Hello! I am using a nordic chip to hopefully driver a speaker trough PWM ( not for this sole purpose ), for this, i use a TPA2006D1DRBR with a 2nd order low pass filter on its input to "turn the pwm into analog". I have now tried it, and i do get sound, but very, very subtle. i do get the appropriate readings on every pin, and i do hear a change in pitch when i change frequency. According to chatgpt, it could be due to the fact that i am powering the amplifier from 3.3v, but using the differential input with 1.8v this causes it to not get the full reach of the 3.3v thus the sound is very low. What do you think? I will have to redesign the board anyway, what is my best take to stay in the same cost range? I have thought of simply using a level shifter to bring the 1.8v up to 3.3v, i dont know if thats the best solution tho. I have linked a picture of my setup.
Which speaker - there are a lot of possibilities here!
what frequencies are you expecting?
You are generating PWM, "converting to analog", then connecting a class D (ie PWM) amplifier to drive the speaker.
Why not use the KISS principle? ie just drive "the speaker" directly from the PWM?
Whay are you using 1.8V?
Why are you using an amplifier with differential inputs?
an op-amp to replace the whole circuit?
CMS-151103-088SP
I was simply testing to see if the speaker was working so i programmed it to output in a range of 440hz to 10khz without sepcific sound, just the tone.
I don't think the mcu would be able to output enough power on its gpios to power the speaker on its own, right?
I am using 1.8v because of other low power sensors on the board, the speaker is pretty much the only one to require 3.3v so it would make no sense to switch all other sensors just for this purpose.
This is the one i found that was reasonable in terms of footprint,price and output power. I didnt choose it specifically for differential, i checked wether or not i could drive it with pwm and it said i could so i went along with it.
Isn't PWM a digital signal?
so an 8 ohm tiny spk.
You could use this circuit - remember a loudspeaker is a motor.
Drive it from FAST PWM (ie class D), and make sure the fet is normally OFF
If you want "hifi" sound (oh yeah - from a micro speaker?) you could also look at this circuit
I’m not looking for high quality, just audio indicators that are a bit more complex than buzzers so I definitely do not need hifi. So the circuit you shared is the whole thing? I feed the pwm from mcu to the mosfet? Will it output 3.3 or 1.8v to the speaker?
I am not very good with audio stuff , does the circuit you provided act as an amplifier too?
Would I be better using something like the max98357a for playing audible audio like voice or something, basically not just beeps. Is it 1.8v compliant? i saw it has 1.3v VIH, but it was the same for my current amp and it doesnt work so...
Why don't you increase the gain of the TPA2006D1
it's fixed to 6db i believe
No.
It depends on the input resistor. With 150K it'a gain of 2, with 75K it would have a gain of 4
where Ri is the input resistor.
by increasing gain, i would eliminate the effect of the lower voltage? It says the lowest resistor i can use is 15k for a 26db.
my problem, in summary, is that my input signal is centered at 0.9 V, and the amp’s internal reference is 1.65 V, so the output swing is asymmetric.
I don't follow your logic. The input bias is unrelated to the output.
Add a coupling cap...
really? isnt that the reference point of the center of the pwm ? As i said, i know very little about audio stuff.