Fosi ZA3 Amplifier Deep Dive Review (with Data) 

I reviewed a Burson amp a long time ago where they sent opamps that I could change. I could "sense" some kind of difference, but didn't have the tools to measure the change. See if you can get some of those or if they even still sell them. I'm curious as to what they did.

On the op-amp thing... (edit: In relation to these mini-amps) Get a copy of the NE5532 spec sheets and look at the various curves and ratings. Two things become readily apparent ... the chip has uneven gain below about 1 volt of input (which is where we use them!) and they don't even specify differential input specs because the two inputs are diode clamped together giving a maximum of 1.2 volts across the two inputs. (On the TI data sheet take a look at graphs 5, 6, and 7) Now you have your difference... a cheap as chips chip with uneven gain and very uneven input characteristics vs. any other chip that does not suffer these limitations. The real difference in op-amps with these small amps is NOT putting in an expensive chip ... it is getting rid of the stock chips. I commonly pull out all the NE5532 chips supplied and put in either RC4558 or TLO72 ... under $1.00 each in lots of 10. Yes that first step does make a noticeable difference ... beyond that, not so much.

As long as both channels of the amp have the same "uneven gain" for any given input voltage, what's the advantage of swapping one type of cheap op-amp for another? Given that (e.g.) the NE5532 and TL072 are equally elderly designs, why is the NE5532 the overwhelming favorite for inclusion in mixing desks and other professional recording gear? I'm pretty certain the engineers of enormously pricey studio gear from the likes of Studer and Neve could have designed in any op-amp chip they wanted, so there must be something about the NE5532 that made it preferable, right? Then there's TI's decision to equip their own TPA3255 eval. board with NE5532s -- if the RC4558 is better for that application, then why that op-amp choice when TI makes both ICs?

@@Douglas_Blake_579The NE5532 and TL072 are pin compatible, but _not_ interchangeable. The 5532 is a bipolar input dual op amp, the TL072 is FET input. Wildly different input impedance.

@@editorjuno If I feed in 1 volt on a gain of 10 circuit... I expect 10 volts out. Not 7 or 13... 10. it's not rocket science. The advantage is that the 4558s and tl072s are linear gain, their input impedances are higher and their inputs are not diode clamped at 1.2 volts peak to peak.

On it!!!! I am talking with MiniDSP about getting a FlexHT to do a video (or series) on how to use them to tune your system. I'm excited about it!

@@ErinsAudioCorner Wow, that's incredible news! I find their software a high value addition to a system. I'm looking forward to gaining a deeper understanding of any aspects you decide to cover. Thank you sincerely Erin!

I"m really happy with my SHD. Dirac made a huge difference in my system. I have tso of the ZA3s. I biamp with them using my SHD. They work fine in my system.

​@@ErinsAudioCornerOh, awesome. I love my Flex HT, but I'm sure you can help me extract much more from it.

100%

Pull them bum cheeks apart.

Facts

So for the most in depth reviews on youtube. Love this channel

Thank you!

And the conclusion is that load dependency is a thing, but it might just be barely audible. And anyone that's interested in audio should have a DSP anyways so those small frequency response deviation can easily be fixed together with the rest.

Listen, this is the best reply I will get on this video. By far.

We all appreciate toilet paper and your reply 😂

Hi, you do very good tests. I would love to see the Topping pa7 in the future. Regards​@@ErinsAudioCorner

But as someone who has been a professional writer for his entire career, I would suggest looking closely at the meaning of the cliche. The grain of salt is to guard against the negative qualities of what is being consumed. So if you really want to distance yourself from an assertion, you use MORE salt, not less. Rather than a grain, you might suggest applying a spoonful of salt or even a salt mountain. Or just say "I'm highly skeptical about ... ."

I don’t know about anyone else, but I have found the aforementioned roll of toilet paper to being quite useful!😁🧻

That's great! I'd vote variable for sure.

PFFB is location bias. Depending on where you live in the world. Most electrical circuit feeding homes in Europe won't need PFFB and so the average consumer will not necessarily see any improvement anyway...apparently.

@@Panslapper measurements show significant improvement

@@ubacow7109 So it appears...I can't sat either way myself as I have no way of testing but if offered with or without I would likely go with...

No way! Thank YOU!

I appreciate the sentiment. I definitely put a lot of time into this to figure some things out that had me stumped for a bit. LOL

It's not weird. It's the industry standard. Idk what Erin is on about

@@NickSim86 Exactly .... you gotta climb the price scale quite a bit before you find high pass filters on the mains.

I've got Monitor Audio Radius 45 ( 3" Cubes ) so not only do a need High Pass but it has to be flexible enough to go to 120hz. I get that vendors want to "tick boxes" but just save the Sub out money , if you're not adding a HPF to the mains "you're doing it wrong" I don't care how cheap your product is if it's designed inflexible.

@@hmurchison8123 you have very niche speakers and should use an AVR

I was using the subwoofer output in mono mode and it was quite weak, i did the same in the stereo mode and it was fine, now i have gone back to the mono mode using two ZA3'S with the P3 Preamplifier, running two subs from my P3 Preamplifier using a 3.5 mm to rca y splitter, lovely results, enjoying the music. 😊

I'll look into it. Thanks for the suggestion.

I have one, purchased before any independent reviews came out and am not impressed. I find the interface clunky at best. In stereo mode, the power into 8 ohms is only 50 watts and has worse distortion than average these days. It has a lot of power as a monoblock into 4 ohms--over 400 watts---but at the expense of even more distortion.

@@ErinsAudioCorner rsl also has a cheap one with hpf. Sabaj a30a is a digital amp with that option also.

@@stephenjarzombek2903 yes, just checked asr's review. He wasn't impressed either. It's a shame because the featureset and price are great!. I don't understand why there aren't more amps with hpf. If one's into streaming only there's always something like bluesound node or the upcoming wiim ultra that have bass management, but other options are much pricier.

I disagree. The use of a subwoofer alleviates the mains from playing lower frequencies so those speakers then improve in both a) distortion and b) dynamics. (These are objective things, btw.) "You want your speakers to cover as much of the full range as possible, and for the subwoofer to only come in at the very low-end as the speakers are already rolling off. You don't want to pick a crossover point. Your crossover point is determined by where your speakers bottom out.... personally don't agree with this logic" ^ I agree with your disagreement here but maybe for a different reason because this particular item doesn't offer control of where the filter is implemented. Maybe if they did I might be more apt to not be bothered by not having a HPF (i.e., mains roll off naturally at 60Hz so you apply the subwoofer to ~60Hz and below). But that's a moot point here, unfortunately.

​@@ErinsAudioCorner You mentioned the -3db point on the sub out as being 180Hz. I'm assuming Fosi picked this as a safe maximum for the highest people want their subwoofers to play. The idea being that you adjust the bandwidth of the signal yourself using the LPF built into your subwoofer. I've seen this same approach on $1000+ integrated amps that have a sub out: Speaker level outputs always play full range, and the sub out is not adjustable on the amp. You're expected to low-pass the signal on your sub. It's not ideal, since as you said, I can't pick a crossover point that minimizes distortion, improves dynamics, etc. It also assumes you're using a subwoofer that has a LPF built-in, and it leaves you SOL if your room is creating a null in the crossover region forced upon you. But for whatever reason, this has become the "common wisdom" that hi-fi amp and speaker manufacturers have settled on. To be clear, the lack of this functionality in this amp annoys me just as much as you. I feel like I'm taking crazy pills when I see integrated amps costing thousands of dollars that lack the most basic of bass management features,

@@ErinsAudioCorner Not to mention, that although a driver/crossover combo has a point where it will no longer reproduce linearly, the speaker will still be forced to "overlay" frequencies above the crossover point on top of lower frequencies. Unloading the speaker by removing lower frequencies lets the upper-bass/mid frequencies drive the speaker's cone with less "interference".

its a desktop amp for 150 bucks , what do you expect?

You have to be careful and not generalize. There's different chips which make a significant difference. The newer one ones use TI TPA3255 amp chip, whereas older D amps used TPA3116 or others, and the newer ones are by far superior; much more comparable to 'real' amplifiers in terms of performance. But yeah, I agree, older chips weren't quite up to snuff. Latest generation though? Debatable at best, my personal experience; great performance for very little $$. Really not the same as older amps which were audibly inferior.

@@Rumburak852Exactly what I wrote.

@@literalghost929 Newer ones are better, but still not the miracles that a lot of people claim they are. I do enjoy some of them though. I just don't pair them with speakers that need more than what they offer or use them for ultra high fidelity situations. For my retro consoles though they work well. As a little integrated amp for some small speakers in the dining room that will never be more than background, they also work well. As a PC speaker that you're not looking for monitor class audio fidelity, again, they're fine. Attached to a pair of KEF R3 Metas in a properly conditioned listening room? Not so much.

@@literalghost929fosi amps are garbage

Totally disagree as someone that owns both. Dual ZA3’s with a wiim pro is SO much better than a wiim amp. It just doesn’t have near enough power.

Agree. Even some of the bigger hifi companies don’t include high pass filters. It seems like that’s something if want, but you have to hunt around to find that. Or buy a WiiM amp 😂.

@@steveh545 I mean its not hard to make an amp with a fully adjustable high pass, dual sub outs, load indepedent and hdmi out. Lol! Infinite PEQ would be nice too. 😜

@@AbsoluteFidelity no idea how hard it is to make an amp… just many many more expensive amps don’t offer high pass filters, so I wouldn’t ding this amp for that. YMMV.

@@steveh545 its like amp manufacturers do not know what the consumer really wants, they keep churning out same ol products. It could be easily done by china, or any manufacturer for that matter.

What you all are talking about is easily and thoroughly solved with a minidsp. Pair a 2x4HD with Any amp (including this fosi) and you have the power to time align your individual subs and mains to maximize amplitude then dial in the frequency response as desired for multiple listening positions. It's a real game changer. Far superior to an amp with "dual sub outs".

Good point! Thanks!

That’s good news.

Glad it was helpful!

Thanks for watching! Glad to have you back (in the hobby!)!

Any review methodology that chases a single number in its rankings will result in companies trying to game that one spec while ignoring everything else.

@@sk9592 And then the 'measurement' type folks will scoff at people hearing differences which they didn't measure. I'm much more on the measurements side, but sadly, people take it to extremes, and ASR, although I hate to say it, might have gone ab it over the edge in that respect.

@@sk9592 I do agree, though most on ASR don't really do the SINAD chasing most people think they do. I would say the biggest problem is that the power numbers are greatly exaggerated on ASR in comparison to what actual music would take.

Because they can actually work against each other depending on the rolloff profile of the mains. Aside from that, I answered the question in the video.

​@@ErinsAudioCornerSo my big 10 inch woofers would go dormant?

@@albe1796 not necessarily. But if you look at my reviews of many budget 10” subwoofers you’ll see they typically don’t have much extension past 40Hz. Most of them are just midbass modules. Obviously there are exceptions. And more expensive 10” subs typically use DSP to extend the bass response (but do run out of excursion at high volume).

My point wasn't about the subwoofer. It's about how the mains don't have a high-pass filter applied to them to limit the bass frequencies and letting the subwoofer play that. Limiting the bass frequencies from the mains lowers distortion and improves dynamic range of your speakers.

@@ErinsAudioCorner I understand all of that. Still not sure how that's that big of a deal. I understand that's the ideal situation to be able to control what frequencies come out of your Mains and the subwoofer but I've got three rigs in my place add all three have full frequencies being played to the main speakers and then I'm controlling the frequencies for the subwoofer at the subwoofer. All three rigs sound very good.

@@ryanschipp8513 okay, then you don't see the need for mains to be high passed. I do. That's cool with me.

@@ryanschipp8513 I have rarely, if ever, been able to PROPERLY integrate the main speakers to subwoofers in regards to ideal phase alignment/phase coherency if/when the main speakers are still playing full range w/o a HPF so that they are "overlapped" with the subwoofer(s). Ideal phase alignment between the main speakers and the subwoofer(s) is required in order to achieve a perfect "up front bass" experience where ALL of the low frequencies from the subwoofer(s) appear to be locked up front to the main speakers only and the bass will actually image properly within the soundstage (if that is part of the recording....play the intro to Malia & Boris Blank's "Celestial Echo" from the Convergence album). Sure, at particular frequencies, you will have some "constructive interference" where both the subwoofer & main speaker sum together in order to provide a boosted response. HOWEVER it is very unlikely for that constructive interference to be both Phase AND Time aligned. So while that scenario might provide MORE BASS at a certain frequency, it will not be "articulate/detailed/clean/fast" bass, and will not provide that perfect "up front bass" experience that is locked to the main speakers. This will be most noticeable with instruments such as an acoustic upright double bass which has lower fundamentals combined with lots of upper harmonics and transients, or fast "slap bass" on an electric bass. So the phase alignment through the crossover region and beyond will rarely be ideal between the subwoofer and main speakers when both are significantly "overlapped". In fact, you will usually want a Half Octave UNDERLAP between the subwoofer and mains center crossover frequency. As a starting point to achieve the easiest integration between the subwoofer and the main speakers, and as a most-likely scenario example, I would set the main speaker's HPF to 80Hz and the Subwoofer's LPF a half-octave lower to 60Hz, and use 4th Order Linkwitz-Riley (LR4) filters for both. While this is not a "set in stone" configuration that is guaranteed to always work perfectly every time, it is the most likely configuration to get the response and integration of both the subwoofer and the main speakers at least close or "in the ballpark" more so than others, especially if your subwoofer has a "Variable Phase" control knob rather than just a 0°/180° phase switch, i.e. "normal/reverse" polarity switch. What really matters is the actual Acoustic Crossover response of the system and the Phase Alignment between the mains and subwoofer(s) within and throughout the crossover region. And to achieve that, your crossover network filters (either passive or active) may need to be asymmetrical with completely different center frequencies and orders/slopes for the HPFs & LPFs that are applied to each driver. Achieving proper summation between a subwoofer and the main speakers is going to be difficult to achieve with the mains playing full range, at least if they also have response down into the subwoofer region below 50Hz-40Hz.

@@ryanschipp8513 the normal way is a big deal. It can add mid range to speakers that don’t have that in the mix when they are trying to handle lows AND mids. Basically adding a sub and high passing can make mediocre speakers sound better. So huge deal.

class D has interesting history. maganovox employee discovered long time ago early 90s ? and was really unrecognized / misrepresented. but i don’t recall the deets very well and if chip was grown up yet n stuff

Kinda... but there is a slight power increase in the mono channel. FWIW, I did measure the other channel in mono and nothing changed, so you have to make sure you are using the channel labeled as "mono" to get the extra power output.

@@ErinsAudioCorner I do have everything set correctly. I was looking for a difference in sound quality and there's none that I noticed on my CSS, ELAC, or my PSB speakers.

@@991carreras3it may just be that the difference is so small that it might not sound louder. Otherwise, yeah, I dunno, man. Just weird stuff.

The problem here is one of expectations ... many people are stuck on the notion that bridging an AB amplifier does increase output power. Not so with Class D chip-amps. They are in fact, already bridged in the stereo mode. The TPA3255 chip as 4 independent power amplifiers in it. They can be combined in various ways to provide differing system topologies... 4 x SE (single ended) ... giving 4 independent outputs 2 x SE + 1 x BTL (bridge tied load)... your typical 2.1 stereo setup 2 x BTL ... two pairs of bridged SE outputs... the most common use. 1 x PBTL ... parallel connecting the 2 x BTL outputs for more current. They are *not* just turning off one channel. They are switching the chip into PBTL mode which allows it to drive loads as low as 1.5 ohms. The two stereo BTL channels are paralleled so there is no increase in speaker voltage (and thus power), it's all about the current needed to drive low impedance loads. If you're expecting some magical transformation in mono mode, you're clearly at the wrong alter. Above 4 ohm impedance it makes little or no difference at all. The only real advantage to running 2 monoblocks is that each amplifier has it's own power supply.

A few others have asked about that RSL as well so I will send them an email and see if they are interested in sending me one to review.

Thanks for sharing!

@ErinsAudioCorner I've got the B50 SE which is a great value if you don't need the extra features. Awesome review as always 👏

which one do ya mean ?

haha. Wasn't intentional. Not for this ginger, at least. :D

Cool. Thanks!

My personal take is this: as someone who does have a bit of a following (albeit, quite small) I will always default to not talking about hearing things that I can't at least point to evidence with. In the case of amplifiers, many say they hear different things. A regular ol' dude talking to his buddies is one thing. A fellow with 100k subscribers is entirely different. I want the latter to be able to prove he heard these results. A blind test is an easy way to do this. So just offer up those results and that's that. None of this means that differences can't be a) heard and b) measured. It just means that I take a reserved stance and am cynical of these claims without some sort of backing. If someone can so easily hear differences in amplifiers, opamps, etc then it really should be easy to prove via a simple blind listening comparison. No one is doing that, though. At least not in the review space. So it's just conjecture until then. That doesn't make me a non-believer. It just makes me someone who is careful with what I say and believe. 😉

@@ErinsAudioCorner 100%! BTs, DBTs & measurements are hard and take a lot of effort. In a perfect world, there would be no subjective opinion and non-validated claims, but ASR just went a bit too far in that direction. (No measurements = does not exist = imaginary = censored.). In that particular case, ASR failed as it actively censored individuals who were possibly correctly reporting their experience; they failed to appropriately measure amps, and censored reports of experience with said gear, if the experience didn't match the review's conclusion and measurements. In this case, there were shortcomings in measurements, and ASR effectively censored the truth by moving comments to a generic to snake-oil thread.

Well. That’s the point. Why have a subwoofer pre-out if it doesn’t high pass the mains. Use your own logic; that’s the preamps job.

@@ErinsAudioCorner I get it, but for argument sake, let’s pretend that it didn’t have one-Are you saying you’d more likely recommend the amp? Because that makes even less sense. I hope you see what I’m getting at… It’s a $150 dollar amp that is pretty darn impressive for the money (I own one, along with tons of other expensive amps). Hard to fault something this inexpensive that performs this well in my opinion.

Glad I could help

If you search for "RSL Amplifier" you will see that they have a $110.00 USD amplifier with a "Selectable 90Hz high-pass filter"

Hey, thanks for asking. This is the 3rd electronics review I've done. I just started within the past few weeks. Here are my other two. NAD C 3050: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-8gQB21TuFrI.html WiiM Amp: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-RcJAcVbq5xQ.html

Thanks for the feedback. I definitely have some others that I want to look into so I will take those considerations in mind.

I disagree with this. The basis for a subwoofer being used is to alleviate the load of the mains. I'll go to my grave over this.

They are so anal about subs and mains being fed the same signal, they want you to pass speaker-level material through the subs (plural important here) and onto the mains. I can see arguments being made on both sides of this issue, but personally I tend to be in the high-passing mains camp too and I own both PS Audio and Rel equipment. Maybe because I don't play in Paul and John's level of high $ equipment. 🙂

I’m honestly not sure about the differences. I haven’t looked into it yet.

Given the crappy thermal solutions on these fully enclosed amplifiers, I would not be surprised. A quick check of pin 7 on the chip itself will tell the story. Table 4 on page 19 of the Spec Sheets will tell you what to look for.

I know! I shined (shone?) a flashlight through it and … nothing. Booooo!!! 😂😂

Not currently.

Thank you!! For the longest time I thought that I made the wrong choice buying a used Creek 4240 amp over a Class D TPA 3255 amp like the Aiyima A07.

That’s a really good question, I’d be interested to see him test this.

Not snake oil, best upgrade you can do to your amp.

@@wric01 If it really is such a good upgrade then most manufacturers would be preinstalling more expensive op-amps. The fact that they don't tells your something. The reality is that Erin blind tests amps and can't tell the difference in sound between then (when the gains are matched between the two) so how is an op-amp going to change things so materially?

Same here mine acts glitchy sometimes, but I use external Bluetooth so I don't even use it for that, other than that it's great for the money. Especially with 12 db of bass gain and 6 db if treble gain and tube swap options. I like it well for the money, just waiting on a balanced preamp to be released next, gives you roughly another +6db of gain.

Definitely not normal.

If the volume control in these is a cheap potentiometer it's possible there's a lot of channel imbalance at lower volumes. I've ran across this with multiple products. In this case the channel imbalance decreases as you up the volume.

Yes that's pretty normal at extremely low volumes. It's in the way volume control 'pots' are made. They use mechanical wipers on a strip of carbon resist material to divide the input voltage onto their outputs. If those wipers are not tracking perfectly together you can have a noticeable low level balance problem. And no this is not Fosi or even a chifi thing... I've seen it on some pretty high end stuff, over the years.

I do the same with a Fosi V3

Scale matters .... take a close look at those charts ... Erin is pointing out differences of fractions of a decibel. Now I don't care how good your hearing is (or you think it is) literally nobody is going to notice a difference of less than about 2db and especially not across differing frequencies. If you're listening at about 80db that divergence everyone is freaking out over amounts to a couple of milliwatts of power...

I am not sure yet. I tend to only talk about things that I have direct experience with.

My take on this... The even amplitude multitone test Erin is using is very torturous for an amplifier. It's average power is way higher than single tone. Plus he's feeding in full power signals at 5khz and above. In normal use with music an amplifier would rarely if ever encounter such a signal. First there are no music fundamental tones above 4,200hz, everything above that is harmonic content that decays in amplitude quite rapidly until at 10khz there's almost nothing there. This is normal musical decay as you find in the "Fletcher Munson" body of curves. This is placing an abnormal load on the amplifier. While in average day to day use a 36 or 48 volt 5 amp supply would be adequate, this test is triggering the current limiting protection in the amplifier chip and/or the power brick resulting in artificially low power readings. The TPA3255 has current limiting logic that can be set by a control voltage on pin 7. by manipulating this pin the current limit can be set between 17.5 amps and 12.9 amps and it can be programmed to either self-restart or latch into fault mode. (This is all explained in the "TPA3255 Data Sheet" which you can download for free) A single sine wave at clipping, with a 48 volt power supply of adequate current capacity will produce about 40 peak volts on a speaker. For a 4 ohm speaker that turns into 40 / 4 == 10 amps (200 watts rms)... double that for stereo and you've got 20 amps. Which does exceed the chip's maximum safe current. Increase the average power with a multi-tone signal and you're well beyond the safety limit of the chip and it will limit the output, even at the maximum setting of pin 7. It might even overheat and go into full on protection. I would also not be at all surprised to find out with a 5 amp power supply, the power brick itself is going into an overcurrent panic and reducing voltage to protect itself. In fact, I would suggest that Erin needs to monitor the DC supply during this test. Some bricks just shut off, others will reduce the DC voltage. So... double whammy. But to be honest, I would much prefer it self-limit like this rather than upright burning itself out. There can be multiple reasons why the NAD passed this test where the TPA3255 is driven into protection ... Perhaps the NAD has a higher limit, perhaps it has no protection against over current, etc.

The explanation was very clear, great， brother@@Douglas_Blake_579