Ways of Driving a High Power Stepper Motor at 24V, from a 12V Battery

[sweaterlife]

Hi,

I'm an electronic engineering student currently working on a project for my university's student racing team.

Then goal is to provide an electrically actuated gear shifter to operate the cars gear box, i.e. a paddle shift gearbox.
The gearbox on the car is a sequential shift gearbox taken from a honda motorbike.

We require 10Nm of torque to execute a shift, by turning the gear selector shaft where the bikers foot would usually kick a lever. We had considered using a solenoid, however this is a common solution, so we are attempting to solve this problem with a stepper motor. This will offer greater positional control and efficiencies.

In order to deliver the required torque, we must drive our stepper motor at24V, 6A per phase. This equates to 24V, 12A which is a huge power demand of 288W. We have already built a H-Bridge circuit capable of operating under these conditions, and constructed a microcontroller system to drive it, however we are struggling to produce the required voltage and current from the cars on board 12V battery.

Attempts have been made to create a switching power supply (classic PWM with in line inductor style) however the size of inductor required in order to prevent saturation is not practical, and also there is no reason to supply a constant 12A output current as a single shift operation will take 200ms. We also estimate that the driver will not shift more frequently than once every second or so, so we have a duty cycle of 20%.

This has led us towards the possibility of charging a capacitor bank to drive the motor. Are there any other suggestions as to how we might create a 24V 12A supply from a 12V lead acid battery.
N.B. The car also has an alternator which will be recharging the main battery during operation.

Power calculations reveal that we require somewhere in the region of 60-70Joules to execute a shift, which is equivalent to using a 216mF Capacitor. This would be the size of a couple cans of coke and also impractical.

does anyone have experience with supercapacitors (double layer capacitors) on the forum?
Is it a feasable solution to charge a series bank of supercapacitors up to 12V off the car battery (when not executing a shift), then drop them in series with the battery when required using heavy mosfet switching? This series combination would give 12V and both elements should be capable of delivering 12A.
Three 5.5V, 1.5nF supercaps would give a series capacitance of 500mF, which should allow us to store double the energy required.

Is there a fundamental reason not to use a bank of capacitors in series with a lead acid battery?

This is a first post, but I've tried to give as much information as possible.

Thanks in advance for any help/advice!

Cheers, Paul

 

[MikeMl]

If direct driving the gearbox shaft, how many steps of the stepper are required to change from one gear to the next? Can you put a gear reduction between the shaft of the stepper and the gearbox shaft?

 

[sweaterlife]

Ah yes, sorry I forgot to include how we are driving the shaft.

We require 18 degrees of angular rotation in each direction (down and up).
We are already gearing down the motor with a leverage system at a 2:1 ratio in order to get the most torque out of the motor possible.

So at 2:1 this is 36 degrees rotation in 200ms

with a 1.8 degree step angle this is 18 steps, i.e. 11ms per step

We really need the improved torque performance of a 24V driving voltage over simply using 12V, and can't just increase the current at 12V as the motor windings are rated for 7A.

 

[MikeMl]

So if you charge a capacitor to 24V, it would have to supply enough current to power the windings of the stepper motor continuously for 200ms for each gear change. How long does it have to recover before the next time you change gears again?

 

[sweaterlife]

Assuming the driver of the car isn't doing anything crazy, he shouldn't shift more than about once a second. So we need to supply the 24V for 200ms, then we have 800ms to recover minimum.

Obviously in some cases he wont try and change again for a good few seconds, e.g. if he's on a straight/in top gear

 

[MikeMl]

Here is my suggestion: Get a small 12V (say 4AH) SLA battery. Wire it in series with the vehicle's battery. That will give you a 24V supply with good peak current capability, and the ability to sustain that current for several minutes.

Get a isolated switch-mode power supply; 12Vin and 14Vout and connect it's output to the "floating" battery. To keep the floating battery charged, the switcher will only have to supply the "average" current required by the stepper.

 

[sweaterlife]

Sounds good to me, thanks Mike.

The Alternator in the engine block will push the car battery voltage up to 14 V, do you think there's any way to use this rather than a switcher? This is why we'd planned on trying to switch the secondary battery into a charging configuration with the car's battery in the down time?

Also, weight is a serious problem, as our stepper motor is already 3Kg, and this system's meant to be lightweight for racing use! We thought a stored energy solution such as the capacitor bank might be lighter?

 

[MikeMl]

**broken link removed** will not weigh even as much as the capacitors. If you add the equivalent of a Double-Pole Double-Throw switch, you can have the SLA in parallel with the alternator for the 80% of the time; and then switch it in series with the main battery during the time it takes to shift gears. The switching can be done with FETs.

 

[Hero999]

Capacitors have a lower energy density than batteries so you're not gaining anything there.

A DC-DC converter designed to produce 288W for a short period of time isn't likely to be that bulky.

Why did you go for the 24V motor? Couldn't you get hold of a 12V motor?

Did you consider using continuously variable transmission so you don't need a huge clunky stepper motor?

 

[sweaterlife]

Thanks for all your suggestions


We have tried to produce a DC-DC converter, however under testing any current demand over 5A saturates the inductor and the circuit fails.
We can run the stepper motor at various voltages, however the 24V torque/rpm response is the best one to get the torque we require.

This is the motor family:
Stepper motors - ST8918-Nema 34 - Nanotec
And it's the ST8918L4508

We need to drive it at 60rpm to meet the speed requirements of our shift. We're looking to use the 24V parallel configuration, which gives about 7Nm of torque. At a 2:1 gearing ratio this is 14Nm which should comfortably turn our shaft. Once we get it working we can optimise it and, for example reduce the current requirement etc.

The SLA you suggested Mike is over 1Kg, we'd planned on trying these supercaps which are no more than 10g each:
PANASONIC|EECS5R5H155|CAPACITOR, 1.5F, 5.5V | Farnell United Kingdom

Hero, We have no control over the type of transmission, that's down to the mechanical engineers, and we just need to work with their choices!

 

[MikeMl]

CΔV = q = It,

rearranging gives C = It/ΔV.

If you are willing to let your capacitor sag from 25V to 20V in 0.2s with the stepper drawing 12A, you will need 12x0.2/5 = 0.5Farad.

To get 25V, you will need 5ea 2.5 Farad supercaps in series. You need to check what the peak allowed charge/discharge current for the supercaps is? You will also need to check the requirements for limiting the per supercap voltage (due to imballance of the supercaps) during charge/discharge (Zeners in parallel?)

 

[Hero999]

The inductor needs to be larger to avoid saturating it. Using a push-pull transformer driven topology should increase the available current.

Did you read the datasheet for the capacitor you've just linked?

The internal resistance is far too high for what you want.

 

[sweaterlife]

Hero:
I will research the push-pull transformer topology
the 30 ohm resistance stated was for 1kHz, I read on another datasheet that the ESR was 200-350mOhm which I thought was acceptable?

Mike:
As We'll be connecting the supercapacitors in series with the battery, we only require about 15V from the capacitor bank, which means we can use three in series which gives 0.5F, and could experiment with parallel combinations of the series caps to increase total capacitance?
I have already got a plan for a voltage balancing circuit to maintain the appropriate voltages on the capacitors to avoid overcharging.

The datasheet for the supercaps I linked to was quite sparse on information, so there is quite a bit of guess work at this stage unfortunately!

 

[Hero999]

It's a memory backup capacitor so is designed for low current applications.

The ESR won't be much different at lower frequencies.

You estimated capacitor size is far too low.

You might nee 70J per shift but as Mike said, the motor is going to stop working when the voltage drops below a certain point.

You say you only require 15V from the capacitor bank? Does that mean the motor will work from 15V?

If so an SMPs is much easier, or you could even use a small 6V SLA in series with the 12V battery and charge it using an SMPS as Mike said.

 

[MikeMl]

Go to DIGIKey. I think they have supercaps with full data sheets on line. Carefully check the series resistance (or peak charge/discharge currents).

 

[sweaterlife]

I was planning on 'floating' the capacitor bank on the 12V car battery, as though It was simply another SLA Battery.
So as long as I could design the cap bank properly to provide the 13A, it should work with the battery to give me 24V at 13 A, and then I could recharge it off the battery (and alternator) on the downtime from shifting.

Yes, Digikey have a much better selection than farnell etc thanks, I'm searching there now.
What kind of ESR would be recommended for this amount of current? Simply as small as possible?

 

[MikeMl]

The ESR dictates the maximum charge rate or discharge rate, as well as self heating...

 

[shortbus=]

If you use one of these; TSFNA57 - Permanent Magnet (PM) Non-Captive Linear Actuators

You would put a clevis on the end of the motor shaft to connect to the shifter arm. Mount the motor body so half the screw travel is the resting position of the shifter.

 

[sweaterlife]

I'm afraid we're too far down the design route to change actuator...we have everything working apart from this Power supply system so it's the final piece of the puzzle!

 

[Oznog]

I think you need to rethink this.
The motor you have is beyond huge. Get a more appropriate motor! This is actually less costly than the components you're thinking of to "fix" the problem. Most steppers are like 3v-6v, and a better gear ratio would help.
The problem with steppers is they're high inductance, and move very slowly when excited with a DC voltage based on the DC rating of the coil, because the current rises too slowly for each step. Try to step it fast and the current won't reach the specified target current in each phase before transitioning to the next step, thus torque drops off sharply and the coil will stall in use.

Instead, for CNC use, we use like a 48v supply and the current rises very quickly. However, the current is regulated (via PWM) so it won't exceed the coil rating. The coil's rating is actually what won't overheat the coil due to simple (I^2)*R heating. As long as the current is regulated, using a higher voltage will not increase the coil heat. Just make it more powerful at higher speeds.