Fosi Audio's First Dual-Mode Balanced Amplifier ZA3

[nathan]

The frequency response is not independent of the impedance at high frequencies.

Better implementation (more money)

I think I understand.

The response can vary by +1db/-0.5 db in the top octave, the 10k-20khz range, in the cheaper of those two amps, if the speaker impedance is 8ohms, right?

And that is why the Wharfdale is a problem for the cheaper Class D amp, right? The Wharfdale has a 10 to 12 ohm impedance in that top octave. The KEF is down below 6ohms so much less likely to create issues right?

And something like the following would probably be "ok" but not quite as linear as the KEF.....

 

[charlielaub]

YEs

I think I understand.

The response can vary by +1db/-0.5 db in the top octave, the 10k-20khz range, in the cheaper of those two amps, if the speaker impedance is 8ohms, right?

And that is why the Wharfdale is a problem for the cheaper Class D amp, right? The Wharfdale has a 10 to 12 ohm impedance in that top octave. The KEF is down below 6ohms so much less likely to create issues right?

Yes, that is more or less correct.

The output filter interacts with the load. It is typically designed so that the response is flattest when driving a 4 Ohms load. When the load impedance is higher, there is peaking. When the load impedance is lower, the response will sag. This effects happens near the upper end of the audio band for the filters that are found in class-D amplifiers and it is only around that frequency that there is any change in the amplifier's response with changing load impedance. If you don't use the amp above a couple of kHz or don't care that there is a 1dB or so slow rise in the response, there is nothing to worry about.

So, for example, in the plot above of the Revel M55X there is a peak in the impedance of almost 18 Ohms (max occurs around 2kHz) but this is below 5kHz and so there will be only a small impact on the frequency response (but there will be some). There is a small rise above 5kHz from 6.5 Ohms to around 8 Ohms that will drive some sort of rise in the response there.

 

[doug s.]

YEs


Yes, that is more or less correct.

The output filter interacts with the load. It is typically designed so that the response is flattest when driving a 4 Ohms load. When the load impedance is higher, there is peaking. When the load impedance is lower, the response will sag. This effects happens near the upper end of the audio band for the filters that are found in class-D amplifiers and it is only around that frequency that there is any change in the amplifier's response with changing load impedance. If you don't use the amp above a couple of kHz or don't care that there is a 1dB or so slow rise in the response, there is nothing to worry about.

So, for example, in the plot above of the Revel M55X there is a peak in the impedance of almost 18 Ohms (max occurs around 2kHz) but this is below 5kHz and so there will be only a small impact on the frequency response (but there will be some). There is a small rise above 5kHz from 6.5 Ohms to around 8 Ohms that will drive some sort of rise in the response there.

i'd think that if the frequency response increases only 0.5-1.0db (or even 2db?) above 15khz, it would be virtually unnoticed by the vast majority of listeners.

doug s.

 

[maty]

With KEF Q350 impedance graph maybe you are right but with others loudspeakers.. That is why it i important to buy hard that we know the measurements of, especially the speakers, to minimize the risk of error in the purchase.

Excellent measurements do not guarantee that the sound is great, but at least we will know that things have been done right.

 

[doug s.]

If you are desperate to buy an amplifier at a lower price then indeed it will be at the sacrifice, at one point or another, of performance)
I would have a hard time comparing a 3E Audio module or even a Sylph Audio to this Fosi with all due respect.

View attachment 336026

i have no dog in this fight; not having heard either this amp, or the 3e audio iteration. but with all due respect, specs are a guide. when the sinad is good enough as to be inaudible, then perhaps a less expensive amp with not quite as good specs, may be just perfect.

i do know that a well respected speaker designer (cube audio and qualio) prefers the za3 amp with his qualio iq speakers over the $5.7k naim supernait 3. he has no reason to push these inexpensive amps. of course, synergy is key. and the rising frequency response in the treble, w/o pffb - possibly a db or so - is most likely not going to be heard by the majority of listeners w/the majority of speakers.

ymmv,

doug s.

 

[etherealshadowkaizer]

Good Day All,

New to the hobby, and currently waiting on my ZA3. Can I run my Wharfedale Diamond 12.2 on biwire/bi-wire using only 1 unit of this amp? Sorry, not sure if the question makes sense. Thanks.

 

[doug s.]

In all seriousness your hyper focus here on one little aspect of my statement, is a very big indicator of Autism Spectrum Disorder.
I mentioned about 5-6 changes myself and many other would like to see any you just can’t seem to take your mind off toogle switches.

All those unused inputs you seem to love covering your devices well they cost money. Frankly I don’t like paying for something that goes unused and I’d rather it not even be there. As a audio device those inputs could potentially introduce noise into the device. Better they aren’t there.

I would like Fosi to have as many amps as they wish to have. They seem very responsive to us customers. I would like them to spend less time with those amps that measure poorly and focus on innovative and growing into the future. With the v3 and this Za3 they seem able to accomplish that.
I believe that they can and they can do it without your precious toggle switch!

Frankly I don’t like paying for something that goes unused and I’d rather it not even be there. As a audio device those inputs could potentially introduce noise into the device. Better they aren’t there.
- i agree w/this statement. and i like your other suggestions regarding this amp. but regarding the stereo/mono switch you are mistaken. everything gygess said about it is correct. were fosi to have two separate products, you would pay more for the one you wanted, w/no switch, vs having to suffer with a toggle switch that you have to set once and then leave untouched. and who knows? maybe something in your set-up needs a temporary fix of one sort or another, and it turns out you're glad that switch was there after all.

doug s.

 

[doug s.]

I feel like most of the people who plan to do active crossovers aren't the people who only have $150 for a 2-channel amp.

IMHO the niche for this type of amp is for a nearfield/smaller setup (e.g. PC/budget monitoring) where you have 2 bookshelf + 1 sub and you're looking for a low-complexity, medium cost system with "good enough" fidelity. Thus, sub-out saves a ton of setup time and a decent chunk of $$, while balanced helps significantly with grounding issues that would be more noticeable up close.

PFFB also would definitely push the amp into "good enough" territory. But, since you lose an appreciable amount of gain because of it, it would increase the cost of the amp since you now need an extra internal (very low noise) pre-amp stage to get back that gain. Fosi for sure would not want to sell an amp with less than 20 dB gain, probably closer to 24-26 in order to snag customers with sources maxing out around 1.5 V or so.

I feel like most of the people who plan to do active crossovers aren't the people who only have $150 for a 2-channel amp.
- you might be surprised. i have an active quad-amped system, with a pass f2 clone and an upgraded almarro a205a mkii doing the duties over 175hz. (full range driver, and ribbon tweeter >7khz.) as they both idle at ~300w combined, (120vac), i love the idea of replacing them w/something like the fosi (or aiyima) amps for these prices, if they sound good, and idle at ~12w combined, and will probably rarely if ever draw more than 20-30w total. that's why i'm following these threads.

doug s.

 

[daniboun]

rising frequency response in the treble, w/o pffb - possibly a db or so - is most likely not going to be heard by the majority of listeners w/the majority of speakers.

PFFB really makes the difference with TPA32X amps... this is why Fosi and Aiyima are in the process of developping their new amps with PFFB this year. (I know it from a reliable source)
BTW > Fosi ordered a Sylph Audio module with PFFB probably for retro engineering.

Another good example is the Sabaj A30A that is using PFFB too....f you deprive it of its Axign chips its performance would collapse.

 

[somebodyelse]

Good Day All,

New to the hobby, and currently waiting on my ZA3. Can I run my Wharfedale Diamond 12.2 on biwire/bi-wire using only 1 unit of this amp? Sorry, not sure if the question makes sense. Thanks.

Yes, you can. https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/biwire/Page1.html has some diagrams showing the connection. If you don't already have the extra cables it probably isn't worth the cost and effort though - see the analysis in the later pages from that link, as well as this thread with some measurements.

 

[OldTimer]

Good Day All,

New to the hobby, and currently waiting on my ZA3. Can I run my Wharfedale Diamond 12.2 on biwire/bi-wire using only 1 unit of this amp? Sorry, not sure if the question makes sense. Thanks.

the high and low frequencies travel only in their designated cables.
It’s like feeding the dog with cat food and vice versa if you do single wire

 

[maty]

Good Day All,

New to the hobby, and currently waiting on my ZA3. Can I run my Wharfedale Diamond 12.2 on biwire/bi-wire using only 1 unit of this amp? Sorry, not sure if the question makes sense. Thanks.

Wharfedale Diamond 12.2 but the impedance graph must be similar (same tweeter I suppose). You will have a slight rise in the treble but you may not notice it.

 

[jooc]

PFFB really makes the difference with TPA32X amps... this is why Fosi and Aiyima are in the process of developping their new amps with PFFB this year. (I know it from a reliable source)

I'm waiting for a fully bridgeable PFFB implementation from these guys, which I'd like to use as monoblocks into some Wharfedale Lintons.

 

[CamRector]

I'm waiting for a fully bridgeable PFFB implementation from these guys, which I'd like to use as monoblocks into some Wharfedale Lintons.

My original plan too. Then @daniboun got me building a better amp with the 3e modules and micro audio powersupply.
I don’t think anything fosi does will come close to these 3e amps

 

[doug s.]

Good Day All,

New to the hobby, and currently waiting on my ZA3. Can I run my Wharfedale Diamond 12.2 on biwire/bi-wire using only 1 unit of this amp? Sorry, not sure if the question makes sense. Thanks.

if you have the room to connect two sets of speaker cables to the tight spacing of the speaker connectors on this amp, you should be able to biwire, if your speakers allow it. but honestly? i think your money would be better spent on something than a 2nd set of speaker cables.

doug s.

the high and low frequencies travel only in their designated cables.
It’s like feeding the dog with cat food and vice versa if you do single wire

no it's not - the same frequencies travel through both cables when bi-wiring; and they're split at the speaker binding posts; just as the signals are split with single-wire.

doug s.

 

[doug s.]

PFFB really makes the difference with TPA32X amps... this is why Fosi and Aiyima are in the process of developping their new amps with PFFB this year. (I know it from a reliable source)
BTW > Fosi ordered a Sylph Audio module with PFFB probably for retro engineering.

Another good example is the Sabaj A30A that is using PFFB too....f you deprive it of its Axign chips its performance would collapse.

yes - it makes a difference. but one that is not audible in many, if not most instances. yes, fosi and aiyima are developing pffb amps. but most won't be able to detect a sonic difference in most applications. i'm sure marketing has much to do with it.

doug s.

 

[doug s.]

I'm waiting for a fully bridgeable PFFB implementation from these guys, which I'd like to use as monoblocks into some Wharfedale Lintons.

based on the impedance curve i've seen on the lintons, if you were to compare amps whose only difference was pffb implementation in one, it's highly unlikely you'd be able to hear any difference. and any measured difference of the frequency response of the speakers would be minimal, if any, as well.

doug s.

 

[daniboun]

yes - it makes a difference. but one that is not audible in many, if not most instances. yes, fosi and aiyima are developing pffb amps. but most won't be able to detect a sonic difference in most applications. i'm sure marketing has much to do with it.

doug s.

PFFB is nothing to do with marketing. It s an implementation which was wrongly put aside.
You will be able to hear a difference between a 87db Sinad and a 102db Sinad....

 

[doug s.]

PFFB is nothing to do with marketing. It s an implementation which was wrongly put aside.
You will be able to hear a difference between a 87db Sinad and a 102db Sinad....

i don't think so. maybe you will be able to hear it. but most people will not.

doug s.

 

[somebodyelse]

I'm waiting for a fully bridgeable PFFB implementation from these guys, which I'd like to use as monoblocks into some Wharfedale Lintons.

Which implementation of PFFB have you seen that isn't bridged already?