Topping PA5 Review (Amplifier)

[dcolak]

Would this be a good little amp for something like the JBL Stage A130??

I just put in the shopping cart both! :-D

 

[Ro808]

Earlier this year I sent an email to Topping asking if they might also have balanced (power) amplifiers in the pipeline.
They replied that this was indeed the case.
The result was well worth the wait as it turns out.
It's a confirmation that implementation matters.

The noise levels are a blessing for owners of sensitive loudspeakers (≥100 dB) that don't require the power of the ICEedge, Ncore and Purifi amplifiers.

 

[TheWalkman]

We can also infer from the fact that 4 audio grade 10uF electrolytic caps are used between the pot and 4 more between the D-01 block and the amp that this is very close to a TPA32XX implementation. The fact that there is a DC blocking cap shows that the opamps are running single rail and at 24vpp (since the rail cap of the DCDC buck converter followed by linear voltage regulator are 25v rated).

Now I believe the potted block D-01 contains opamps arranged as a balanced line receiver / buffer, followed by more opamps for a gain stage (9dB), followed by a opamps arranged as a balanced line driver. Probably OPA1642 (could be almost any low distortion, low noise, high performance dual package opamp) . All running 24v single rail. The TPA32XX has modified PFFB for about 10dB intrinsic gain. Overall gain is 19dB. For amp to output 68Vpp requires that the opamps swing 21.5Vpp (assuming amp is 10dB gain). Hence 24v rail should be sufficient to hit clip. TPA32XX provides built in DC offset protection, thermal protection, shorting protection. The relays between the input jacks and the pot provides anti pop turn on delay protection.

Since a 4 gang pot (Alps RK09 series) is used (5th section is just base for main knob shaft, 4 pins on left side) linearity of volume tracking and CMRR for bal input will not be perfect at partial volume levels. Best performance will be achieved at max volume and use preamp to control volume.

Edit: heatsink removed and all is revealed. That is TPA32XX for sure. Even the 22k current limit setting resistor is there to allow max current to be achieved.

Edit: after studying the underside of the PCB I see that the PFFB traces go to the end pins on the D-01. So it looks like the PFFB is implemented via AC coupling where the distortion is more important. At lower DC frequencies, the feedback is not important. I take back what I said about where PFFB loop is. It does look like the amp may actually be a composite amp design now. Please see feedback traces below.

View attachment 170001

So XRK,

You clearly understand the nuances of what you’re seeing, more so than probably most here. Are you seeing anything in this design or implementation that should raise the manufacturing costs by 5X over another 32XX based amp (like the Aiyima 04 or 07)?

In other words, could a bit of better engineering take a $50 amp and raise its performance to the same level, raising the bar to best in class performance for an MSRP of, say, $125 - $175?

FWIW, I‘m happy (and did) pay the $350 for the Topping but if manufacturers can use these techniques and take the category from, really, a toy amp product to a serious, audiophile level product, could the lower price point open up the category to a new generation of listeners?

 

[xrk971]

I think Topping has used the best output inductors available (same as what I use on my commercial TPA3255 amp). These four inductors alone cost $24 (in quantity) and clearly, there are about 8 premium opamps at about $4ea so another $32 there. So I see at least $56 in upgraded parts and if we add the 4 gang pot, two high quality relays, four TRS jacks, two MOSFETs for remote power switch, I see another $30 in components. So we are at $80. Typical markup for retail is at least 3x so $240. Add that to usual $100 cost for a cased TPA3255 amp and we are at $340. The cost is fair and reasonable. You can’t sell it for less without skimping on features. Sorry $350 is a fair and excellent price for the level of performance you are getting here.

 

[Billy Budapest]

So XRK,

You clearly understand the nuances of what you’re seeing, more so than probably most here. Are you seeing anything in this design or implementation that should raise the manufacturing costs by 5X over another 32XX based amp (like the Aiyima 04 or 07)?

In other words, could a bit of better engineering take a $50 amp and raise its performance to the same level, raising the bar to best in class performance for an MSRP of, say, $125 - $175?

FWIW, I‘m happy (and did) pay the $350 for the Topping but if manufacturers can use these techniques and take the category from, really, a toy amp product to a serious, audiophile level product, could the lower price point open up the category to a new generation of listeners?

I really cannot understand why people are bellyaching about the $350 price when performance is equivalent to $1000-$2000 Purifi-based amps and approaches the $3500 Benchmark AHB2.

I guess because of the small size people are calling it a “toy amp” and cannot wrap their heads around its SOTA performance. What is this place, Stereophile?

 

[TheWalkman]

I really cannot understand why people are bellyaching about the $350 price when performance is equivalent to $1000-$2000 Purifi-based amps and approaches the $3500 Benchmark AHB2.

I guess because of the small size people are calling it a “toy amp” and cannot wrap their heads around its SOTA performance. What is this place, Stereophile?

Billy,

It appears you missed my point: I bought the Topping, and as stated in my post, I’m not complaining in the least.

What I was asking of XRK, who lists himself as a manufacturer and is clearly knowledgeable of the design aspects of the Topping, is there anything driving the 6 - 8 X price of the Topping or could these techniques be replicated inexpensively in other 32XX offerings.

From what I’ve read, younger folks aren’t embracing audiophilia as they once did and the demographic gets older. If you believe in simple economics, less expensive, audiophile amps could entice many new people (particularly younger hobbyists with less disposable income) into the hobby, creating the next generation of audio aficionados and ultimately helping the entire industry.

Though there will always be the Benchmark/ McIntosh/ Genelec crowd, for a lot of folks, $350 is a lot of money. A $150, killer amp is much more attractive to them, they fall in love with it and 15 years down the road, they want more power, more buttons, more glitz and become the high end customer that sustains the industry.

 

[merelyok]

I really cannot understand why people are bellyaching about the $350 price when performance is equivalent to $1000-$2000 Purifi-based amps and approaches the $3500 Benchmark AHB2.

I guess because of the small size people are calling it a “toy amp” and cannot wrap their heads around its SOTA performance. What is this place, Stereophile?

I think when someone has paid 2-4K for an amplifier that doesn’t measure as well ( and on paper shouldn’t sound as good - if audible ), they would be hard pressed to justify their purchase, and thereby other intangibles come into play such as the amplifier having a “soul” or “body”. Honestly If you gave me a blind test between the PA5 and other more costly amplifiers, I swear I wouldn’t be able to differentiate them based on SQ.

It ultimately boils down to the way the industry is i guess, where one assumes that an expensive boutique product will invariably sound better than a mass produced ( but well engineered ) product from China. The dinky Chinese amp with its small case has *no* right to sound better than a hand made artisanal product. It doesn’t do timbre or capture the atmosphere well enough because it only costs 350 USD and comes with an external power brick.

But there have been too many a time that I voted with my wallet and got burnt. Something that is more expensive isn’t necessarily better. ( ok fine, a huge ass amp with fancy casework with a fancy R2R dac that weighs almost as much as a small car will look got any rack, but will it really sound better ? ) If I’ve already spent that much money on it, you bet your ass my brain would tell me it sounds better!

 

[DanielT]

It seems to be really good but I am most interested in how it performs / measures from 6 kHz at half power then up to clipping, measured at bandwidth up to around 80 kHz plus how it handles tough loads from demanding speakers.

I am interested in how it performs on the rally track and not in the depot, out at sea and not in a windless harbor....and so on. You understand the point. I'm very curious. This amplifier seems to have potential. Good job Topping!

 

[pma]

@amirm , any chance you would measure a CMRR and input impedance at several positions of the volume pot?

 

[Rottmannash]

You might want to look at Hifigo who have it for a out £250 ish and take care of tax and delivery. I have bought a out three times from them and never any issues.

$349 on Hifigo in the US market.

 

[daniboun]

I think we should wait to see the price of the next Allo TPA3255. I shared the first picture + measures in another thread.... This new amp has some innovative spec too : air core inductors / ultra low noise psu / new op amps / PFFB etc
The amp will come with XLR and RCA inputs and probably in a full alu case. I expect to see it @ less than $ 400.

I am pretty sure that Allo s team is reading us and is aware about the perf. of the PA5 )

Wait and see ))) good match to come

 

[pma]

@amirm , any chance you would measure a CMRR and input impedance at several positions of the volume pot?

I am asking because years ago I designed and built a balanced preamp (both I/O balanced) with 4 gang blue Alps potentiometer and I know very well where are the parameters limitations, it is especially in CMRR.

 

[Sheriff1972]

Here a TPA3251 with differential OP amps, could help to compare... (OPA1632 + OPA1612)

That's pretty. What is it?

 

[daniboun]

That's pretty. What is it?

This is a TPA3251 with differential OP amps. It has never been tested and only a few people know it.

 

[TimF]

I'm happy this was introduced and made available before the end of 2021. It is a swell way to end the year. This is pleasing. Why not good things for the masses--or at least the middle class. I see no reason to object to Topping making some money. Otherwise why bother. Last year was a blowout year for DAC's, now we are seeing developments in class d offerings. By the way, your combined detective work clearly beats Columbo. So now we have witnessed western interests reverse engineer Chinese products. As the world turns.

 

[daniboun]

there are 18 welds under the D-01 secret box including 10 at the top and 8 at the bottom .... It leaves a lot of room for mystery)

 

[Papaya_X]

Amazing, thank you guys for deciphering the mystery, and good job from Topping to push the limit of this TPA325X, they are now winning again against infineon's chips !
Amazing time to live for audio tech enthusiasts

Would it be worth to put a beefier PSU ? (Like 36V 6A or more for example ?)

 

[daniboun]

Amazing, thank you guys for deciphering the mystery, and good job from Topping to push the limit of this TPA325X, they are now winning again against infineon's chips !
Amazing time to live for audio tech enthusiasts

Would it be worth to put a beefier PSU ? (Like 36V 6A or more for example ?)

If you consider the great measurements the amp is just fine with the stock brick psu )

About Infineon chips vs Tpa325X it is pretty like AKM vs Sabre.... Matter of taste ) I personnaly prefer the AK4499 chip.... The same for Infineon )

 

[TNT]

Run it on 3 batteries?

What is absolute maximum on the DC inlet?

//

 

[KSTR]

I am asking because years ago I designed and built a balanced preamp (both I/O balanced) with 4 gang blue Alps potentiometer and I know very well where are the parameters limitations, it is especially in CMRR.

View attachment 170046

Normally, one would place an kind of volume pot after the input diff-amp (the "dereferencer") so that it never compromises CMRR. The whole idea of balanced input is that the difference signal gets referenced to the local GND as soon as possible.

In case of a balanced signal path after that, the dereferencer would simply have a balanced output, ususally done by replacing the 3rd opamp (the subtractor) of a standard instrumentation amp with a "super-balanced" version (as D.Self named this configuration).

But we don't know, so you are right this should be checked for.