DRV8412 bootstrap capacitor at 4kHz or 8kHz

Steve,

What are the issues with driving the DRV8412 below 10KHz?

We are trying to use this device to drive stepper motors at frequencies <1KHz.  Do you have a device to offer at drive frequencies that low?

Best Regards,

Jim

Jim,

Can you tell me what your current and supply voltage requirements are?  I can direct you to a more suitable part with this information before we go down the road of DRV8412 at low frequency operation.

Regards,

Ryan

The stepper drivers being used are rated at 2A, .8A and .7A.  All are operating at 27vdc.

Jim

Check that...the stepper motors are rated as described.

Hi Jim,

For those currents, the DRV84x devices seems too powerful. We have integrated solutions which will offer better performance/(cost_size) ratio. These should also solve your low frequency PWM input requirement, as they do not require PWM inputs. For example, here are a few suggestions I would start with if I were designing the three motors specified above:

With regards to Voltage:

At 27V, any of our integrated stepper drivers will work, so no need to detail this aspect of the design

Low Current Steppers (0.7A and 0.8A)

I would consider the DRV8824 and DRV8811 in that order. DRV8824 will give you up to 1.6A per phase and DRV8811 is rated up to 2.5A per phase, but 1.7A is far much more realistic. To reach the 2.5A you would need to decrease the thermal impedance as much as possible by adding extra heat sinking and air flow. Where the DRV8824 excels is at the fact that it requires less external components and it offers up to 32 degrees of microstepping, versus up to 8 with the DRV8811.

If you need more than 32 degrees of microstepping (one of the features why the DRV84xx family is used for with steppers), we can use the very same technique with the DRV8812 (1.6A per phase) which is a single device with dual H Bridges, but would require an external microprocessor to perform the microstepping commutation. We have an application note on the matter you can find here: http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=slva416&fileType=pdf.

High Current Steppers (2A)

When it comes to the 2A range, I would then consider the DRV8825. Although DRV8811 can reach the 2A mark, it is quite hard to do so in terms of thermal impedance. The DRV8825 has considerably better RDSon, making this the better fit. The other added bonus is that the DRV8825 is literally identical in footprint to the DRV8824 so all you need is one design and then by choosing which device to populate, you get the current required by the application. Everything else, of course including the 32 degrees of microstepping, operate 100% identical.

Going back to the need of more than 32 degrees of microstepping, the DRV8812 will not be able to tackle this higher current motor. This is why we have a 100% identical both in functionality and pinout configuration device, the DRV8813 which is capable of up to 2.5A per phase and like the DRV8812 is a single device with dual H Bridges.

Other considerations:

All of our devices operate with digital logic levels and do not require a PWM input like the DRV84xx devices. Hence, your low frequency problem is non existent with these devices. If you are going to go with the DRV8824/25 family, you do not even need as much MCU resources as all of the control is performed internally. The microstepping is taken care of by the internal indexer so all you need to provide is a STEP input to coordinate stepping rate and a DIRECTION input to determine direction of rotation.

If you prefer to cause the stepper commutation with your own firmware and are looking for the DRV84xx interface style (namely INx inputs), the DRV8841 offers this interface style. It is virtually identical to the DRV8812/13, except that instead of PHASE/ENABLE interface, it gives you IN1 and IN2 interface, or the ability to control each H Bridge independently. This would only apply if you have an existing firmware you need to reuse which revolves around this interface style.

All of our devices include internal current regulation which is configurable through a VREF analog input and the selection of the SENSE resistor, unlike the DRV84xx devices in which the DSP code would need to generate the current control algorithm.

Hope the info helps. Best regards,

Jose Quinones

Thanks for that valuable information Jose.

We also have a need to drive DC motors, both with and without external encoders.  What do you recommend for those applications?

Best Regards,

Jim

Hi Jim,

For driving DC motors, we have some very nice drivers on the same line with the DRV88xx devices that I mentioned before. For example:

Dual H Bridge Drivers with PHASE / ENABLE interface (DRV8802 and DRV8814):

Once you look at the DRV8812/13 to control your stepper motors, you may start to wonder, "why not use these devices to drive two DC motors as well?" Seems doable, but the truth table for the DRV8812/13 is not practical for driving DC motor loads as it does not treat slow decay after disablement, in the correct fashion. Namely, it is not possible to brake a DC motor with the DRV8812/13.

This is why we have virtually identical devices such as the DRV8802 (up to 1.6A per phase) and DRV8814 (up to 2.5A per phase) which in fact allow for braking to take place. These two devices are tailored for DC motors and are not recommended for steppers because of what happens when the ENABLE line is de-asserted. If on slow decay, both low side FETs will be enabled continuously. On a DC motor this means the collapse of the back EMF and the bracking mechanism. On a stepper, this would mean rotor cogging, which may not be desirable. If the user does not care, then the DRV8802/14 could be used to drive steppers as well, to be honest. This is the only difference.

Dual H Bridge Drivers with INx interface (DRV8841):

I mentioned DRV8841 as a stepper driver on my previous post, but the truth is it can be used to drivre DC motors as well.

Single H Bridge Drivers (DRV8840 and DRV8842) 

We took the DRV8814 and paralleled both H Bridges so that you could enjoy from half the RDSon and hence close to twice the current. The result is the DRV8840 which is a single H Bridge capable of delivering up to 5A. If enough heat sinking is provided you could easily do this continuously as in forever. Otherwise, with a 2 layer board you can easily supply a continuous 4A. The device will get hot, but it will deliver the current to the load for pretty long times. With DC motors, however, the current is only proportional to loading and it is usually high only when approacing stall. So this device is more than plenty for a fair number of DC motor applications.

For smaller DC Motors you can also consider the DRV8800 or DRV8801. The device is rated at 2.8A of current, although a more realistic number is 2A. This is the only device, from all the ones I have mentioned, which does not include current regulation. It has current protection and will disable the H Bridge for about 1.5 ms when the current is too high, but this mechanism is too coarse for it to provide any meaningful current regulation. However, unless you require torque control, this device will do perfectly fine at the specified current levels.

Other Considerations:

Like I said, all of these devices (except the DRV8800 and DRV8801) will have current regulation built in. All you have to do is select the VREF and the SENSE resistor and the system internally regulates the current. Only meaningful if you require some sort of torque control.

The DRV8802/14 are pin to pin compatible so you can decice which one to populate on your design later on and according to how much current you require. I have some customers startingo out with the low current version but because for their application the device runs too hot, they then decide to run with the higher current version just to add system reliability. Is up to you really, but there will be no need to tweak the layout, firmware, etc.

I must not forget to mention that all of these devices have short circuit protection. If the power outputs become shorted between themselves or any power rail (VM or GND), the device will protect itself by shutting down. In other words, no flames and completely normal operation once short is removed and power is recycled. DRV8800/01 has retry, so even with the short applied, it will try to push current. The other DRV88xx devices can be made to try by connecting the nFAULT output to nRESET.

These devices are dumb H Bridges in the sense that you need to close the loop to control the speed. Whether the application calls for a shaft encoder or not, is transparent to which device to use. Whichever is chosen, the microcontroller will need to close the loop by modulating the ENABLE, PHASE or INx PWM duty cycle. But this is pretty much industrial standard nowadays, so I assume you should have some knowledge into servo implementation. If not, we have some code hoepfully can help you to get started.

BTW, none of these devices require a PWM input by definition, as with DRV84xx. 0% or 100% duty cycle are perfectly legal conditions.

Hope the info helps. Best regards,

Jose Quinones 

Thanks much for all the cogent and insightful information Jose.

Best Regards,

Jim

Jose,

Since we are pretty far down the path with the DRV8412, we wanted to use it to at least drive two of our DC motors.  But we are having problems driving in one direction.  We have smooth operation in one direction only.  We know that our proto has omitted some of the capacitors which you have indicated are critical in previous posts (namely the .1uF decoupling caps near the PVDD feeds).

We have ordered a DRV8412 evaluation module to patch into our design, but until it arrives, wanted to get an understanding of:

1. How critical some of these capacitances are for correct operation in DC mode (successfully driving in both directions), namely the .1uF decoupling caps on the PVDD line, the overall PVDD capacitance (1660uF indicated on another post), how necessary the EMI circuit is, etc.

2. The possibility that operation without having the correct capacitance in place might have compromised the 8412, now requiring its replacement.

3. Whether or not others are having issues driving DC motors in both directions using the 8412, and why.

Any assistance you could offer would be very welcomed.

Best Regards,
Jim