Have you tried to bypass the pot already?
May you try a different power supply for the amp?
Can you try powering the rpi+DAC from battery or an isolated supply?
I suspect some phase-issues here.
Does a phone or mobile player works with the amp?
May you try a different power supply for the amp?
Can you try powering the rpi+DAC from battery or an isolated supply?
I suspect some phase-issues here.
Does a phone or mobile player works with the amp?
Ok, the culprit was the mute circuit I was using for the amp.
I used this: https://www.allaboutcircuits.com/textbook/semiconductors/chpt-5/transistor-switch-jfet/ in order to control the mute via an FTDI board connected to the NUC driving the first DAC. I have the resistor connected to the ground of the FTDI chip instead of the JFET source (in an attempt to isolate the mute switch from the amp as much as possible) which apparently works when the grounds of the NUC and amp are joined via the audio source.
But as soon as I use another audio source (like my phone or the isolated RPI) while still using the NUC for the mute it stops working. Or at least that's my interpretation on what's going on here.
Thanks for help!
I used this: https://www.allaboutcircuits.com/textbook/semiconductors/chpt-5/transistor-switch-jfet/ in order to control the mute via an FTDI board connected to the NUC driving the first DAC. I have the resistor connected to the ground of the FTDI chip instead of the JFET source (in an attempt to isolate the mute switch from the amp as much as possible) which apparently works when the grounds of the NUC and amp are joined via the audio source.
But as soon as I use another audio source (like my phone or the isolated RPI) while still using the NUC for the mute it stops working. Or at least that's my interpretation on what's going on here.
Thanks for help!
So the amp is in mute all the time, makes sense.
Can you post a drawing what you exactly built?
Can you post a drawing what you exactly built?
The resistor should be close to the fet to properly pull down the gate, otherwise it can pickup charge from the wire and switch on.
The amp will play if you remove this at all, for sure. What value is this resistor? 10k should be fine.
The amp will play if you remove this at all, for sure. What value is this resistor? 10k should be fine.
I was using around 560 for the resistor, upped it to 6R49K and now feeding directly from a GPIO pin of the RPI. Resistor is few mm from the FET and connects the gate and source directly. Still seems to need the ground formed via the audio input to function but doesn't really matter.
Seems to work just fine and much quieter than the Intel NUC and USB DAC.
Thanks again for your help!
It's finally begun ...
https://m.intl.taobao.com/detail/detail.html?spm=a1z5f.7632060.0.0&id=555203865367
https://m.intl.taobao.com/detail/detail.html?spm=a1z5f.7632060.0.0&id=555203865367
Good find! I've been looking out for one of those for some time. Here's the same thing on domestic Taobao - https://item.taobao.com/item.htm?sp...tIaxL&id=555203865367&ns=1&abbucket=12#detail
You think Yuan Jing didn't make the 1 million boards they show as available ? 🙂
They took long enough time after announcing the boards.
They took long enough time after announcing the boards.
build in LDO, single ended input, simplified design
Definitely more quality and power than tpa3116
http://m.ti.com/pro
duct/tpa3221
Definitely more quality and power than tpa3116
http://m.ti.com/pro
duct/tpa3221
I can only find their 3251 which is "sold out". The link to 3255 is redirected to 3251.
Nice find. 🙂
But THD is shown as 0.03% compared to 0.005% for TPA3251/55/45 etc.
So this needs some "magic" PFFB. 😀
Last edited:
I think published 1w distortion for 3221 is higher than 3118/28, but maybe they'll expand 3221 datasheet a little. Also noise 3221 is higher than 3118/28.
I think there's changes in temperature, tpa3116 is using ambient , tpa3221 is using case .
Not sure how significant it affect THD
There's no direct comparison in figures of distortion as well. Tpa3116 is using half power at 0.1%. Tpa3221 is 1w at 0.03%
Last edited:
If you check other 32xx datasheet you can find how they get the 0.03% and in the complete 3116/3128 datasheet that 1w distortion(thd+n) is lower. The 3221 inductors do seem lower distortion than the ones chosen for 31xx, so with lower externally added distortion it already seems to perform a little worse at that 1w than 31xx, but 3221 datasheet isn't finished, it may look different when completed.
There's no shortage of 'Taobao agents' who will handle the order for you, for a percentage fee. Google for one.