Understanding Amplifier Noise in IEMs

[MulattoKid]

Hello everyone!

Disclaimer: this is my first week of properly diving into all of this, so bear with me

Some years ago I bought a Schiit stack: Modi 2 Uber + Magni 3 (while I'm new here I've noticed Schiit not always getting the best mentions here, but anyway...bought is bought...for now at least). I've used this stack to drive a few different headphones without ever giving it much thought.

However, last week I received my pair of custom-fit JH7 IEMs! I was so excited to try them out, but I could hear a lot of noise when using them with the stack. When connecting them to my Mac M1 Air there was no audible noise. I was a bit puzzled...

Back on my desktop and Schiit stack, I had my volume in Windows turned quite low. I therefore had turned the volume knob on the AMP up quite high. After doing some reading I read that it's better for audio quality to turn the volume in Windows all the way up to avoid Windows' mixer adjusting the amplitude in the signal it sends out. So, I did that, and then naturally had to turn the volume knob on the AMP down. The noise was no longer audible in the volume-knob-range I use the AMP in, so in a way all was good.

But, being who I am I started trying to understand what caused the noise. I disconnected the AMP from the DAC, plugged in my IEMs, and turned the volumne knob up. I could again hear the same noise. The more I cranked up the amplifier, the louder the noise became. This tells me that it's the AMP that's introducing most of the nosie I heard originally. So, onto the specs
JH7 IEM:

• Input sensitivity: 124dB @ 1mW
• Impedance: 17 Ohm

Magni 3:

• Maximum Power, 16 ohms: 3W RMS per channel
• THD: Less than 0.001%, 20Hz-20KHz, at 1V RMS, less than 0.02% at 5V RMS into 32 ohms
• SNR: Greater than 108db, unweighted, referenced to 1V RMS
• Output Impedance: Less than 0.3 ohms at either gain
• Input Impedance: 25k ohms
• Gain: 1.4 (3db) or 7 (17db), selectable via rear switch

Here's what I've been able to put together so far (I've spent a lot of time on this, but still struggle a bit):

1. I've set the AMP to low gain - that's good for my situation, as I'm in no need of extra amplification
2. The Magni's input impedance is quite high, so there should be a big voltage drop and little actual current flowing from the DAC into the AMP - that's good AFAIK
3. The Magni's output impedance is low, so there should only be a small voltage drop and most of the output current will flow to the IEMs - that's good AFAIK
4. The Magni's THD and SNR seem reasonable to me compared to what I've seen for other AMPs
5. The JH7's have relatively low impedance (17 Ohm), so the voltage drop will be relatively small
6. The JH7's sensitivity is pretty high (124dB @ 1mW), so as I understand it: less power is needed to create loud sounds

Given all of this, here's what I'm thinking: the IEMs have relatively low-impedance, so the output voltage from the AMP will result in more current flowing through them, compared to some other high-impedance headphones. Next, the IEMs are very sensitive, so it doesn't take much power for them to generate loud sounds. Thus, when my volume in Windows was low and I had to turn up my AMP volume, the amplified noise from the AMP became a bigger part of the final signal because the audio signal received from the DAC was pretty small in terms of voltage. Am I making sense here? Also, please correct me if I made any incorrect statements - I really want to learn

Thanks, Daniel

 

[staticV3]

There are some misunderstandings WRT impedances and voltage drop.

The Magni's input impedance is quite high, so there should be a big voltage drop and little actual current flowing from the DAC into the AMP - that's good AFAIK

High Amp input impedance -> low load imedance for the DAC -> minimal voltage drop -> this is good!

The JH7's have relatively low impedance (17 Ohm), so the voltage drop will be relatively small

Low load impedance -> Amp will clip sooner. Voltage drop will only occur once you turn up the Amp to 100% and start increasing the load.

Am I making sense here?

Yes.

 

[DVDdoug]

Back on my desktop and Schiit stack, I had my volume in Windows turned quite low. I therefore had turned the volume knob on the AMP up quite high. After doing some reading I read that it's better for audio quality to turn the volume in Windows all the way up to avoid Windows' mixer adjusting the amplitude in the signal it sends out. So, I did that, and then naturally had to turn the volume knob on the AMP down. The noise was no longer audible in the volume-knob-range I use the AMP in, so in a way all was good.

You are describing "gain staging". Generally you want as much signal as early in "the chain" as possible to get the best signal-to-noise ratio (as long as nothing gets pushed into clipping). Every analog stage will add some noise and anything amplified later boosts the signal and the (accumulated) noise together.

The published signal-to-noise specification of any device is usually measured with the maximum output signal. (Amir may use a standardized signal level.... I think he uses 5W for power amplifiers.)

Given all of this, here's what I'm thinking: the IEMs have relatively low-impedance, so the output voltage from the AMP will result in more current flowing through them, compared to some other high-impedance headphones. Next, the IEMs are very sensitive,

It's really just sensitivity but that's (somewhat) related to current.

More current means more electrical power (Watts or milliwatts) which is converted to acoustic power. But there is also an efficiency factor so although a lower impedance consumes more electrical power it doesn't always mean "louder" if the IEM, headphone, or speaker is less efficient at converting electrical energy to sound.

 

[MulattoKid]

Thanks for the replies! I'll try to tackle these one by one

High Amp input impedance -> low load imedance for the DAC -> minimal voltage drop -> this is good!

I don't know how I ended up thinking a big voltage drop was good...this makes more sense. I did some searching and I think what's at the start of an AMP is a voltage divider, correct? If so, why is that needed; what would happen if it was removed?

 

[MulattoKid]

I think I understand know. The voltage divider allows minimal voltage drop into the Amp while simultaneously limiting the current flowing through the circuit. My mistake was forgetting the the current is constant in the circuit.