Problem with IC: SSM2305

Good evening

I have problem to use devices SSM2305. I have attached the scheme implemented to connect the device, as suggested in datasheet. I have also linked the datasheet where you can find the single ended configuration that I used, at page 12.
The speaker that I used, is 4 Ohm. At this moment I haven’t realized part dedicated to the second channel, but it will be like the first one.

I have found an unexpected behavior in trimmer. When trimmer is turned to half scale, I can measure about 0.5V in input pin (+) of device. But when I exceed this position and tension level thus is over 0.5V, the audio out is not reproduced correctly. In this circumstance I can hear only a crackle.

Measuring the resistance between pin 2 and 3 of trimmer, it is about 13.4K Ohm. I can find the same value, checking between pin 1 and 2.

So I have control the resistance value (acquired in pin 1 - 2 and 2 -3), varying the trimmer position. I can measure the same value from 0 to half position. The measures over half position decrease from 13.4 to 0 K Ohm.

If I eliminate the capacitors, I reveal zero mean value in input signal, otherwise it increases.

I hope I was clear.

Datasheet AD.pdf (397 KB)

Try bias the neg-input to 1V+ (voltage divider)
..or feedback via ca. 20k from out+

I don’t know anything about that chip, but your schematic looks OK compared to the datasheet.

I have found an unexpected behavior in trimmer. When trimmer is turned to half scale, I can measure about 0.5V in input pin (+) of device. But when I exceed this position and tension level thus is over 0.5V, the audio out is not reproduced correctly. In this circumstance I can hear only a crackle.

Capacitor C1 should block DC (or any DC effects) from the input or volume control. So, maybe C1 is shorted (internally or externally).

It may be normal to have a DC “bias” voltage on Pin-3. Most audio amplifiers operated from a single-ended power supply have bias on the inputs & output (before being filtered with a capacitor) so the audio signal can swing positive & negative relative to ground. Normally, it would be biased at half the supply voltage.

With a bridged-output (as you have) you don’t need an output-capacitor because both outputs are biased at the same DC voltage (so no DC current flows through the speaker) and there is no ground connection to the speaker.

P.S.
Try adding a bypass capacitor across the power supply, close to the chip. m (Page 13 of the datasheet and good practice with all analog or digital ICs.)

Thanks a lot, to your answer.

Should I try to connect via feedback obtained by 20K (like in the attached)?

But if it was like that, is it unbalanced with signal “In +” ? This signal has mean value less than 1V.

Connect R directly to neg. input. Keep the small cap to GND.

see page 11http://www.farnell.com/datasheets/662037.pdf

DVDdoug:
Try adding a bypass capacitor across the power supply, close to the chip. m (Page 13 of the datasheet and good practice with all analog or digital ICs.)

True.
This is a digital IC (class-D), not analogue.
It needs very good decoupling with two different caps.
Electrolytic and ceramic, as on the datasheet.
VERY close to the IC.
If you don't do that, it most likely will do strange things.
Leo..

Ok, sorry I spent a lot time in other activities. Now I have just resumed my project.

I desired to verify this saturation condition, so I want to estimate the out power when saturation occur. I want guarantee that power is less than decalred limit (about 2.8W).
But now the question is: how can I evaluate the power delivered to speaker, that is in PWM form? And what is the power of PWM signal (with variable duty cycle)?

Perhaps is it evaluated like PAM wave?