Douk Audio A5 stereo amplifier review
HTML-код
- Опубликовано: 8 фев 2025
- This new TPA3255 based amplifier features an adjustable high pass filter for powering the main speakers in a subwoofer equipped system. This product was sent to me for free to review.
It is due out in Feb. 2025.
Product links:
www.amazon.com...
doukaudio.com/...
Patreon www.patreon.co...
Email: johnaudiotech@gmail.com
Thank You!
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
As always, another great review from John!
At least they nailed the linearity, and almost got the high-pass perfect.
May I suggest testing the high pass at 80ish Hertz? As at either end of the hp, the filters perform differently. Also, 80hz seems to be a more commonly used frequency when high-passing mains (small) for subwoofers.
Excellent video as always braddah!!🤙🤙
@johnaudiotech Agreed, it would be great to see on the dial, like a clock, where ~80Hz is. Thank you!
You're spot on about not pushing the amp.
I read the power rating on the website, 65 Watt at 8ohms with 32V@5A seems reasonable, i'm not convinced with running 120+ Watt, saw a few cases of Fosi V3 smoking (thermal overload) online with this kind of small chassis so i wanted people to be aware of how these Chi-fi brands are way too overconfident with god-knows what kind of test they did to came up with that numbers, they might got the sounds okay but many of these chifi slop has a really bad thermal capacity with no cooling at all, just to be safe.
my e3audio tpa3255 amp boards have heatsinks but doesn't seem to be enough as they got got at 48v! got fans running above them to help. hard to find china tpa3255 amp board that can operate at full off 48v.
The numbers quoted are for the "Absolute Maximum" power. That is, the point beyond which the amplifier chip would self-destruct. There are built in safeties for overheating and over-current so what generally happens is that when pushed too hard the thing just shuts down. Allow it to cool for a while (10 minutes or so) and it will resume normal operation.
Yes you are correct in pointing out these tiny amps use horrible cooling solutions. They will run just fine with occasional clipping on reasonably dynamic music (>10db dr) while getting only warm to touch. But if you push them with today's over-compressed garbage or test tones they will overheat and turn themselves off. For normal "living room" levels of 5 watts or so, these are a very good choice.
(And FWIW, some of the trash music from the "loudness wars" is actually more punishing than test tones!)
Like John I would not suggest running these things with 48 volt supplies. The Sweet Spot seems to be around 36 volts (6 amp). There you get a reasonable amount of power without excessive heat.
@@Douglas_Blake e3 audio did cost more and was sold as ok for 48v . Others with the boards mention surprise at how hot they get. Its in summer when its a problem. Ok rest of year. They where some of the very first boards with tpa32** series. I am tempted to set boost converter to a lower voltage. Mean to use more and see how well fans hell it to cope.
@@revolvingwonder3777
I've used some of the earlier 3E boards, the ones with the big fanning heat sinks. They do just fine in home audio unless you're trying to make people's ears bleed. They work well enough in PA systems where gain staging is used to limit clipping in the output amplifiers. But, they really don't do so well on the test bench.
This is, in part, because of the tightly regulated SMPS used with these amplifiers. Where unregulated rails, like you find in class AB, produce an apparent "overhead" that is rapidly used up, as the supply sags and not calculated into sustained power, the regulated supplies just go to maximum power and stay there which can produce a lot more heat. If you put an AB amp of similar ratings on a regulated supply, you would get the same results as you get from Class D.
The big trick here is to not clip a class D amplifier. This is easily arranged when you have volume/gain controls in the final power stage and at least one preceding stage in the system. What I typically do is to crank the pre-amplifier wide open, set the power amplifier to the loudest safe level (which is seldom full power) and then in routine use I will adjust the actual listening level with the pre-amp. This gives you a system that simply never clips... and runs a lot cooler.
The older A07 and A3001 amps, with their separate power switches made this real easy. But, unfortunately with these newer mini-amps and the power switch on the volume control, this isn't always practical to do.
thank you for the detailed review and test and explanation of the graphs.
Nice review! A 48v 10a Power supply would nice to see measured with that :)
Hello John .... FWIW ... The TPA3255 by itself has 21.5 db gain. The extra 6 db probably comes from the op-amps that are used as phase splitters to drive the BTL outputs. With 2 volts of input (from a DAC, Streamer, CD etc.) it should just reach clipping at 48 volts.
Checking the spec page... this thing does have PFFB, which seems to be the current fad for these amplifiers.
Miss the old Scope Power Measurement you always do in the past
Me too, so I go and watch an old one to get that old fashioned vibe going, get me some Feeeeeeeeltech.
Be nice to see a bit of square wave squirted into it with a scope hooked on !
@@mrb.5610 :)
The scope's feeling lonely since I got the QA403
@@JohnAudioTech
Actually the scope version of the test is better than the QA version. With the scope you can fine tune to "just below clipping" and get a more accurate reading.
My hookup is a scope to show the output, get it to clipping then reduce the output by 1% and take a voltmeter reading across the speaker terminals.
The WARM tips are probably due to poor cooling.
What are you using to test the amplifier?
Very interesting information.
The distortion versus frequency shown is at only 2 watts, and the chart shows 0.1% distortion at 5khz which is a highly audible frequency.
What is the distortion at 5khz, 20 watts?
With most speakers, 20 watts at 5khz is ear shattering levels though.
You were having so much fun with the QA403 I got one - but cannot find the frequency response chart dbr/hz curve you generate. Do you use the "automated" charts for this?Tx
I use the frequency response button under Generators on the control panel.
If you compare the distortion vs power with that of datasheet, you can see the 8 ohms should be lower than 4 ohms. This makes me doubt about the authenticity of the Chip. I ordered a similar one from Ali for less than $4 all Inc. I will compare it with original one that costs $9 +Ship.
Hi John, I'm just wondering if the aux output is affected by the high pass filter or if it bypasses that?
I didn't check but it should not be since that would go to the subwoofer amp.
Where the opamp's it came with the ubiquitous NE5532 please ?
As always.
@ :)
The output capacitors 1uF63v Wima MKS2 are too small. Maximum 20khz is 4V rms.
I have doubt about the 3255 to be original. The original ones have the printed numbers wider than the heatsink tab. You can try to clean the print with alcohol, if it get erased, it is fake.
I think you may misunderstand the output circuit in a Class D amplifier. The output coils are in series with the speaker, the cap runs to ground as part of a low pass filter used to take out the 400khz carrier wave used to amplify the signals. There is actually very little current going though those capacitors.
@Douglas_Blake look at the datasheet, the maximum admissible voltage curves vs frequency.
It is 4v rms for 20khz. It makes 0.5A .
Instead of proper heat sinks, just let the end user replace the OpAmps when they give up...
It's not the op-amps that get fried. It's usually the TPA3255 chip itself.
We can not make class d amplifiers from scratch so i dont like them.😊
You absolutely can build a Class D amplifier from a bag of parts, no problem.
Do some searching, there are even kits out there you can build.
@Douglas_Blake i want to build from discrete parts.i dont have that much knowledge to do that.i think it is very difficult.
@@HimanshuSharma-dm8ol
In my experience, technology is only "very difficult" until you understand it. Most times there's this "Ahhhhh" moment when it sinks in and from there it gets a lot easier to understand.
If you truly want to build from discrete parts (which is entirely possible) the first step is learning. You need to investigate how Class D works, understand the circuitry used and then, when you have that moment, you can begin experimenting and building.
BTW... that's how most of this stuff gets done... first you get curious, then you learn, then you experiment and finally you do your build. It's no different for a one-off personal project or a manufacturing run of millions... it all starts with learning.
@Douglas_Blake thanks for motivating.
I would like an option to purchase without powerbrick for cheaper, I already got enough power bricks from chinese class d amps.