Announcements

# irs2092s Self-oscillation Start-up Condition

Level 2
Level 2

AN-1138 is described as follows:

1. How is condition 3 calculated, can you give an example? Is Vin an rms or peak-to-peak?

2. Is R1 related to the input voltage? If PWM is determined to be 360khz, how is R1 calculated? What is the effect of the value of R1 on the startup of 2092?

3. The following figure is the parameter I set, is it reasonable? Whether it affects the 2092 self-oscillation。

4, as shown in the figure below, I see that IRADAMP7s and IRADAMP5, IRAUDAMP9 are different in the feedback network, IRAUDAMP7s does not have Rfb1 and CFB, IRAUDAMP5, IRAUDAMP9 there is Rfb1=1k and CFb=150pf, what is the role of this RC for 2092 self-oscillation? IRAUDAMP7s are not required?

1 Solution

# Re: irs2092s Self-oscillation Start-up Condition

Moderator
Moderator

Hi,

1.  How is condition 3 calculated, can you give an example? Is Vin an rms or peak-to-peak?
>> Vin is a peak voltage. Vin/Rin is input current to the summing node. Assuming IN+ is grounded, this input current goes to feedback resistor. If output voltage is not high enough to provide same current through Rfb, the control loop saturates.
For example, Vin_peak=1.41V, Rin = 3.3k ohms, Rfb = 47k ohms., supply voltage is +/-35V,
(Vin-peak/Rin) = 427uA < (V+B/Rfb) = 745uA, this means the output will not clip.

2. Is R1 related to the input voltage? If PWM is determined to be 360khz, how is R1 calculated? What is the effect of the value of R1 on the startup of 2092?
>> R1 is not related to the input voltage. As there are many combinations of component values to run a self-oscillating frequency, finding optimum combination between Rfb, C1, C2, and R1 is very complex. We don't have a equation for it. The reason is that there is a optimized carrier amplitude in COMP pin that gives best audio performance. We made a table to guide you to find a combination in application note AN-1138.

For 360kHz, use C1=C2=2.2nF and adjust R1.

3. The following figure is the parameter I set, is it reasonable? Whether it affects the 2092 self-oscillation。
>> It look reasonable but not sure how THD+N goes. We recommend to start with the values in the table above. The difference is the carrier signal amplitude in COMP pin. Smaller the amplitude the higher the loop gain but there is a cap to it due to noise immunity and jitter.

4. As shown in the figure below, I see that IRADAMP7s and IRADAMP5, IRAUDAMP9 are different in the feedback network, IRAUDAMP7s does not have Rfb1 and CFB, IRAUDAMP5, IRAUDAMP9 there is Rfb1=1k and CFb=150pf, what is the role of this RC for 2092 self-oscillation? IRAUDAMP7s are not required?
>> The Rfb1-Cfb is a noise filter and a delay. This is originally added for noisy MOSFET switching stage. With current better MOSFET generations, we would not need to have this network unless PCB layout is not well done.

Best regards,
Pranava
3 Replies

# Re: irs2092s Self-oscillation Start-up Condition

Moderator
Moderator

Hi,

1.  How is condition 3 calculated, can you give an example? Is Vin an rms or peak-to-peak?
>> Vin is a peak voltage. Vin/Rin is input current to the summing node. Assuming IN+ is grounded, this input current goes to feedback resistor. If output voltage is not high enough to provide same current through Rfb, the control loop saturates.
For example, Vin_peak=1.41V, Rin = 3.3k ohms, Rfb = 47k ohms., supply voltage is +/-35V,
(Vin-peak/Rin) = 427uA < (V+B/Rfb) = 745uA, this means the output will not clip.

2. Is R1 related to the input voltage? If PWM is determined to be 360khz, how is R1 calculated? What is the effect of the value of R1 on the startup of 2092?
>> R1 is not related to the input voltage. As there are many combinations of component values to run a self-oscillating frequency, finding optimum combination between Rfb, C1, C2, and R1 is very complex. We don't have a equation for it. The reason is that there is a optimized carrier amplitude in COMP pin that gives best audio performance. We made a table to guide you to find a combination in application note AN-1138.

For 360kHz, use C1=C2=2.2nF and adjust R1.

3. The following figure is the parameter I set, is it reasonable? Whether it affects the 2092 self-oscillation。
>> It look reasonable but not sure how THD+N goes. We recommend to start with the values in the table above. The difference is the carrier signal amplitude in COMP pin. Smaller the amplitude the higher the loop gain but there is a cap to it due to noise immunity and jitter.

4. As shown in the figure below, I see that IRADAMP7s and IRADAMP5, IRAUDAMP9 are different in the feedback network, IRAUDAMP7s does not have Rfb1 and CFB, IRAUDAMP5, IRAUDAMP9 there is Rfb1=1k and CFb=150pf, what is the role of this RC for 2092 self-oscillation? IRAUDAMP7s are not required?
>> The Rfb1-Cfb is a noise filter and a delay. This is originally added for noisy MOSFET switching stage. With current better MOSFET generations, we would not need to have this network unless PCB layout is not well done.

Best regards,
Pranava

# Re: irs2092s Self-oscillation Start-up Condition

Level 2
Level 2

1、According to the description of condition 3, | Iin| <|ifb| is one of the conditions for 2092 self-oscillation.

My supply voltage is ± (75V-85V).

Calculated according to condition 3,

e.g. (1.2Vrms*1.414)/1.5k=1131.2uA > 75V/75k=1000uA Does this cause 2092 to be unable to oscillate?

2. Can you explain what this sentence means?

“Note that this condition also limits the maximum audio input voltage feeding into R1. If this condition is exceeded, the amplifier stops its oscillation during the operation period”

# Re: irs2092s Self-oscillation Start-up Condition

Moderator
Moderator

Hi,

e.g. (1.2Vrms*1.414)/1.5k=1131.2uA > 75V/75k=1000uA Does this cause 2092 to be unable to oscillate?

>> Yes, you should make sure that condition 3 is met. You can try RIN 3.3k Ohms which will bring down the Iin value.

2. Can you explain what this sentence means?

>> From condition 3 and formula for Iin you can conclude that Iin is directly proportional to Vin. Therefore at some Vin the value Iin increases such that | Iin| <|Ifb| is no longer satisfied. This limits the maximum audio input voltage to a certain value.

Best regards,
Pranava