New Pololu DRV8825 Stepper Motor Driver: A4988-compatible with more power!

Hello.

This is Ryan from Pololu Robotics and Electronics. I am excited to announce the new DRV8825 stepper motor driver carrier, a higher-power alternative to our A4988 drivers that can be used as a drop-in replacement in many applications. The DRV8825 module can deliver approximately 1.5 A per coil without a heat sink (rated for up to 2.2 A with sufficient cooling), and it offers a wider operating voltage range (8.2 - 45 V) and 1/32 microstepping.



Though this carrier has the same dimensions and virtually the same pinout and interface as our A4988 carriers, there are a few key differences. Please see the bottom of the DRV8825 product page for more information.

We will also be releasing a version of this board that uses larger current sense resistors for improved microstepping performance at lower currents. That board will be able to deliver up to 1.2 A per coil and should be available next week.

Sample Giveaway

We are giving away some free drivers to qualified testers who can give us feedback on how this driver performs as substitute for the A4988 carrier. For more information (and to apply), please see our forum thread.

- Ryan

Perfect timing!

I was looking for something like this. I've just tested (and liked) GRBL shield from Synthetos and I wondered why there was not a higher power Pololu. Unfortunately it is a three drivers only board, not suitable for 3D printers but for 3 axis CNCs.

DRV8825 is an awesome driver! That is the one I am using on my 3D Printer and also on my CNC Plasma Cutter. Will eventually use it on my CNC router as well.

On my 3D Printer I am running Sprinter and Replicator G. I am too knew to this 3D Printing phenomenon so went with what I had heard, but will eventually jump to Repetier and Repetier Host, I think.

Here is a video where you can see my DRV8825 based board which plugs on top of an Arduino Mega 2560.

[www.youtube.com]

JIQ

Hello Ryan,

What kind of tests do you want us to do, and more accurately, what kind of data return do you expect ?

Thanks for your time.

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

DeuxVis,

We are trying to get more information about how it is to use the DRV8825 as a substitute for the A4988. We are also interested in general feedback about using the new drivers from experienced users. I've already verified the basic functions of the board in isolation, but it is good to get feedback about how it performs in actual applications, and what, if any, differences or problems are encountered compared to the A4988s.

- Ryan

I've just received my units in the mail. I'll be reporting soon.

My first impression is that this new driver from Pololu is the response to those of us that were unhappy with the performance of our direct drive extruders because we could not get more current out of our electronics without active cooling. This new board does a great job delivering 1.5A with no active cooling.

I see it as a good replacement for the A4899 drivers for the extruder. Other axis may not need the oomph.

Just remember that if you replace you driver, the new one will be configured to x32 microstepping unless you change your jumpers, so a change will be needed in the steps/mm of your firmware.

Hi all!

Just a reminder that all of Pololu's stepper motor drivers and some of our stepper motors will be available at a big discount during our Black Friday Sale, which starts tonight and runs through Monday. Here are some of the more relevant deals:

A4988 driver (original/green version) for $7.95 (39% off)
Black Edition A4988 driver for $9.95 (41% off)
new DRV8825 driver for $10.95 (45% off)

For more information, please see our Black Friday Sale page

- Ben

Misan,

What's the resistor attached to the FAULT pin for? The website says it was designed so it won't be bothered by a 5V reference signal, and I thought that most of the boards out there just directly attach that pin to the reference pin on the MCU. Did you hack the board so you could read from the pin, or is there some other advantage I'm missing?

Thanks,
Chris

PS. nice heatsink on the one beside it.

I've used a 4K7 resistor from /Fault pin to /Sleep pin (which is connected to /Reset in my board). According to the specs the /Fault output is connected through a resistor to the pin, so that output is not damaged when connected to +5v. Unfortunately, I am not taking advantage of the Fault signal, but this way this makes a drop-in replacement for A4988. I am very satisfied with the result, that clearly runs cooler than A4988 for the same current.

Oh, okay. My understanding from the website was that you could just drop it in to slots designed for A4988 and it would work, without any additional modifications.

I won't do that, as then pins /SLEEP and /RESET would be on a floating state which could cause problems unless the circuit you plug it in does set a stable logic level for these pins. In Sanguinololu board there is just a connection between /SLEEP and /RESET (as A4988 does have a /RESET pull-up).


the RESET and ENBL pins are both pulled low by default through internal 100kΩ pull-down resistors. This means that RESET and SLEEP will be preventing the driver from operating if left disconnected. Both of these pins must be pulled high to enable the driver (they can be connected directly to a logic “high” voltage between 2.2 and 5.25 V, or they can be dynamically controlled via connections to digital outputs of an MCU).

Aah, I didn't make the connection about the /RESET pull-up on the A4988s. RAMPS is built the same way, with /RESET and /SLEEP connected, so when I get one of these I'll be sure to add the resistor.

Thanks!

TopherMan: If you look closely at the picture on the first post you will notice there is a component missing between a capacitor and a resistor near the bottom edge of the board. If you can solder a SMD 4K7 resistor there, you will get the same result as I did but I am not sure what size it is (603? 402?)

Edited 1 time(s). Last edit at 12/27/2012 05:20PM by misan.

Misan and TopherMan,

What Misan says in his last post is not correct. The exposed pads connect SLEEP to 3V3, but the 3V3 line does not turn on unless the board is taken out of sleep mode. We are coming out with a revision to the board that will make it unnecessary to have that external resistor.

The not populated pad is for 0603 components.

- Ryan

Oops, my mistake. I assumed the 3v3 would bo on to act as a source for pull-up for /SLEEP and /RESET (as they both are connected in RAMPS board) but I did not realize of that catch 22.

Thanks Ryan.

I am building my first reprap prusa mendal and had some trouble getting the step sticks to work properly with my oversized stepper motors so I bought some of these, hopefully they work as expected. My steppers are 2.3 amps/coil.

Most RepRaps are not powering their motors at their specs current. (ie. recommended motors are 1.6A/coil specs, but they are usually powered below 1A).

I think you can go with 1.5A and see what happens.

I have use them for powering a homemade CNC machine 2000x1200 mm using nema23 motors successfully (reaching reliably speeds of up to 8000 mm/min).

I cannot see why your oversized printer won't work (unless your design calls for the maximum torque of your motors). I do like to work whenever possible without the need of active cooling, but if you provide active cooling youy can reach 2.2Amps :

Quote
The DRV8825 driver IC has a maximum current rating of 2.5 A per coil, but the current sense resistors further limit the maximum current to 2.2 A, and the actual current you can deliver depends on how well you can keep the IC cool. The carrier’s printed circuit board is designed to draw heat out of the IC, but to supply more than approximately 1.5 A per coil, a heat sink or other cooling method is required.

I got 2 of the drivers to work properly, the third just vibrated and didn't turn, and the last didn't respond to input but initialized during connection to pc.
Also they wouldn't turn backwards but I think that's because my endstops weren't installed yet? I have ordered some arctic silver thermal expoxy and MOSFET heatsinks if I need them.

I am not familiar with the failure mode of these modules (thankfully all mine are still alive), but usually a stepper vibrates mostly for one reason: only one coil is working. This could be due to a cabling problem (one coil short or open circuit) or to one of the H-bridges on the chip is no longer working.

Sometimes, if current setting is too low for the motor to turn, it will vibrate too (but not loud). If current is low enough, the motor will do nothing.

You should check the current settings for each motor that it is adjusted with the on board pot. Voltage on pins 12 & 13 on the chip (not on the carrier) will tell you the current setting: Current Limit = VREF * 2

Make sure all your drivers have the same current setting that you find on those axis that work. If you need more than 1.5A then coolers and fans will help you keep the electronics from overheating.

BTW, if a chip is overheated, thermal protection will disconnect the motor to prevent the electronics to release the magic smoke inside. After a moment, the chip will cool down a bit, so the driver is reactivated, and the heats up again and it is shutdown again. This cycle may happen quite quickly and it may look like the motor is vibrating too. You know this is the case because the chip on the driver will be burning hot.

Thanks for that info,

I set the current using the pot to the same level as the working drivers and it just vibrated, increasing or decreasing the current simply increased or decreased the intensity of the vibrations. I used a known working motor on it as well. Same result.

It is worth noting that I had applied the heatsinks that came with it with some brand name thermal compound, but that did not stick and some thermal compound ended up touching the leads on the chip. This happened to a certain degree to all 4 drivers. I tried cleaning it as best I could, then I tried to make the drivers work. After some troubleshooting, drivers 1,3 were working and the rest unresponsive but initializing (twitching) during connection to pronterface. I then washed the boards with soap and dried them in front of a heater. After trying them again drivers 2,3 are sort of working (moving only in 1 direction and not periodically not obeying commands), driver 4 is vibrating as described previously and driver 1 is completely unresponsive save for the initializing twitch. I spent another hour or so trying to make them work but then decided to order some pololus as I have done now. I never saw or smelled any magic smoke.

Keep in mind I didn't have any endstops, thermistors or bed attached at the time and cannot work on the printer until I return home on Monday night. All the missing parts have arrived now so I should be able to finish her up

Edited 2 time(s). Last edit at 12/29/2012 10:01AM by Inakto.

Some thermal compounds are electrically conductive too which, as it seems your case, may cause shorts.

As the chips are thermally protected, my advice would be to used them without a cooler in the first place. Extra cooling may or may not be needed depending on the drive current needed. Don't assume it is a must. I do not know of a good binder for sticking the heatsinks (I've used superglue in the past but I do not think it will last for a long time nor I am sure which are the risks involved if any).

Please note that Sprinter or Marlin firmware will refuse to move backwards if end-stops are not wired.

Marlin firmware may not respond at all if there is not a thermistor connected... though a proper message is sent to the host software stating so.

Ahh we'll that probably explains why those 2 working drivers were behaving weirdly... I even started a print but only the z axis seemed to respond according to the Gcode.
If I end up using the heat sinks on the new drivers I will mask off the top of the driver chip.

Hey, I just wanted to update this, I am using 4x these drivers on my sanguinololu and they are amazing, whisper quiet on 1/32 microstepping. Not using a heatsink, just blowing some air on them. Honestly, my fans are louder! Totally recommended

Hi, all.

Thanks for the positive comments Inakto!

We recently released a new version of the DRV8825 carrier that adds a resistor between fault and sleep, removing the need to add the external resistor when used where you normally use an A4988 carrier. We also have a version with the DRV8824, which is less expensive and provides less current, but still has 32nd-microstepping and should have better performance with low current microstepping than the DRV8825.

- Ryan

Edited 1 time(s). Last edit at 01/14/2013 05:30PM by RyanTM.

I've got the new DRV8825 (model b) with my A4988 I would get 12V in reference to ground on my RAMPs 1.4 but with the DRV8825 I don't any suggestions? I tried the Tying the Fault to the Sleep.

Hello, pastlifeuknow.

This isn't a particularly good thread for technical support. I recommend posting your technical questions on the Pololu forum. That said, I do not understand what problem you are facing. What two nets did you expect to measure 12 volts between?

- Ryan

Edited 1 time(s). Last edit at 02/05/2013 12:36PM by RyanTM.

Hey, anyone use this drivers? any reviews? will they work plug n play with ramps?(b version)

Not a longish review but ... [fightpc.blogspot.com.es]

New version (which I have not tested) adds the resistor I have to add externally to make it totally a direct replacement of Allegro-based driver. Just remember that the micro stepping will be by default then set to 32 (instead of 16). If you do not want to chance your software configuration, use the jumpers to set it back to 16 too.

Must. Have. Many.

Great lil device. I'm sad I'm just now learning about it. But happy to have found em.