MakerBot Industries

Hi
Thanks for your answer. I had a look at the page you linked to, it looks promising. Which specific one did you have in mind?

And just to get it all straight, am I right in my assumptions in my first post? That the driver can only deliver 1,25A? Because the guy I bought them from told me that a 2,5V voltage over the REF pot would give me 2,5A to the stepper motor.

Thanks.

http://wiki.makerbot.com/thingomatic-doc:stepper-driver-3-3

The chip is rated at an absolute max of 2.5 amps (Makerbot says 2.8????) but that's properly heatsinked to the thermal pad on the bottom of the chip. The PCB Makerbot designed is NOT properly heatsinked to the specification. Thus, even if you add a heasink to the top of the chip, it will not be effective since the heat is actually going down towards the PCB. There SHOULD be tons of throughhole vias (just like under the mosfets on the 3.6 extruder controller) to connect through the layers of the PCB underneath the center of the chip and then a heasink on the bottom of the PCB, or a hole through the PCB that allows the heatsink to physically touch the metal pad on the bottom center of the chip but that's not how the board is made. You will overheat and the chip will fail or go into thermal shutdown.

Now, as far as the ref voltage and current relationship. Ed Nisley deserves the credit for writing it up first but we discussed it in this thread http://wiki.makerbot.com/forum/t-358724 Be sure to click on the link in Ed's post.

Basically the ref voltage is double what the current is in Amps output, thus 2.0 volts REF is 1.0 Amps. 2 Amps would be 4 volts. The 2.4 amps you want to set it at would be 4.8 volts Ref, which is not achievable without changing the pot, and even then will not work because the chip isn't heatsinked.

Good job on your research from the data sheet, testing has shown the REF voltage to current realtionship to be what you calculated and thus YES, Makerbot designed the pot to current limit the driver since the chip is not heatsinked. Oddly, I think I have set one of these to REF of 3.38 volts to drive a larger NEMA17, thus think it's a fluke yours only reach 2.5 volts REF, but MBI might have changed the resistor values during the run of them to prevent users from putting too much current. These were designed to run a Makerbot motor, and those are all way less than 1.5 Amps.

Now to the linked page. THIS driver is the cheapest one rated at 2.5 amps http://www.lightobject.com/1-Axis-CNC-Laser-Machine-Stepping-Motor-Driver-TB6560-P451.aspx
It will properly drive your steppers at just slightly over the rated current of the motor (they have a fixed 2.5 amp output by switch, not a pot). 2.5 AMPS is only 4percent over the motor specificaton. Whiel the motors would be warm, they should be just fine and definitely reach full torque.

However, you may want to ust get even large drivers to "make sure" so go with these
http://www.lightobject.com/2-Phase-45A-Stepping-Motor-Driver-Support-to-Nema-1634-Leetro-replacement-P453.aspx rated at 4.5 Amps, or extreme overkill 7.8 amp drivers http://www.lightobject.com/2-Phase-78A-Stepping-Motor-Driver-Leetro-replacement-P454.aspx

Hi again
Thanks a lot for your thorough answer. I made a new measurement over my REF pot, and sure enough, it went all the way to 3,38V. I don't know why I only got 2,5 in the first test. But 3,38V (1,7A) is still not enough because I need to squeeze every last drop of torque out of my motor, so I have ordered a new driver from LightObject (the cheapest one you linked to). It's still on its way over here to Denmark so I haven't tried it yet. I might write you back when I get it regarding the setup of the optocouplers if I can't figure it out.
Again, thanks a lot for your help.

Glad I could help. Basically, from gen4 motherboard, steal 5v power from the endstop connector and feed that as a bus to all 3 of the stepper opto inputs on the driver, then the low pins of each input on the stepper driver get attached to the step and direction pins since they "sink" low. Thus you've "sourced" the opts with 5V power and sinked them through the outputs of the arduino.

Ahh, that makes sense. Since I'm using an ATX power supply with both 12V and 5V in the same plug, is there a reason why I couldn't just use the 5V from there?

ANY 5v source will work as long as the grounds are tied. It's not likely to cause any issue but just be aware of ground loops. A ground loop occurs in an electrical system when not all components are tied to the same ground. In other words, you are powering the motherboard and it's grounded to the common ground of the PSU. In your case, the source 5V is fed to the stepper driver opto input and the return path goes to the arduino pin which then is sunk through the common ground plane of the motherboard back to the power supply. This is perfectly fine. What would get very troublesome is a separate supply for the stepper driver power being used if you didn't connect the two grounding systems together. In other words, when I hooked up the Gecko servo drivers, I used some HP printer power supplies that had multiple voltages and could have powered the opto input from that and would have had to tie the 32volt servo supply ground to the main board supply ground, otherwise, there would be no return path.

I'm assuming your circuit it much simpler and you are just using a single supply and won't have any issues at all.

Hi again
So, long story short, I recieved the new driver, mounted it and everything worked fine. So thanks for your help on that:)
Now I seem to have another problem that you might be able to help me with, you seem to know your way around stepper motors:) With the new driver from LightObject attached to my motors, the motors emit a sort of hissing/humming/buzzing sound that seems to vary a bit in frequency (it sounds a bit like a high pressure air leak). That kind of sound was not present at all with the MakerBot drivers. The motors also have some slight mechanical vibrations that follow the sound. I tried with both a Nema 17, 2,4A, and a Nema23, 2,8A, with the same result. The noise is only present when the motor is enabled and standing still. The driver is set to 2,5A, no microstepping and 0% decay mode.

I hope you can help me, if it's is actually a problem at all, as the motors seem to run fine, although a bit hot.

That hissing is the current limiting portion of the driver. In other words, the motor coils are X resistance (on "good" motors, really low), thus the source voltage would exceed the current rating so the driver is PWMing the outputs to limit the current to the set level. The reason we do this is the motor is an inductor and by pushing higher voltage and fighting the "knee" curve of current rise in an inductor over time. Basically, the hiss it what gives you the torque and speed of the motor. More expensive drivers like those from Gecko Drive http://www.geckodrive.com/geckodrive-step-motor-drives.html might switch the PWM at a rate higher than human hearing. Basically your dogs and your wallet would be upset with the more expensive drivers.
Geckos are twice the price each of the drivers you got. Sorry, that's just the tradeoff of cost. I honestly am annoyed by it too, but between all the other hardware running, I barely notice it. That said, the noise only happens when the stepper is locked still, and really it should be disabled (not locked) by the enable pin when the bot isn't building. I think you just need to fix the enable pin logic. As stated, these drivers are reverse logic on the enable pin only, thus no current into the optoisolator=enabled. I thought most if not all the logic coming out of the motherboard is logic low, meaning they short to ground for a high signal. So for the enable pin only, the wiring would be connect -side of the opto input to ground (motherboard ground) and the + side of the enable opto to the enable output pin on the motherboard for each axis.This effectively inverts the logic signal from the motherboard as required by the driver. In other words, I think when you boot the motherboard all steppers are disabled so the enable pin should be at 5v compared to motherboard ground in this state. Then if you use either the LCD control panel or the CP in Rep-G, the steppers become enabled after the first movement and so the logic level on the enable pins should be at ground.

Hey, FYI, I just got an amazing X-Y gantry Friday from LightObject.com for a song!!!
These are amazing http://www.lightobject.info/viewtopic.php?f=45&t=1155
So yes, I actually used the same drivers and setup you are last night for my intial motion tests. Think of the print you could make with this system (and probably cost equal to or less than the replicator).
For me, the drivers really weren't that loud but maybe I've lost that part of hearing from my other toys, namely the turbine engines.
<iframe width="420" height="315" src="http://www.youtube.com/embed/zPTvoabcU-k" frameborder="0" allowfullscreen></iframe>

If you want silent, servo is about the only way.
http://www.mouser.com/Electromechanical/Motors-Drives/AC-DC-Servo-Motors/_/N-uwj6?P=1yztqh1

Or, I used Gecko 320s and motors from HP printers here.
http://www.thingiverse.com/thing:6814

As you can see, the cost is several orders of magnitude over even expensive stepper based systems. Plus, don't forget you either need linear, or rotary position sensors.