I am using a tpa 3123 board that has easy to assess AVCC pins. The sound improvement is much more the bootstrap snubber mod after implementing the isolation of the Avcc pin with resistors. It works with 3.3 R and 10R. I measure the curent draw of
the avcc of each chip is about 12ma. The board uses 2TPA3123 chips.
the avcc of each chip is about 12ma. The board uses 2TPA3123 chips.
That current draw for AVCC is a useful figure to have, thank you - so it was a reasonable first guess it turns out to aim for a total of 20mA, which leaves 8mA for the TL431 to regulate with.
wo0t! I fixed my SMSL 36Apro. Finally was able to adequately reflow all the output connection points and got it playing again! 😀
I'm going to bring it to my brother's place tomorrow for a birthday party in an antique barn. Viva la TPA3118!
I'm going to bring it to my brother's place tomorrow for a birthday party in an antique barn. Viva la TPA3118!
I think it was merely a weak solder joint that became intermittent and then broke (on my first try to re-solder 😀 ) on the left negative output.
You may get in trouble doing when you switch on/off the amp, as then the PVCC/AVCC rails are going to have different charge-/discharge time constants which may lead in a huge voltage difference between both inputs.
Just measured a TPA3118 - there's no diode drop between AVCC and PVCC in either direction. Gonna try running a wiener card with a dual output bench supply shortly, to see what happens.
Perceval,
Sorry🙂. My interpretation is that one has to join the solder bridge at each of the connection point (S2, S3 ....., S6) and break the S1 joint. As both xrk and I pointed out, each of the those connection point consists of 2 separate solder points. I do not think you have to link all the connection points. I think I am correct but please do not hold me responsible for any mishap. Please check out this link from the manufacturer for illustration.
TPA3116Êý×ÖС¹¦·Å°å12V ´ó¹¦ÂʳÉÆ·°å/´ø¿ª¹ØµçλÆ÷/²¢½Óµ¥ÉùµÀ-ÌÔ±¦Íø
I presume you can write Chinese, you can contact the manufacturer for confirmation.
In general, for Class D amps, one should not power on the amp without a load (someone please correct me if I am wrong). The output filter could be damaged.
This is a nice amp. Hope you enjoy it as much as I am.
Regards,
With no load the output inductors rings with high voltage.
This can kill the output filter capacitor.
I always over rate the output capacitor voltage in my designs to get around this.
The irs2092 can act oddly with no load.
It either has the output sitting at 17VDC or it oscillates OK.
I have seen both states.
Interesting. I hooked my up to a 24 V SMPS. I cannot hear any hisses at all even when I placed my ear next to the tweeter with the volume pot turned on full. However, I did hear a light pop when I turn on the amp using the built-in on-off switch - none of you guys experienced that
.Would you mind explaining " analog power in RC mod"? Thanks!
Regards,
Interesting. I hooked my up to a 24 V SMPS. I cannot hear any hisses at all even when I placed my ear next to the tweeter with the volume pot turned on full. However, I did hear a light pop when I turn on the amp using the built-in on-off switch - none of you guys experienced that
Would you mind explaining " analog power in RC mod"? Thanks!
Regards,
I did not put my ear next to speaker, but was about 1 ft away from an 85dB sensitive driver and listened carefully and heard no audible pop when turning on off via pot knob. I suspect that the knob actuates either the SD or MUTE functions rather than truly shutting off.
Lo_Tse, have you applied any mods to yours or running it bone stock still?
xrk,
You may be right. I wonder whether the pop has anything to do with the higher gain of my amp. I recalled when I was using a TPA3123 board, the gain setting (it has a dip switch for selecting gain levels) has a direct impact on the hiss. As expected higher the gain, higher the noise background.
I suspected that during the manufacture of the amp, the control of components quality is non-existent - whatever components are available (cheap), they will use it. Thus, the quality of the resulting product varies over a wide range. It happens quite often that one board from one vendor is good whereas the "same" amp from a different source displayed some problems.
I have not modded anything yet. I am still contemplating how to approach it.
Increasing the value of the input caps should (theoretically) increase the base response. However, depending on the type of music one listens to, this mod might not provide much practical result. I do agree that using better input caps (e'g' Wima MKPs), if you can afford, is generally a good thing.
Swap out the power caps to better ones (e.g. your favorite OSCON) should improve things too. This is something I am thinking. However, the max voltage rating of the polymer caps are kind of low (typically ~25 V). I am thinking of using the Panasonic FRs, the ESR of these caps are quite low and one can get higher voltage rating.
I would also like to try the snubber mod since you reported good results. However, that's a bit more involved and depending on the board. However, this Sanmin board seems to provide much better access than the YJ blue/black (no need to scrap out anything to find the trace).
Personally, I think the out put filter (the inductor and cap) should have a big impact on the sound (just my hunch) and I would like to play with that. I have some decent Coilcraft coils in my parts drawer. Very few people has done that and they reported good results.
These 3116 amps seems to be quite insensitive to PS mod. Be it a brick, SMPS, or linear PS, the results are all good. Although some people in Audiocircle mentioned that the Anstron linear PS seemed to afford "better results" than SMPS. I recalled that you have good experience with some inexpensive 19V bricks. Besides, the PSRR of the 3116 is quite good according to the data sheet (> 76 db I think). So I am going to leave this alone.
Would appreciate your thoughts on modding since you have done a few.
Regards,
Please define "things" and why do you think so. Do you know the the actual original parameters?
By things, I mean the general aspects - lower leakage, tighter tolerance etc. Better quality caps should perform more closely to the design parameters. Am I right. The TI data sheet described the recommended parameters, the input cap values has a direct impact on the low frequency response of the amp. However, as I mentioned before, practical experience indicated that the result may not be significant, depending on your music taste and of course , your speakers response. I would not worry about the cap values too much be it 1 or 3.3 or 5 uf, if you speaker only can do -3 db at 45 Hz.
Regards,
When a stock no name Chinese cap is swapped out for an OSCON, the ultra low ESR seems to improve "things" such as more articulate bass, higher definition of sharp percussive sounds, improved clarity, more bass peak headroom. One of the best bang for the buck mods worth many times the prices of $1.80 SEPF 330uF 25v cap.
Firstly, you don't know the quality of the actual caps, secondly, why bother for leakage current which wont be in the >= mA range.
From your your saying, i would understand that you're thinking the "engineer" who designed this board were/is a noob not knowing what he is doind regarding the specs.
This is just by 1/2*pi*R*C where R = input impedance which is directly related to the amplifiers gain. So changing the gain will increase/decrease your lower cut-off-frequency when not changing the caps as well.
So just changing anything for "known to be good in some way" is not the way to go. (Imho)
🙂
I would really appreciate if someone could send me 12 of those 25SEPF330 OS-CON-s to Croatia (EU) for a price like $1.80 a piece + shipment fee 😀
Really, no joking, I can't find decent supplier with cheap delivery prices for such a small package 🙁
What do mean by you do not know the quality of the actual caps? Doesn't tolerance reflects quality, to a large extent? When Wima sells me a MKP claiming 2.5% tolerance, it is a lie? So it is better to use a ceramic that claims +- 50% tolerance and no one knows what is the "actual" value.
From your saying, I believe you meant I do not know anything. You may be right, or not. That's OK, everyone entitles his/her opinion. I can take insult. Would you elaborate your view/opinion with details instead of just giving negative remarks. You might be the one who "knows it all", but instead of just criticizing, you can utilize facts and data and theory to support your opinion and view. I know, you believe that there is no point in wasting time with people who apparently less knowledgeable than yourself. And it is so much more enjoyable just to criticize and show people how good you are. I have seen quite a few in this forum.
If you read my post, I was specifically talking about caps and what I wrote is correct, is it not?
Would you elaborate on what is "the way to go" then.