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

cross mob
VGS_MLN
Level 1
Level 1
First like given First reply posted First question asked

Ladies and Gentlemen,


We are using the IR2103SPBF to realise a totem pole stage for ohmic/inductive loads with two N-channel MOSFETs. This is working very well so far. However, we have noticed a behaviour that we can reproduce, which is causing us some concern. The IR2103 "pulls down" the input voltages through the microcontroller or function generator. Both tested simultaneously down to below two volts or lower, as soon as the system is switched off and on again briefly.

If we switch the system voltage off for a few seconds and then on again, the drivers show this behaviour. It appears as if the HIN and /LIN inputs are no longer high-impedance and even load the microcontroller or the function generator. We have documented this behaviour with the help of our scope and would like to show our circuit together with the attached circuit diagram section and ask whether you may know the problem or whether you know whether the behaviour could be remedied by an external circuit => e.g. controlled discharge of all capacitors when switching on/resetting?

Let's switch the entire system off for a longer period > 20 seconds and then switch it on again. The system restarts correctly and the logic levels of the microcontroller and the function generator are not affected.

Here are our boundary conditions:

Vcc = 12 V (buck/boost converter for IR2103 supply)
Vlogic = 3.4 V (microcontroller & all other ICs in the system)
GND same GND potential for all devices
Vtotem_pole = 11.8 V (designed for 9 V to 30 V) Totem pole supply (MOSFETs) => all tests with 11.8 V

Scope:

Channel 1 => yellow => Output Totem-Pole

Channel 2 => green => Input from function generator

Channel 4 => blue => Input from microcontroller

 

Kind regards

0 Likes
1 Solution
SivaBommadevara
Moderator
Moderator
Moderator
First like received 25 replies posted 10 solutions authored

Dear @VGS_MLN ,

Greetings from Infineon.

As I per my understanding your  buck/boost dc dc is not regulating the voltage to the Gate drivers in the situation you have seen and informed to us.

meanwhile your microcontroller is powered up with minor unregulated DC DC output or may be with working with another source, can generate PWM pulses to gate driver to operate the Half bridges.

please correct me if i understood in different way.

in this case please maintain some delay in PWM generation from the controller / function generator to the half bridges untill your buck/boost converter regulates the stiff voltage.

I hope this will works to solve the issue.

Thank you.

 

Best Regards,

Siva Naga Malleswararao.

View solution in original post

7 Replies
SivaBommadevara
Moderator
Moderator
Moderator
First like received 25 replies posted 10 solutions authored

Dear @VGS_MLN ,

Greetings from Infineon.

I understand the issue. but i did not completely understood about your circuit.

by looking at your waveforms I did not get why both Function generator and Micro controller has connected to the driver IC.

also you for get to mention the operating switching frequency of the Driver IC.

At the same time i observed that micro controller signal  frequency and totem pole signal frequency both are not matching.  could you please explain why is it so?

in my opinion the Driver IC never pulls down the source supply, because it has internal high resistances are incorporated for HIN and LIN pins

SivaBommadevara_0-1706780392091.png

please share your complete schematic diagram of your test setup including driver IC.

 

Thank you.

Best regards,

Siva Naga Malleswararao

0 Likes
VGS_MLN
Level 1
Level 1
First like given First reply posted First question asked

Hello Silva Naga,

we have >10 totem-pole output stages(mor than 10 IR2103). As a test, we decoupled one stage and controlled it with a function generator and the other >10 with a microcontroller (PWM output).

I also assumed that the HIN and /LIN inputs are so high impedance that they do not affect the signal. Unfortunately, the voltage also changes by +0.3 V when we switch the microcontroller to high strength mode (GPIO). This means that if we increase the signal power (higher current), the voltage at the input increases in the case of a reproduced error. In my opinion, the input is not as high impedance as it should be.

0 Likes
VGS_MLN
Level 1
Level 1
First like given First reply posted First question asked

The test circuit

Test_Circuit.png

 

0 Likes
SivaBommadevara
Moderator
Moderator
Moderator
First like received 25 replies posted 10 solutions authored

Dear @VGS_MLN ,

Greetings from Infineon.

As per the information provided,

In my opinion, the Inputs are definitely tied with high impedance.

However, due to multiple circuits (as mentioned >10) at the input side connected to Microcontroller.

This means all inputs of the IC are getting paralleled and behaving as low impedance and might be in the marginal border case of the microcontroller capacity.

I suggest you use a high series resistance (in terms of 10s of kilo ohms) at each HIN and LIN of all the ICs will solve your issue.  

Thank you.

Best Regards,

Siva Naga Malleswararao

0 Likes

Dear @SivaBommadevara 

Of course, we have not connect all MOSFET drivers in parallel on one GPIO of the microcontroller. Each stage has its own GPIO and the GPIOs are distributed across several ports (Axx, Bxx, Cxx, Dxx, ...of the microcontroller. In addition, the GPIOs can drive several milliamperes in high strength mode. That is not the problem.

This is also confirmed by the externally connected function generator. This is completely independent and self-powered. Only the GND reference potential is identical. This cannot be the cause.

Do you have another suggestion? This phenomenon only occurs when switching off and on again briefly. I would also be happy to include the voltage of the buffer capacitor.

 

Kind regards

0 Likes

We discovered that the power supply for the MOSFET driver was not stable. A delay in the microcontroller ensures that the PWM outputs are not switched before the buck/boost converter supplies the IR2103 properly. Further investigations are now being carried out.

The buck/boost converter supplies enough current (1 A) to supply the MOSFET driver. Then there are the buffer electrolytic capacitors. Max 3 or 4 totems poles running simultanious.

But only if it is working in the control range and is regulated.

0 Likes
SivaBommadevara
Moderator
Moderator
Moderator
First like received 25 replies posted 10 solutions authored

Dear @VGS_MLN ,

Greetings from Infineon.

As I per my understanding your  buck/boost dc dc is not regulating the voltage to the Gate drivers in the situation you have seen and informed to us.

meanwhile your microcontroller is powered up with minor unregulated DC DC output or may be with working with another source, can generate PWM pulses to gate driver to operate the Half bridges.

please correct me if i understood in different way.

in this case please maintain some delay in PWM generation from the controller / function generator to the half bridges untill your buck/boost converter regulates the stiff voltage.

I hope this will works to solve the issue.

Thank you.

 

Best Regards,

Siva Naga Malleswararao.