I have tried a number of tpa amps. I liked the filterless Sure tpa3110 and the ybdz green version the most. I tried your really small amp today with a meanwell 19v supply, schiit modi dac on some ls3/5a clones. Sounds great. Just a small turn on pop. I am impressed by this tiny amp. Lowering the gain to 20db gave the ybdz more body. That is perhaps the only mod i would like to try on this board. I will use this tiny amp for a few weeks to see if it grows on me.
Thanks for the feedback. The 32dB gain was chosen as it is most suited to being run from a phone or something (my phone at least). But changing it is only a matter of swapping one resistor.
Well that is true of any half decent amplifier. As this amp is designed primarily to be squeezed into a portable speaker system, a CD player or Vinyl player is likely not going to be used.
My two amps from Sheldon arrived today.
Sheesh, those things are really tiny - I mean postage stamp-size tiny... 🙂
Could't wait and quickly soldered a pot I had at hand (50K) and some RCA cinch inputs, in order to try it out.
First impressions are very good. Haven't done any critical listening yet, but already I can say that it's better than unmodded original YJ 2.0 red board. I'm not quite sure, but it could be better than HIFI (Breeze audio) TPA3116 stereo (2.0) amp which I have. No glare in the higher registers, bass is nice, stage is detailed, there's really nothing amiss... Powered by 12V 4A laptop "brick", sound coming from my Raspberry Pi + Wolfson DAC kit, speakers are Infinity SM 85 (98dB, 8Ohm). The sound is really nice... And the quality and workmanship is better than the usual stuff that we discuss here. Solid connectors, good soldering, no complaints here.
So, after brief first impressions, I can say that I'm very happy with these little buggers and that I expect a lot of good stuff from Sheldon in the future 🙂
Sheesh, those things are really tiny - I mean postage stamp-size tiny... 🙂
Could't wait and quickly soldered a pot I had at hand (50K) and some RCA cinch inputs, in order to try it out.
First impressions are very good. Haven't done any critical listening yet, but already I can say that it's better than unmodded original YJ 2.0 red board. I'm not quite sure, but it could be better than HIFI (Breeze audio) TPA3116 stereo (2.0) amp which I have. No glare in the higher registers, bass is nice, stage is detailed, there's really nothing amiss... Powered by 12V 4A laptop "brick", sound coming from my Raspberry Pi + Wolfson DAC kit, speakers are Infinity SM 85 (98dB, 8Ohm). The sound is really nice... And the quality and workmanship is better than the usual stuff that we discuss here. Solid connectors, good soldering, no complaints here.
So, after brief first impressions, I can say that I'm very happy with these little buggers and that I expect a lot of good stuff from Sheldon in the future 🙂
Well, after some more listening and testing, there is one minor issue - turn-on pop... But I guess it will be resolved in the next version, I think Sheldon is working on it...
Good to hear you're happy with them.
The next revision won't have the turn on pop fix, I'm still working on that. I wanted to get this batch done before xmas so the gerbers are sent off already. Next version will have anti pop sorted.
The next revision won't have the turn on pop fix, I'm still working on that. I wanted to get this batch done before xmas so the gerbers are sent off already. Next version will have anti pop sorted.
I am keeping this tiny amp in my main system. I think it sounds very good. ( It shouldn't, elco's in the signal path and to little capacitance in the power supply ) And it really makes a lot of fun showing of this amplifier. Can you show me the gain setting resistor?
Glad you like it. As I said before this is not my first board with this chip so the components were chosen before the board design. Next version (PCBs on their way) has the turn on pop fixed, and I switched to ceramic input caps mostly for assembly ease.
There's also no problem with electrolytics in a signal path, provided they are used correctly.
It has 440uF total on the power rails, which is fine IMO for something like this. A switcher like this won't care if you have 440uF or ten farads on the supply rail, it'll still have plenty of noise at the pins. Also this board has separately filtered AVCC... Also this power output is quite low. This isn't a class A or B remember.
One thing that helps this amp is the star ground and the tight board layout. So many of these class Ds use big boards that are spaced out, most definitely not the right way to do things. Class D amps are switching circuits and should be treated as such.
I have attached an image of the gain setting resistors. Look up the datasheet schematic, in relation to that schematic the resistor on the left is R1, and the one to the right of it (middle of the three) is R2. Use the table to select your values. Here, R1 is 39K and R2 is 100K.
Last edited:
Using MLCCs as input caps might be an issue. They don't like being voltage biased and depending on on voltage rating can lose 60% or more of it's rated capacitance. And then there's the piezo effect on an audio range signal. It will most likely have significantly higher THD. Usually by a factor of ten.
PMLCs (polymer multi-layer capacitor) by Rubycon would be the preferred compact outline choice for signal input, decoupling of power supply and output filter caps but especially for signal input caps.
See especially page 9 of this overview www.rubycon.co.jp/en/products/topics/img/t001_22.pdf
PMLCs (polymer multi-layer capacitor) by Rubycon would be the preferred compact outline choice for signal input, decoupling of power supply and output filter caps but especially for signal input caps.
See especially page 9 of this overview www.rubycon.co.jp/en/products/topics/img/t001_22.pdf
Last edited:
I'm aware of this. The ceramics I have chosen are 50V, so the capacitance nonlinearity is almost a non-issue.
Those PMLCs are interesting. The main issue with those is the price - I want to keep these things cheap. The new boards have 1206 caps for the input coupling caps and the LC filters. The large footprint was chosen for those components as they are the most likely to be changed by the end user, so it's easier to solder. I'll see if I can get hold of some and give them a go.
- Status
- Not open for further replies.