Can you elaborate on this a bit?
I was trying to use this article, Understanding output filters for Class-D amplifiers, to make sense of the output side of this board. I thought the combination of three capacitors, one inductor and one resistor looked most like "Figure 5: Common-mode filter response with bypass capacitors in parallel with the damping resistors" from the article. But to be honest that was just a guess on my part.
I want to add that labeled picture to the Wiki but thought it best to check it here before posting.
I believe, if you read through this entire thread (as well as the lengthy ones on AudioCircle and AudioKarma), you'll find just about everything capable of outputting the correct voltage has been used.
IIRC, the range is 6 to 24 Volts.Lots of people are using laptop power supplies, of which 19V is very common. These are cheap and readily available, although quality is also all over the place.
There are a few reports of folks getting great results using batteries: sealed lead acid (SLA) and LiFePO4 being the most common suggestions.
Based on anecdotal information, high-current, high-quality regulated linear power supplies seem to give the best sound quality. In particular, the big supplies built for amateur radio such as the Astron RS-12A or Ten Tec 937. These are semi-expensive to buy new, but if you're patient, you ought to be able to find one in good shape at a reasonable price on Craigslist or eBay.
But the whole point (in SMPS as well) is the higher the frequency the smaller the inductors.
If I push the 3dB point from 40kHz to 80kHz it will still get rid of the switching frequency to a level that is acceptable to me.
And if 400kHz is too low the 3116 can go to 1200kHz.
true you can use smaller filters when going to higher switching frequencies.some people seemed confused what filter components into a load do.
the table (excerpt from) you responded to shows how certain components filter, those components in signalpath filter that way 20 years ago and also in 20 years. the fact that they seem to be aiming at -3dB for ~40khz means that switching frequency they had in mind was indeed 400khz(decade=40khz).
some might be looking or experimenting with increasing freq. to 1 Mhz, thats why they are so interested in PCB layout. that would mean you could drive higher loads with smaller filters LOL don't forget many people are using 8 ohm speakers where 20khz impedance is above 16 ohm sometimes far above...
According to the data sheet the gain setting is marked wrong. It is formed by the two resistors and the Input DC-Coupling caps, not the small cap marked on the photo.
Thats not how I read it. The section describing gain resistors does mention the input caps but purely because the gain setting determines impedance which in turn has an effect of the filter characteristics of the DC blocking cap. Gain setting is done just by the 2 resistors but the adjacent cap is associated with that part of the circuit according to the schematic.
gainsetting is 1 or 2 resistors
Ok, we just misunderstand each other. Gainsetting is off course only the two resistors, nothing else. But, as soon as the gain is changed the input capacitors also need to be changed because of the lower corner frequency of the high pass filter they form together with the gain setting resistors.
Note that the gain resistors act as a divider and only set a voltage which is then read by the 3116's on-board course-ADC at startup. This voltage then tells the 3116 to set the gain internally - so the input impedance change is internal and not directly tied to the resistor that sets the gain.
I want to chime in as I have been reading this thread and finally got my YJ blue board. I was initially following Rhing's TDA7297 thread and then to here for the 3116D2.
My blue board is stock, no mod. Supply is a Dell 19v SMPS claiming 2.1A output (I know it is weak). Only thing added is 3x4700uF Mitsubishi tank caps between the SMPS and the board. Signal source is my usual ipod touch 4g line output.
I have about 6-7 hours run time and I have to say the board sounds pretty good without mod. Bass is good. I have another Sure TK2050 heavily modded with the exact same 10uH toroids Rick uses (I posted pics before and I wonder if he got this idea from me? "hehe"). http://www.diyaudio.com/forums/class-d/224206-ideas-make-use-space-left-inside-enclosure-2.html
The YJ blue board sounds at least as good as the modded TK2050 board. Female vocals definitely warmer and more upfront compare to the Tripath.
Next plan is to use linear supply. I have one already but the thing is too big to fit inside the chassis so I have to take it apart, swapped out the TO-3 2n3055 with something like TO-220 so I can mount with the transformer turn 90 degree etc (pls disregard what I just said). Will also get the toroids, I know how much improvement they can make.
-AC
My blue board is stock, no mod. Supply is a Dell 19v SMPS claiming 2.1A output (I know it is weak). Only thing added is 3x4700uF Mitsubishi tank caps between the SMPS and the board. Signal source is my usual ipod touch 4g line output.
I have about 6-7 hours run time and I have to say the board sounds pretty good without mod. Bass is good. I have another Sure TK2050 heavily modded with the exact same 10uH toroids Rick uses (I posted pics before and I wonder if he got this idea from me? "hehe"). http://www.diyaudio.com/forums/class-d/224206-ideas-make-use-space-left-inside-enclosure-2.html
The YJ blue board sounds at least as good as the modded TK2050 board. Female vocals definitely warmer and more upfront compare to the Tripath.
Next plan is to use linear supply. I have one already but the thing is too big to fit inside the chassis so I have to take it apart, swapped out the TO-3 2n3055 with something like TO-220 so I can mount with the transformer turn 90 degree etc (pls disregard what I just said). Will also get the toroids, I know how much improvement they can make.
-AC
Last edited:
I added a note to my original annotated picture, calling it "version 1.0" and noting that there were some inaccuracies. I didn't want to just modify the original picture in-place, in case people reviewing the thread get confused (i.e. retain context).
So, here's version 2.0, how does this look?
So, here's version 2.0, how does this look?
An externally hosted image should be here but it was not working when we last tested it.
I found the Bourns accidentally from a local surplus store. Too bad they don't have it anymore. I was about to junk the Sure TK2050 board. Seriously, the Sure TK2050 was a big disappointment in stock form and it did not sound right. I put it on the TK2050 board thinking I have nothing to loose and have a big surprise for improvement. Definitely worth a try IMHO.
which 10u CC inductors? rest of board is standard unmodded? do you know high freq impedance of your speakers ?
Hi, today I received my Feixiang 2.1 amp and before to plug it in I saw a lot of thermal paste on the TPA3116 chip pins. I want to ask if its safe to plug it in or I have to clean the thermal paste from pins? As well I ordered a LT1083 linear adjustable power suply to power this amp board.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Not all thermal paste is conductive, some is though. If you have a multimeter you could try and measure it. Cleaning it off can't hurt of course (well, unless you damage something while cleaning
). Btw, the specs say the LT1083 is a voltage regulator with a 5V output, that is about the minimum you can use with a TPA3116 (protection kicks in below 4.5V) and will give you very low power output (something like 5-6W into 4 ohm).
Last edited:
This is adjustable version of the regulator, see here LT1083CP LM317 LM1086 Power Supply DC 5V | eBay, but I don't know how original is the regulator because I checked radionics and the price for the regulator only is 15 euro.Not all thermal paste is conductive, some is though. If you have a multimeter you could try and measure it. Cleaning it off can't hurt of course (well, unless you damage something while cleaning
Btw, the specs say the LT1083 is a voltage regulator with a 5V output, that is about the minimum you can use with a TPA3116 (protection kicks in below 4.5V) and will give you very low power output (something like 5-6W into 4 ohm).
@eplates,
may I respectfully disagree? Did the same thing, replaced the Würth I had put in some days ago by the CC - any compared to those I find the Würth mere toys. The CC present a lot more detaily and the bass becomes pretty outstanding.
To sum up: original to Würth: veil lifted, Würth to CC: a whole curtain has been removed (to my ears, of course).
may I respectfully disagree? Did the same thing, replaced the Würth I had put in some days ago by the CC - any compared to those I find the Würth mere toys. The CC present a lot more detaily and the bass becomes pretty outstanding.
To sum up: original to Würth: veil lifted, Würth to CC: a whole curtain has been removed (to my ears, of course).