SparkFun Forums

[strokedmaro]

Setup: Arduino Uno, Sparkfun MP3 shield, self made "simple" line out protection PCB.

This setup works like a charm with both headphones and one of those self powered external iphone/ipod portable speakers but its not loud enough and way to "twangy"

Ordered actual speakers: http://www.amazon.com/Yamaha-NS-AW150W- ... 628&sr=8-6

Ordered and assembled the STA540. This time however as soon as power is applied to the amp I get the power light, STBY light, and the Peak light at the same time. switch to "on" and the STBY light goes out like it should but the peak light remains. Disconnected the speaker and the peak light is still on. (both POTs completely counter clockwise during all of this.) I metered all points listed on page 10 of the instruction manual and everything was really close to the figures listed (prior to and since applying power) but I can't find anything else wrong. I'm a hobbyist and not an expert so any info much appreciated. The only other things worth mentioning is that the amp is being powered with a regulated Astro power supply (13.8V at 12.5 amps max) currently dialed to 12.1vdc and I'm currently only trying to use one channel and one speaker output. The peak light is half as bright as the other 2 LED's and I read 10.35vdc at the DIAG test point behind the heat sink. I was unable to read Hz at that point however (thought it might have been PWM causing the dim LED)

In any case I get no output at the speaker at all (whether its on the left or right output with the input switched respectivly) and Im at a loss. I cant see any solder bridges or any defects. Any help much appreciated!

[Mee_n_Mac]

Read the comments on the product page and see if some of the problems described there sound like yours.
http://www.sparkfun.com/products/9612
Going by those comments I might suggest trying a different power supply, a car battery perhaps ? Also be careful about how the "ground" (GBUF) signal from the MP3 shield is tied into your system. It should NOT be connected to ground in your amp as GBUF is not really a buffered ground. It's really a DC voltage output, intended to be used to remove the DC offset on the right and left outputs. Your amp has blocking caps that perform an equivalent function for the R/L inputs.
What I think is happening is your amp is oscillating (hence the peak LED) and this might be due to a number of reasons, ground loops being one of them. What, if any, true ground tie is there between the MP3 shield, or Uno, and the amp ?

See page 14 re: the GBUF signal if you're curious.
http://www.vlsi.fi/fileadmin/datasheets/vlsi/vs1053.pdf

[strokedmaro]

The uno/mp3 shield were powered with USB and the amp separately from the power supply. I don't think a ground loop is the problem with this setup. The only connection between the two is the mp3 shield audio out to the signal in of the amp. This condition also occurs with power connected to the amp with no inputs or outputs at all. I did read about power supplies causing problems so I'll report back after I try a few others

[strokedmaro]

Still no luck Tried a 19v laptop power supply with the exact same results. Rechecked all components to make sure they were oriented correctly and in the right spots...re-metered all the points described and I found all checked perfectly again.

[Mee_n_Mac]

Still no luck Tried a 19v laptop power supply with the exact same results. Rechecked all components to make sure they were oriented correctly and in the right spots...re-metered all the points described and I found all checked perfectly again.

Hmmm, and this is w/o anything input or speakers on the outputs ? It might be interesting to know what the L&R outputs are doing ... are they stuck solidly at either the + or - rail or oscillating between them. If you have a battery (hence floating) powered DMM then perhaps you could measure the DC (and AC) voltage of the L and R outputs. I'm not sure what to make of this ATM but I'll drink, errr ... think on it a bit and see if something smart comes as a result.

FWIW : Does the same thing happen when you power it up in either "standby" or "amplify" modes ?

[strokedmaro]

The attached picture is with no inputs or outputs connected and the switch in "on" just so its less blinding in the picture...only power from a 19v laptop supply. I couldnt read any vdc at either output pair but I had .057 vac on both outputs if that means anything. Position of the switch on power up has no effect on the lights or operation. The pots also have no impact at all. Thanks for drinking about it ^

picture: http://s1229.photobucket.com/albums/ee4 ... 6002499349

[Mee_n_Mac]

I couldnt read any vdc at either output pair ...

Do you mean no DC between the output terminals (+ to -) or from any output to ground ? Given there's no caps on the outputs, I'd expect to see VCC/2 on each output (relative to ground) with no input signal(s).

[strokedmaro]

I only tried (+ to -) at each output pair but you may be on to something. I did some further readings and think I may have found something. By "relative to ground" do you mean between power connector ground to each output? If so, then Im reading it exactly opposite. When I check between power connector ground and each output + or - I only read about 3mV. However when I check from power connector VCC to each output + or - I read VCC. Its as if each + and - output is a perfect ground with the board powered. With power off Im reading in the realm of 8-9K ohms between power connector ground and each output and not "10's of K's as described in the final check on page 10 of the manual. Thank you so much for your efforts so far!

[Mee_n_Mac]

I only tried (+ to -) at each output pair but you may be on to something. I did some further readings and think I may have found something. By "relative to ground" do you mean between power connector ground to each output?

Yes, that's what I meant. The STA540 IC will output a low on the DIAG pin (making the peak LED come on) whenever it detects either :
- a too hot condition (clearly not your problem)
- a too small resistance to ground on either the + or - output pin (again, seemingly not your problem as 8K >>> tens of ohms)
- the output running very close to either ground or VCC. (Hmmmm ?)

If so, then Im reading it exactly opposite. When I check between power connector ground and each output + or - I only read about 3mV. However when I check from power connector VCC to each output + or - I read VCC. Its as if each + and - output is a perfect ground with the board powered. With power off Im reading in the realm of 8-9K ohms between power connector ground and each output and not "10's of K's as described in the final check on page 10 of the manual. Thank you so much for your efforts so far!

Well 8K isn't that bad, it's not a short to ground ... which is a semi-likely fault condition. That may well be your DMM looking into the output transistors of the IC and biasing them ever so slightly on. Did you use a "diode" setting when doing the measurement ? If so switch to a 'k ohms' setting (if your DMM is not autoranging) and see what you get.

I've yet to figure out 2 apparently contradictory measurements you've made. The DIAG pin DC voltage is something betwen Vcc and 0V, indicating that it's switching between the 2 states fast enough that your DMM reads something of an average. This says the output is not shorted to VCC or ground and the amp is oscillating for some reason. That said, I'd expect a measurement at either the + or - terminals (to ground) would also read some average as the output banged from VCC to 0V ... and that average should be VCC/2 or in your case about 6V. Yet you've measured 0V. FWIW I'd expect the AC output signal, if oscillating, to be too high in frequency to be measured by almost all DMMs. That's what I think you measure ~0V on the ACV setting.

Perhaps you could measure (DCV) the STA540 IC inputs, pins 4,5, 11 and 12 to see if they look odd. My guess is that the IC will bias these at VCC/2. Also check the DC voltage at pin 7. Lastly I still wonder why people were reporting problems at turn on with various power supplies. If you have a 12V car/ATV/PWC battery handy ...

EDIT : I don't know what page 10 of the manual says for debugging but if you haven't already it may be worth ensuring good voltage (VCC) at the IC itself on pins 3 and 13. And (w/o power) a hard short to ground at pins 8 and 9.


http://www.sparkfun.com/datasheets/Comp ... STA540.pdf
http://www.sparkfun.com/datasheets/Kits ... %20v11.pdf

[strokedmaro]

- a too small resistance to ground on either the + or - output pin (again, seemingly not your problem as 8K >>> tens of ohms)

not "10's of ohms"...."10's of k's of ohms"
(With power off Im reading in the realm of 8-9K ohms between power connector ground and each output and not "10's of K's )

Pin 4 and 5 read about 335 mV to ground while pins 11,12 and 7 each read about 550 mV. These readings were not very stable but hovered in these areas. Pin 3 and 13 had good VCC.

I also tried a 12 volt battery (12.8 exactly) and again the results were the same.

On a good note however, SparkFun picked up on this post and has shipped me a replacement free of charge! I will do some comparisons and report back once it arrives!