DRV8231A: Solenoid and Motors

Jacob Thompson

Thank You for your reply.

So, in regard to sensing the current, is it not possible to simply provide a Vref, e.g. 2.5V and monitor the output voltage using an ADC, or do you need an amp?

Does TI offer something similar to this IC but with an additional Enable? If not, I can create a 3-bit binary driver, ideas i2C GPIO, as suggested.

When driving a small pump using the DRV8231, do we need to use any additional protection circuit, as we had to use diodes and fuses with IXDI630MYI?

Yes, we are wasting half a FET, but ideally, we want to consolidate our BOMs so that we have common parts across all our PCBs; in addition to that, the price for DRV8231 is far less than IXDI630MYI, therefore even with the FET being wasted, it saving us an overall cost on the PCB.

In addition to this, would there be any issue if this device were to be used as an LOAD switch as well?

In regard to the diagram, so with the DRV8220 , can both outputs a high voltage at the same time?

Dear Jacob Thompson

Any update my additional questions

Thanks for the reply Jacob Thompson

Great, i will connect it to the ADC, this would give a good indication that the pump/ motor is working.

We aim to place about 8 of them on a different PCB design, so ideally, we would like to control them all with a common Signal + Enb for each device.

Well, the pump is 12V max, about 1A. The connector is used to connect a long wire with the pump connector, and we will simply insert the pump into that connector; therefore, would we need ESD protection for that?

Well, we sometimes need an odd load switch in our design, but to keep our BOM list to a minimum and cost done, I was thinking that I could simply use the DRV8231A and keep one input forced to GND. 

We aim to use the SOIC version, but with the exception of connecting the GND pad to copper pours, how can we work out the heat dissipation and how long can we run a motor continuously?

Dharmesh Joshi said:

We aim to place about 8 of them on a different PCB design, so ideally, we would like to control them all with a common Signal + Enb for each device.

Hmmm potentially you could use a digital signal for VREF and set it LOW/HIGH (0V / 3.3V), or if a lower voltage was desired put that through a resistor divider to lower it but still digitally control it.  

Dharmesh Joshi said:

Well, the pump is 12V max, about 1A. The connector is used to connect a long wire with the pump connector, and we will simply insert the pump into that connector; therefore, would we need ESD protection for that?

Would that insertion only be during initial assembly in an ESD controlled environment, or would the end user be connecting/disconnecting the pump?  If it's only in a controlled environment then no protection should be needed.  

Dharmesh Joshi said:

Well, we sometimes need an odd load switch in our design, but to keep our BOM list to a minimum and cost done, I was thinking that I could simply use the DRV8231A and keep one input forced to GND. 

Yeah absolutely, that would work.  

Dharmesh Joshi said:

We aim to use the SOIC version, but with the exception of connecting the GND pad to copper pours, how can we work out the heat dissipation and how long can we run a motor continuously?

Unfortunately we don't have any good thermal simulation for these parts, the best we have is the 9.3 Current Capability and Thermal Performance section in the datasheet.  Check that out and run the numbers and it should get you a ballpark.  Beyond that, you can order an EVM and test it out.  Looks like we only have a DRV8251AEVM, so you would have to order that and order DRV8231A devices and swap out the chip on the EVM.  If you aren't already see if you can join the TI Sample program to be able to order free samples of devices and EVMs - Ordering FAQs | TI.com 

Best,

Jacob 

Jacob Thompson

We aim to replace the DRV8837CDSGR with the DRV8231A. Previously, we had 8 devices driving eight solenoids. Each device had its own enable signal, with a common IN1 and IN2. Since this new device does not have an enable signal, I could modify the logic so that all solenoids' positive wires are connected to a common drive, and then use an additional 4 drivers to connect to the negative wires of each solenoid. This should give me the option of driving 9 solenoids, if I am correct. The solenoids are not controlled simultaneously, so I believe this should work logically.

Therefore, is it possible to connect multiple IPROPI signals to a common ADC? Is this feasible?

Hi Dharmesh Joshi,

Dharmesh Joshi said:

Therefore, is it possible to connect multiple IPROPI signals to a common ADC? Is this feasible?

Yes, this should be feasible. As you pointed out, as long as the solenoids are not being driven simultaneously this should work. 

Best,

David

David Medis

We aim to place multiple devices of about 8/9. Does each device need a Bulk Capacitor, or could I place one 100uF Capacitor for all the devices? Each device will have its own 100nF capacitor.

Hey Dharmesh,

You could potentially share one bulk capacitor if all the motor drivers were in a star arrangement (*) so that the distance from each motor driver VM pin to the bulk cap was exactly the same, and was the shortest distance for the VM line (ex. it can't find going to another device to be shorter than going to the bulk cap).  

This might work for you in pairs or groups of 3-4, but would be difficult for 8 or 9 drivers I would think.  

Best,

Jacob

Jacob Thompson

I am thinking about using the DRV8847 as it has two independent H-Bridges, so 4 half-bridges. If I am correct, I could use just 2 IC to drive 16 solenoids.

So, could I ask my questions here, or do I need to open a new post?

Hey Dharmesh,

You can ask here, that's similar enough to this post topic. 

Yes, that would work, though you'll only be able to drive 4 solenoids with one IC since each 1/2 bridge can only drive 1 output.   

You can also check out DRV8908-Q1 - it can drive 8 solenoids.  DRV8912-Q1 could drive 12 solenoids.  Sadly no 16-channel version

Best,

Jacob 

Jacob Thompson

This is what i intend to do. The solenoids that i will be using are all latching, so they do not need continuous power.

Ahh interesting, creative! I like it, seems feasible to me. 

If you want, you can order 2x DRV8847EVM and try it out pretty easily.  You can remove the column of 0Ω resistors and jump in your own control signals 

Jacob Thompson

I think it would be better for me to create the PCB .

Before I resolve this issue, I have the following Questions:

1 . Can the Isen and Fault pin be left float? I think as pulse are going to be very short max 1 second, it might not be very usefull.

2. If, in the short pulse, the current was to go a little higher than 1A, would they cause any shutdown in the device?

Hello,

1. The ISENx pins should be connected to GND if current regulation is not required. The nFAULT pin is an open drain output so yes, it can be left open (no connect).

2. No this should not shutdown the device. The outputs will become Hi-Z if the current exceed IOCP trip point which is greater than 1.6A for longer than tOCP = 3us.

Best,

David

David MedisJacob Thompson

Thank you for your input.

I plan to connect INSEx to GND as indicated in the datasheet.

Another aspect that came to mind is that, with the exception of the solenoid being operated, all other solenoids will have both their positive and negative wires set to identical polarities to prevent current from flowing and inadvertently activating multiple solenoids. For example, when J10 has both + and -, J11, J12, and J13 will need to be either +/+ or -/- to prevent them from changing state. Based on the datasheet, it doesn’t seem possible to force an output to High-Z while another output is either High or Low. Therefore, to ensure proper operation, my plan is to first set all inputs to the desired logic states, and then set the SLEEP pin to High.

I may have missed this in the datasheet, but is there a specified duration when the sleep mode is enabled during which we need to be mindful of the output states? At any given time, only the solenoid being activated will have a + or -, while all others will need to be set to either +/+ or -/- to prevent any undesired current flow.

Hi Dharmesh,

When the MODE pin is Hi-Z (as shown in your schematic) then the device is in independent 1/2-bridge mode. This should give you independent control over each output. Each input state will reflect its corresponding output state.

Once again, this is a good reason we suggest evaluating designs with the DRV8847EVM.

DRV8847EVM Evaluation board | TI.com 

Best,

David

Thanks David Medis

Before we resolve this case.

The Bulk Capacitors, I presume these need to be ceramic Caps, and how can we determine the values to use, i am thinking about placing two 10uF for each IC