Topping PA5 Review (Amplifier)

[daniboun]

He would have told already if he was allowed, and the marketing pics also don't tell, plus they had the trouble to cover the actual chip, so no way we will know... Unless a brave soul decides to open his/her unit, uncover the chip and take a picture. Of course losing waranty in the process.

You open the case and you look at the PCB from below and you count the number of pins ... and we should guess the chip)
In my opinion one or more smart people should be able to guess soon ....

 

[Universal Cereal Bus]

Maybe that’s one more standard test or graphic Amirm can add to his usual tests. A high bandwidth O-scope screenshot of a 10kHz sine wave at say 8Vpp into 8ohms (1W). We will see the fuzz if it’s there.

What's wrong with presenting it this way?

 

[PeteL]

You open the case and you look at the PCB from below and you count the number of pins ... and we should guess the chip)
In my opinion one or more smart people should be able to guess soon ....

Most of them are surface mount.

 

[restorer-john]

What's wrong with presenting it this way?

Nothing.

We all know that square waves and HF sine waves look terrible on Class Ds. That's why nobody shows them.

I use square waves all the time as they can show you a whole lot in one waveform.

 

[Bruce Morgen]

I'm curious where you got, "notoriously poor thermal management" from.

From the undersized heat sink and the absence of ventilation holes or slots in the case. When I was using one to drive my main L+R speakers -- 4 ohms and 93dB sensitivity -- the case got very warm to the touch and I suspect that little heat sink was lava-hot inside the unvented case. I now use two A07s for the (8 ohm, 90dB-sensitive) ambience/surround speakers of my 6.1 system and they run much cooler, so I'd say the thermal management is pretty marginal for applications where the load impedance is low and the music is loud for sustained periods. Some means of thermally coupling a suitably larger heat sink -- even a big block of aluminum would do -- to the case would probably be a good idea, but of course such a measure would make a bargain amp more expensive.

 

[TimF]

I am prone to forget that this site is a (mostly) men's journal and typical of the genre the members are want to go on and on about such things as the performance differences between (example) Suzuki Hayabusa and Ducati Panigale, and especially at the extremes of use. Not that it matters to any but a few members, if any. It is fun to do so and fun is good for its own sake. I tried once to downhill ski on a pair of skis designed for and actually used in competitive Slalom and Giant Slalom. It was a waste of a good ski; but the ski was good relative to specific parameter. Otherwise the ski was lousy. We measure and analyze and make determinations and argue specifics that are largely if not entirely in the realm of the inconsequential for most of us in the real world. And that is why it is and can be fun, the consequences won't hurt us--unless you start playing around inside the workings of a tube amp. We should discuss the qualities of the various cinnamon grown around the world. Would you pay an extra $100 to $200 to get an outstanding looking and larger oversize case for this amp? I would for $100 but not $200 more.

 

[Blake Klondike]

Before someone starts complaining that this is not a "real" amplifier as it does not have enough power at 50w into 8 ohms and 80w into 4 ohms, I'd just like to point out a few things:

• For almost any office or other near/mid field listening, this is plenty of power even for inefficient speakers.
• If used with speakers with high but not extreme sensitivity of 89 dB SPL / 1W@1m, it will be the equivalent of a 200/320 watt amp when used with low sensitivity speakers of 83 dB SPL / 1W@1m.
• Similarly, if used at a listening distance of 6', this will be the equivalent of a 200/320 watt amp when used at a distance of 12' to the speakers. Living rooms in many parts of the world are a lot smaller than those in America. This will be fine in them.
• Even in a typically sized living room for someone in America who would be likely to consider an amp in this price range, this will be plenty of power for average sensitivity stand mount speakers being used with a subwoofer.
• It has a decent amount of reserve power for peaks, similar to class AB amps, and it has no problem controlling the moderately difficult to drive Infinity R153s.

If you need a phono input or multiple RCA inputs, this probably is not the amp for you unless you are willing to use a preamp.

While it is totally unnecessary for desktop use, with such low noise and low distortion even at low power, I can't help wishing this came with a remote. However, that might push the cost up another $25-$50.


Yes, this would be great for the JBL Stage A130, in my opinion, although I personally might be inclined to go with a somewhat less expensive amp and somewhat better speakers (Wharfedale Diamond 12.1, maybe?) for the same price.


That is my favorite feature, appearance-wise!

Speaking of pre-amps, is there a candidate for a good match here?

 

[PeteL]

Speaking of pre-amps, is there a candidate for a good match here?

Toppings own Pre90 looks like a good candidate, but if you really want to have a decent set of inputs you are adding 850$ to the cost. With still a limited swt of functionalities by todays standard and form factor that don’t match. I feel this is really tought of to have a balanced dac plugged in it and not much else.

 

[xrk971]

What's wrong with presenting it this way?

Nothing, but show the same FFT of a Class AB amp with a linear supply and none of those peaks are there. -40dB attenuation is 100x so at 5Wrms is 17.9Vpp into 8ohms. -40dB would be about 180mV pp carrier hash. you would be looking at a fuzzy trace if it has 180mV of 600kHz on top of it.

 

[xrk971]

Speaking of pre-amps, is there a candidate for a good match here?

I plan on using my own design DIY preamp. It takes single ended or balanced in and puts out SE and Balanced. Has built in selectable gain via DIP switches) of 0dB/6dB/14dB/20dB and built in low noise +/-15v PSU all fed from a simple 12v wall wart.

Alternatively, I also have a Focusrite 2i4 that has balanced TRS output jacks! So a pair are of TRS to TRS male-male patch cables from Hosa for $10 is all I need to get going. It has a built in DAC of course. Probably not the best but quite listenable. These are about $150 to $200. They make nice audio interfaces for doing your own amp distortion measurements too. That’s actually what I mainly use it for.

 

[Universal Cereal Bus]

Nothing.

Nothing

I could be wrong, but it seems you're merely advocating for a time-domain representation of "fuzz". Amir already provides a frequency-domain representation of "fuzz", and his time is valuable so we should not ask for redundant information if the alternative presentation provides no additional value.

 

[Walter]

Toppings own Pre90 looks like a good candidate, but if you really want to have a decent set of inputs you are adding 850$ to the cost. With still a limited swt of functionalities by todays standard and form factor that don’t match. I feel this is really tought of to have a balanced dac plugged in it and not much else.

I agree. The preamp functions of the E50 or a similar balanced DAC/preamp are the intended match.

As has been pointed out before, there is a real shortage of high performing, lower cost analog preamps--quite possibly because most source components are digital so it just makes sense to combine these functions with a DAC.

 

[PeteL]

I agree. The preamp functions of the E50 or a similar balanced DAC/preamp are the intended match.

As has been pointed out before, there is a real shortage of high performing, lower cost analog preamps--quite possibly because most source components are digital so it just makes sense to combine these functions with a DAC.

Sure but stuff like Sub out is pretty basic stuff, and in many cases a must. Do you know a Balanced DAC that offer that? I'd say it would make sense to add functionalities to DACS, add DSP, an analog or two ins, but who does it? Definitely not Topping. Seams like you either have full fledge AV receivers that offer too much but and perform poorly, or completely nude and raw stuff like this amp. I'm OK with the power part being separated, I think it makes sense it's got a very specific job but then we want great pre upstream and the offer is very weak if we compare to the multitude of straight up DACS that all does the same thing.

 

[xrk971]

I could be wrong, but it seems you're merely advocating for a time-domain representation of "fuzz". Amir already provides a frequency-domain representation of "fuzz", and his time is valuable so we should not ask for redundant information if the alternative presentation provides no additional value.

Although it would be useful if a square wave was shown. That shows slew rate and overshoot that are more appropriate for time domain. But I agree, it adds effort and time required so may not be practical. Also, it will make most Class D amps look rather “ugly”.

 

[TheWalkman]

From the undersized heat sink and the absence of ventilation holes or slots in the case. When I was using one to drive my main L+R speakers -- 4 ohms and 93dB sensitivity -- the case got very warm to the touch and I suspect that little heat sink was lava-hot inside the unvented case. I now use two A07s for the (8 ohm, 90dB-sensitive) ambience/surround speakers of my 6.1 system and they run much cooler, so I'd say the thermal management is pretty marginal for applications where the load impedance is low and the music is loud for sustained periods. Some means of thermally coupling a suitably larger heat sink -- even a big block of aluminum would do -- to the case would probably be a good idea, but of course such a measure would make a bargain amp more expensive.

Interesting. I’m running both an A04 and A07 into Elac Reference DBR62s and neither amp is even warm to the touch when driving to 90+/- dbl.

What voltage power supply are you using? I’m powering these with a Mean Well 36 V power brick. I’ve read other threads where people are powering the A07s with a 48V power supply and complaining of heat but I’m certainly not seeing it.

 

[Billy Budapest]

I don’t want to count any chickens before they hatch, but the PA5 is making me psyched for a larger amp companion to the PRE90, presumably with a built in power supply, full size XLR and RCA inputs, and possibly greater performance, perhaps overtaking the ABH2.

 

[Doodski]

I don’t want to count any chickens before they hatch, but the PA5 is making me psyched for a larger amp companion to the PRE90, presumably with a built in power supply, full size XLR and RCA inputs, and possibly greater performance, perhaps overtaking the ABH2.

Whew! That's a tall order. The AHB2 is pretty full of 2 layers of PCBs and electronics and to make something similar would require a major effort. I wonder how many engineers and such worked on the ABH2.

 

[Bruce Morgen]

Interesting. I’m running both an A04 and A07 into Elac Reference DBR62s and neither amp is even warm to the touch when driving to 90+/- dbl.

What voltage power supply are you using? I’m powering these with a Mean Well 36 V power brick. I’ve read other threads where people are powering the A07s with a 48V power supply and complaining of heat but I’m certainly not seeing it.

I currently use a single 48VDC 10A SMPS throttled down to 46V to power my two A07s. As I noted, the amp cases get just barely warm in that application, so the PSU voltage is not an issue in and of itself -- and the SMPS stays almost cold despite its fan being disabled. It's when I was pushing my former 4 ohm L+R main speakers to room-filling volume that I noted substantial heat on the A07's case -- but there was never any sign of the chip's internal protection mode kicking in and that setup always sounded fine at such levels.

 

[amirm]

I could be wrong, but it seems you're merely advocating for a time-domain representation of "fuzz". Amir already provides a frequency-domain representation of "fuzz", and his time is valuable so we should not ask for redundant information if the alternative presentation provides no additional value.

The original motivation for that graph was to determine the switching frequency. Someone asked about it and I started to measure it. It has usefulness to show how far out the switching frequency is as that can impact the filter design. Recently we had a GaN design that used very high frequency which we determined through that graph.

That said, yes, it is a bit of work as I have to remove my AES filter out of the signal path and then measure this.

 

[pma]

We all know that square waves and HF sine waves look terrible on Class Ds. That's why nobody shows them.

They look the way they look, John. And they have to be interpreted properly, which is not always the case in a non-professional forum.
Below is the step response of Hypex UcD180. It reflects 2 things, switching frequency residuals and output LC filter response which has cut off near 50kHz. Thus 7us rise time. It does not reveal slew rate, slew rate is a nonsense parameter in class D because slew rate is related to input stage topology and in class D input is the comparator with huge slew rate.
With PA5, step response will be similar and a bit shorter, as the LC cut off is tuned a bit higher.