Go Down

Topic: MSGEQ7 printing out weird values (Read 287 times) previous topic - next topic

Snidgetmist

Hi everyone,

I'm trying to make a LED strip blink to the frequency of music by using the MSGEQ7 chip and an instructable i found here https://www.instructables.com/id/Blinking-LEDs-to-the-Frequency-of-Musi/ and when it didn't work out my first thought was the current wasn't enough or the mosfets were damaged etc. After trying out everything i decided to test the chip alone following this page: http://nuewire.com/info-archive/msgeq7-by-j-skoba/ but all the column values were 20 so i decided to go and buy a new one and now i get this:

354 354 354 354 353 353 353
 353 352 352 352 352 351 351
 350 349 349 348 347 346 345
 344 343 342 341 340 339 338
 337 336 335 334 333 332 331
 330 329 328 327 326 325 324
 323 322 321 320 320 319 318
 317 316 315 314 313 313 312
 311 310 309 308 307 307 306
 305 304 303 303 302 301 301

... and after a while it gets:

198 198 198 198 198 198 198
 199 198 198 198 198 198 198
 198 198 198 198 198 198 198
 198 198 198 198 197 198 197
 198 198 198 197 198 198 198
 199 198 197 198 198 198 198
 198 197 198 198 198 198 198



When there's nothing playing and i've also tried to connect the input to the ground to get all 0's but the values stay like that.

I've attached a picture of my breadboard circuit. I've also read a lot of threads on this same subject and nothing seems to work for me (i've checked ALL the values from the resistors an caps like three times to make sure they're okay)

I appreciate all the help :)

Grumpy_Mike

Quote
When there's nothing playing and i've also tried to connect the input to the ground to get all 0's but the values stay like that.
It is a noisy chip. When I have used it I get base noise levels around 150 but occasionally peaking to 300.

I am assuming you are getting variation when you are playing something. I found the audio output of my iPad had to be at full to get a full scale reading so I used an a amplifier between chip and audio input.

You could subtract the base noise level from your readings before you use them. You could also make the circuit so that it is not sprawling all over the place like you have it. That is throw away the bread board and solder it all up on strip board. The supply decoupling capacitor leads are far too long, they should be as short as possible.

But you will always get noise out of the chip as well as the signal you want.

These chips are now obsolete, that is they are not manufactured ny more. It looks like those on the market at the moment might contain a lot of chips from the reject bin.

Snidgetmist

Hey!!
Thanks for your help, i tried to make the capacitor leads as short as possible, and whenever i plug the pin to the audio jack the values go from this:

172 169 161 165 171 168 160
 165 171 162 158 166 169 160
 158 167 164 156 159 166 164
 153 162 166 158 153 162 165
 153 156 164 161 152 157 164
 156 153 161 164 156 155 162
 160 154 157 165 161 155 159

to this:

545 703 664 568 534 603 628
 629 642 636 572 478 578 629
 666 622 711 629 535 568 617
 715 561 598 649 576 502 591
 790 573 593 714 603 615 626
 620 320 440 591 630 636 646
 642 642 646 652 646 642 643
 647 641 218 414 584 591 521
 697 671 584 502 591 735 601
 569 652 581 500 579 625 618

when there's nothing playing and they stay around the same values when there's something playing as well.

Is there another ic that i can use for this project? or this seems to be workig just fine? I'm just really confused bc i thought that the columns should display rather different values and in my case they display values within the same range.

DVDdoug

#3
Sep 21, 2018, 07:41 pm Last Edit: Sep 21, 2018, 07:56 pm by DVDdoug
Quote
when there's nothing playing and they stay around the same values when there's something playing as well....

...I'm just really confused bc i thought that the columns should display rather different values and in my case they display values within the same
So when you ground the inputs again the readings don't go back to around 160?

For testing, I'd suggesting generating some single-frequency sine wave files with Audacity

Quote
Is there another ic that i can use for this project?
I've never seen anything else like it.   There are filter chips or you can build filters with op-amps, but you'd have to build 7 of them, bias the outputs, and connect them to 7 analog inputs.

There are FFT & FHT libraries for frequency filtering/analysis in software.    You do need to bias the input (so you can read the negative half of the AC audio waveform) and it may still help to amplify.   Or, if the signal is low you can try the 1.1V reference and adjust the bias accordingly to increase the sensitivity.

Grumpy_Mike

So you make the capacitor leads as short as possible and things get worse? I think that shows the circuit is not solid and stable in its wiring, possibly a problem with the bread board.

Go Up