Fosi Audio's First Dual-Mode Balanced Amplifier ZA3

[antcollinet]

We agree on the theoretical advantage. I have been in the home audio profession my entire life and have almost never (1/2 of 1% would be too high) run into hum, buzz, noise caused by normal length correctly built RCA audio cables in a home. If that 1/2 of 1% noise exists it is very easy to detect by ear and fixed by substituting a better built/shielded RCA cable. So, our difference is in the frequency of the problem, you posit "minority" (49% ?) and I say less than 1/2%.

- Noise is not a recently discovered issue: RCA cables have existed since the 40s, and XLR Cables Since the 50s. XLR was almost never found in even the best home gear until very recently. If the problem was real/common in homes and the solution already known, it would have been in use decades ago. Now tiktok insists XLR is necessary even for sub $200 components.

- Point to an Amir test that shows audible benefits from XLR vs unbalanced connection. Balanced always has a slight measured noise advantage, but that advantage is well below audibility. Fully balanced headphone amps CAN provide more power which could provide audible benefit in a few hard to drive headphones, but the same benefit is found in powerful unbalanced amps.

The real story: - Expectation bias for XLR true believers means 95% of them will hear a definite improvement in sighted comparisons. Cryogenic solid silver speaker wire believers will hear similar improvements. XLR is snake oil of the tiktok post-truth age. It is popular at ASR because you can attach numbers to it. Listening bias is real and effects everyone.

For some ASR members XLR connections make them feel better even if there is no benefit. I can respect that. R2R DACs, rolling op amps, and German capacitors make others feel better, and ASR members like me are into the gear itself. That's fine as long as the rent is paid and no animals are harmed.

XLR exists in home components to solve a sales issue, not a noise issue.

The difference between the past and now are subs, and powered/active speakers - oh, and PC's. In the olden days we only needed to run rca cables from one unit to the next in the stack, and the sources were relatively benign hifi components.

Even in these cases it is possible, and I have experienced ground loops.

Now many people are running them from source to speakers on the other side of the room. Worse often a pc is used as a source, which take great delight in injecting noise into ground circuits, especially from USB ports.

And the noise is ground noise - shielding doesn't help - in fact it is the shield conducting the noise.

 

[GGroch]

The difference between the past and now are subs, and powered/active speakers - oh, and PC's. In the olden days we only needed to run rca cables from one unit to the next in the stack, and the sources were relatively benign hifi components.

Really?
Powered subs are not new. They were popular when component audio started in the 1970s olden days. Subs were far more common in home theaters of the 80-00s than they are today. If balanced signal cables cure sub hum, then why are the sub outs on all amps unbalanced RCA connectors? I am not saying sub hum does not exist; its common. I am saying it's almost always caused by AC plug ground loops, not some radiation subwoofers give off that infiltrates signal cables. RCA cables do not cause sub hum.

In the olden days home systems were far more complex than today requiring many more RCA cables. Look at the inputs and outputs on the back of any 1980-2000s AV receiver and compare that to the 1 or 2 inputs on Fosi amps/preamps. Many systems back then did have component stacks, but with longer cables than today because of component size and quantity. Most had additional long RCA run connections to the VCR, DVD, TV across the room. Today, signal cables on Fosi type gear will almost always use even shorter cables less prone to electronic interference.

I don't remember the last time I hooked a PC's analog audio out to an amp, but I do know its an unbalanced connection.

I have never heard noise from my delighted PC's USB connection through my DAC. Plenty of companies do sell $$ USB cleaner devices and none of the ones Amir tested do anything. Good DACS, even cheap ones, are designed for PC hookup and isolate USB noise. Laptops and Mini PCs produce far less electronic noise than a 1980s tube TV does. Phones/Tablets produce almost none, and Bluetooth is wireless. Tech dinosaurs like me might still use a tower PC. If you wrap your analog RCA signal cables around it there may be an issue. I suggest not wrapping.

 

[antcollinet]

I am saying it's almost always caused by AC plug ground loops, not some radiation subwoofers give off that infiltrates signal cables. RCA cables do not cause sub hum.

I think you are not understanding how ground loops inject noise into the signal, and it has nothing to do with radiation (Though it can get onto the ground loop in the first place via magnetic coupling). No, RCA Cables don't cause the hum - but they allow the hum to be added to the signal.

The whole point of balanced connections is to reject this type of noise.

Ground loop (electricity) - Wikipedia

en.wikipedia.org

 

[GGroch]

I think you are not understanding how ground loops inject noise into the signal, and it has nothing to do with radiation. No, RCA Cables don't cause the hum - but they allow the hum to be added to the signal.

Perhaps you are not understanding what I meant. RCA cables do not allow the AC ground hum you hear in powered subs to be added to the signal.

Please explain to me as you would to a child. If RCA cables, especially long ones, allow subwoofer AC hum to be added to the signal, why are they universally used to connect powered subs?

And, how would short XLR cables between the source/preamp/amp have any impact on this issue at all?

I have fixed sub hum on hundreds of systems. In no cases did it involve replacing RCA cables with XLR, either on the sub or on other components. No one has ever even suggested that.

In all cases it was fixed by connecting the sub to a different AC outlet, preferably the same one used by the amp; or by putting an AC ground loop noise eliminator on the AC plug. Problem solved, RCA cables still intact.

 

[V.b.]

It is also 5x the price

Actually, it is 2.5x the price, pa7 is dual TPA3255.

 

[Guddu]

@Fosi Audio

This is from Audiophonics, but same may be on other platforms too, requires a correction for opamps. See the picture, right bottom corner.

 

[Zek]

It also requires a correction for block capacitors - Wima, not Weima.

 

[MrOtto]

@Fosi Audio
I would like to use my expensive and advanced DAC or headphone amp as preamp. What about making TPA3255 stereo poweramp or monoblocks with a selectable gain or just one gain setting?

 

[Toku]

I had a chance to use Fosi Audio's ZA3 prototype, and I immediately did a quick test. The included power adapter is 48V/5A.
I used Topping's E30ll and E50 connected to the PC via USB as input sources. Quad S-2 is connected to the speaker.

The appearance of ZA3 is quite large compared to V3. The ZA3 has a design with quite a bit of room in its case size, but I think a slightly smaller case would be fine.

Since there is no power switch mark displayed on the front panel, I thought it was a type of power switch that turns the volume knob like the V3, but it turned out to be a type that requires pushing the volume knob. I press and hold the knob to turn it on and off, but the operation is unstable and sometimes the power won't turn on unless I press it multiple times. Also, after turning off the power, I had to wait 1 to 2 minutes before turning it on again before it would work. I don't know if this is the specification.
When I turn off the power, I can hear the sound of the relay operating, but the power LED lamp remains lit for a while, making me wonder if it has really been turned off.

I don't hear any pop noise when I turn the power on/off. Also, the noise from the speaker during silent periods cannot be heard even if your ear is placed close to the unit.

When switching between RCA and XLR inputs, there is a level difference between the two. Normally they are designed to be at the same level, but this is not the case with ZA3. The difference is about 3dB, but I feel a little strange every time I switch.

Stereo <-->mono mode switching can be operated smoothly. When switching, you can hear the operating sound of the relay, but there is no pop noise that I was worried about. In monaural mode, only R-ch is enabled.

The ZA3 uses a push-type power switch and a taper A volume pot, making the volume rotation angle wider than the V3, making it easier to adjust the volume. The volume changes are very natural and there are no gang errors.
But there is one thing I don't like. There are times when I feel that the rotational backlash of the volume shaft is a little large and difficult to operate. I think this is simply due to manufacturing variations in the volume pot parts.

The main body does not get very hot even with a 48V power supply, probably because the casing is large. Obviously it generates less heat than V3.

In sound quality tests, even difficult sound sources were clearly separated and expressed without blurring. However, it is the perfect characteristic for me who prefers monitor-like sounds, but it may not be suitable for people who prefer analog-like soft sounds.
When comparing the sound quality with the same Fosi V3, I feel that the ZA3 has slightly better sound separation. But the sound quality of both is very similar.
Compared to the SMSL A300/AO200, which I often use, I get the impression that the ZA3 has more subtlety, but that is just my personal opinion and that is only a slight difference.

I tried SUB output, but unfortunately I could not get an output signal from the RCA terminal. I checked the connection many times, but it seems that the SUB output is not working.
It is necessary to open the case and check whether the SUB OP amplifier is properly inserted into the socket on the board.
The SUB output circuit uses two 2-ch OP amplifiers, and I wanted to hear the sound to see if it was a circuit with an LPF, but I couldn't. Even after reading the manual, there is no explanation at all.

I felt it when I purchased the Fosi V3, but even with amplifier products using the same TI TPA325x series, the design technology has gradually evolved with each new model, and I feel that this is slightly reflected in the sound quality.
Fosi is also progressing with plans to install PFFB, and we look forward to further evolution in the future.

What I just noticed while testing is that a device called ZD3 that resembles a DAC appears in the manual's connection diagram. After this, something in series with ZA3 may be announced.

(Please forgive any mistakes in English sentences.)

 

[Guddu]

What I just noticed while testing is that a device called ZD3 that resembles a DAC appears in the manual's connection diagram. After this, something in series with ZA3 may be announced.

ZD3 - Probably a DAC, but an amplifier with DAC would be even better.

 

[MCH]

I had a chance to use Fosi Audio's ZA3 prototype, and I immediately did a quick test. The included power adapter is 48V/5A.
I used Topping's E30ll and E50 connected to the PC via USB as input sources. Quad S-2 is connected to the speaker.

The appearance of ZA3 is quite large compared to V3. The ZA3 has a design with quite a bit of room in its case size, but I think a slightly smaller case would be fine.

Since there is no power switch mark displayed on the front panel, I thought it was a type of power switch that turns the volume knob like the V3, but it turned out to be a type that requires pushing the volume knob. I press and hold the knob to turn it on and off, but the operation is unstable and sometimes the power won't turn on unless I press it multiple times. Also, after turning off the power, I had to wait 1 to 2 minutes before turning it on again before it would work. I don't know if this is the specification.
When I turn off the power, I can hear the sound of the relay operating, but the power LED lamp remains lit for a while, making me wonder if it has really been turned off.

I don't hear any pop noise when I turn the power on/off. Also, the noise from the speaker during silent periods cannot be heard even if your ear is placed close to the unit.

When switching between RCA and XLR inputs, there is a level difference between the two. Normally they are designed to be at the same level, but this is not the case with ZA3. The difference is about 3dB, but I feel a little strange every time I switch.

Stereo <-->mono mode switching can be operated smoothly. When switching, you can hear the operating sound of the relay, but there is no pop noise that I was worried about. In monaural mode, only R-ch is enabled.

The ZA3 uses a push-type power switch and a taper A volume pot, making the volume rotation angle wider than the V3, making it easier to adjust the volume. The volume changes are very natural and there are no gang errors.
But there is one thing I don't like. There are times when I feel that the rotational backlash of the volume shaft is a little large and difficult to operate. I think this is simply due to manufacturing variations in the volume pot parts.

The main body does not get very hot even with a 48V power supply, probably because the casing is large. Obviously it generates less heat than V3.

In sound quality tests, even difficult sound sources were clearly separated and expressed without blurring. However, it is the perfect characteristic for me who prefers monitor-like sounds, but it may not be suitable for people who prefer analog-like soft sounds.
When comparing the sound quality with the same Fosi V3, I feel that the ZA3 has slightly better sound separation. But the sound quality of both is very similar.
Compared to the SMSL A300/AO200, which I often use, I get the impression that the ZA3 has more subtlety, but that is just my personal opinion and that is only a slight difference.

I tried SUB output, but unfortunately I could not get an output signal from the RCA terminal. I checked the connection many times, but it seems that the SUB output is not working.
It is necessary to open the case and check whether the SUB OP amplifier is properly inserted into the socket on the board.
The SUB output circuit uses two 2-ch OP amplifiers, and I wanted to hear the sound to see if it was a circuit with an LPF, but I couldn't. Even after reading the manual, there is no explanation at all.

I felt it when I purchased the Fosi V3, but even with amplifier products using the same TI TPA325x series, the design technology has gradually evolved with each new model, and I feel that this is slightly reflected in the sound quality.
Fosi is also progressing with plans to install PFFB, and we look forward to further evolution in the future.

What I just noticed while testing is that a device called ZD3 that resembles a DAC appears in the manual's connection diagram. After this, something in series with ZA3 may be announced.

(Please forgive any mistakes in English sentences.)

View attachment 324376

Hey @Toku thank you for this.
Have you had the chance to test the trigger input?
If yes, does it make any pop noise when switching on or off?
If not tested, no need to test it for me.
Thanks!!

 

[Toku]

Hey @Toku thank you for this.
Have you had the chance to test the trigger input?
If yes, does it make any pop noise when switching on or off?
If not tested, no need to test it for me.
Thanks!!

Sorry, I don't have a suitable 12V power supply here right now, so I haven't tested the Trig terminal yet.

 

[space_duck]

I had a chance to use Fosi Audio's ZA3 prototype, and I immediately did a quick test.

How does the sound quality compare to the A07 Max?

 

[Toku]

Correction regarding ZA3 SUB output

Fosi provided an explanation about SUB output.
The SUB output is a variable output linked to the main volume control. And as I predicted, the signal has passed through the LPF, and the cutoff frequency is set to 200Hz.
Because I didn't have this information in advance, I made a mistake in how to check the operation of SUB output.

 

[Toku]

How does the sound quality compare to the A07 Max?

Although there is a slight difference in sound quality, the two are very similar and it is difficult to tell which one is better than the other. Although both products use the same TPA3255, they have definitely evolved from previous TPA3255 amplifiers.
I have A07 and A04, which were the representative models until recently, but I won't be using them anymore.
The biggest difference between these two units is the amount of heat generated. The A07 Max is used with the voltage lowered to 40V for safety, but it still gets quite hot. The ZA3, on the other hand, uses a 48V power supply, but it doesn't feel that hot.
The advantage of ZA3 is the SUB output that is linked to volume adjustment. It also has a built-in 200Hz LPF. Also, for those who value balanced input, having both XLR/RCA is an advantage.
Pop noise and noise from the speakers are well suppressed on both the ZA3 and A07 Max, so you won't hear them.

 

[mike70]

I checked the impedance graphs maded by Stereophile and the Quad S2 have impedance above 8 ohms in high frequencies ... how do you feel the load dependency effect?

Supposedly, is at low impedances that the effect is minimal.

 

[jayapple]

take over PA5, we needed !

 

[Toku]

I checked the impedance graphs maded by Stereophile and the Quad S2 have impedance above 8 ohms in high frequencies ... how do you feel the load dependency effect?

Supposedly, is at low impedances that the effect is minimal.

I think your question is about how the rise from around 20kHz in the frequency response graph with an 8Ω load affects the sound quality.
The rise in this part is about 3dB and is at the upper limit of the audible band. I have never actually felt the effects of this part. In actual music, the level of this frequency range is low and the increase is about 3dB, so it is difficult to recognize. Moreover, as I am old, I can no longer detect frequencies above 13kHz.

 

[NoMoFoNo]

Although there is a slight difference in sound quality, the two are very similar and it is difficult to tell which one is better than the other. Although both products use the same TPA3255, they have definitely evolved from previous TPA3255 amplifiers.
I have A07 and A04, which were the representative models until recently, but I won't be using them anymore.
The biggest difference between these two units is the amount of heat generated. The A07 Max is used with the voltage lowered to 40V for safety, but it still gets quite hot. The ZA3, on the other hand, uses a 48V power supply, but it doesn't feel that hot.
The advantage of ZA3 is the SUB output that is linked to volume adjustment. It also has a built-in 200Hz LPF. Also, for those who value balanced input, having both XLR/RCA is an advantage.
Pop noise and noise from the speakers are well suppressed on both the ZA3 and A07 Max, so you won't hear them.

Do the Aiyima A07 Max and Fosi ZA3 both use the same slab-style heat spreader bolted to the bottom case member, like the Fosi V3? I've encountered minimal heat from my V3 with 48V power supply. The slab-style bottom-bolted heat solution seems highly effective.

 

[NoMoFoNo]

I think your question is about how the rise from around 20kHz in the frequency response graph with an 8Ω load affects the sound quality.
The rise in this part is about 3dB and is at the upper limit of the audible band. I have never actually felt the effects of this part. In actual music, the level of this frequency range is low and the increase is about 3dB, so it is difficult to recognize. Moreover, as I am old, I can no longer detect frequencies above 13kHz.
View attachment 324579

The load dependency looks dramatic on graphs but is likely minimal-to-meaningless in the real world for many/most listeners. Up to 20K, the threshold of human hearing, there is a rise (only at 8 ohms) that is less than .5db up to 15KHz. Most of us over the age of 40 or so cannot hear at 15KHz or beyond, so it becomes a mute point. Even for the young bat-ears among us, 1db rise at 20Khz, again only at 8 ohms, is small.

If one has 6 ohm or 4 ohm speakers, you're looking at a full-range frequency response that's ruler-flat to 10Khz, and a very slight dip thereafter, never breaching .5db.

In other words, the response from amps like the V3 already closely approach the ruler-flat load-independent response we call for when looking at graphs. This is not very different from proclaiming the superiority of a DAC with a SINAD of 120 over one with a SINAD of 110. There isn't a genuine difference between the two for actual listeners.