Close

The amplifier

A project log for 3D printed Bluetooth speakers - LittleJet!

My goal is to design a 3D printed standalone speaker based on transmission line theory - info: Martin J. King - http://www.quarter-wave.com/

davide-ercolanoDavide Ercolano 02/04/2020 at 16:5811 Comments

Hello!
Since this speakers are meant to be "stand-alone", I'm searching for a little amplifier with wireless connection and this features:

I used to follow some YouTubers working in this field then I knew which IC amplifier to search, two of them actually: the Texas Instrument TPA3116D2 and the Yamaha YDA138-E.
For the "budget" point, I'm searching on Aliexpress (a good platform indeed).
Anyway, almost 2 months are elapsed now and I have some informations to share with you.

FIRST BOARD: WUZHI AUDIO ZK-1002 - $13.09 shipped to Italy

Really good PCB layout, pretty decent capacitors and inductances on the RLC filter and power line, volume POT with ON/OFF, two TPA3116D2 in single-channel mode... "Perfect!" do you would say... NO.
The Bluetooth IC is a wild one, named AS19AP22702-25A4 from a company with a "greek Pi" as a logo (edit: is "JL" logo, by JieLi company); the input audio is managed by this chip, also the input from the AUX jack goes inside to this IC, then, is that bad?! Quite YES


Using the AUX input  the amplifier is NOT linear and the signal get smaller from around 2.7kHz becoming really silly at 22kHz.
Usually this behavior is caused by an incorrect RLC filter, but is this the case?

Connected on Bluetooth you get a linear output till 19kHz... What happen later?...

This! What is this? I really don't know, a noise I would say, but anyway it is crazy output therefrom 19kHz to 22kHz and up.

So the Bluetooth IC it's not reliable and the AUX input is junk.

Anyway, the TPA3116D2 in this single-channel configuration (two, one per channel) it's really good and actually listen to it by Bluetooth is nice, with hi-power and no distortion.

SECOND BOARD: HAZY AUDIO XH-A354 - $14.86 shipped to Italy

When I have chosen this board I was attracted by the YDA138 single-channel configuration, so to have more power on a 4ohm driver, like the one into LittleJet.

The first impression on this board when it's arrived was so terrible: the codes on the Bluetooth IC have been erased, also on the controller for the SD and two more wild IC that I think are some OP-AMP (do someone of you know??). Also the volume POT doesn't have the on/off switch and the worst part: the inductances are ridiculously small. Finally the AUX input goes inside the Bluetooth chip, like the Wuzhi, then... Let's see...


0.5 per division, almost 2v p/p and it is already REALLY distorted (1khz).
And by bluetooth is the same...
At full power the inductances get really hot and also the linearity (from 4kHz to 22) is a shame

Measuring along the source path I found that the signal is good later on the decoupling filter (RC) and also later on the volume POT, then till the BL input, but it was absolutely distorted on the output of the Bluetooth IC.
Listen to this amplifier is a pain, this board is simply garbage. DON'T BUY IT.

Anyway I want to listen to a good YDA138 board, so I'm waiting for a new one (without Bluetooth connection).

BLUETOOTH 5 CONSIDERATION

The actual integration of the standard by these boards (everyone you can find now) is wild and not reliable then I will adopt a different approach: amplifier board (TPA3116) + Qualcomm CSR8675 module. The budget will grow up a little but we will have more audio quality and reliability.
Anyway now everything is going very slow in China because of Corona Virus, so I don't know when I will have this configuration. Meanwhile I will design the integration for the Wuzhi ZK-1002.
See you soon!


Discussions

jockebq wrote 10/27/2020 at 19:34 point

For some reason I cannot reply to your comment. I use a 24V 5A power supply and maybe 1/4 of the volume on the amp. When I raise the volume around 3/4 on the Chromecast Audio I can hear the noise. First I thought it was because the volume was to low on the amp. But raising the volume did not help. According to the manual the Bluetooth should be automatically disabled when it detects a 3.5mm aux input. It does seem to do this at first, but after a while the Bluetooth is back and you can connect to it. Wish to disable BT completely.

By the way, from your post it seems like you recommend this for Bluetooth over AUX?

Do you know any other amp which will fit my needs better? For using with AUX. With less issues.

  Are you sure? yes | no

Davide Ercolano wrote 10/27/2020 at 22:21 point

I would not go over 19v on that board, that sayed, the volume from the souce must be always 100% then use the amplifier POT.
As I can remember if sound come from AUX then this have the priority but the BT will always be on without playing.
In the post I do recommend BT over AUX because of the frequency rolloff but I would use it most by AUX without the aux-issue. Bluetooth chip here doesn't use any of Hi-Fi protocols made by Qualcomm.
If you only need AUX you could buy this amplifier: I selected it for the LittleJet desktop version (I will publish it here): https://www.aliexpress.com/item/32815823913.html

Stay safe!
Davide

  Are you sure? yes | no

jockebq wrote 10/28/2020 at 07:04 point

But the specsheet says that 24V is recommended to get as much power as possible.

I cannot do 100% from the source, because I need the ability to control the volume from the phone when connected to the Chromecast Audio. I have no way to control the volume using the amp's volume knob, as this is placed in my attic.


The amplifier you posted, does it have the issues you mentioned in this article? Or does the AUX work well?

  Are you sure? yes | no

Davide Ercolano wrote 10/29/2020 at 17:29 point

Specsheet apart, it's different from 8ohm load and 4ohm load: I would use 24V only with 8ohm load, but the heat sink would be small anyway for that wattage.

This amplifier doesn't have AUX issue: https://www.aliexpress.com/item/32815823913.html

  Are you sure? yes | no

Davide Ercolano wrote 08/16/2020 at 11:18 point

Hi Claudio, thanks for your tinkering about this. I was thinking to something about those same capacitor, the circuit there was quite suspicius, but anyway I can't confirm because I don't have anymore that board, sorry!

  Are you sure? yes | no

claudio Silva wrote 07/22/2020 at 16:02 point

Hi Davide, regarding WUZHI AUDIO ZK-1002 board I have notice exactly the same issue on the line In input. yesterday I have analysed the circuit and found that the problem is due capacitor C41 and C42 that are connecting que input signal after pass through a resistor to a ground before connect to the bluetooth IC. I have just removed these capacitors and surprise, now the sound is full of bright.

Could you please remove the same capacitors and test again the circuit on the osciloscope?

I am curious how about the frequency response have changed after these changes.

Thanks

  Are you sure? yes | no

bobhavlin wrote 10/09/2020 at 13:20 point

Thanks for the review.  I have the same Wuzhi and noticed the Aux was bad.  So removing C41 C42 works well?

  Are you sure? yes | no

Davide Ercolano wrote 10/27/2020 at 19:33 point

Others people had tryed with success :)

  Are you sure? yes | no

jockebq wrote 10/27/2020 at 18:48 point

I bought 2 of this board to use together with Chromecast Audios. Having two issues, first I get a noise on higher volume. The next bad thing is that the Bluetooth function is not disabled when AUX is plugged in. This means they are always available for anyone to take over and I don't like this idea. Will your solution solve both these issues?

  Are you sure? yes | no

Davide Ercolano wrote 10/27/2020 at 19:27 point

Hello!

Witch kind of noise do you listen? Probably it's the high frequency switching from the power supply, you should try a different one.

Anyway, about the AUX, I remember that the sound from there is the default over the bluetooth one, if isn't there is not much you can do about this because the chip is programmed like so.

Removing the 2 capacitors C41 and C42 only resolve the frequency rolloff issue.

Keep yourself safe! Bye!

  Are you sure? yes | no

Davide Ercolano wrote 10/19/2022 at 16:02 point

Hi Claudio, long time since this comment I know but, hey! It's never late for science! I bought again the amplifier and then measured around: the problem as you found, is the RC low pass filter on the input; with a resistor of 4.6Kohm and a capacitor of 10nF the filter is around 3.5kHz. Probably they just cloned this circuit with the capacitor's wrong value of 10nF instead of 1nF, which would set the low pass filter to a better 34kHz (around) for auto-oscillations purposes.
I can confirm that by removing the C41 and C42 capacitors you are removing the filter so no more falloff in the frequency response (tested on the oscilloscope).
If anyone wants, could swap the C41 and C42 with 1nF for a safer amplifier operation.
Bye and thanks for your help

  Are you sure? yes | no