Budget passive preamp with stepped attenuator?

[Deleted member 58865]

There is no ground loop. Only my computer is grounded through a 3-prong AC inlet. Everything is grounded through my computer.

But, if I buy a new computer monitor, I'd like to buy a monitor that has a 3-prong AC inlet for grounding. Everything is connected to a power strip. But, there can be a low-intensity ground loop current flowing in the HDMI cable between the monitor and the computer. I don't know whether that will affect audio.

 

[Deleted member 58865]

I ended up buying FV3 SMD Dact 21-step stepped attenuator from ebay.com.

I didn't order a power supply because I concluded I can reduce a lot of noise in the entire computer system by buying a new computer monitor with a 3-prong AC inlet.

The current computer monitor is sending a lot of noise to the entire computer system through the HDMI cable because the noise created by the monitor is grounded via the HDMI cable. I think it may be increasing the noise floor a lot. A monitor with a 3-prong AC inlet grounds its own noise through the AC cable instead of the HDMI cable.

If you have a computer monitor without a 3-prong AC inlet, buy one with it. It will reduce a lot of noise in the entire audio/computer system.

If I buy a power supply, I may purchase GST60A24-P1J because the minus side of DC plug is not grounded. GSM60A24-P1J has the minus side of DC plug connected to AC FG(frame ground). It may create a low-intensity ground loop even though everything is connected to a power strip.

 

[Deleted member 58865]

GSM60A24-P1J has the minus side of DC plug connected to AC FG. Does that mean an amplifier can be grounded through GSM60A24-P1J?

 

[solderdude]

Yes, it also means you could be bothered by a ground loop.
Mains SG -> PC gnd -> USB gnd -> DAC GND -> audio Gnd (common) -> power supply -> mains SG
When using the same wall socket there is no ground loop issue. Only when you connect to 2 different outlets that both have safety ground and have currents flowing through ground.

 

[Deleted member 58865]

When using the same wall socket there is no ground loop issue. Only when you connect to 2 different outlets that both have safety ground and have currents flowing through ground.

Fortunately, everything is connected to a power strip.

 

[Deleted member 58865]

Upon further contemplation, I realized increasing SNR by 3~4dB with an expensive power supply is foolish.

There are two ways to increase SNR by 20dB or more without an expensive power supply.

• Buy an amplifier with a low gain or an amplifier with selectable gains. One of selectable gains should be low.
  • With a low gain, you don't need to attenuate nearly as much.
• Attenuate the output of an amplifier.

Is there a (stepped) attenuator for amplifier output? Even a fixed 20dB attenuation at the amp output can be good.

 

[notsodeadlizard]

Unfortunately, the concept of "budget" for such a device may not correspond to the generally accepted perception of the level of "budget prices".
If this is an honest device, a resistor divider is required with a constant output resistance, for example.
There is no secret in such dividers, they have been well known for a long time (at the same time they also have a logarithmic scale): https://en.wikipedia.org/wiki/Logarithmic_resistor_ladder
In addition, very good relays are needed, it is highly desirable with sealed contacts.
In addition, you need to manage all this, reliable and convenient.
In general, if you try to do this for yourself, you will be surprised at the resulting price and effort.
I did it and I know what I'm talking about.
The result is of course very good, if the resistors are of high quality and everything is done as it should be.
But I wouldn’t call it budget, it can be said to be “high end”, if these words mean anything at all.
I say this to the fact that such budget devices are simply impossible, their customer audience is very small, and the manufacturer must receive at least some profit in order not to go bankrupt.

 

[mdsimon2]

Upon further contemplation, I realized increasing SNR by 3~4dB with an expensive power supply is foolish.

There are two ways to increase SNR by 20dB or more without an expensive power supply.

• Buy an amplifier with a low gain or an amplifier with selectable gains. One of selectable gains should be low.
  • With a low gain, you don't need to attenuate nearly as much.
• Attenuate the output of an amplifier.

Is there a (stepped) attenuator for amplifier output? Even a fixed 20dB attenuation at the amp output can be good.

Why are you so obsessed with improving SNR? Do you have audible noise currently?

Michael

 

[Deleted member 58865]

Why are you so obsessed with improving SNR? Do you have audible noise currently?

AIYIMA A07 has 21.5dB of gain. I probably have a previous iteration of AIYIMA A07 that has more than 21.5dB of gain. This is too high. I already keep the amp volume pot at 9'o clock. Because it has channel imbalance below 9'o clock, I set the DAC volume to -20dB.

I have probably lost about 40~60dB of dynamic range because the amp's gain is too high.

The problem is that if the gain is too high, you will lose a lot of dynamic range through attenuation before the amplification.

I want a low gain, no gain, or adjustable gains. The ultimate solution is for an amp to have a volume knob that adjusts the gain instead of the attenuation before the gain.

 

[mdsimon2]

AIYIMA A07 has 21.5dB of gain. I probably have a previous iteration of AIYIMA A07 that has more than 21.5dB of gain. This is too high. I already keep the amp volume pot at 9'o clock. Because it has channel imbalance below 9'o clock, I set the DAC volume to -20dB.

I have probably lost about 40~60dB of dynamic range because the amp's gain is too high.

The problem is that if the gain is too high, you will lose a lot of dynamic range through attenuation before the amplification.

I want a low gain or adjustable gains. The ultimate solution is for an amp to have a volume knob that adjusts the gain instead of the attenuation before the gain.

Do you currently have audible noise?

Michael

 

[Deleted member 58865]

Do you currently have audible noise?

No. But, I don't like too much gain which has to be compensated for with a lot of attenuation.

Are you appealing to subjective perceptions? Or, do you have any scientific argument? What's your argument?

 

[Jimbob54]

No. But, I don't like too much gain which has to be compensated for with a lot of attenuation.

Are you appealing to subjective perceptions? Or, do you have any scientific argument? What's your argument?

I think he is asking why you have embarked on this quest. If you have no audible noise, why seek to minimise it?

 

[mdsimon2]

No. But, I don't like too much gain which has to be compensated for with a lot of attenuation.

Are you appealing to subjective perceptions? Or, do you have any scientific argument? What's your argument?

If you can't hear it there is no point in minimizing it.

As explained to you earlier I've done a good amount of research comparing noise measurements to my own listening thresholds. I know that noise in the low 100 uV range at the amplifier output is inaudible for my hearing, speakers and listening distance. Therefore I don't get too excited that one of my setups has ~80 uV residual noise and one has ~130 uV residual noise as both are inaudible.

Michael

 

[Deleted member 58865]

Anyway, upon further research, I found out that attenuating the output of an amplifier is impractical because it wastes a lot of energy.

Thus, next time I buy an amp, I will buy one with a low gain, no gain, adjustable gains, or a volume knob that adjusts the gain level.

That's the simple lesson I learned through research. The less gain there is, the less you have to attenuate. The less you attenuate, the more SNR you have.

Reducing the gain is the simplest way to improve SNR substantially.

The conclusion is that looking for other ways to improve SNR before reducing gain is foolish. That's why I'm going to stop researching for now.

 

[Deleted member 58865]

I know that noise in the low 100 uV range at the amplifier output is inaudible for my hearing, speakers and listening distance. Therefore I don't get too excited that one of my setups has ~80 uV residual noise and one has ~130 uV residual noise as both are inaudible.

How did you find out the noise level in uV? How can I figure that out for my amp?

 

[NTK]

21.5 dB gain will give 70.6 W output (8 ohm load) with a 2 Vrms input. That is not too high.

 

[mdsimon2]

How did you find out the noise level in uV? How can I figure that out for my amp?

To measure it you need an ADC -> https://www.audiosciencereview.com/...a-with-topping-performance.24768/post-1008785

I've already shown you how to calculate it based on ASR measurements -> https://www.audiosciencereview.com/...mp-with-stepped-attenuator.43381/post-1538281

Michael

 

[Deleted member 58865]

21.5 dB gain will give 70.6 W output (8 ohm load) with a 2 Vrms input. That is not too high.

I listen at low volumes. At 9'o clock, I still need to set DAC volume to -20dB. Below 9'o clock, I have channel imbalance.

If the volume pot is at max position, I find myself setting DAC volume to something like -50dB.

 

[Deleted member 58865]

Without a lower noise / lower gain amplifier you will not be gaining much from using a higher dynamic range DAC or an analog volume control.

This. Next time I buy an amplifier, I will buy a low gain amplifier or an amplifier with selectable gains. That's the lesson.

 

[Deleted member 58865]

Even if I have no problem with noise, too much gain makes the volume pot pretty much unuseable, and the volume pot has channel imbalance below 9'o clock.

I never go above 9'o clock on the volume knob. I am 50 centimeters away from my speakers. Anything above 9'o clock is too loud.