THE Bipolar Stepper Driver. Have you seen it?

[irving2008]

... I don't understand what your talking about with the JEDEC standards.

Thats exactly my point... the TISSOP package heatsinking needs are very critical and the 28degC/W thermal rating is measured under strict conditions, one being the area of copper ground plane used as a heatsink and the number of vias that act as heatpipes from the contact pad to the ground plane. By shrinking the board size you've compromised your heatsink performance dramatically.

I looked at the A3979 and I finally chose the A3977 simply because it stopped at 1/8th step as I feel like 1/16th step is way overkill. As you would need a machine that was very precisely made for it to be worth it.

Disagree, depends on desired resolution and screws used, nothing (directly) to do with how well machine constructed.

I however didn't check out the Rds on...that does make its mosfets more efficient I'll have to think about that one.

Rds is your enemy, minimising it is always a good strategy, Esp on the brushless motor controller I'm designing at the mo for 250A continuous, 300A peak! The MOSFETs for that are 0.02ohm and I'll still need 10 or more per channel if the heatsink is going to be
smaller than a football pitch

[~andy]

This will be of more interest to the 3D printer market I think.

It's worth sticking some optocouplers on that breakout board to cover your arse.

[Jonathan]

the brushless motor controller I'm designing at the mo for 250A continuous, 300A peak! The MOSFETs for that are 0.02ohm and I'll still need 10 or more per channel if the heatsink is going to be
smaller than a football pitch

You're not the only one. What's that for? I'm guessing electric car, or reasonably high power mobility scooter?

[JAZZCNC]

You're not the only one. What's that for? I'm guessing electric car, or reasonably high power mobility scooter?

. . . I'm guessing a very fast Kettle. ..

[irving2008]

You're not the only one. What's that for? I'm guessing electric car, or reasonably high power mobility scooter?

Let's call it a mobility aid... Scooters can't take rehab seating, go off-road and have 4 wheels...

What's yours for and would you be interested in collaborating or at least sharing ideas?

[Jonathan]

What's yours for

Hard to say, it keeps changing.

would you be interested in collaborating or at least sharing ideas?

Certainly..

[NorDevX]

Wow 250A sounds like a fun project! I'de like to change your mind about needing 1/16microstep for a cnc or similar machine.

Lets pretend I have a screw that takes 10 revolutions to complete 1inch. A stepper motor that takes 200 full steps to complete one revolution.

If I use 1/16th step it will take 200(steps)x16x10=32000steps to complete 10 revolutions and go an inch. Now you just take 1/32000 =.000031inches which is your resolution. Lol that's a really high resolution. Even at full step resolution is good at .0005 inches...this will vary with your threads per inch/meter. Microstepping is definitly needed though as it makes everything much smoother and you have control over your resolution but I see no reason to go super high with microstepping. Unless you have 2tpi and want .0001 then it is very usefull...now I'm curious what microstep are all of you using on your machines right now?

[Jonathan]

Whilst I agree that the higher resolution microstepping being available with the newer IC is not much especially useful, that statement is largely irrelevant - the lower Rds alone is the reason to use the A3979, especially when you're trying to make this compact. Lower Rds means lower heat dissipation, which means less space is needed to extract the heat, i.e. the heatsink is smaller.

Unless you have 2tpi and want .0001 then it is very usefull...now I'm curious what microstep are all of you using on your machines right now?

Not really, as if you really need extra resolution you'd change the drive ratio between the screw and the motor as the accuracy of micro-stepping is limited. A lot of people here use 10mm pitch screws driven 1:1, so that's effectively 2.54tpi...but the main factors in choosing the microstepping valua are the maximum frequency your controller can output, the maximum feedrate you require and the value which causes the least resonance in the machine.

[John S]

Every time you microstep you loose torque.
So by the time you drop down to 1/16th and need to overcome stiction it will miss steps unit it has enough to make a sudden move so basically anything over 8 is a waste of time.

Also remember systems can't remember microsteps so a shut down and start up will go to the nearest full step regardless of where it was.