AIYIMA A07 TPA3255 Amplifier Measurements and Review - LM4562 (and OPA2134) option

[375HP2482]

The noise floor exists, though. Yes, I know it's cheap. If you expect the purest of noise void, $65 won't do it. For now. Or will it?

If you're driving the typical 86 dB/w speakers used nowadays, it's fine.

If you're thinking of driving a 106 dB/w compression driver, you will need something with a lower noise floor.

 

[LtMandella]

I tried swapping the original 5532 op amps for OPA2134 (purchased from Amazon) and amp sounded noticeably worse with the 2134s. I double-checked the orientation to make sure the op amps were installed properly. I have no way to know if any of the op amps - original or 2134 replacements - are genuine.

In any event, I put the 5532s back in and bingo, sound improved back to how it was originally (pretty damn good). My speakers are considered to be easy to drive and fairly sensitive: Reference3A DeCapo I. So ends my experiment with replacing op amps

 

[pma]

Test of AIYIMA A07 for overload recovery time according to IEC 60268-3. The test is described here. Supply voltage 48V, load 4ohm.

AIYIMA A07 fails to pass the test and shuts down after 0.2s. Turn off-on sequence needed to make it running again.
---------------------------------------------------------

The measurement was possible with shorter overload of 0.2s, however it is not according to the standard.

 

[pma]

Maximum short-term power is 148.6W/4ohm for 1s with 48Vdc power supply (current capability 10A).

 

[pma]

THD+N vs. power into 4ohm, measurement bandwidth 45kHz. Supply voltage 48V. An attempt to measure the plot at 20Hz switched the amplifier off at 103W.

 

[LtMandella]

I have another A07 arriving tomorrow for mod experimentation, and then I had a thought. Why don't folks buy the TI eval board for this amp chip - with presumably guaranteed all authentic and high quality components rather than the Chinese boards with probably counterfeit components?

In TI's spec sheet for the eval board, they specifically tout the sound quality. And by the time one upgrades the Chinese board to authentic components the cost is almost the same if not greater. Plus the TI board skips the cheap, low quality preamp section.

What am I missing?

 

[pma]

I have another A07 arriving tomorrow for mod experimentation, and then I had a thought. Why don't folks buy the TI eval board for this amp chip - with presumably guaranteed all authentic and high quality components rather than the Chinese boards with probably counterfeit components?

In TI's spec sheet for the eval board, they specifically tout the sound quality. And by the time one upgrades the Chinese board to authentic components the cost is almost the same if not greater. Plus the TI board skips the cheap, low quality preamp section.

What am I missing?

I do not think that those who are serious in audio would buy this A07 as a main amplifier and would assume quality+reliability. In my case, I bought it for the only reason - I was curious how such “ASR recommended” component would behave like and where are its limits.

 

[TheWalkman]

I have another A07 arriving tomorrow for mod experimentation, and then I had a thought. Why don't folks buy the TI eval board for this amp chip - with presumably guaranteed all authentic and high quality components rather than the Chinese boards with probably counterfeit components?

In TI's spec sheet for the eval board, they specifically tout the sound quality. And by the time one upgrades the Chinese board to authentic components the cost is almost the same if not greater. Plus the TI board skips the cheap, low quality preamp section.

What am I missing?

I’ve always been a bit perplexed by TI’s pricing for the eval boards. If I were a TI product manager (and I have some peripheral experience in that area), I’d price these eval boards at or below cost and subsidize the price with my marketing budget, particularly when the chips are early in their life cycle.

Though a serious manufacturer wouldn’t blink at spending a couple of hundred dollars on development boards, getting folks like ASR types excited about the next greatest chip amp (as we see in this thread) has the potential to create viral talk about my product and buzz = BU$$. Look at all the interest in the new Hypex Nilai.

There’s no question that a number of audio manufacturers - at least the smaller ones - monitor ASR.

Though I’ve never worked for a semiconductor company, I guess they would prefer to court a handful of traditional OEMs and not bother with enthusiasts like us (and deal with all the questions, etc.) which I can also appreciate. If I can spend my finite time with a JBL sound bar product manager or the GM audio engineer to sell my amp on thousands of next years car model, that’s time better spent.

 

[Joe Smith]

We've now got a host of options under $300 for decent lower power Class D amps, with a wide range of feature-sets. SMSL seems to have the most higher-power and rich feature options that appeal to me, but the A07 amp gets in a good sweet spot for my needs - just enough power, just an amp - and for my next amp, I'm probably going to need to get a Hypex amp from Buckeye to get into that comfortable 150 wpc @ 8 ohms power range.

It's hard to say if there is a niche for an A07 like amp with the TI board, improved internal power etc. as that would probably push cost up to close of the Hypex 252 turnkey makers for 2-channel amps...?

 

[HappyMetalGuy]

I do not think that those who are serious in audio would buy this A07 as a main amplifier and would assume quality+reliability. In my case, I bought it for the only reason - I was curious how such “ASR recommended” component would behave like and where are its limits.

What a phenomenal work, thank you.

 

[Schlippwhip68]

Try the BRZHIFI 3255XLR, add a couple of decoupling boards with a pair of Muses 01, 02 or 8820d or even opa2604 which sound great using an E50 dac, Mac Mini with Apple Music and a pair of Monitor Audio speakers and tell me you wouldn’t take that one seriously because you will.

 

[HappyMetalGuy]

I've managed to get a japonese Cosel ADA600F-24-F for free to power the Aiyima A07 all good and no difference to my ears (against the 32V5A that came with the Aiyima).

 

[375HP2482]

I do not think that those who are serious in audio would buy this A07 as a main amplifier and would assume quality+reliability. In my case, I bought it for the only reason - I was curious how such “ASR recommended” component would behave like and where are its limits.

While I appreciate the diligent curiosity of PMA in his tests, it may not be reasonable to expect continued reliability of a product that is stressed beyond its expected performance -- namely, the limits implied by the bundled 32-volt 5-amp power supply, as opposed to Aiyima (and others) cutting-and-pasting Texas Instrument's chart of absolute power limits of the TPA 3255 chip when configured under more optimum conditions (heat sinking, ventilation, convection) than the A07 presents.

In my experience what you get with the A07 when used with its bundled supply is performance whose shortcomings fall outside the boundaries of human perception with the current crop of bookshelf speakers that the majority of music lovers use these days.

 

[HappyMetalGuy]

While I appreciate the diligent curiosity of PMA in his tests, it may not be reasonable to expect continued reliability of a product that is stressed beyond its expected performance -- namely, the limits implied by the bundled 32-volt 5-amp power supply, as opposed to Aiyima (and others) cutting-and-pasting Texas Instrument's chart of absolute power limits of the TPA 3255 chip when configured under more optimum conditions (heat sinking, ventilation, convection) than the A07 presents.

In my experience what you get with the A07 when used with its bundled supply is performance whose shortcomings fall outside the boundaries of human perception with the current crop of bookshelf speakers that the majority of music lovers use these days.

Totally agree. However I didn't mind spending 4€ in two LM4562 (as they measure "a little, little better" with the A07 and as I expected inaudible differences). The power supply Cosel ADA600F-24-F was free so why not (it's 24V 24A 600W and it's set for 29V).

 

[Schlippwhip68]

Totally agree. However I didn't mind spending 4€ in two LM4562 (as they measure "a little, little better" with the A07 and as I expected inaudible differences). The power supply Cosel ADA600F-24-F was free so why not (it's 24V 24A 600W and it's set for 29V).

I would try the Muses 8820d, not much difference in price and very tonally rich.

 

[ToMyr]

No offence but the difference between these two op amps is very much distinguishable. I am a little baffled as to why so many people think that rolling op amps is a waste of time, in my experience is most certainly is a beneficial practice. Sound character can even be slightly different on the same op amp but different manufacture. For example I have a few NE5532 opa from Philips that actually do sound slightly superior to the Texas Instruments version and again it's noticeable when using the A07. I also believe in 'complimentary balance' which means having certain opa in one part of the chain and another type in another part of the chain. I have done this with great results using the Aiyima A07 and the Aiyima T6 Pro. I went through a few different opa in search of a satisfactory sound with either a complimentary balance or a pure line of one sort. In the end it was a complimentary balance between the BB opa2604 and the LM4562 that won the day. You see the opa2604 has quite a thick sound, easy to listen to, not bright or analytical in any way and the LM4562 leans in the opposite direction, not too much mind but enough to liven up the opa2604 and put some sparkle in there while the opa2064 beefs up the low end of the LM4562 quite nicely and the end result is a fuller rich and detailed sound. The LM4562 sitting in the A07 and the BB opa2604 sitting in the preamp output stage. I also used a pair of Mullards 5654 8100 valves instead of the Jan 5654 stock that came with the T6 pro which can be a little bright at the top end but the Mullards are like the opa2604, nice and thick sounding with a strong low end. Overall a very fine result in sound quality.

The tubes in T6 Pro only works as filters and warm cosy glowing lights.

 

[ToMyr]

I don't understand this opamp rolling. To my knowledge the circuit around the opamp needs to be designed specificly for the opamp used. Just because it fits mechanically and has the same pinout doesn't mean it fits the electric circuit i sits in. People don't go around 'rolling' different to-3 trannies in their amps, do they?

 

[ToMyr]

Early 80s speakers! Maybe I should go vintage...

You should always go vintage

 

[Doodski]

I don't understand this opamp rolling. To my knowledge the circuit around the opamp needs to be designed specificly for the opamp used. Just because it fits mechanically and has the same pinout doesn't mean it fits the electric circuit i sits in. People don't go around 'rolling' different to-3 trannies in their amps, do they?

Yes, I agree but it does create a enthusiasts section of MODers and OP amp rollers that learn to solder and read a data sheet even if it makes little to no sense. Everybody starts somewhere and for many this is the beginning.

 

[LtMandella]

I don't understand this opamp rolling. To my knowledge the circuit around the opamp needs to be designed specificly for the opamp used. Just because it fits mechanically and has the same pinout doesn't mean it fits the electric circuit i sits in. People don't go around 'rolling' different to-3 trannies in their amps, do they?

I tend to agree but then again AIYIMA offers an "upgrade" opamp version of the A07 on their website. Do you think they tweak the circuitry for the "upgrade" opamp? I doubt it. Do they just do this for the uninformed?