TPA6120A2: Where does that integrated noise spec come from?

Hello Michael,

Below you will find the test configuration for this spec:

Bandwidth used here looks to be from 22Hz to 22kHz.

Best,

Hasan Babiker

I am surprised that anyone had a clue - the spec line says gain of 1? Over what frequency span for integration?

Ok, here you go using the web model - And I will tell you this model is way low on the current noise terms. How do we get 0.9uVrms out of this, 

Hello Michael,

Yes this testbench seems to be different than what is indicated on the datasheet. Load in datasheet is also indicated to be 16-ohms. 

I was unable to find the details for the 0.9uVrms measurement, however the above configuration that I attached measured around 0.98uVrms between 22Hz & 22kHz. This doesn't seem to deviate too far from TINA sims.

Best,

Hasan Babiker

Well Hasan, the load does not matter for a noise sim

1. I was guessing the 0.9uVrms was a TINA sim of the THS6021 model (I doubt the audio guys went through the trouble of hand calculation from the spot noise plots)

The gain of 1, I was guessing just a 1kohm feedback, which if I go 10Hz to 20kHz I get the 0.9uVrms. This is of course deceptive since the current noise in the model is way low. 

Hello Michael,

Yes I agree the current noise for the THS6012 to be around a magnitude lower than what is indicated on the device datasheet. I was highlighting the load to confirm the the TPA6120 test configuration I provided wasn't what was used in the datasheet. Can't say for certain how the 0.9uV measurement was done but the gain =2 configuration along with the 0.98uV measurement were done on bench.

Best,

Hasan Babiker 

Really, is that something you did or in some notebook? 

I tried that test bench with all thos 800ohm R's through 22kHz, with the corrected current noise I get 2.2uVrms. Hard to read on your plot but looks like it might be 1.5uVrms. Is that really a bench measurement? or maybe a sim. What instrument emulates the total output noise sim in TINA.  

I removed the external current noise sources and ran the original model with your circuit, yes I get 0.9uV in sim, you should not get that on the bench as the current noise is physically much higher, 

Hello Michael,

No I didn't take that measurement, was documented around the time of device release where an AP analog analyzer was used. Some more notes on that specific measurement:

• Ferrite cores were added to reduce effects from power supply
• Radiation shield box was used

Best,

Hasan Babiker

Interesting, Well I am mystified, I can't get the sim using the corrected current noise models to get anywhere new that. 

Ok I see now, I need to take your test bench at face value and include the attenuator at the output - I have been looking only at the op amp output pin, 

Once you add that, and the current noise correcting elements, I get good agreement to the 0.9uV number

How would anyone know all this was there for that spec line???