[Markvdv]

Mark, thanks for the info on the inductors on Audiobah board - I was just wondering that as I have some other 10uH inductors and was wondering whether to swap 'em. I'm guessing the filter will be compromised if I do...? I'm listening to the board now on speakers with CSS FR125sr (drivers 7ohm nominal) but will be also using the boards with Alpair 7.3 and probably others too.

Gain on Audiobah board is currently 26db according to EBay listing.

I would think you found that 22uH inductors are best choice for both Full Range speakers mentioned above. i received a blueblack first time for me, outputfilter caps are 560nF, not 680nF, YJ did that earlier with redboard I read, perfect 3ohm filter now I think, first impression, distorted like oversteered (is this english?) but will let it play for few days before serious opinion.

edit: dutch "overstuurd'' is probably overdriven in english not oversteered?

[dboy]

There is informative discussion on diyaudio at the moment about bad cap choice (voltage ratings) on those boards. Maybe Audiobah spec'ed his better and that is why that board sounds more refined(?).

[Markvdv]

All chinese parts could be something else then expected, with smd's you are really blindly depending on that they put the correct part in, when part shows value 560nF and pcb silkscreen tells you 680nF you can see they changed something. I have seen pictures on diyaudio were different sized SMD capacitors were used in L and R channel, then it is obvious too:)

Diyaudio wiki on this amp seems to differ completely from what the TI designers themselves tell about the bootstrap part BTW, if you look at TI evaluation module, were all distortiondata in datasheet originate from, on that TI EVM the bootstrap is the lowest spec you find on the entire module(all other ceramics used have better classification), BOM is in guide for EVM. There also aren't any layout recommendations regarding the bootstrap caps that I found. Lowest specced part choice by TI and no special considerations mentioned suggest the part is not critical. It could be it is the point were you could improve upon TI EVM.

Will try to measure bootstrap on blueblack later today, there is still bad SQ I hear from new blueblack. If I use unmodded sure3110 as a reference this blueblack seems much worse, not close. So something must be wrong.

[dboy]

I found the blue/black sounded excellent with less good speakers. With my better speakers (still diy) it still sounded very impressive in the detail and in the way the music jumped out at you, especially after the limited whatever-to-hand mods I made. I would not have called it distortion, but something there in the sound made it less than involving however - impressive but not entirely listenable long-term. I assumed, perhaps incorrectly, that it was the input capacitors. After giving the Audiobah board only an hour or so to warm-up, the sound just did not seem to have that issue. As I said before, it just seemed more 'right'.

Without wishing to offend anyone or drag the discussion down in any way, I wonder if the valve pre-amps are serving to mask or smooth over that subtle distortion that you (Markvdv) are hearing and that I perhaps sensed subliminally without being able to put my finger on? Only a thought, and I speak as someone who ordered a C&C unity-gain valve buffer kit to try out (still waiting for delivery).

[dboy]

In defense of the Chinese makers of these boards, I imagine they have quite a Darwinian struggle to survive with us lot constantly comparing the prices of their products and purchasing the cheapest. I guess they adapt their practices in order to succeed in that particular marketplace. 

[Markvdv]

I wonder(ed) about the tubepre too:) The version I got originally only added loads of distortion and noise and compressed dynamics. With that solved differences are much much smaller. Noise, only high frequency now, is louder, but so is total gain a little, there is a little more then just added gain I guess however, unhearable from listening position, but closing in on loudspeaker audible when stopping for some time at 1 meter. High frequencies are actually a little louder with my tubepre then without it, no rolloff, high frequency detail is lost a little, Panasonic FM/FR ecap lose more detail on my board for example, tube doesn't sound harsh when distorting, some bad recordings sound a little nicer not much. Real low frequency now is about same strenght, but again little less detail in lowest frequencies. Bass/mid in general does sound warmer, fuller and that might be part of sounding a little more dynamic. I use Twaron all-over the tube pcb, it is a big bulb of yellow ultrafine kevlarlike wire I see, no magical glow

[Markvdv]

boottrap capacitor measured on pcb's:

sure3110: 155/167/175/163 nF ceramic original
sure3110: 171/185/173/181 nF ceramic original

hiampmini: 224/220/219/221 nF film
hiampmini: 218/224/218/223 nF film

blueblack: 234/223/216/224 nF film

so not value related unfortunately, blueblack does have longest path

[dboy]

Now I am curious - what does the Twaron do?

[Markvdv]

Basicly they finish what the rubber rings around the tubes missed
Ceramics are also microphonic, but not quite as much as these tubes

[ClefChef]

All chinese parts could be something else then expected, with smd's you are really blindly depending on that they put the correct part in, when part shows value 560nF and pcb silkscreen tells you 680nF you can see they changed something. I have seen pictures on diyaudio were different sized SMD capacitors were used in L and R channel, then it is obvious too:)

Diyaudio wiki on this amp seems to differ completely from what the TI designers themselves tell about the bootstrap part BTW, if you look at TI evaluation module, were all distortiondata in datasheet originate from, on that TI EVM the bootstrap is the lowest spec you find on the entire module(all other ceramics used have better classification), BOM is in guide for EVM. There also aren't any layout recommendations regarding the bootstrap caps that I found. Lowest specced part choice by TI and no special considerations mentioned suggest the part is not critical. It could be it is the point were you could improve upon TI EVM.

Will try to measure bootstrap on blueblack later today, there is still bad SQ I hear from new blueblack. If I use unmodded sure3110 as a reference this blueblack seems much worse, not close. So something must be wrong.

TPA3116D2 is a digital amplifier working with square impulses at close to 400 Khz. At these frequencies any extra length of signal path results in distortion of digital signal, loss of signal shape, timing problems, noise, etc. etc. Components have to be located as close to the chip as possible and only SMD components may be suitable.

The "Blue" board, with all its "through the hole" components, is good for modifications, but it's bad for sound quality. All those extra "legs" on the caps and resistors add parasitic capacitance and inductance. In my experience the "Green" board from Audiobah on eBay may be the only decent board available.

[wushuliu]

TPA3116D2 is a digital amplifier working with square impulses at close to 400 Khz. At these frequencies any extra length of signal path results in distortion of digital signal, loss of signal shape, timing problems, noise, etc. etc. Components have to be located as close to the chip as possible and only SMD components may be suitable.

The "Blue" board, with all its "through the hole" components, is good for modifications, but it's bad for sound quality. All those extra "legs" on the caps and resistors add parasitic capacitance and inductance. In my experience the "Green" board from Audiobah on eBay may be the only decent board available.

lol. that's hilarious.

[Markvdv]

So I am comparing left blueblack with right hiampmini. From looking at ebay pictures most will not see the difference in size:) The hiampmini only has through hole components surrounding the tpa3116 chip and for powersupply decoupling that does mean the 0.1uF capacitor has a longer signal path. This might be reason high frequency detail disappears when using Panasonic FM/FR, not sure, but could be. I actually have 2 hiampmini's in PBTL that I compare to 1 BTL blueblack and none of the components on my hiampmini boards came with the boards, they came from Mouser or Digikey.

When listening to blueblack I get the feeling I need to turn up the volume, and when I turn the pot higher it feels like nothing happens till a violin for example splits my brain, auwch, highs are edgy. All frequencies are there, real low bass detail is missing, but it is there. Cymbal detail is there. Whatever I focus on seems to be OK, except high frequencies getting ugly, but I do got the 3 ohm filter version it seems, yellow 560nF caps in output, not the blue 680nF like the picture shows. Everything seems heldback, it is comparable somewhat with connecting a low efficiency small loudspeaker to an amp that can't really drive them, the small loudspeakers often need a big amp to come alive. Clarity, transparancy I guess are words coming close. Funny thing is listening for more then 20 minutes and question comes up: what was I missing??? I can't hear the heldback effect any more, the brain seems to have corrected it, because switching to another piece of music immediatly brings the feeling back, I want to push it, music doesn't sound open and after a while it is corrected again between my ears Switching to hiampminis does give a sense of relief, it is much easier to listen to, it sounds open and there is clarity without the high edge.

First thing I will change will be the yellow 560nF, I hope that cleans up the high frequencies and edginess

[ClefChef]

lol. that's hilarious.

Just take out your scope and function generator and measure the TDA3116 output at various points, look at the signal shape.

I trust you do have a scope, bro 

[wushuliu]

I'm more inclined to believe the audiobah sounds better stock because the guy actually uses properly specced *genuine* high quality components as well as a quality pcb. There's a reason why audiobah costs more than the other boards. IIRC he got direct feedback on his design at diyaudio and followed/follows the thread there.

I've listened to other boards just as compact and they didn't sound a whole lot different from larger ones.

[HaroldHill]

So I am comparing left blueblack with right hiampmini. ... Switching to hiampminis does give a sense of relief, it is much easier to listen to, it sounds open and there is clarity without the high edge.

Thanks, Mark. Aside from loosing the top end glare and gaining openness, are there other audible differences? I'm thinking bass quality, vocal quality, etc.

[Markvdv]

It is like driving with the parkingbrake on. Energy is lost somewhere. On low volume blueblack loses everything, music partly disappears, hiampmini sounds the same whatever volume is, so does sure3110, you still get a complete soundpicture.
Everything sounds like the brake is on with blueblack here, except maybe distorted high frequencies, but that is the 560nF caps I mentioned I think/hope.

(Sound of 3110 is close to and clearly related to hiampmini, my blueblack is totally different, I think I will check if it is a 3116 under the heatsink, blindly you would never ever guess it is the same amp)

[wushuliu]

It is like driving with the parkingbrake on. Energy is lost somewhere. On low volume blueblack loses everything, music partly disappears, hiampmini sounds the same whatever volume is, so does sure3110, you still get a complete soundpicture.
Everything sounds like the brake is on with blueblack here, except maybe distorted high frequencies, but that is the 560nF caps I mentioned I think/hope.

(Sound of 3110 is close to and clearly related to hiampmini, my blueblack is totally different, I think I will check if it is a 3116 under the heatsink, blindly you would never ever guess it is the same amp)

Sounds like your blueblack board is defective.

[dboy]

I'm more inclined to believe the audiobah sounds better stock because the guy actually uses properly specced *genuine* high quality components as well as a quality pcb. There's a reason why audiobah costs more than the other boards. IIRC he got direct feedback on his design at diyaudio and followed/follows the thread there.

I've listened to other boards just as compact and they didn't sound a whole lot different from larger ones.

The Audiobah board is not as compact as the Hiamplifier board in the picture Mark posted - looks closest in size to the red YJ board although slightly smaller. Where Audiobah differs is in using all SMD components with the exception of 2x470uF electrolytic caps.

Are you mixing the two up? I may be wrong, but I seem to remember seeing Hiamplifier posting at diyaudio but not seeing Audiobah. I have not read the whole thread though so not sure.

[wushuliu]

The Audiobah board is not as compact as the Hiamplifier board in the picture Mark posted - looks closest in size to the red YJ board although slightly smaller. Where Audiobah differs is in using all SMD components with the exception of 2x470uF electrolytic caps.

Are you mixing the two up? I may be wrong, but I seem to remember seeing Hiamplifier posting at diyaudio but not seeing Audiobah. I have not read the whole thread though so not sure.

Yeah I meant hiamplifier. Certainly if I was looking to diy but not do any mods the hiamplifier is the one I would get. I just don't believe the trace lengths have much to do with why hiamp board may sound better. In Markvdv's case it sounds like the there is a problem with the board.

[ClefChef]

The Audiobah board is not as compact as the Hiamplifier board in the picture Mark posted - looks closest in size to the red YJ board although slightly smaller. Where Audiobah differs is in using all SMD components with the exception of 2x470uF electrolytic caps.

I didn't have Hiamplifier board, but looking at the pictures I'd say that Audiobah board may be similar size if not smaller. It is less than half size of the blueback. Audiobah is the only board available (as far as I know) that has mute circuitry implemented.