Audio circuit- high frequency background hiss

[Jester]

I just completed initial testing of an audio device that I redesigned for someone. This device sends a low level signal to a PA or guitar type amplifier. Everything works as expected except that when the volume of the amplifier is set above say 1/2 volume, there is a noticeable high frequency hiss. I would describe the hiss as very similar to what you would hear when you adjust the tuner of an AM radio to a frequency with no signal.

This device has a very limited bandwidth 175 - 2400Hz, the noise is well beyond 2400Hz, sounds like > 7kHz or above.

The 2nd last stage is a 5th order SC filter (MAX7426) set to fc=2800Hz. Using a signal generator I plotted the filter response and the signal is 40mVp-p or less outside the pass band all the way up to 10kHz, the signal level in the pass band is adjustable and presently set to about 2Vp-p.

The output of the MAX7426 drives an op amp with a gain of 2 (that uses a 2.5V pseudo ground) that then drives the external amplifier with a series combination of 220 Ohms and 150nF to ac couple the signal output to the external amplifier.

Is 40mVp-p too much noise?

If so recommendations, to quiet it down?

 

[Anna Conda]

no doubt you can hear it at 40mVpp, try substituting an LC filter for the MAX7426, and re-listen to see if it is the culprit?

Also power supply noise can get in at every stage, especially the external amp!, try shorting the amp input with a 1k resistor and then wind up the volume - if hiss still there = amp problem... (not uncommon).

 

[BradtheRad]

IEverything works as expected except that when the volume of the amplifier is set above say 1/2 volume, there is a noticeable high frequency hiss. I would describe the hiss as very similar to what you would hear when you adjust the tuner of an AM radio to a frequency with no signal.

Does the hiss come on suddenly? As though it could be ultrasonic oscillations? You might not hear a pitch, but you might hear noise.

Does the hiss stop suddenly, even when you reduce volume gradually?

Have you examined the output with an oscilloscope?

If your desired audio signal is good, then the hiss is not likely to be parasitic oscillations. Also, you can disregard this idea if gain is very low, or if your circuit contains precautions against parasitic oscillations.

 

[Jester]

Does the hiss come on suddenly? As though it could be ultrasonic oscillations? You might not hear a pitch, but you might hear noise.

Does the hiss stop suddenly, even when you reduce volume gradually?

Have you examined the output with an oscilloscope?

If your desired audio signal is good, then the hiss is not likely to be parasitic oscillations. Also, you can disregard this idea if gain is very low, or if your circuit contains precautions against parasitic oscillations.

No, and No

Here is what the noise looks like.

I stated earlier that the noise was likely > 7.5kHz, it's definitely higher than that, I injected a signal using a signal generator and swept the frequency up, anything less than 10kHz is more of a whine (pretty unpleasant to the ear), around 10kHz my hearing starts to roll off, and if I crank up the signal generator at 10 kHz it sounds somewhat like the noise I'm trying to get rid of.

 

[Audioguru]

Maybe you have a very noisy opamp. Maybe an active part is defective. Maybe the power supply filtering is poor.

 

[Jester]

Maybe you have a very noisy opamp. Maybe an active part is defective. Maybe the power supply filtering is poor.

The background noise looks to be about 4mV with no signal (see scope capture above)

Power supply ripple is 30mVp-p

The opamp is low noise (19nV/√Hz): https://www.digikey.ca/product-detail/en/OPA2170AIDGKR/296-36694-1-ND/4341325, PSRR is ~50dB@10kHz, so I don't think its the opamp

The clock feed-through of the SCF (MAX7426): https://www.digikey.ca/product-detail/en/MAX7426EUA+/MAX7426EUA+-ND/1513945 is 5mVp-p, however that should be at > 100kHz

- - - Updated - - -

no doubt you can hear it at 40mVpp, try substituting an LC filter for the MAX7426, and re-listen to see if it is the culprit?

Also power supply noise can get in at every stage, especially the external amp!, try shorting the amp input with a 1k resistor and then wind up the volume - if hiss still there = amp problem... (not uncommon).

Does not appear to be the external amplifier, If I short the output of my signal the noise is gone, if I just load it, say with 1k, the noise is still there perhaps attenuated slightly.

 

[KlausST]

Hi,

a schematic would help to understand the circuit.

Are you sure the output OPAMP is stable to drive 200R and 150nF load? Especially the 150nF is a big capactive load.
do you use a switch mode power supply? (Try to use a battery to power the circuit)


Klaus

 

[Jester]

Hi,

a schematic would help to understand the circuit.

Are you sure the output OPAMP is stable to drive 200R and 150nF load? Especially the 150nF is a big capactive load.
do you use a switch mode power supply? (Try to use a battery to power the circuit)


Klaus

It does have a switcher, however it's well filtered, I tried a linear supply and no difference.

I found it.... the last stage before the external amplifier is an inverting summer, and the non inverting input to ground resistor was incorrect, much better now.

How I found it: I connected a probe to the external amplifier and then listened at various points in the circuit, there was a noticeable increase in noise at the last stage.

Thanks for all the help.

 

[D.A.(Tony)Stewart]

Since your x2 Op Amp has a bandwidth of around 1MHz driving a capacitive cable load plus termination, it could be oscillating and you are hearing the thermal noise in the output stage.

Load the output with something like 0.1uF in series with 100 Ohms to ground ~ 50kHz
Although the 220 Ohm series R ought to prevent this. So layout, decoupling etc and choice of Op Amp and phase margin all play a role.