Hi

I have designed an acoustic guitar preamp. It has 1 NE5532 and 3 TL074s running at +/-15vdc.


Many good preamps list their noise floor, like this:


I have a good scope (Siglent SDS804x hd) and a AWG. Could I replicate some of those type of tests? I also have some OPA4134s I thought about trying.

here’s a picture of me looking at bode plots of it, one of the knobs is a notch filter

 

I don’t think that your setup can achieve a 128dB range.
There’s a reason why audio analyzer equipment like the Audio Precision is expensive, and it is because of the wide dynamic range.
Additionally you must provide a bandpass circuit function which should not introduce noise of its own.

I may be wrong, I am not familiar with your scope, but perhaps your best bet is to lease an analyzer equipment.

 

0 dBm is 1 mW into 50 Ω, that is 224 mV.
-120 dBm is 0.001 pW or 224 nV, that is 0.2 μV.

Can your test equipment measure less than 1 μV?

 

No my scope only does 0.5 mV.

are there simpler tests I could perform to compare it running TL072s vs OPA4134s?

 

No my scope only does 0.5 mV.

are there simpler tests I could perform to compare it running TL072s vs OPA4134s?

Yes, use your ears.

 

I have a good scope (Siglent SDS804x hd) and a AWG. Could I replicate some of those type of tests?

Most of the tests could work, maybe even the noise test. You can certainly try by switching to the frequency domain. The scope is only a 12-bit model but the huge 2Mpoint FFT capability, >10Mpoints aquisition, time domain/FFT averaging, windowing and oversampling adds a lot of processing gain. On my 12-bit Rigol, depending on settings, I can reach a -140dBV noise floor and have a usable dynamic range of 120dB at audio. You obviously can't compare a 12-Bit scope to a decent 24-bit ADC but a lot of audio tests can be done with it.

 

Returning to your project...it looks really beautiful!
And I expect that it will sound as good as it looks

And a cautionary tale: Texas has decided to re-design the venerable TL074 and replacing the JFET input stage with a CMOS input stage. Renamed it the TL074H. Those devices offer several improvements, one of them definitely NOT being noise. In particular the 1/f noise has been reported to be significantly worse.

 

TL074 is a JFET, ne5532 is bipolar. The new TL074H is CMOS. I don’t understand the difference.

for audio at relatively low frequencies (<20khz), requires low or very low noise and high impedance. Current draw isn’t especially important since most use a PS.

 

Returning to your project...it looks really beautiful!And I expect that it will sound as good as it looks

thanks! If you’re interested, I could send you a set of PCBs. Requires probably $90 in parts.

 

Unfortunately, I don’t play the guitar.
But I appreciate the looks.

Is that a custom enclosure?

 

Tayda will drill/paint/UV aluminum enclosures. This one is a 1590XX, $13 for the enclosure + $13 drill/paint/UV. Smaller ones, like for Q-tune, are $6+6 for 1590b and 125b sizes.

 

Most of the tests could work, maybe even the noise test. You can certainly try by switching to the frequency domain. The scope is only a 12-bit model but the huge 2Mpoint FFT capability, >10Mpoints aquisition, time domain/FFT averaging, windowing and oversampling adds a lot of processing gain. On my 12-bit Rigol, depending on settings, I can reach a -140dBV noise floor and have a usable dynamic range of 120dB at audio. You obviously can't compare a 12-Bit scope to a decent 24-bit ADC but a lot of audio tests can be done with it.

ok. I’m an ME, not an EE. A lot of this is new to me. I am still a novice scope user. What’s an easy test or 2 to try?

Plug the DUT output into Ch1. Leave DUT input open or apply a load? What would I measure?

 

I am also waiting for 0ri0n to elaborate his technique.

 

I made some plots last night on my Siglent SDS804xhd, still learning how to use it.
AWG providing 200 Hz.

Ch2 is my preamp that I designed. It runs at +/- 15v
Ch3 is a commercially available pedal running at +9.5v single sided.