[pminmo]

Paul,
What is your power supply voltage? What is your coil voltage or resistance?

Phil

[andyew]

spookie Paul,

as mrbean Terry said, Farnell (Newark actually) has em in stock..... A3977SED is listed there quite cheap... from my part of world, it takes 2 weeks from order to postman reaching my house since it's from Newark US.

Good luck and phil is really helpful, he will guide u the right direction, just bear with him sometimes, look at his signature u will know...... :P
Andy

[sp00k1e]

Yes Farnell has the 3977 in stock, but since they are from Newark the postage is another £13, not too bad i suppose if they are what I want. My power is a rectified smoothed 25v toroid so about 35v, but that could easily change, and the X and Y motors are Sanyo Denki 103H7126-0440, I cant find the specs at the moment but I seem to remember about 3.5V, rated 2A on the case. The Z at the moment is a 6V printer motor.

[pminmo]

Under 2.5 amps and 35V the 3977 is a great chip. But you have to be able to solder well and very fine. The L297-8 is my second choice. One the pricey side is the picstep board, but the driver is a gem.

Phil

[MrBean]

Yes Farnell has the 3977 in stock, but since they are from Newark the postage is another £13

That's not good. All the orders I've placed with them have not had postage charged. Maybe I've just been lucky.

Regards Terry.....

[sp00k1e]

I think then I will try and sneak a few 3977 on my next order and try and hide the postage cost :devious:

[sp00k1e]

Well I have managed to get my 3977, almost double the cost due to the postage from the states, but delivered in less than 48 hrs.
I have milled one board to try, a slightly modified version of phils circuit so that I could transfer it onto the PCB mill at work. only a few more components needed and then we will see what it can do. You can see I lost a couple of pads from the top of the chip when I was trying to clean up the remaining copper, a couple of wire links will sort that out. Next one I will do a full mill to remove the excessfrom around the chip.

Paul

[pminmo]

spook1e,
Looks good, keep us posted.
Phil

[H500]

The board looks good. 35v might be bordering on the high side. When the motors are decellerating they put out back emf that could potentially cause the supply voltage to go too high and zap the 3977.

[sp00k1e]

Well ive built the 4 axis breakoutboard and one drive, and now have it almost working, when I enable the drive the motor locks as expected, but it just very quietly whispers and wont move, it is probably a fault on my remanufactured boards but just to check, and before i can get my scope from work, the double sided version circuit has a 10k pullup on the dir signal but the single sided doesnt. I understand that in general the printer port is TTL but it may be open collector, but i guess the 7414 should sort that out shouldnt it?
Paul

[pminmo]

No R5 on the DS version shouldn't be there, the SS version is correct. It was intended to be a pullup for the Dir signal, but was drawn on the wrong side of the 10K series resistor. It should have pulled up pin 3 of JP1. I just eliminated it on the ss version, and don't install it on the double sided version. Printer ports vary, but usually have an active high side drive. My thought process was a fall back postition if I needed to have a termination.

Phil

[sp00k1e]

Pminmo.
I found the problem after I got the scope. Running Mach2 the minimum step pulse width was too small, looking at pin 5 of ic1 the filter of R17 and C14 was causing a very small triangular wave which wasnt allowing the input to the inverter to be enough to operate. Is there a reason to have such a large RC time as this must limit the max step rate available ( not in my case as 5uS pulse width is fine). I put a 330R in parallel with the 10k and managed to get the motor whistling at 1uS pulse width.
On a side note what is the best setting to have the microstepping? on full step the motor whistles although on less torque, im testing with a loose motor on the bench, on 1/8th step the motor top speed is significantly slower, I assumed that since Mach2 operates at such a speed the max motor speed wouldnt change or maybe increase and the torque would increase.
anyway thanks for the circuits, this forum is great, plans for everything, JGro and you deserve a medal !

Paul

[pminmo]

The RC constant on the step input is not considered long vs a stepper motor. The reason to keep it longer is improved noise immunity. Most stepper motors really drop off torque wise above 1khz at full step, so even if you are screeming along at 16K pulse rate with 1/8th microstep, thats only 62.5us pulse rep rate and a 20us pulse flys through the filter.
Sounds like your board is running? Can you post a picture? Please, Please nobody has posted a completed picture of the ss board, make my day. :-)

Thanks for the medal. :-) The old saying "giving is better than receiving" is so true, I do enjoy helping! Big kudo's to ynneb (Benny) though, he through out the line, and I swallowed hook, line, sinker and all.

Phil

On the microstep question, depends on what you wind up with as a problem torque vs resonance. Microstepping reduces the resonance problem but lowers the torque.

[sp00k1e]

I was just about to go to bed but here it is!
Please understand that i am just getting this running and it wont be this untidy in the end, and I have just removed the previous circuits from my box, .
Regarding the microstepping, I connected the machine up and discovered that the top speed doesnt seem to change when under load irrespective of the microstepping rate but the smoothness does and presumably the accuracy.

Paul

[pminmo]

Paul,
Thanks, great work. Untidy means your doing something.

Phil