low cost ADAU1452 China board...

[torgeirs]

Could not get mine to go under -40 dB 2. harmonic with DC on signal. From -20 dB. I can't measure the DC on the opamp as it is hidden under the ADAU board. Best result about -2dB FS DC added to signal.

 

[Dmitrij_S]

You can make a very simple angle adapter

 

[torgeirs]

Smart!
Why have I not thought of that idea?

 

[torgeirs]

Seems at some point they sold with the rigth opamp. Wonder how they measured?
https://www.diyaudio.com/community/threads/does-opamp-offer-gain.372018/post-6646189

 

[torgeirs]

Ah the OP in that thread could have bought two boards and used them differential. Then 1.8 volts out.
The inverting of the negative diff leg can be done in the ADAU

 

[torgeirs]

Think I will just go forward with the CS42448 codec card. It is easy to desolder the output components and make my own with components that dont have worse spec than the codec. And using it differential 4 channel out.
Just 2 x 40$ for two combos.

 

[Dmitrij_S]

To verify the measured results, I performed a simple LTSpice simulation of the DAC output filter with opa1612. The simulation qualitatively confirms the measurements. Therefore, opa1612 needs to be replaced!

Whereas opa2340 looks fine !

 

[torgeirs]

Hm tried a soic OP1678, thinking it is a opa2340 equivalent.
I’m not a good smd solderer but checked for shorts and connections.
Terrible noise floor at-80 dBu. Distortion masked by the noise
The original opamp has about -120 to -130dBu noise in same passive components. (But terrible HD as discussed)

How is that possible. What can be the user fault😝

 

[theAnonymous1]

Maybe because OPA2340 is rail-to-rail input and output and OPA1678 is only rail-to-rail output?

 

[torgeirs]

Would expect large distortion then. Which is’t there.
But why large noise with no signal?
And guess it should handle +-1volt input?

 

[theAnonymous1]

Specs say 2V below positive rail. So with a single 5V supply, I don't think it can handle a 1V input.

 

[torgeirs]

Voltage​	Supply voltage, VS = (V+) – (V–)​	​	40​	V​
Input voltage​	(V–) – 0.5​	(V+) + 0.5​	V​

Common mode is (V+)-4 in my setup

VCM​

Common-mode voltage range​

​

(V–)+0.5​

​

(V+) – 2​

V​

 

[torgeirs]

Still dont see why common mode voltage should create white noise on one opamp but not another

 

[torgeirs]

 

[theAnonymous1]

 

[theAnonymous1]

I simulated the circuit with OPA1678. With a 5V supply, it does not like anything more that a measly 4-5mVrms before the waveform distorts. For a 1Vrms input, the supply voltage needs to be close to 11V.

 

[torgeirs]

I think my board/codec has some damage.
Do you get the same result with 1.5 volt common mode? That is adding a 1.5 volt dc to your signal.
That is what a working board outputs from the codec

 

[torgeirs]

Mine is 0.2 volts on new opamp. And -80dBV noise on amp input. So no need to blame amp for the noise;-)
But I got almost a volt of signal out of it.
So maybe something happens that I dont know
Sorryaboutthat

 

[torgeirs]

But nice to know if the opamp is suitable for cheapdacs as it is cheap itself

 

[theAnonymous1]

I think my board/codec has some damage.
Do you get the same result with 1.5 volt common mode? That is adding a 1.5 volt dc to your signal.
That is what a working board outputs from the codec

Sorry, I had put a cap on the input for some dumb reason and also forgot sig gen is Vpp not RMS, so ignore my previous post. It can put out around 0.5Vrms with 1.5V common mode on 5V supply. OPA2340 can put out 1Vrms, so twice as much.