Yep, agreed, I love the Murata 1206 common-mode chokes and ferrite beads. I don't knock them in general, just in regard to their X5R/X7R ceramic performance under bias. Their C0G caps are also just as good as TDK's as far as I can see.
'Blah' wasn't meant to be unfriendly/offensive/whatever, I meant to imply "meh, not a difficult problem to practically solve..." or something along those lines. Apologies if it was misinterpreted.
The amp doesn't have to be driven with a "80% duty all the time" to cause rail droop at the switching frequency, just driven with a signal loud enough for the music envelope to reach that point. 80% is an arbitrary number also, the same effect still exists though much less pronounced at lower duty cycles, and obviously greater when you're pushing the amp the whole way to the rails.
The effect I don't like is that PVCC's "nastiness" is modulated by the music envelope - with rhythm, with a pronounced bassline, etc. The TPA has pretty decent PSRR (-55dB at 20KHz) which should keep the "directly audible" nastiness on the PVCC rail from showing up on the TPA's output, but I wonder whether higher frequency crap entering the chip can end up mixed/modulated/whatever down into the audible range within the TPA itself.
I think as a general rule, keeping your rails as stable/clean as possible on a class D amplifier is prudent. Even if you make them unnecessarily clean, it certainly can't hurt anything.
Thanks for the clarification. 🙂
I wonder if the influence of supply voltage variation is a "big" problem due to the feedback.
To my knowledge, a lot of people who "tune" their amps with caps and "magic" do just hear an inprovement because they're biased to this. But actually nearly none of them do real measurements/simulations to compare. An Ear isn't an accurate measuring device..
But well, "you need to believe".
From my experience, the TPA series is good, sound wise, but really bad with their power-on/off sequencing. (I.e. the Maxim amps got internal POR-delay - adjustable via timing-C, no need to bootstrap, no need for snubbers, easier to layout, has better thermal design.. but more expensive and abit lower efficiency.)
I wonder if the influence of supply voltage variation is a "big" problem due to the feedback.
To my knowledge, a lot of people who "tune" their amps with caps and "magic" do just hear an inprovement because they're biased to this. But actually nearly none of them do real measurements/simulations to compare. An Ear isn't an accurate measuring device..
But well, "you need to believe".
From my experience, the TPA series is good, sound wise, but really bad with their power-on/off sequencing. (I.e. the Maxim amps got internal POR-delay - adjustable via timing-C, no need to bootstrap, no need for snubbers, easier to layout, has better thermal design.. but more expensive and abit lower efficiency.)
As far as I know, we still haven't found the full Wiener.
Weiner vs wiener:
http://dictionary.reference.com/help/faq/language/s39.html
I recall 'braxy mentioning improvements in the input grounding and input routing for that board (as well as other improvements from others.)
...
As far as I know,
Still lookin' for The Wiener.
Last edited:
My modified YJ blue black board has 330uf 25v silmics right now. Afaik silmic has ~600mA ripple current. I have bunch of 470uf 63v nichicon caps, rated approx. 1.8A ripple current. Is it good idea to replace silmics with nichicons?
2.1 Progress
Hi Alontuch - I wondered if you received your replacement boards & if yourself or Danzz had made any progress with the 2.1?
I am wondering why there isn't a board available that provides proper sub XO frequency control. I am kinda surprised - but perhaps there isn't sufficient a market to justify developing a 'fully functional' 2.1 amp?
I see this guy has developed a base board for two 3116 amps with filtering & controls - but he doesn't appear to have reached the stage of testing with the 3116's yet.
I also found this guy had modified a YJ 2.1 board to provide a variable sub frequency control and improve low frequency response.
It would be really nice to have a 'proper' 2.1 board available for us (well me at least) to tinker with 🙂
Hi Alontuch - I wondered if you received your replacement boards & if yourself or Danzz had made any progress with the 2.1?
I am wondering why there isn't a board available that provides proper sub XO frequency control. I am kinda surprised - but perhaps there isn't sufficient a market to justify developing a 'fully functional' 2.1 amp?
I see this guy has developed a base board for two 3116 amps with filtering & controls - but he doesn't appear to have reached the stage of testing with the 3116's yet.
I also found this guy had modified a YJ 2.1 board to provide a variable sub frequency control and improve low frequency response.
It would be really nice to have a 'proper' 2.1 board available for us (well me at least) to tinker with 🙂
Quoting machineage
"It would be really nice to have a 'proper' 2.1 board available for us (well me at least) to tinker with" 🙂
plus One here....
"It would be really nice to have a 'proper' 2.1 board available for us (well me at least) to tinker with" 🙂
plus One here....
maybe 'proper' and 'functional' depends too much on what speakers are used.
maybe you should try an active crossover and two 3116-modules?
Well for fun please remember also new tpa3116 datasheet dated january 2015 forgot to include what was already long known in september 2013, datasheetvalues for synch do not function properly...so slave has random switching frequency if Chinese just copied datasheet...guess what, they do just that
I have been thinking about how to get a low cost 2-way passive line level XO to use with a pair of TPA3116 amps. The problem has been that most RCA level outs don't have enough oomph to drive them properly. I then noticed some fairly inexpensive 1:4 headphone amps with independent volume knobs. That headphone amp can easily drive a 2nd order or even 4th order PLLXO comprised of nothing more than low cost caps and resistors designed for say a 30kohm input inpedance of the 26dB gain 3116.
For about $25 to $45 an 8 channel buffer driver with gain can be had. With this many channels you could even do a 3 way or 4 way speaker.
Take this for example:
Amazon.com: rolls HA43PRO 4 CH Headphone Amp: Musical Instruments
For about $25 to $45 an 8 channel buffer driver with gain can be had. With this many channels you could even do a 3 way or 4 way speaker.
Take this for example:
Amazon.com: rolls HA43PRO 4 CH Headphone Amp: Musical Instruments
I read Post 6341 by GMASH and manage to solder a 0.1 UF 805 x7R cap on the PVCC pins of the red YJ board, I scrape away the mask to access the ground plane.
I can report that the improvements is as much as the bootstrap snubber mod.
Thanks to GMASH.
My next target is to solder a 10 UF MLCC cap nearby but not enough space as the heatsink is too big.
I can report that the improvements is as much as the bootstrap snubber mod.
Thanks to GMASH.
My next target is to solder a 10 UF MLCC cap nearby but not enough space as the heatsink is too big.
As far as I know, we still haven't found the full Wiener.
Weiner vs wiener:
Dictionary.com
I recall 'braxy mentioning improvements in the input grounding and input routing for that board (as well as other improvements from others.)
...
As far as I know,
Still lookin' for The Wiener.
Since you know what a "weiner/wiener" looks like, why don't you design and make one? I am sure some people on this board will be more than happy to buy it from you when the "weiner/wiener" is available, myself included.
Regards,
So you are basically telling someone else what to do ? If I am not mistaking you both have the same goal. OK, in the same way of thinking: why don't you combine forces and work together making the Wiener ?
BTW Wiener comes from Wien, the Austrian capital. I wouldn't know why letters are switched but I can guess who do this.
I think I posted picture earlier, parts aren't too big if you sand down part of heatsink bottom that doesn't touch heatpad tpa3116
referring to post #6362,
Not wishing to cast any aspersions, or start a flame war.
You miss the point.
A decently done 2.1 board would allow US to then design a complete system with a great deal of simplicity, and flexibility.
For EXAMPLE: I'd like to have a semi portable system for my laptop, that beats the SQ of the little crap speakers that you typically find available. If there were a decent 2.1 board available, then I could build speakers to suit the application.
There are all different types of folks here, and not everyone following this thread is here to tweak endlessly. Some of us just want to build things to use...
John
Not wishing to cast any aspersions, or start a flame war.
You miss the point.
A decently done 2.1 board would allow US to then design a complete system with a great deal of simplicity, and flexibility.
For EXAMPLE: I'd like to have a semi portable system for my laptop, that beats the SQ of the little crap speakers that you typically find available. If there were a decent 2.1 board available, then I could build speakers to suit the application.
There are all different types of folks here, and not everyone following this thread is here to tweak endlessly. Some of us just want to build things to use...
John
Last edited:
2.1
The car / auto amps have had these features for - well - decades. Switchable filter types with variable frequency controls are the norm. The Vibe Slick Stereo 4 is just one of many Class D examples - offering 4 channels (2 stereo pairs) - bridging - LP or HP 50 - 750Hz variable filter for one stereo pair & HP 0 - 200Hz for the other.
With regards to the speakers - I should imagine that market has great diversity in the range of drivers / setups & enclosures - not so dissimilar to our market?
The car / auto amps have had these features for - well - decades. Switchable filter types with variable frequency controls are the norm. The Vibe Slick Stereo 4 is just one of many Class D examples - offering 4 channels (2 stereo pairs) - bridging - LP or HP 50 - 750Hz variable filter for one stereo pair & HP 0 - 200Hz for the other.
With regards to the speakers - I should imagine that market has great diversity in the range of drivers / setups & enclosures - not so dissimilar to our market?
For sure.. What PVCC, what speakers, what cables, what PSU?
Any measurements? So 100n changes everything? Speaks for a bad system design, IMHO.
Regards.
100nF on redboard is like 35mm away from tpachip, even past electrolytics on red ampboard left side, putting them near on on tpachippins should/could give a result audible 🙂
Hi doctormord.
I use ordinary household wire for speakers and power supply.
The power supply is a transformer based and regulated with LM 388.
No measurements and no interest to invest in equipment to do that.
I am just a solder gun slinger trying out suggestions from senior members here during my spare time.
When compared to you, I am obviously a novice with badly designed systems.
I wish to thank other members here for making my life a bit more interesting with caps here and there.
cheers
kp93300
I use ordinary household wire for speakers and power supply.
The power supply is a transformer based and regulated with LM 388.
No measurements and no interest to invest in equipment to do that.
I am just a solder gun slinger trying out suggestions from senior members here during my spare time.
When compared to you, I am obviously a novice with badly designed systems.
I wish to thank other members here for making my life a bit more interesting with caps here and there.
cheers
kp93300