LM4991: Timing issue with shutdown

Joachim Hoppe

Part Number: LM4991

Hi,

we have an issue with the LM4991. In our design the startup time is very important. We want to switch off the amp via the shutdown-pin between audio signals. 90 ms after we pull the shutdown to GND (to activate audio) the audio signal is applied.

We now have LM4991-ICs from different two suppliers that show a different startup and shutdown behaviour.
One chip ("MJAB LM4991MA") has a startup time of about 84 ms.

Here is the switch-off time of the same chip:

The other chip ("62AC LM4991MA") needs 250 ms to start.

Here is the behaviour of the same chip when shutting down:

At the moment we can only use the first chip in our design, as the second needs too long to start. We are now wondering if one chip is faulty or if the chips are from different fabs and behave differently but as intended?

The measurements were taken in the same circuit, where only the LM4991 was swapped.

Cb = 1 µF

Ri = 20k

Ci = 1 nF

Rf = 16 k

Cf = 220 pF (parallel to Rf)

Vcc = 3.3 V

The input signal was applied all the time for the tests. Only in the final application it will be triggered 90 ms after the shutdown-pin.

Thank you very much in advance.

Best regards,
Joachim Hoppe

over 6 years ago

0 José Luis over 6 years ago

TI__Mastermind 20085 points

Hi Joachim,

The start-up time should be very similar between different devices. It is difficult to determine the cause, but that could be caused by a faulty chip, process error in a fab, etc.

Best Regards
José Luis Figueroa
Audio Aplications Engineer

0 Joachim Hoppe over 6 years ago in reply to José Luis

Prodigy 20 points

Hi José,

for clarification with our customer, can you please confirm the following points explicitly:

a) The first LM4991 in my post with about 85 ms startup time is a correctly functioning unit.

b) The startup time of the LM4991 in general should be around 84 ms, with Cb = 1uF and Vcc = 3,3 V.

c) The startup behavior of the LM4991 in general should be a linear ramp of the output to Vcc/2 and after a delay of 30 ms the output gets activated.

d) There are no different fabs / processes that can generate a different startup time and behavior of the LM4991. There was no change in technology that changed the startup time and behavior.

e) The second LM4991 in my post with a startup time of about 250 ms is a faulty unit.

Thank you so much!

Best regards,
Joachim Hoppe

0 José Luis over 6 years ago in reply to Joachim Hoppe

TI__Mastermind 20085 points

Hi Joachim,

Sorry for the delay. I am trying to get an EVM (Evaluation Module) to make the tests. But it is quite difficult due to this device is very old and it seems that there are no available EVMs.
I will inform you if I can get an EVM.

Best Regards
José Luis Figueroa
Audio Applications Engineer

0 José Luis over 6 years ago in reply to Joachim Hoppe

TI__Mastermind 20085 points

Hi Joachim,

Sorry for the delay. I have performed some tests to measure start-up time of some chips.
a and b) I performed the tests under the following conditions: Vdd=3.3v and Cb=1uF and I got that the average value of start-up time was between 90ms to 100ms. A solution could be to reduce the bypass capacitor to get lower start-up time.
c) That's right. The chips had the behavior like that in all the tests I did.
d) Unfortunately, I don't know the process or how this device is built.
e) It could be a wrong behavior, but I am not sure if it is a failure.

Best Regards
José Luis Figueroa
Audio Applications Engineer

Audio

Audio forum

LM4991: Timing issue with shutdown