IRS2092 with IRFB4227 problem!

[tantal.lpt]

Hi

I have a big problem with the IRS2092 Amp from International Rectifier and the MOSFET "IRFB4227" .

The Amplifier works well ... when it works ...

It does not start up!

On the one hand when I drive the high input (U_eff ~3V) with no load then it starts up and works correctly. On the other hand when the class-d amp drives a load I need only to touch the input and the amplifier is going to work correctly.
Can you give me an advice what I am doing wrong???

In the attachment I enclosed our class-d schematic and some snapshots from the scope (scope on IRS2092 PIN13).

[DarkOne]

Hi,
you have an error in schematic. Feedback resistor should be connected after input resistor. Your's is connected between the DC blocking cap and input resistor.

[tantal.lpt]

Hi,
thank you very mutch for this tip .. now the amp works perfectly!

But now I have a new problem ... ich have create a second amp. This amp like realy the same as my first amp ... the same PCB layou and the same components. But the output signal is different and whem I power it on it dosen't start

The porotection comes every ~0,5 sec and the amp makes a restart.
The amp starts wenn I gave in a big signal <-- like the first amp.

Can it be that the DT is too small or the protection is too sesitiv?

[DarkOne]

Originally posted by tantal.lpt
Hi,
thank you very mutch for this tip .. now the amp works perfectly!

But now I have a new problem ... ich have create a second amp. This amp like realy the same as my first amp ... the same PCB layou and the same components. But the output signal is different and whem I power it on it dosen't start

The porotection comes every ~0,5 sec and the amp makes a restart.
The amp starts wenn I gave in a big signal <-- like the first amp.

Can it be that the DT is too small or the protection is too sesitiv?

Probably I can't help you more, but useful information for others (experts) may be what's the difference between this and the other amp?

[mag]

Hi,
I had the same problem with IRS2092+IRFB4227.
When I was pushing the amp the current limit trips in and mutes the amp for around 0.5s.

Check the voltage on the IRS2092 pin 5 (CSD). If you find it going down to -5V (the VSS rail) when it mutes it means that the current limit trips in.

Current limit thresholds are set trough resitors. Check on the IRS20955 datasheet (there is no information on the IRS2092 datasheet...) how to set these thresholds.

If you have a too inductive path between the low side mosfet source and the COM terminal (pin 10) of the IRS2092 the current limit can trip due to voltage spikes, muting the amp far before the calculated current limit. (I have the same problem on my design).

My amp works fine with 5k6+4k7 for dead time (40nS). If it is not enough increase it to 65ns by putting 8k2+3k3.

Why did you put diodes in parallel with the gate resistors?
By doing this you will change the behaviour of the internal dead time circuit. If you really want to put a diode don't use the MURA120, 35ns recovery time. Use a small diode as the 1N4148 or a schottky diode


Check also my thread

Class-D Amp with IRS2092


probably you can find some useful information about this amp

ciao

-marco

[sz_ahong]

The current sense routing path is very important.
I use IR20955 as a full bridge amp,it can output 1000W

[tantal.lpt]

Hi,
thank you very mutch for the informations!
I hope that I find the error now!

[txelleta]

Hello,

I am using IR20955 as the same schematic as application note iraudamp4. In my output there is a resistor that limits the current and a transformer. I works perfect, but if i use a resistor of 15 ohms or higher. but if I want a lower resistor to supply more current at the output, the driver starts to work and stops without complet all the signal that i send to the input.

Can anybody help me???

[Benshee]

I had a similar problem with starting amplifier. The problem was resistor connecting Vbus+ with bootstrap capacitor.
First I advise You to disable high MOSFET OCP, and increase treshold of low one. As sz_ahong wrote it can be problem with current sense routing, the driver is waiting 500ns and then starts sensing. If inductance of Vs is too big, it can generate spikes triggering overcurrent protection.
Now I'm preparing to build BTL with IRS2092 for 1400W

[sphinx]

If your intention is to implement soft clip diodes as in the IRS2092 reference design, they're connected to the wrong node. They should be connected to the node after the low pass filter ( speaker + ).