CSD19532Q5B: The most suitable MOSFET selection for solenoid consuming 40W peak energy

Hello Alev,

Thank you for you interest in TI FETs. I am not familiar with solenoid drive requirements. Can you share more details on how the FET is used to drive the solenoid? Is the FET ground referenced (i.e source is grounded and drain is connected to the solenoid) or floating (drain connected to supply, source connected to the solenoid)? For 48V, you would need to use at least a 60V FET. TI's FET product line does not make any devices with integrated drivers. The gate driver would need to be external. TI does make gate drivers in multiple configuration and also integrated solenoid drivers. You can find more information at the links below. I look forward to your response.

https://www.ti.com/power-management/gate-drivers/overview.html?keyMatch=gate%20driver&tisearch=search-everything

https://www.ti.com/motor-drivers/solenoid/overview.html?keyMatch=solenoid%20driver&tisearch=search-everything

Best Regards,

John Wallace

TI FET Applications

Hi John;

Actually, an H bridge with an internal fast diode to drive the solenoid in both directions would also do the job. I don't know, does Texas Instruments have such products?

Or can the 48V DRV drivers produced for stepper motors do this job?

But we need a mosfet and a driver to drive in at least one direction.

Can you suggest us a GATE driver (I don't know if there is a single one) that we can use directly with CSD19532Q5B and a sample schematic configuration including the pin connection of the microcontroller?

Hi Alev,

Thanks for the updated information. The CSD19532Q5B is a 100V FET in 5x6mm SON package. It is the lowest on resistance, 100V device in this package and may be higher performance and cost than your requirements dictate. Also, all of TI's 80V and 100V FETs require minimum VGS >=6V.  Because the current is only 0.8A, the conduction loss in the FET is very small and a lower cost FET with higher Rds(on) can probably be used.

As an alternative, I'd recommend the CSD18543Q3A 60V N-channel FET in 3.3x3.3mm SON package. A lower cost, smaller alternative would be the CSD18541F5 60V N-channel FET in F5 FemtoFET package. What is the PWM frequency? The CSD18541F5 has 1.2kΩ nternal gate resistance which limits its switching frequency. TI makes many single low side drivers as shown at the link below. I'm only responsible for the FETs. I can reassign this thread to the gate drivers applications team to assist with your selection.

Your other alternative, is a DRV type driver with integrated FETs. These products fall under a different applications team. Please let me know how you would like to proceed and I will reassign this thread to the appropriate applications team.

https://www.ti.com/power-management/gate-drivers/low-side-drivers/products.html#480=1%3B1&sort=1130;asc&

Thanks,

John

Hi John;

Can you recommend us a driver for both CSD19532Q5B and CSD18543Q3A?

We can use both. Are there single drivers for these products?

Thanks...

Hi Alev,

I'm going to reassign this thread to the gate drivers applications team for a recommendation.

Thanks,

John

Hi John;

I would like to ask one more thing, for the D1 ultra-fast diode in the schematic, considering the current, if CSD19532Q5B is used, what kind of specification do you suggest?

Thanks...

Hi Alev,

I'm an applications engineer covering gate drivers. It looks like you need a low-side gate driver, and I see that you are using 48V for your supply; however, we do not have any low-side gate drivers in our portfolio that supports that high of a voltage. Our highest VDD low-side driver is the UCC27531, which supports a maximum of 35V, and our newer generation drivers, UCC27614 and UCC57108, support a maximum of 30V. Do you plan on implementing a separate voltage line for the gate driver?

Thanks,
Rubas

Hi Rubas,

We also have 8V supply voltage. So how can we use the CSD19532Q5B mosfet circuit with the UCC27614 gate driver?

Can you provide us with a schematic with the recommended resistor and capacitor values? We are preparing for high volume mass production.

Thanks.

I have a few questions;

a) What values ​​do you recommend for the RC filter on the line from the MCU to the input pin of the UCC27614 driver? MCU output pin -> Driver IN+ pin

b) Sometimes drivers are driven with a single PWM output (IN+) and sometimes with two PWM outputs (IN+ and IN-). Is it possible to drive this driver in both ways? Actually, we prefer to drive it with a single PWM (IN+) signal.

c) Doesn't the UCC27614 product need a bootstrap capacitor? And Why ?

d) Do you have another suggestion for the line from the gate driver to the MOSFET in the attached picture?

Hi Rubas;

Before you, dear John, he told us about the CSD18543Q3A product. This is a 60V transistor.

I would like to ask a few questions about this subject;

a) Is it possible to drive the CSD18543Q3A product with the same schematic as the UCC27614? Without making any changes to any electrical elements?

b) Since a solenoid is driven, it is considered to use an ultra-fast 1A diode to protect it from reverse current. In this case, do you think it would be more useful to use a 100V transistor instead of 60V to protect the transistor from reverse current?

c) The solenoid is driven with 48V momentarily (about 1/4 of a second) to pull back its stroke at the first opening. This is the energy required to pull back its stroke without difficulty. Then, an average of 5V energy is enough to keep its stroke inside. When first powered, it draws about 0.8A current when supplied with 48V. Considering that it is driven with PWM, is both a fast diode and 60V a sufficient precaution for reverse current? Or does the 100V transistor provide a longer life and more durable circuit?

Thank you.

Hi Alev,

Our gate drivers can drive most FETs on the market. It looks like the CSD18543Q3A could be an even faster option due to the lower gate resistance, input capacitance, and gate charge.

For questions b and c, I would be more inclined to use the 100V FET for extra precaution, but I believe John would be able to answer those questions more accurately, so I will loop him back in the thread.

Thanks,
Rubas