Announcements

From sunburn to sun earn – we’ve got the power! Watch our #poweringgreen videos now.

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
RPD123
Level 1
Level 1
First reply posted First question asked Welcome!

I'm working with a mix of IFX007Ts and the older BTN/BTS7960s (latter on dev boards, former for custom designed PCBs).

I'm working well below their actual current limits and using them to run a very sharply spiking waveform (surge of increasing current enters a low resistance inductive load (BLDC coil) with the driver in the HIGH ON state for tens of microseconds, then the river goes to the LOW ON state and the current recirculates for hundreds of microseconds until it drops to near zero, then the cycle starts again). The peak of the waveform is at 15A for tens of microseconds, well below the 43A spec of these half-bridges. The mean average over time is more like 4A, a tiny fraction of what these chips are rated for. Inhibit is permanently held high.

On the dev boards though there are 10K resistors on the slew rate pins, limiting how fast the half-bridge lets the current rise when it has just switched state.

Would, with a high PWM rate being applied to the Input pin of such a half-bridge, using smaller slew rate resistors, or directly grounding the slew rate pin, further reduce any heating inside the half-bridge chips which may be occuring? When it comes to the IFX007T custom board, should I be fitting much smaller Slew Rate resistors to reduce the potential for overheating the half-bridge chips (I'm not in a situation where I'm concerned abut the increased EMI possible from high slew rates with large currents, but I'd like to keep the slew rate limited as a way to make by low resistance load have a shallower gradient of current rise if doing so does NOT increase heating in the IFX chips)?

I would assume the only way that an IFX007 can limit the slew rate is by gradually taking the internal MOSFETS through higher resistance conducting states before reaching the minimal resistance of being fully ON, and this would surely generate more heat than switching straight to being fully ON??

 

I've seen rumoured discussions of IFX and BTN/BTS half-bridge driver chips being prone to overheating when run at high PWM rates, is this a condition which only occurs when PWMed or currents at close to the 43A? Or is this a condition which specifically occurs, potentially even at much lower time averaged currents , when a large resistor is used to set a slow slew rate?

Thanks

 

P.S. is there any reason that most exmaple IX/BTN/BTS boards including Infineon's own Arduino form factor triple IFX007T BLDC shield seem to use series resistors on the Input and Inhibit pins? Can't those just be driven directly from a 5V microcontroller's output pins? Aren't series resistances on these lines liable to cause the voltage at the Input and Inhibit pins to jump around somewhat when in electrically noisy environments when compared to the more tightly fixed voltage that would be achieved at the pins with a direct connection to an MCUs outputs?

 

Thank you

0 Likes
1 Solution
Xiangrui
Moderator
Moderator
Moderator
500 replies posted 250 solutions authored 50 likes received

Hi,

1. The IN and INH resistors in series are using for MCU protection in worst case. Of course they can be driven directly by 5V.

2. It's hard to say overtemperature occurs in high PWM and low current condition or in low PWM and high current condition. We can only say they both has risk to lead an overtemperature. You know, normally the losses of MOSFET consist of two part: conducting loss (current related) and switching loss (PWM frequency related).

3.  Normally, low slew rate --> long switching time --> more switching loss, but better EMC and lower voltage spike during switching off. For a higher slew rate and regardless of EMC, you can set RSR=0.

Regards,

Xiangrui

View solution in original post

0 Likes
3 Replies
Xiangrui
Moderator
Moderator
Moderator
500 replies posted 250 solutions authored 50 likes received

Hi,

1. The IN and INH resistors in series are using for MCU protection in worst case. Of course they can be driven directly by 5V.

2. It's hard to say overtemperature occurs in high PWM and low current condition or in low PWM and high current condition. We can only say they both has risk to lead an overtemperature. You know, normally the losses of MOSFET consist of two part: conducting loss (current related) and switching loss (PWM frequency related).

3.  Normally, low slew rate --> long switching time --> more switching loss, but better EMC and lower voltage spike during switching off. For a higher slew rate and regardless of EMC, you can set RSR=0.

Regards,

Xiangrui

0 Likes
RPD123
Level 1
Level 1
First reply posted First question asked Welcome!

Thanks for such a prompt reply.

 

Regarding "2.", you can confirm that any risk of overheating in high PWM rate low current situations would be worsened by a slower slew rate, as this WILL(?) atleast to some extent always increase switching losses when compared to the slew rate possible with the slew rate pin grounded (directly or via some minimal resistance like 10 ohms).

 

Also, as switching losses go, does putting diodes in parallel with the IFX007's gates help, a beefy high power rated one pointed from Gnd to the half-bridge's output, and a smaller one pointed from the half-bridge's output up to the positive voltage supply? This would help divert some proportion of the flowing current along a different path during the time switching takes??

Also to clarify, when the slew rate is being limited, does this have a significant effect on fast the current in to an inductive load will rise* when the upper mosfet is turned on, or is the gradient of that rise mostly going to be determined simply by the inductance of the load, with the slew rate acting on much faster time scales.

*I can get the numbers for this for my application, but not until a few days time when I've rigged up a circuit to give a current dependent voltage waveform that my o-scope can plot

 

Lastly, regarding (1.), the worst case which those series resistors are protecting against is one where the IFX007's input pins become direct shorts to ground, and therefore threaten the MCU pins by overcurrent? Or is the worst case which the series resistors are there to mitigate one where the IFX007's input and inhibit pins somehow get raised to a high voltage (like the 12V I'm using asmy upeprpower rail) above that which the MCU pins can survive?

 

Thank you

0 Likes
Xiangrui
Moderator
Moderator
Moderator
500 replies posted 250 solutions authored 50 likes received

Hi,

Regarding 1, The digital inputs IN and INH need to be protected against over-currents (e.g. caused by induced voltage spikes) by a series resistor of typical 1 kΩ. Against overvoltage need additional devices to protect, like clamping diode in parallel. 

Regarding 2, total switching losses is proportional to frequency. It makes sense because each time you turn on and off MOSFET causes switching loss and the more you switch, the greater the losses in a certain time. However, if applied PWM is identical, the longer switching time (i.e lower slew rate) leads greater switching losses for each time.

Paralleling external diodes provide more paths during MOSFET switching off, they work with body diode together which reduces the heat inside the chip. However, it doesn't reduce the total losses and increases cost. I'm not sure if it worth on your application. 

Regarding slew rate effect on current rise in inductive load, there is a simply to explain. As you mentioned, the internal MOSFET resistance will be higher with a limited slew rate, then the charging current for inductive load will be decreased as well. The MOSFET need more time to fully switch ON, and the inductive is charged by lower current. Then the voltage spike is better in comparison with higher slew rate.

However, all above are theoretical qualitative analyze, people need find proper parameters based on their own board.

Regards,

Xiangrui

0 Likes