Topping B100 Amplifier Review

[DWPress]

Not fully convinced many failures are not "user error" but glad I didn't pull the trigger on getting a set of these for my tweeters - which would probably be fine since they are 8ohm and power requirements are smaller. Still feeling the pain of the PA5 v1 that toasted itself pulling the same duty though.

 

[Chr1]

Aye, but surely there is protection circuitry for this?
Clearly it doesn't work.

Edit: Perhaps I am wrong here, but surely a competent amp should not just expire when exposed to a difficult speaker load?

 

[miero]

The protection circuit is possibly implemented in the firmware. The MCU has 108 MHz, which is high enough. Which would mean that it could be adapted by a software update.

 

[Chr1]

Fair enough. But it should not really be necessary. And I doubt those with broken amps and difficult returns will find that comforting.

High SINAD. Low reliability.
Bad product imho.
Again, my 2c.

 

[Old_School_Brad]

“The rubber tears more easily than expected.”

Oh, you're a father too?

 

[Deleted member 72219]

Oh, you're a father too?

Well, yeah, actually…a 6-year-old and a 4-year-old, both highly destructive little boys.

-Ed

 

[pogo]

a competent amp should not just expire when exposed to a difficult speaker load

Topping would actually only have to extend its note, perhaps to 'at least 4 ohms: purely ohmic or an equivalent EPDR value'. I assume that the amp performs well with a purely ohmic load of 4 ohms and was designed accordingly.
Does anyone from topping actually read along here and could confirm one or the other or not?

 

[miero]

They might read it here, but they might not be allowed to reply here.

For a direct contact with Topping you might use the “service email” specified here:
- https://www.toppingaudio.com/about

 

[Old_School_Brad]

They might read it here, but they might not be allowed to reply here.

For a direct contact with Topping you might use the “service email” specified here:
- https://www.toppingaudio.com/about

or
@TOPPING
@JohnYang1997

 

[KSTR]

The protection circuit is possibly implemented in the firmware. The MCU has 108 MHz, which is high enough. Which would mean that it could be adapted by a software update.

I think we have several factors here at work
- the sensing circuits
- the algorithms used for modelling the device state
- the robustness at the choosen limits when they are violated constantly
- the general design and execution quality, and production quality control

There might be issues with all of the above. Topping seem to have a very aggressive policy, trying to make profit by cutting as much cornes as they can, everywhere and all the time, except concentrating to land in the top 10 of the SINAD charts.

 

[Sokel]

If the protection circuit follows the action of this bimetal switch or if it cuts the circuit of its own I would look at this first.
They are notoriously inaccurate but cheap I guess.

(I hope they didn't implemented this after my suggestion at the fosi mono thread,this was a desperate measure I proposed to some users to control fans,not to track critical temps)

 

[MAB]

Meant IC integrated circuit (chip) ...if it was fully discreet class B amp (not chips not firmware not USB c ...just linear PSU, capacitors, transistors, resistors, inductors... It would MUCH more durable, it would perhaps last forever...but perhaps internal programmed obsolescence is involved. My TVC all discreet line stage line preamp no only inside transformers wounds and point to point aerial internal wiring and some switch made in Switzerland, that preamp is 8 years old daily use and it will last forever, on other size , my friend Julio loved her Topping line preamp... Till like our b100 it blew up too. it was many technician even traveled to china, no fix. Topping is great, best actually, for DAC. But preamps and DACs has no reliability at all. I was considering B200... It is actually**** discreet?? Idk. Meanwhile I'm happy with my 3E Audio TPA3255 PFFB module amp board based DIY power amp also reviewed by Armir, not best measurement in world but as good as NAD Purify comercial power amp. I wish there is a Nilai or AHB2 cheaper but there is not. Meanwhile Im happy with 3E Audio PFFB amp diy

Best to minimize anecdote and speculation on what is wrong. We don't know why these particular amps failed, we do know that discrete amplifiers are not intrinsically more reliable than chip-based amps.

In fact, modern chip amps have advanced reliability and safety features built in (temperature and voltage protection, short-circuit protection, soft-start and -stop, clipping, etc.) Discrete require all of this to be designed, built, and tested by the product development team. Also, IC manufacturers offer this in a package that is rel-tested. So long as the product doesn't do anything odd in implementation, or mess up during manufacturing and testing, the IC isn't intrinsically disadvantage to fail. The presence or absence of luxury components, like switches from certain regions, or inductors wound by certain individuals, doesn't fundamentally change the product's reliability.

Serviceability is sometimes better for discrete circuit implementations.

 

[Chr1]

Have to say that the fact that it seems several have failed so soon does not reflect particularly well. Again.
Surely this is the sort of thing that should have become apparent in pre-release testing?

 

[Deleted member 72219]

Have to say that the fact that it seems several have failed so soon does not reflect particularly well. Again.
Surely this is the sort of thing that should have become apparent in pre-release testing?

I was one of the first people to get these and I put a good several hours of play into my system daily (background music while I work), so perhaps it takes a lot of thermal cycles before things go awry.

-Ed

 

[transformer-coupled]

Best to minimize anecdote and speculation on what is wrong. We don't know why these particular amps failed, we do know that discrete amplifiers are not intrinsically more reliable than chip-based amps.

In fact, modern chip amps have advanced reliability and safety features built in (temperature and voltage protection, short-circuit protection, soft-start and -stop, clipping, etc.) Discrete require all of this to be designed, built, and tested by the product development team. Also, IC manufacturers offer this in a package that is rel-tested. So long as the product doesn't do anything odd in implementation, or mess up during manufacturing and testing, the IC isn't intrinsically disadvantage to fail. The presence or absence of luxury components, like switches from certain regions, or inductors wound by certain individuals, doesn't fundamentally change the product's reliability.

Serviceability is sometimes better for discrete circuit implementations.

Oh! So an all discreet output Hypex/Purifi/Benchmark amp is not more robust and reliable regarding low impedances or even power surges than a chip-amp like Tripath or Texas Instrumments based projects? fascinating
Said that, contradictorily, im super happy with my, reviewed here,
3e Audio PFFB stereo amplifier module for DIY , with a toroidal power supply, box, connectorr wiring and power switch, it sounds really clean detailed and nice -as measured by Amirm, and that a chip amp based tpa34255 by Texas Instruments... Wish to known the budget for do it with a Hypex Nilai module and mainly the knowledge to set it with a full multivoltages linear power supply (i know they "measures worse" but i like it) but i don't, meanwhile im super happy with my 3e Audio DIY module based amp and my high impedance floorstanders

 

[MT_bassbone]

Could these mono blocks power larger towers like CSS Audio 2TD-X's or Neil Blanchard MLTL-6's? I am not understanding the safety mode/shut down thing? Does it reset? How often does it happen?

 

[miero]

This depends on the usage. Typically it seems to be one of the following:
1) too much peak power into a low impedance
2) overheating due to inadequate cooling

 

[restorer-john]

Surely this is the sort of thing that should have become apparent in pre-release testing?

You would think so wouldn't you? I'd say the designers probably thoroughly tested the amplifiers on typical small bookshelf speakers and all was well. So they pushed this thing out the door and let the real world beta testing be done by their paying customers.

Then you get reviewers merely testing amplifiers on an analyzer and dummy loads and not actually listening at medium-high levels for an extended period on a pair of large multi-way loudspeakers. Loudspeakers with potentially challenging impedance curves, and subsequently, they don't experience these failure/shutdown events either.

They might be Topping the silly SINAD chart, but they are Bottoming the reliability chart in my books.

 

[restorer-john]

we do know that discrete amplifiers are not intrinsically more reliable than chip-based amps.

We know do we? I didn't get that memo.

Amplifiers deliver high voltages and high currents on a constant basis into very low impedances. That's what they are supposed to be designed to do. Well designed amplifiers can do that, but IC (chip-amps) based amplifiers need a whole lot of housekeeping to keep them from not blowing up when the going gets tough. They are always a compromise, because their entire existence is driven by minimizing costs.

Discrete amplifiers can parallel up output devices, beef up the power supplies and improve the heat sink efficacy. They can be scaled easily, a basic design can serve multiple price points and performance demands, all while being extremely reliable and low cost, especially when it comes to repair.

A chip-based amplifier (we are talking here OPTs on the substrate, not a 'driver' chip for conventional OPTs) is a take-it or leave-it design, mostly.

Take the biggest and most widely used "chip" based HiFi amplifier devices I can think of. Historic, for perspective in this discussion. Sanyo's thick film hybrids (STKs) from the 1970s through the 1990s. They had many hundreds of STK SKUs. Ranging from single darlington output packs to an entire 150W+ stereo amplifier in a SIL package. Millions and millions sold to OEMs over 3 decades. Pretty much every brand used them at some point. Be it class A driver stages for a real output stage or basically an entire power amp in plastic package.

STK based amplification was sold on cost, ease of assembly and guaranteed performance (it was average at best). Not SOTA, but good enough for Joe Average. They also are renowned for blowing up in the face of difficult loads. You cannot short an STK and expect it to survive. They overheat easily and fail, especially when installed on heatsinks that are too small. They can oscillate and destroy themselves. They sound fine when they are working, but Sanyo discontinued the entire STK thick film range in December 1995 and left everyone out in the cold. Now there are only fakes and very old NOS or Chinese SMD copies that don't work. People have resorted to building discrete equivalents to keep their gear alive. Same thing happened in the 70s/80s when Sanken killed their "amp on a chip" hybrids. People had to reverse engineer the IC or throw their amplifier on the trash. We know most went to landfill...

Here's a nice little complimentary pair of Toshiba outputs I scavenged from a 1980s amplifier I found on the side of the road. About AU$7 per piece to buy today. Yes, you can still buy them!

High voltage : check! 180V Vce
High current : check! 12A Ic max
High dissipation : check! 130W each.
Junction temp max: 125 degrees max
High fT? : 30MHz.

You blow up you amp, you can fix it if it's discrete. In the 1980s, 90s, 2000s and today in 2024. Big difference.

Long post, I know, but amplifiers, particularly power amplifiers, need devices and design that can withstand abuse and real world demands. That's why they are called "power" amplifiers.

 

[MAB]

We know do we? I didn't get that memo.

You didn't get the context either.

And thanks for conflating serviceability and reliability in your manifesto!
Poorly designed discrete products have extreme unreliability.

Agree on the ruggedness part though. But that doesn't appear to be the issue here.