A4983 Stepper Motor Driver with Boost Regulator

I am planning to build a robot that will use two stepper motors (each 1A, less than 5V) using A4983 Stepper Motor Driver. I am using single LiPo battery (2C, 3.7 V, 860mAH) for power supply, and Arduino Pro 3.3V as Microcontroller. The problem is, in the specification I saw, A4983 needs 8-35V as VMOT supply. My motors run on less than 5V. Is this a good idea to use a single Pololu Adjustable Boost Regulator 2.5-9.5V to drive both motors? But this gives output up to 2A, is that okay?


That is not a good idea. The stepper driver will take care of the current not getting too high in the motor even if the supply voltage is much higher, so you should really try to get a higher voltage to the driver (e.g. use a 3-cell, 11.1V pack).

Your last question makes it seem like you don’t understand current draw. If you have a power supply that can supply up to 2A, say, and you connect something that wants to draw 1A, the power supply is not going to force 2A into your device. However, in the case of our boost regulator, 2A is the input current side. If you are tripling the voltage, you would get 1/3 the current out, even with 100% efficiency, which you generally can’t get.

You should also keep in mind that you probably need a supply that can deliver in excess of 4A since your 1A rating is probably per coil, and you have two coils per motor.

- Jan

Thanks Jan. I am not sure which are the wheels that fit with THE POLOLU stepper motors. Does Pololu have those sort of wheels? Could you please list all of them? Or is there any special technique to stick a wheel with stepper?

What about this https://www.pololu.com/catalog/product/688 ?

We just started carrying those steppers, and we don’t have a wheel solution for them yet. I don’t think the wheel you linked to is likely to be easy to mount.

- Jan


This isn’t a complete solution but, the smallest stepper motor has a 4mm shaft which means it works with our 4mm hub.

- Ryan

Thanks Jan and Ryan. I hope this all-battery.com/111volt-700m … ypack.aspx LiPoly (3-cell) battery is good enough to drive two steppers. Isn’t is able to supply output current up to 10.5A? Do you have any better suggestion?

I am still looking for a stepper motor (length <= 26mm, NEMA 14/16/17) that has a low rated voltage around 3 - 5 V. I will definitely power them up with 11.1V and limit the current. I learned that higher voltage than rated one gives better resolution.

This one is perfect pololu.com/catalog/product/1208 is good in size. But has a larger rated voltage 10V.

Please see my thread at http://forum.micromouseonline.com/index.php/topic,442.msg2771.html#msg2771. You will see what I meant.

I don’t have a specific better suggestion for the battery, and 10.5A is indeed 15 times 700 mA. However, you should keep in mind that the capacity might go down with higher load, and if you do actually draw something close to 4A, your running time will be under 10 minutes.

- Jan


I have a similar question to Reefat, except in my case, power supply is equal and not lower. And while I understand
the current supply issue, in my application, precision is my priority - not torque.

I have 6 x 2700mAh AA NiMhcells producing about 7.2-7.8V. I recently purchased 1x 7.4V 280mA stepper motor from Pololu
as well as the A4983 Stepper Motor driver. I intend to the power straight from the battery. And jumper the pins for VCC
at 5V as per the tech guide for control logic voltage. I am about to set this up, however, the voltage difference issue (particularly
as the battery gets flat over time) and peoples comments on this forum raised question marks in my head.

Is my power supply sufficient (7-8V)? Or does the A4683 really need to be supplied with 8-35V? What if it is a tiny bit under?
The stepper motor will be used as part of a PAN-TILT servo mechanism. However, trying to position a continuous rotation servo
proved difficult, hence the switch to stepper motor - at least for the panning.

I intend to use Port D2,3 on the AT324P for Step & Direction outputs. I read about the using a low frequency rate from your previous forums.
That is to create a 100 milli second delay to generate the duty cycle. For precision control (1.8º step), is the voltage issue sufficient? Will
current really matter here, assuming there is torque. Or would you advise I use the timer pins, as the SVP has three hardware timers? frequency?

Is there anyway of turning off the power to stepper motor when I dont need to use it. (is the ENABLE pin ?? ) Otherwise, the stepper motor
will draw large current to keep into position when it doesn’t really need to.

Does there need to be a power capacitor (100uF) at the power lines, due to back EMF when then stepper produces. Or does
the driver offer protection in that respect? As I have a host of other analog&digital sensors sharing the same power supply.

Some info for others who may be using the same stepper (no query on this bit:
For the 5mm shaft on this 7.4V stepper, I used the 6mm alum hub, and with a Dremel I shaved slightly the end where the slug presses on the shaft.
to make it straight, rather than round. It makes a good retro fit. Widening the hole of a 4mm Alum hub was also an option, but I didnt want to ruin the beautifully CNC’d smooth finish it came or risk having the hole not absolutely vertical. Thats a good quality stepper though. Good one Pololu!

Any advice would be appreciated.


We haven’t played with the A4983 under 8V much, so I’m not sure how well it will perform. In general, behavior might be unpredictable and vary from unit to unit, so I recommend going to 8 cells if you can. If you proceed with six, please share your results with us.

Any comments in other threads about 100ms delays were probably for people who were not getting any success at all; the point of going so slowly is that it is on a human scale, and you can still distinguish the individual steps. In actual use, you’ll probably want to go a lot faster. I don’t really understand the rest of your questions in that paragraph, but it seems like a lot of them would be addressed by just playing around with your motor and system.

Yes, you can use enable to cut down the power you use when idle.

In general, I recommend big caps on power lines to reduce noise from switching the coils on and off. Back EMF isn’t really relevant for typical stepper use. If you have noise on your power lines that is affecting other circuitry, adding capacitors shouldn’t hurt.

By the way, we have 5mm hubs in the works.

- Jan


You can also use the SLEEP pin to decrease power consumption. See page 8 of the datasheet for more information on ENABLE and SLEEP. ENABLE is pulled down through a 100k resistor on the carrier board and SLEEP is pulled up through a 100k resistor; these resistors will draw around 50 uA when you use these pins to disable the A4983. If current consumption is really important to you, you could desolder the pull-up resistor on the SLEEP line and use an I/O line to drive the pin high or low depending on the state you want the board in.

- Ben

Ben & Jan,

Thanks again for your advice.

I will play with the power supply and let you know. The RP5 chassis only has room for only 6 x AA cells(space restriction)
If it didn’t work, would a 9-11V 560mAh LiPo battery pack be sufficient for driving two coils 280mA at 7.4V?

In terms of generating PWM signal for the STEP input pin, would the SVP324 inbuilt hardware timers be a reasonable option?

50uA is not costly for the ENABLE or SLEEP pins.

5mm Hub in the works. Sounds good - when will this be available?


560 mAh should be okay; just make sure you don’t over-drain your battery.

You could do something based on the hardware timers on the Orangutan SVP, though most of the hardware PWM pins are taken up for other hardware. By the way, you don’t really need a PWM signal; you need to vary the frequency according to the step rate you want.

I think the hubs are already made; they just need final packaging and then to get shipped from China, so it will probably still be a few weeks.

- Jan

Timer 0 on the Orangutan SVP’s ATmega324/1284 is available for you to use, and so are its two PWM outputs: PB3 and PB4. You can read about it in the ATmega324 datasheet to figure out whether its frequency range is suitable for your application. --David


Thank you for your advice. Some findings tested and shared for anyone is interested.

The A4983 Stepper Motor driver, worked with as low as 6.5V, despite the minimum 8-35V on the data sheet.
At around 6.5V, the stepper motor did not move anymore (for a 7.4V stepper motor)
At around 7.4V with DIR, STEP & ENABLE pins active, the stepper motor and driver circuit draw about 0.65 A.

There were no side-effects noticed on the controls side nor any fluctuations on the driver side between 6.5V - 11.1V
as DC lab power supply was varied. Current didn’t change. The current draw would obviously depend on
the stepper motor current draw. The current adjustment screw on the driver board was not changed (ie as per manufacture preset)

I couldn’t figure how to use the internal timer/counter at PB3, as per Davids suggestion though. So instead I used a simple WHILE loop
with a calculated number of steps, and the duty cycle generation code followed by a i-- statement.
(Note: the for (i=0, i=__; i++) { …} loop doesn’t work as the program cannot exit the loop. (ie. motor keeps turning indefinitely)

A 4 millisecond (250Hz - 50% duty cycle) gave the best result - fast but smooth. 100 Hz is rough and causes the motor to vibrate its base. 500Hz
is about the maximum and did a full 360 in about 1 second. The direction setting, however, stopped working. I have yet to diagnose why this
is the case. It was working before.

This duty cycle was achieved by initialising the STEP I/O line as zero first, then driving to high for 2ms and then
LOW for 2ms. Works well. The enable pin also worked very well.

I am happy to provide the code if you have the time to look at it and can suggest improvements. I also noticed that after several tests, the
stepper motor did get very warm. I am guessing this normal, given that is consuming about 0.65 A at 7.4V (~ 5W). Please advise if you think


Getting hot is normal for the driver, but it’s not dissipating 5W since that is the total that goes into the driver and motor. Also, the current setting is not really preset by us; the pot just gets left somewhere around the middle when we test the units.

- Jan