Fosi Audio ZA3 Amplifier Review

[FanatNsk]

During the experiments it was possible to establish that for Fosi ZA3 and V3 the optimal input signal is no more than 900 mV, with such voltage TD+N 0.7%, if the input voltage is more than 1 V for ZA3 via RCA or via TRS/XLR then the distortion is more than 1%, for V3 the same threshold is 1 V, optimally 900 mV. At the same time, my sound card gives out 2.24 V, and various DACs with TRS/XLR outputs from 1.9 V to 5 V. Why did they make a balanced connection in ZA3 if its input sensitivity is so low?


ZA3 Input TRS 1.1Vrms, out 1Vrms for 8Om TD+N 1.51%

ZA3 input TRS 2.24Vrms, out 1Vrms for 8Om TD+N 19.6% (this good?)

 

[LSPhil]

can you tell us where to measure the output, and on what load.

If you use voltage values, please write what voltage you mean, whether Vss or Vrms.

The relationship is as follows: e.g. 1.41Vss corresponds to 0.5Vrms
2.42Vss corresponds to 1.0Vrms
The TPA3255 can have a maximum of 4Vss at input voltage, of 51V DC and a 4.0 Ohm output of approx. 230Wrms.

 

[FanatNsk]

can you tell us where to measure the output, and on what load.

At the amplifier output, the resistive load is 8 Ohm, the voltage on the resistive load is 5 Vrms. At the amplifier input, I fed a signal from the E-MU1616M sound card with a voltage of 2.2 Vrms and 900 mVRms.

screen 2.2Vrms TD+N 19.5%

screen 900mVrms TD+N 0.73%

 

[LSPhil]

The standard output voltage for HiFi devices is 2.0Vrms. My card gives exactly every 2.0Vrms. I don't understand your problem. Write what's going on with this amplifier. You can write to me privately, because that would probably be the best. You can write in Russian, Polish or German.

 

[Joe0429]

Did they state which model amplifier they were talking about? V3 Mono is balanced internally, so there is an op-amp performing a signal conversion for single-ended inputs on those. For ZA3, it is the reverse, where internally the amplifier is not balanced, so there is an op-amp performing a signal conversion for the balanced input, which means it DOES take advantage of fully balanced differential input signals, but it gets converted down to single-ended for amplification.

-Ed

If I remember correctly, Fosi said ZA3 went like this: BAL -> SE (volume control) -> BAL -> TPA3255.

 

[EddNog]

If I remember correctly, Fosi said ZA3 went like this: BAL -> SE (volume control) -> BAL -> TPA3255.

Wow, this is just dumb. Sheesh.

-Ed

 

[FanatNsk]

The standard output voltage for HiFi devices is 2.0Vrms. My card gives exactly every 2.0Vrms.

I have no complaints about the RCA inputs of the Fosi ZA3, they will probably be used to connect low-power sound sources, such as a laptop (1.2Vrms) or a phone (200uVrms) - so I have no complaints about the unbalanced inputs.

I am more concerned about the TRS inputs of the ZA3, are they not able to correctly receive a 5Vrms signal?

Perhaps not all manufacturers comply with the HiFi standard. For example, here is the specification for the topping DAC, as you can see, the TRS output has 5Vrms.

 

[somebodyelse]

Wow, this is just dumb. Sheesh.

-Ed

Please explain. It seems the correct way to me. A balanced input can achieve its ground loop immunity and CMRR without balanced output. Volume control is easier to do accurately with SE. Actively generating the opposite phase to drive the bridged amps is necessary because there is a single ended input, and because a balanced input can legitimately have a passive leg, usually the cold. A similar arrangement is used in the preamp example at the end of The G Word, but the balanced output there has a passive leg.

 

[antcollinet]

There is a big difference in volume between RCA and XLR with Topping DX5. I read on one forum that the XLR Fosi input is more decorative. It is converted into an RCA input later. That is why I doubt using XLR.

The benefit of XLR is on the connection between devices. Once it is inside the receiving device, it makes no difference if it is converted to single ended - in fact most devices will do this.

Another benefit is the (typically) higher voltage transmitted on XLR - but that is secondary.

 

[antcollinet]

The Texas Instruments says in its catalogue that the maximum input voltage should be 4.0Vpp.
You can calculate your 1.3Vrms and the answer will be self-explanatory

That is for the input to the chip. Many amps will implement a buffer between the input and the chip which may alter the voltage. Volume control will also lower the voltage between input and chip.

 

[antcollinet]

During the experiments it was possible to establish that for Fosi ZA3 and V3 the optimal input signal is no more than 900 mV, with such voltage TD+N 0.7%, if the input voltage is more than 1 V for ZA3 via RCA or via TRS/XLR then the distortion is more than 1%

Correct me if I am wrong, but isn't that because at those voltages the amp is already at max power and you are driving it into clipping?

 

[Sokel]

At the amplifier output, the resistive load is 8 Ohm, the voltage on the resistive load is 5 Vrms. At the amplifier input, I fed a signal from the E-MU1616M sound card with a voltage of 2.2 Vrms and 900 mVRms.

screen 2.2Vrms TD+N 19.5%
View attachment 390985

screen 900mVrms TD+N 0.73%
View attachment 390986

Seems like you're using a multitone signal to test,right?
If that so you must calculate it,during my tests I haven't seen anything useful above -10 or -7dB or so.
We can't use multitone the same way with single sines.
I think even REW warns about it.

 

[LSPhil]

I am not specialized in PA. I asked Ai and got an answer. You can see that you shouldn't mix PA with Hi-Fi Konsument.
AI reply here: Sure! Audio devices with a maximum output voltage of **5V RMS** are typically found in professional or high-end audio equipment. This higher output voltage is often used in **studio gear**, **high-end Hi-Fi devices**, and certain **car audio components** where a stronger signal is required, especially over longer cable runs or when driving amplifiers more effectively.

Here are some types of devices that may provide up to **5V RMS** output: ### 1. **DACs (Digital-to-Analog Converters)** - High-end or professional DACs, used for converting high- resolution digital audio into analog signals, can often output up to 5V RMS or more. Brands like **Benchmark**, **RME**, and **Chord** have DACs with higher output levels to ensure optimal performance.

2. **High-End CD Players** - Some high-end CD or SACD players offer higher output voltages to provide a cleaner signal transfer. Brands like **Esoteric**, **Marantz (Reference series)**, or **Accuphase** often include models with up to 5V RMS output.

3. **Preamps / Preamplifiers** - Professional or high-end preamps, especially those with balanced XLR outputs, can have higher output voltages (up to 5V RMS or more). These are common in **studio environments** or **high-end home audio setups** to ensure better driving capabilities for power amplifiers.

4. **PA and Studio Equipment** - Professional audio gear, such as **mixing consoles**, **audio interfaces**, and **signal processors**, often have higher output voltages (4- 5V RMS) to directly drive amplifiers or other downstream devices in a professional audio setup.

5. **Car Audio Components** - Some **high-end car audio head units** or **car preamps** offer output voltages up to 5V RMS. These are used to feed external amplifiers with a stronger signal, reducing noise over longer cable runs.

In general, **5V RMS output** is more common in professional or high-fidelity applications where maintaining a strong signal-to-noise ratio and minimizing interference are crucial, especially over longer distances.

 

[EddNog]

Please explain. It seems the correct way to me. A balanced input can achieve its ground loop immunity and CMRR without balanced output. Volume control is easier to do accurately with SE. Actively generating the opposite phase to drive the bridged amps is necessary because there is a single ended input, and because a balanced input can legitimately have a passive leg, usually the cold. A similar arrangement is used in the preamp example at the end of The G Word, but the balanced output there has a passive leg.

I think my issue with this whole situation is just that my perspective is that less is better. I’d rather see no conversions (and no volume control) than two back and forth conversions. Essentially, I prefer a, “pure power amplifier,” like V3 Mono, instead.

I don’t tend to see value in a volume control without a remote when it comes to speaker amps, and having a double signal conversion to allow that just makes it worse in my eyes. Worthless added cost at the minimum, even if there is no (audible) impact on performance.

-Ed

 

[antcollinet]

I am not specialized in PA. I asked Ai and got an answer. You can see that you shouldn't mix PA with Hi-Fi Konsument.
AI reply here: Sure! Audio devices with a maximum output voltage of **5V RMS** are typically found in professional or high-end audio equipment. This higher output voltage is often used in **studio gear**, **high-end Hi-Fi devices**, and certain **car audio components** where a stronger signal is required, especially over longer cable runs or when driving amplifiers more effectively.

Here are some types of devices that may provide up to **5V RMS** output: ### 1. **DACs (Digital-to-Analog Converters)** - High-end or professional DACs, used for converting high- resolution digital audio into analog signals, can often output up to 5V RMS or more. Brands like **Benchmark**, **RME**, and **Chord** have DACs with higher output levels to ensure optimal performance.

2. **High-End CD Players** - Some high-end CD or SACD players offer higher output voltages to provide a cleaner signal transfer. Brands like **Esoteric**, **Marantz (Reference series)**, or **Accuphase** often include models with up to 5V RMS output.

3. **Preamps / Preamplifiers** - Professional or high-end preamps, especially those with balanced XLR outputs, can have higher output voltages (up to 5V RMS or more). These are common in **studio environments** or **high-end home audio setups** to ensure better driving capabilities for power amplifiers.

4. **PA and Studio Equipment** - Professional audio gear, such as **mixing consoles**, **audio interfaces**, and **signal processors**, often have higher output voltages (4- 5V RMS) to directly drive amplifiers or other downstream devices in a professional audio setup.

5. **Car Audio Components** - Some **high-end car audio head units** or **car preamps** offer output voltages up to 5V RMS. These are used to feed external amplifiers with a stronger signal, reducing noise over longer cable runs.

In general, **5V RMS output** is more common in professional or high-fidelity applications where maintaining a strong signal-to-noise ratio and minimizing interference are crucial, especially over longer distances.

As usual, AI regurgitates the same old "high end" nonsense that is prevalent all over the internet - which is largely what it is trained on.

 

[HDavidson]

If I remember correctly, Fosi said ZA3 went like this: BAL -> SE (volume control) -> BAL -> TPA3255.

This is exactly what I was talking about. Therefore, it may not make sense to connect via XLR, but to connect via RCA immediately to avoid additional and unnecessary signal transformation? Since the amplifier does not have an input level selection regulator, the RCA input sounds 10db quieter than the XLR input and regulates this level.
With topping DX5, maybe with other models it is different.

 

[EddNog]

This is exactly what I was talking about. Therefore, it may not make sense to connect via XLR, but to connect via RCA immediately to avoid additional and unnecessary signal transformation? Since the amplifier does not have an input level selection regulator, the RCA input sounds 10db quieter than the XLR input and regulates this level.
With topping DX5, maybe with other models it is different.

To that end yes, a sensible point of view, given that there is no avoiding that single-ended signal stage due to the integrated volume attenuator.

-Ed

 

[LSPhil]

As usual, AI regurgitates the same old "high end" nonsense that is prevalent all over the internet - which is largely what it is trained on.

Great, now we can take advantage of the situation to correct the misleading statements generated by AI. It's best to respond in the same order as presented by the AI. Your clarifications will serve as a reference and bring a lot of truth to the discussion about output voltage levels in certain devices.

 

[antcollinet]

Great, now we can take advantage of the situation to correct the misleading statements generated by AI. It's best to respond in the same order as presented by the AI. Your clarifications will serve as a reference and bring a lot of truth to the discussion about output voltage levels in certain devices.

I'm not going to waste my time responding to an AI.

If people are not willing to think critically enough to see how they are being misled by AI/LLMs, that is on them.

 

[FanatNsk]

Seems like you're using a multitone signal to test,right?

I redid the test on a 1 kHz sine wave. On the graph in red, the measurement at 2.24 V of the input voltage to the amplifier and the output voltage of 2.9 V on a resistive load of 8 Ohms. In green on the graph, the measurement at 0.16 V of the input voltage to the amplifier and the output voltage of 2.9 V on a resistive load of 8 Ohms. As you can see on the graph, there are harmonic peaks - in red. The graph of measurements at 0.9 Volts coincides with the graph of 0.16 volts.

I then recorded some music (Audio test - 84 - Robin S _ Luv 4.wav) with 2.24V going into the amp and 2.9V coming out and found RTA analysis a THD of 102.7%

After that I recorded music, with 0.9V entering the amplifier and 2.9V coming out and found found RTA analysis THD 13.5%

Finally I decided to measure how many volts my sound card produces when listening to dance tracks and saw a signal of 1.5V

I have to state the fact - XLR/TRS connection in ZA3 is completely useless Most DACs give a signal to the amplifier input greater than 0.9V