Thingiverse - Things you can make with a Hot Air Rework Station

This is my replacement for the Makerbot Stepper Motor Driver.
Project is still under development. Currently finishing up the layout.

UPDATE @ 10th of May:
Did some little modifications to the design, and it's now off to the manufacturer. Ordered a batch of 30 boards, because it would not be much more expensive than just odering a few single boards to test. They will come in black, and without any overlay printed on them (Text is etched in copper). They should arrive in 2 weeks from now. So, if everything works (knock on wood) i will have some bare pcb boards i'm willing to sell to you.

What changed since the initial release?
- Added some LEDs to indicate 5V and 12V supply status.
- Added a footprint for a 1206 Polyfuse in the 5V line (can be bridged by a 0 Ohm resistor if not needed).
- Moved the "home position indicator" LED more to the left, so it would be less likely getting covered by an optional heatsink on the IC.
- Exchanged the power-connector with the same screw terminal that is used for the motor connector. (Both versions would fit now)
- Tryed to make the top-layer copper-surface under the IC as big as possible to give better heatsinking.
- Some minor graphic and cosmetic layout modifications
- Something else i probably already forgot...
Files in here are not updated yet with these modifications!

Compatibility
The controller-interface (10-pin header) should be compatible with V2.3 and V3 Stepper Drivers of the current MakerBot electronics.
Should offer the same basic interface for two endstops, power and motor connection.
UPDATE: I did just recheck compatibility. For this driver to work with the current MakerBot motherboard it is important to leave away some of the ferrite beads, because some of the signal lines i use on the 10-pin header are connected to GND on the V1.2 motherboard, and that would mess up some settings. (SR, MS1 and MS2 signals)
Not that hard to figure out, but i probably should make a note in the schematics or an assembly variant where these ferrites are not populated.

One thing that is a bit of a trial here is that there is no 12V->5V linear voltage regulator on this board. I wanted to make use of the ATX-powersupplys 5V rail. I know, some people seem to insist that everything should be running from one single powersupply (car battery or something), but for me, i would rather have an external compact, efficient switched DC/DC converter that does the 12V->5V transformation for the entire robot, if it ever comes down to that.

Board Spec
Board Size is at 35x75mm, so it's quite small and compact. (Board prices get calculated by surface, you know...;) ) First i wanted to go full-on with a 4-layer board, but i tossed that idea, because that would instantly double the price of the board. Would have had some advantages, but i think its currently not needed. Another cool thing would be an Alu-Core-PCB, so you could really crank up the max power of the DRV8811 with a nice heatsinking board, but that gets quite expensive too, if you cant order a huge amount of boards.

Via size is 0.3mm drilled, min track space 0.25mm and min width is 0.3mm. Should not be too hard for most manufacturers to do that.

Components
The main part is the DRV8811 Stepper Driver/Controller from Texas Instruments.
focus.ti.com/lit/ds/symlink/drv8811.pdf
It is capable of 2.5A continued winding current (8A peak rating) at up to 38V. You can get if for about 5-6$ per piece (Digikey).
It does have four MicroStepping modes up to 1/8, and well, all the basic goodies you would want to have in one nice package. ;)
Seems to be THE alternative chip to the Allegro circuits if you ask me.
(As far as i can tell, availability is good, and TI is always nice to me when it comes to samples and design advice.

Just for you guys out there thinking that this will crank out the 2.5A to the motor with ease: Wishful thinking, unfortunately. At least not without extra heatsink and/or cooling fan. Package with a good power-pad connection to the copperplane cant dissipate enough energy to make up for the losses. The TI chip has a rather good on resistance of 0.5 Ohm, but still, thats not gonna do it. The datasheet says the package alone can support roughly 2W-3W under normal conditions. Wanna have some simple math?
P = 4 * Ron * (I rms)^2
(Factor 4?: To operate both coils of the stepper motor you have to have a total of 4 FETs in the two H-bridges conducting.)
Now, lets say we have a I rms of 2A, to get some simple numbers, we would already have a whopping 8W of power to dissipate in the circuit. So, good luck getting that power away from that small chip without some solid heatsink. If you really need more power, it's probably better to crank up the voltage to 24V, with the side-effect that you cant profit from the extremely cheap atx powersupply anymore.

The rest of the passive components on this board are mostly 0603 sized. For me this is still very easy to hand-solder. (Revision 001 was originally 0402-size based, but i changed it to make it a bit easyer to correct stuff, should there be any errors in the design.

What else did i change?
Motor connectors are direct (euro-style) screw terminals (3.5mm pitch). Why? Less losses, more accurate feedback to the driver-circuit, more reliable connection, and circuit manufacturer recommends them and said to avoid pluggable terminals if possible, so, there we go...
Power-interface follows the good old ATX layout, but i changed it to some euro-style pluggable interface, uses less space and i actually like them and use them in almost all my projects. Did not want to replace them with the same screw terminals the motor has, but that would be an option to consider.
Endstop-connectors are some cheap generic Tyco connectors with a 2.54mm pitch. Can be replaced with something else if needed, should be plenty of compatible stuff out there.
The controller interface is expanded a little with some extra pads that would allow to use a 14-pin connector and have some aditional GND wires in the cable. (Have to add, im not really happy with the signal-layout of the 10-pin connector, but i copied it for compatibility)

The biggest change over the V2.x and V3.x MakerBot stepper driver electronics is definately that i tryed to include some filtering and protection devices and added a lot of capacitors to keep the entire setup as stable and reliable as possible. Im not shure if it is actually necessary, but having the footprint there in the first place is already worth a lot if there ever is an EMC problem. I did not go full-on here either, as i left out the series inductors for the powersupply and motor connectors, if you would want to get that stuff completely CE certified that would probably be necessary too, as the fast high current switching definately generates quite a lot of noise that would have to be damped to avoid radiating the noise over the wires. Problem is these high current ferrite beads and common mode chokes use up a lot of space and do cost quite a lot, and i had to cut the expenses somewhere. Well, maybe thats something we include in a later revision...

To do...
What i am working on right now? Getting the actual layout to near perfection for the best performance possible. So, we want to have as much as possible of the bottom layer used as GND-plane, no capacitors sitting at "dead ends" of the signal path, and all the important components as close to the controller as possible. Looking good for now, but tomorrow is another day.
Or, maybe someone gives me some feedback and i change some minor stuff.
And, yea, you cant read datasheets and appnotes ever so often. If its your money that goes into a manufactured board, you want to be on the safe side. ;)

License information:
Currently all rights reserved until test of the prototypes has been successful! (Will be less restrictive when i release it)